![sensors
Article
Determination of Optimal Measurement Points for
Calibration Equations—Examples by RH Sensors
Hsuan-Yu Chen 1 and Chiachung Chen 2,*
1 Department of Materials Science and Engineering, University of California, San Diego, CA 92093, USA;
wakaharu37@gmail.com
2 Department of Bio-Industrial Mechatronics Engineering, National ChungHsing University,
Taichung 40227, Taiwan
* Correspondence: ccchen@dragon.nchu.edu.tw; Tel.: +886-4-2285-7562
Received: 26 February 2019; Accepted: 6 March 2019; Published: 9 March 2019
Abstract: The calibration points for sensors must be selected carefully. This study uses accuracy and
precision as the criteria to evaluate the required numbers of calibration points required. Two types
of electric relative humidity (RH) sensors were used to illustrate the method and the standard RH
environments were maintained using different saturated salt solutions. The best calibration equation
is determined according to the t-value for the highest-order parameter and using the residual plots.
Then, the estimated standard errors for the regression equation are used to determine the accuracy of
the sensors. The combined uncertainties from the calibration equations for different calibration points
for the different saturated salt solutions were then used to evaluate the precision of the sensors. The
accuracy of the calibration equations is 0.8% RH for a resistive humidity sensor using 7 calibration
points and 0.7% RH for a capacitance humidity sensor using 5 calibration points. The precision is less
than 1.0% RH for a resistive sensor and less than 0.9% RH for a capacitive sensor. The method that
this study proposed for the selection of calibration points can be applied to other sensors.
Keywords: calibration points; saturated salt solutions; humidity sensors; measurement uncertainty
1. Introduction
The performance of sensors is key for modern industries. Accuracy and precision are the most
important characteristics. Calibration ensures sensors’ performance. When a sensor is calibrated, the
reference materials or reference environments must be specified. For a balance calibration, a standard
scale is the reference materials. For temperature calibration, the triple point of ice-water or boiling
matter is used to maintain the reference environment.
The experimental design for calibration must consider the following factors [1–3].
1. The number and the location of the calibration points.
2. The regression equations (linear, poly-nominal, non-linear).
3. The regression techniques.
4. The standard references and their uncertainties.
Betta [1] adopted minimizing the standard deviations for the regression curve coefficients or the
standard deviation for the entire calibration curve to design an experiment to determine the number
of calibration points, the number of repetitions, and the location of calibration points. Three types of
sensor were used to demo the linear, quadratic and cubic calibration equations: a pressure transmitter,
a platinum thermometer and E-Type thermocouple wires. The estimated confidence interval values
were used to determine the validity of the regression equation. This method was extended to address
calibration for complex measurement chains [2].
Sensors 2019, 19, 1213; doi:10.3390/s19051213 www.mdpi.com/journal/sensors
1](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-16-320.jpg)
![Sensors 2019, 19, 1213
Hajiyev [3] noted the importance of the selection of the calibration points to ensure the accuracy
of the calibration and the optimal selection of standard pressure setters and used an example to verify
the method. A dispersion matrix, →
D
of the estimated coefficients was defined and this matrix →
D
was
used as a scale of the error between the sensor and the reference instruments. Two criteria were used
to evaluate the performance. The minimized sum of the diagonal elements of the matrix →
D
is called
the A-optimality criterion. The minimized of the generalized of determinant of the matrix →
D
is called
the D-optimality criterion. The optimal measurement points for the calibration of the differential
pressure gages were determined using the A-optimality criterion [3] and the D-optimality criterion [4].
Khan et al. [5] used an inverse modeling technique with a critical neural network (ANN) to evaluate
the order of the models and the calibration points. The root-mean-square error (RMSE) was used as
the criterion.
Recently, modern regression has been used as an important role to express the quantitative
relationship between independent and response variables for tests on a single regression
coefficient [6–9]. This technique used to address calibration equations and the standard deviations of
these calibration equations then served as the criteria to determine their accuracy [10,11].
The confidence band for the entire calibration curve or for each experimental point was used to
evaluate the fit of calibration equations [1,2]. The concept of measurement uncertainty (MU) is widely
used to represent the precision of calibration equations [12–14]. Statistical techniques can be used to
evaluate the accuracy and precision of calibration equations that are obtained using different calibration
points [15–17]. Humidity sensors that were calibrated using different saturated salt solutions were
tested to illustrate the technique for the specification of optimal measurement points [18,19].
Humidity is very important for various industries. Many manufacturing and testing processes,
such as those for food, chemicals, fuels and other products, require information about humidity [20].
Relative humidity (RH) is commonly used to express the humidity of moist air [21]. Electric
hygrometers are the most commonly used sensors because they allow real-time measurement and are
easily operated.
The key performance factors for an electrical RH meter are the accuracy, the precision, hysteresis
and long-term stability. At high air humidity measurement, there is a problem with response time of the
RH sensors in conventional methods. The solution for this problem for high air humidity measurement
is to use an open capacitor with very low response time [22–24] and quartz crystals which compensate
temperature drift. An environment with a standard humidity is required for calibration. Fixed-point
humidity systems that use a number of points with a fixed relative humidity are used as a standard.
A humidity environment is maintained using different saturated salt solutions. The points with a fixed
relative humidity are certified using various saturated salt solutions [19]. When the air temperature,
water temperature and air humidity reach an equilibrium state, constant humidity is maintained in the
air space [19].
The RH value that is maintained by the salt solutions is of interest. Wexler and Hasegawa
measured the relative humidity that is created by eight saturated salt solutions using the dew point
method [25]. Greenspan [18] compiled RH data for 28 saturated salt solutions. The relationship
between relative humidity and ambient temperature was expressed as a 3rd or 4th polynomial
equation. Young [26] collected RH data for saturated salt solutions between 0 to 80 ◦C and plotted
the relationship between relative humidity and temperature. The Organisation Internationale De
Metrologies Legale (OIML) [19] determined the effect of temperature on the relative humidity of
11 saturated salt solutions and tabulated the result. Standard conditions, devices and the procedure for
using the saturated salt solutions were detailed.
The range for the humidity measurement is from about 11% to 98% RH. Studies show that
the number of fixed-point humidity references that are required for calibration is inconsistent.
Lake et al. [27] used five salt solutions for calibration and found that the residuals for the linear
calibration equation were distributed in a fixed pattern. Wadso [28] used four salt solutions to
determine the RH that was generated in sorption balances. Duvernoy et al. [29] introduced seven salt
2](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-17-320.jpg)
![Sensors 2019, 19, 1213
solutions to generate the RH for a metrology laboratory. Bellhadj and Rouchou [30] recommended five
salt solutions and two sulfuric acids to create the RH environment to calibrate a hygrometer.
There is inconsistency in the salt solutions that are specified by instrumentation companies and
standard bodies. The Japanese Mechanical Society (JMS) specifies 9 salt solutions for the standard
humidity environment [31]. The Japanese Industrial Standards Committee (JISC) recommends 4 salt
solutions to maintain RH environment [32]. The Centre for Microcomputer Applications (CMA)
company specifies 11 salt solutions [33]. Delta OHM use only 3 salt solutions [34]. The OMEGA
company use 9 salt solutions [35]. TA instruments specifies 9 salt solutions [36] and Vaisala B.V. select
4 salt solutions [37]. These salt solutions are listed in Table 1.
Table 1. The selection of saturated salt solutions that are used to calibrate humidity sensors.
Salt Solutions
OIMI
[19]
Lake
[27]
Wadso
[28]
Duvernoy
[29]
Belhadj
[30]
JMS
[31]
JISC
[32]
CMA
[33]
Delta
[34]
OMEGA
[35]
TA
[36]
Vaisala
[37]
LiBr *
LiCl * * * * * * * * *
CH3COOK * * * *
MgCl2·GH2O * * * * * * * * * *
K2CO3 * * * * * * *
Mg(NO3)2 * * * * * * *
NaBr * * * *
KI * * *
SrCl2 *
NaCl * * * * * * * * * * * *
(NH4)2SO4 *
KCl * * * * * * * *
KNO3 * * * *
K2SO4 * * * * * * * *
Note: OIML, The Organisation Internationale De Metrologies Legale.
Lu and Chen [17] calculated the uncertainty for humidity sensors that were calibrated using
10 saturated salt solutions for two types of humidity sensors. The study showed that a second-order
polynomial calibration equation gave better performance than a linear equation. The measurement
uncertainty is used as the criterion to determine the precision performance of sensors [38].
The number of standard relative humidity values for fixed-point humidity systems is limited
by the number and type of salt solutions. The number of salt solutions that must be used to specify
the calibration points for the calibration of RH sensors is a moot point. More salt solutions allow
more calibration points for the calibration of RH sensors. However, using more salt solutions is
time-consuming. This study determined the effect of the number and type of salt solutions on the
calibration equations for two types of humidity sensors. The accuracy and precision were determined
in order to verify the method for the choice of the optimal calibration points for sensor calibration.
2. Materials and Methods
2.1. Relative Humidity (RH) and Temperature Sensors
Resistive sensor (Shinyei THT-B141 sensor, Shinyei Kaisha Technology, Kobe, Japan) and
capacitive sensor (Vaisala HMP-143A sensor, Vaisala Oyj, Helsinki, Finland) were used in this study.
The specification of the sensors is listed in Table 2.
3](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-18-320.jpg)
![Sensors 2019, 19, 1213
Table 2. The specifications of two humidity sensors.
Resistive Sensor Capacitive Sensor
Model 1 THT-B121 HMP 140A
Sensing element Macro-molecule HPR-MQ HUMICAP
Operating range 0–60 ◦C 0–50 ◦C
Measuring range 10–99% RH 0–100%
Nonlinear and repeatability ±0.25% RH ±0.2% RH
ResolutionTemperature effect 0.1% RH (relative humidity)none 0.1% RH0.005%/◦C
2.2. Saturated Salt Solutions
Eleven saturated salt solutions were used to maintain the relative humidity environment. These
salt solutions are listed in Table 3.
Table 3. The Calibration points for saturated salt solutions to establish the calibration equations.
Salt Solutions
(n1 = 11)
Case 1
(n2 = 9)
Case 2
(n3 = 7)
Case 3
(n4 = 5)
Case 4
uc
LiCl * * * * 0.27
CH3COOK * 0.32
MgCl2 * * * * 0.16
K2CO3 * * * 0.39
Mg(NO3)2 * * 0.22
NaBr * * * * 0.40
KI * * 0.24
NaCl * * * * 0.12
KCl * * * 0.26
KNO3 * 0.55
K2SO4 * * * * 0.45
Note: uc values were obtained from Greenspan [18] and The Organisation Internationale De Metrologies Legale
(OIML) R121 [19].
2.3. Calibration of Sensors
The humidity probes for the resistive and capacitive sensors were calibrated using saturated salt
solutions. A hydrostatic solution was produced in accordance with OIML R121 [19]. The salt was
dissolved in pure water in a ratio such that 40–75% of the weighted sample remained in the solid state.
These salt solutions were stored in containers.
The containers were placed in a temperature controller at an air temperature of 25 ± 0.2 ◦C.
During the calibration process, humidity and temperature probes were placed within the container
above the salt solutions. The preliminary study showed that an equilibrium state is established in 12 h
so the calibration lasted 12 h to ensure that the humidity of the internal air had reached an equilibrium
state. Experiments for each RH environment were repeated three times. The temperature was recorded
and the standard humidity of the salt solutions was calculated using Greenspan’s equation [18].
2.4. Establish and Validate the Calibration Equation
The experimental design and flow chart for the data analysis is shown in Figure 1.
The relationship between the standard humidity and the sensor reading values was established
as the calibration equation.
This study used the inverse method. The standard humidity is the dependent (yi) and the sensor
reading values are the independent variables (xi) [17].
The form of the linear regression equation is:
Y = b0 + b1 X (1)
4](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-19-320.jpg)
![Sensors 2019, 19, 1213
2.6.1. Tests on a Single Regression Coefficient
The criteria to assess the fit of the calibration equations are the coefficient of determination R2, the
estimated standard error of regression s and the residual plots.
The coefficient of determination, R2 is used to evaluate the fit of a calibration equation. However,
no standard criterion has been specified [15,16].
The single parameter coefficient was tested using the t-test to evaluate the order of polynomial
regression equation. The hypotheses are:
H0 : bk = 0 (3)
H1 : bk = 0 (4)
The t-value is:
t = bk/se(bk) (5)
where bk is the value of the parameter for the polynomial regression equation of the highest order, and
se(bk) is the standard error of bk.
2.6.2. The Estimated Standard Error of Regression
The estimated standard error of regression s is calculated as follows:
s = (
(ŷ2 − yi)2
n1 − p
)
0.5
(6)
where ŷi is the predicted valued of the response, ŷi is the response, n1 is the number of data and p is
the number of parameters.
The s value is the criterion that is used to determine the accuracy of a calibration equations [38].
It is used to assess the accuracy of two types of RH sensors that are calibrated using different saturated
salt solutions.
2.6.3. Residual Plots
Residual plots is the quantitative criterion that is used to evaluate the fit of a regression equation.
If the regression model is adequate, the data distribution for the residual plot should tend to a
horizontal band and is centered at zero. If the regression equation is not accepted, the residual plots
exhibit a clear pattern.
For the calibration equation, tests on a single regression coefficient and the residual plots are
used to determine the suitability of a calibration equation for RH sensors that are calibrated using
different saturated salt solutions. The estimated standard error of the regression equations is then used
to determine the accuracy of the calibration equations.
2.7. Measurement Uncertainty for Humidity Sensors
The measurement uncertainty for RH sensors using different salt solutions was calculated using
International Organization for Standardization, Guide to the Expression of Uncertainty in Measurement
(ISO, GUM) [12,13,17].
uc
2
= u2
xpred + u2
temp + u2
non + u2
res + u2
sta (7)
where uc is the combined standard uncertainty, uxpred is the uncertainty for the calibration equation,
utemp is the uncertainty due to temperature variation, unon is the uncertainty due to nonlinearity, ures is
the uncertainty due to resolution, and usta is the uncertainty of the reference standard for the saturated
salt solution.
The uncertainty of xpred is calculated as follows [38]:
6](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-21-320.jpg)
![Sensors 2019, 19, 1213
uxpred = s
1 +
1
n
+
(y − y)2
∑(yi2) − (∑ yi)2
n
(8)
where y is the average value of the response.
The uncertainty in the value of uref for the saturated salt solutions is determined using the
reference standard for the salt solution. The scale and the uncertainty of these saturated salt solutions
are listed in Table 3 that are taken from Greenspan [18] and the Organisation Internationale De
Metrologies Legale (OIML) R121 [19]:
uref = (
∑(uri)2
N2
)
0.5
(9)
where uri is the uncertainty in the humidity for each saturated salt solution and N2 is the number of
saturated salt solutions that are used for calibration.
The calibration equations use different numbers of saturated salt solutions had its uncertainty.
This criterion is used to evaluate the precision of RH sensors.
The accuracy and precision of RH sensors that are calibrated using different saturated salt
solutions was determined using the s and uc values. By Equations (7)–(9), the contrast between the
number of saturated salt solutions is considered. The greater the number of data points that are used,
the smaller is the s value that is calculated by Equation (6). However, this requires more experimental
time and cost and the value of uref may be increased. The uncertainty of each calibration point is
different because different saturated salt solutions are used. The optimal number of calibration points
were evaluated by accuracy and precision.
3. Results and Discussion
3.1. The Effect of the Accuracy of Different Calibration Points
3.1.1. THT-B121 Resistive Humidity Sensor
Calibration equations for resistive sensors using 11 salt solutions:
The distribution of the relative humidity data for the reading values for a resistive sensor is
plotted against the standard humidity values that are maintained using 11 saturated salt solutions in
Figure 2.
Ϭ
ϭϬ
ϮϬ
ϯϬ
ϰϬ
ϱϬ
ϲϬ
ϳϬ
ϴϬ
ϵϬ
ϭϬϬ
Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ
ZĞĂĚŝŶŐǀĂůƵĞƐ͕й
^ƚĂŶĚĂƌĚǀĂůƵĞƐ͕й
Figure 2. The distribution of the relative humidity data for reading values versus the standard humidity
values for THT-B121 resistive humidity sensor using 11 saturated salt solutions (LiCl, CH3COOK,
MgCl2, K2CO3, Mg(NO3)2, NaBr, KI, NaCl, KCl, KNO3 and K2SO4).
7](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-22-320.jpg)
![Sensors 2019, 19, 1213
for the specification of an appropriate calibration equation. The estimated values of standard deviation,
s, is used to define the uncertainty for an inverse calibration equation [35]. The s values for the four
calibration equations are 1.6098, 1.4612, 0.9820 and 0.7719, respectively. It is seen that an appropriate
calibration equation gives a significant reduction in uncertainty.
Calibration equations for resistive sensor using 5 salt solutions:
The estimated parameters and the evaluation criteria for the regression analysis for 5 calibration
points for a resistive sensor are listed in Table 5. The residual plots for four calibration equations are
shown in Supplementary Materials. Similarly to the regression results for 11 salt solutions, the linear,
2nd and 3rd order polynomial equations all employed a systematic distribution in the residuals plots.
These equations are clearly not appropriate calibration equations. For a resistive sensor, the residual
plots for the 4th order polynomial equations presented a random distribution.
Table 5. Estimated parameters and evaluation criteria for the linear and several polynomial equations
for THT-B121 resistive sensors using 5 salt solutions.
Linear 2nd Order 3nd Order 4th Order
b0 −0.970118 −3.1191770 −12.201481 −19.471802
b1 1.0155235 1.12632754 1.8869907 2.743833
b2 −0.001007316 −0.01685101 −0.04766345
b3 9.34623 × 10−5 5.15689 × 10−4
b4 −1.93676 × 10−6
R2 0.9969 0.9974 0.9994 0.9991
s 1.8109 1.7146 0.7984 1.084
Residual plots clear pattern clear pattern clear pattern uniform distribution
The R2 values for the linear, 2nd, 3rd and 4th order polynomial calibration equations are 0.9969,
0.9974, 0.9994 and 0.9998, respectively. However, these higher R2 values do not provide relevant
information about the calibration equations. The s values represent the uncertainty of calibration
equations. For the linear, 2nd, 3rd and 4th order polynomial calibration equations are 1.8109, 1.7146,
0.7954 and 1.084, respectively. The 4th order polynomial equations is:
y = −19.471802 + 2.743833x − 0.047663x2 + 0.0005157x3 − 1.93676 × 10−6x4
(sb = 2.2789 sb = 0.25086 sb = 0.00869 sb = 0.000117 sb = 5.360 × 10−7
t = −8.5447 t = 10.9396 t = −5.4849 t = 4.3946 t = −3.6101)
R2 = 0.991, s = 1.014
The regression results for the 4th order polynomial equations using different calibration points
in different salt solutions are listed in Table 6. The results for 9 and 7 calibration points are similar to
those for 11 and 5 calibration points.
Table 6. Estimated parameters and evaluation criteria for the 4th order polynomial equations for
THT-B121 resistive sensors using four different calibration points.
Case 1
(n1 = 11)
Case 2
(n2 = 9)
Case 3
(n3 = 7)
Case 4
(n4 = 5)
b0 −20.530297 −23.41845561 −23.904948 −19.4718019
b1 2.8051965 3.5861653 3.243023015 2.743832845
b2 −0.04915334 −0.06230766 −0.06426625 −0.047663446
b3 5.39281 × 10−4 7.0951 × 10−4 7.34202 × 10−4 5.15689 × 10−4
b4 −2.07539 × 10−6 −2.81734 × 10−6 −2.92042 × 10−6 −1.93676 × 10−6
R2 0.9993 0.9994 0.9994 0.9991
s 0.7719 0.6951 0.8039 1.084
10](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-25-320.jpg)
![Sensors 2019, 19, 1213
The R2 value is used b to evaluate the calibration equations [27,33]. Even the linear calibration
equation for this study shows a high R2 value. However, the estimated error was higher than that for
other equations. The residual plots all exhibited a clear pattern distribution so the R2 value cannot
be used as the sole criterion to assess the calibration equation. Betta and Dell’Isola [1] mention R2,
Chi-square and F-test to verify the accuracy of a model. This study used t-value for a parameter was
used as the criterion. This method bases on statistical theory.
3.1.2. HMP 140A Capacitive Humidity Sensor
Calibration equations for a capacitive sensors using 11 salt solutions
The relationship between the reading values for a capacitive sensor and the standard humidity
values that are maintained using 11 saturated salt solutions is shown in Figure 4.
Ϭ
ϭϬ
ϮϬ
ϯϬ
ϰϬ
ϱϬ
ϲϬ
ϳϬ
ϴϬ
ϵϬ
ϭϬϬ
Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ
ZĞĂĚŝŶŐǀĂůƵĞƐ͕й
^ƚĂŶĚĂƌĚǀĂůƵĞƐ͕й
Figure 4. The distributions of relative humidity data for standard humidity values versus the reading
values for HMP 140A capacitance humidity sensors using 11 saturated salt solutions (LiCl, CH3COOK,
MgCl2, K2CO3, Mg(NO3)2, NaBr, KI, NaCl, KCl, KNO3 and K2SO4).
The estimated parameters and the evaluation criteria for regression analysis are listed in Table 7.
Table 7. Estimated parameters and evaluation criteria for the linear and polynomial equations for HMP
140A capacitive sensor using 11 salt solutions.
Linear 2nd Order
b0 −0.414520 3.479518
b1 1.031003 0.833274
b2 0.00186718
R2 0.9975 0.9994
s 1.4002 0.6837
Residual plots clear pattern Uniform distribution
The residual plots for the calibration equations for different orders of polynomial equations are
shown in Figure 5.
11](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-26-320.jpg)
![Sensors 2019, 19, 1213
Calibration equations for a capacitive sensor using 5 salt solutions
The estimated parameters and the evaluation criteria for the regression analysis for 5 calibration
points for a capacitance are listed in Table 8. The residual plots for four calibration equations are shown
in Supplementary Materials. Similarly to the regression results for 11 salt solutions, residuals plots
for the linear equation exhibit a systematic distribution. Residual plots for the 2nd order polynomial
equations presented a random distribution.
Table 8. Estimated parameters and evaluation criteria for the linear and polynomial equations for HMP
140A capacitive sensor using 5 salt solutions.
Linear 2nd Order
b0 0.226512 2.911321
b1 1.023088 0.814217
b2 0.00155423
R2 0.9981 0.9995
s 1.4386 0.7890
Residual plots clear pattern Uniform distribution
The R2 values for the linear and 2nd order polynomial calibration equations are 0.9981 and 0.9995,
respectively. The s values for the linear and 2nd order polynomial calibration equations are 1.4386 and
0.7890, respectively. The 2nd order polynomial equations give the smallest estimated errors and listed
as follows:
y = 2.9113205 + 0.864217x + 0.0015542x2, R2 = 0.9995, s = 0.7890
(sb = 0.63806 sb = 0.02925 sb = 0.000278
t = 74.5628 t = 29.543 t = 5.5872)
The regression results for the 2nd order polynomial equations using different calibration points in
different salt solutions are listed in Table 9. The results of R2 values for 5, 7, 9 and 11 calibration points
are similar. However, the calibration equation for 11 calibration points gives the smallest s value.
Table 9. Estimated parameters and evaluation criteria for the 2nd order polynomial equations for HMP
140A capacitive sensors using four different calibration points.
Case 1
(n1 = 11)
Case 2
(n2 = 9)
Case 3
(n3 = 7)
Case 4
(n4 = 5)
b0 3.479580 3.156891 2.871078 2.9113205
b1 0.833274 0.844157 0.862302 0.8142171
b2 0.00186718 0.00176878 0.00161775 0.00155423
R2 0.9975 0.9992 0.9994 0.9995
s 0.6837 0.7127 0.7490 0.7890
3.1.3. Evaluation of Accuracy
The distribution between the number of saturated salt solutions and the estimated standard error
for the calibration equations of two types of RH sensors is in Figure 6. For a resistance sensor, the
s values of 7, 9, 11 calibration points are 0.8% RH. For a capacitance sensor, the s values for four
saturated salt solutions are 0.8% RH. The accuracy of these calibration equations is 0.8% for both
types of RH sensors. In terms a practical application [20,21], the calibration equation can be established
using 7 salt solutions for a resistance sensor and 5 salt solutions for a capacitance sensor.
13](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-28-320.jpg)
![Sensors 2019, 19, 1213
Ϭ͘Ϯ
Ϭ͘ϰ
Ϭ͘ϲ
Ϭ͘ϴ
ϭ
ϭ͘Ϯ
ϰ ϱ ϲ ϳ ϴ ϵ ϭϬ ϭϭ ϭϮ
^ƚĂŶĚĂƌĚĚĞǀŝĂƚŝŽŶ͕Ɛ
EŽ͘ŽĨƐĂůƚƐŽůƵƚŝŽŶƐ
ZĞƐŝƐƚĂŶĐĞ ĂƉĂĐŝƚĂŶĐĞ
Figure 6. The distribution between numbers of saturated salt solutions and estimated standard errors
of calibration equations of two types of RH sensors.
3.2. The Effect of the Precision of Calibration Points
3.2.1. The Measurement Uncertainty for the Two Humidity Sensors
The method that is used to calculate the measurement uncertainty is that of Lu and Chen [17].
Two Types “A” and “B” method are used to evaluate the measurement uncertainty. The Type A
standard uncertainty is evaluated by statistical analysis of the experimental data. The Type B standard
uncertainty is evaluated using other information that is related to the measurement.
The Type A standard uncertainty for the two types of humidity sensors used the uncertainty for
the predicted values from the calibration equations. The Type B standard uncertainty for humidity
sensors uses the reference standard, nonlinear and repeatability, resolution and temperature effect.
The results for the Type B uncertainty analysis for resistive and capacitive sensors are respectively
listed in Tables 10 and 11.
Table 10. The Type B uncertainty analysis for resistive humidity sensor.
Description Estimate Value (%) Standard Uncertainty u(x), (%)
Reference standard, Uref
N1 = 11, uref = 0.3311
N1 = 9, uref = 0.2983
N1 = 7, uref = 0.3151
N1 = 5, uref = 0.3084
Non-linear and repeatability, Unon ±0.3 0.00866
Resolution, Ures 0.1 0.00290
The combined standard uncertainty of Type B = 0.1926
Table 11. The Type B uncertainty analysis for capacitive humidity sensor.
Description Estimate Value (%) Standard Uncertainty u(x), (%)
Reference standard, Uref
N1 = 11, uref = 0.3311
N1 = 9, uref = 0.2983
N1 = 7, uref = 0.3151
N1 = 5, uref = 0.3084
Nonlinear and repeatability, Unon ±0.1 0.0058
Resolution, Ures ±0.1 0.0029
Temperature effect, Utemp ±0.005 0.0043
The combined standard uncertainty of Type B = 0.1924
The Type A standard uncertainty that are calculated using the predicted values for the 4th order
polynomial equation for the resistive sensor and the 2nd order polynomial equation for a capacitive
14](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-29-320.jpg)
![Sensors 2019, 19, 1213
uncertainty for 5, 7, 9 and 11 calibration points is 0.9% RH. In terms of practical applications, this
performance is sufficient for industrial applications [20,21].
The accuracy and precision are 0.80% and 0.90% RH for a resistance RH sensor that uses
7 calibration points and 0.70% and 0.90% RH for a capacitance RH sensors that uses 5 calibration points.
3.3. Discussion
The number of calibration points that are required for sensors represents a compromise between
the ideal number of calibration points and the time and cost of the calibration. The criterion that
Betta [1] used to determine the optimal number of points used the ratio of the standard deviation of
the regression coefficients (sbj) to the established standard error of regression (s).
Accuracy and precision are the most important criteria for sensors so this study uses both
values. Using statistical theory, the best calibration equation is determined using the t-value for the
highest-order parameter and the residual plots. The estimated standard errors for the regression
equation are then used to determine the accuracy of the sensors. The combined uncertainty considered
the uncertainty of reference materials, the uncertainty for the predicted values and other B type sources.
The combined uncertainties for the calibration equations for different numbers of calibration points
using different saturated salt solutions are the criteria that are used to evaluate the precision of sensors.
Two types of electric RH sensors were calibrated in this study. Some calibration works, such as
those for temperature and pressure sensors, are calibrated by an equal spacing of calibration points.
The RH reference environments are maintained using different saturated salt solutions.
It is seen that the optimum number of calibration points that is required to calibrate a resistive
humidity sensors involves 7 saturated salt solutions (LiCl, MgCl2, K2CO3, NaBr, NaCl, KCI and
K2SO4), so seven points are specified. Five saturated salt solutions (LiCl, MgCl2, NaBr, NaCl and
K2SO4) are specified for a capacitive humidity sensor. Considering factors that influence the choice of
salts, such as price, toxicity and rules for disposal, the choice of these salt solutions is suitable.
The calibration equations key to measurement performance. This study determines that te 4th
order polynomial equation is the adequate equation for the resistive humidity sensor and the 2nd
order polynomial equation is the optimum equation for the capacitive humidity sensor. The accuracy
of the calibration equations is 0.8% RH for a resistive humidity sensor that uses 7 calibration points
and 0.7% RH for a capacitance humidity sensor that uses 5 calibration points. The precision is less
than 1.0% RH for the resistive sensor and less than 0.9% RH for the capacitive sensor.
The method that is used in this study applicable to other sensors.
4. Conclusions
In this study, two types of electric RH sensors were used to illustrate the method for the
specification of the optimum number of calibration points. The standard RH environments are
maintained using different saturated salt solutions. The theory of regression analysis is applied. The
best calibration equation is determined in terms of the t-value of the highest-order parameter and
the residual plots. The estimated standard errors for the regression equation are the criteria that are
used to determine the accuracy of sensors. The combined uncertainty involves the uncertainty for the
reference materials, the uncertainty in the predicted values and other B type sources. The combined
uncertainties for the calibration equations for different number of calibration points using different
saturated salt solutions are the criteria that are used to evaluate the precision of the sensors.
The calibration equations are key to good measurement performance. This study determines that
the 4th order polynomial equation is the adequate equation for the resistive humidity sensor and the
2nd order polynomial equation is the best equation for the capacitive humidity sensor. The accuracy
of the calibration equations is 0.8% RH for a resistive humidity sensor that uses 7 calibration points
and 0.7% RH for a capacitance humidity sensor using 5 calibration points. The precision is less than
1.0% RH for the resistive sensor and less than 0.9% RH for the capacitive sensor.
16](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-31-320.jpg)
![Sensors 2019, 19, 1213
The method to determine the number of the calibration points used in this study is applicable to
other sensors.
Supplementary Materials: The following are available online at http://www.mdpi.com/1424-8220/19/5/1213/
s1. The residual plots for the calibration equations for different orders of polynomial equations for resistive
humidity sensor using 5 saturated salt solutions (LiCl, MgCl2, NaBr, NaCl and K2SO4). The residual plots for
the calibration equations for different orders of polynomial equations for capacitance humidity sensor using 5
saturated salt solutions (LiCl, MgCl2, NaBr, NaCl and K2SO4).
Author Contributions: H.-Y.C. drafted the proposal, executed the statistical analysis, interpreted the results and
revised the manuscript. C.C. reviewed the proposal, performed some experiments, interpreted some results and
criticized the manuscript and participated in its revision. All authors have read and approved the final manuscript.
Acknowledgments: The authors would like to thank the Ministry of Science and Technology of the Republic of
China for financially supporting this research under Contract No. MOST -106-2313-B-005-006.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Betta, G.; Dell’Isola, M. Optimum choice of measurement points for sensor calibration. Measurement 1996, 17,
115–125. [CrossRef]
2. Betta, G.; Dell’Isola, M.; Frattolillo, A. Experimental design techniques for optimizing measurement chain
calibration. Measurement 2001, 30, 115–127. [CrossRef]
3. Hajiyev, C. Determination of optimum measurement points via A-optimality criterion for the calibration of
measurement apparatus. Measurement 2010, 43, 563–569. [CrossRef]
4. Hajiyev, C. Sensor Calibration Design Based on D-Optimality Criterion. Metrol. Meas. Syst. 2016, 23, 413–424.
[CrossRef]
5. Khan, S.A.; Shabani, D.T.; Agarwala, A.K. Sensor calibration and compensation using artificial neural
network. ISA Trans. 2003, 42, 337–352. [CrossRef]
6. Chen, C. Application of growth models to evaluate the microenvironmental conditions using tissue culture
plantlets of Phalaenopsis Sogo Yukidian ‘V3’. Sci. Hortic. 2015, 191, 25–30. [CrossRef]
7. Chen, H.; Chen, C. Use of modern regression analysis in liver volume prediction equation. J. Med. Imaging
Health Inform. 2017, 7, 338–349. [CrossRef]
8. Wang, C.; Chen, C. Use of modern regression analysis in plant tissue culture. Propag. Ornam. Plants 2017, 17,
83–94.
9. Chen, C. Relationship between water activity and moisture content in floral honey. Foods 2019, 8, 30.
[CrossRef]
10. Chen, C. Evaluation of resistance-temperature calibration equations for NTC thermistors. Measurement 2009,
42, 1103–1111. [CrossRef]
11. Chen, A.; Chen, C. Evaluation of piecewise polynomial equations for two types of thermocouples. Sensors
2013, 13, 17084–17097. [CrossRef] [PubMed]
12. ISO/IEC 98–3. Uncertainty of Measurement—Part 3: Guide to the Expression of Uncertainty in Measurement; ISO:
Geneva, Switzerland, 2010.
13. National Aeronautics and Space Administration. Measurement Uncertainty Analysis Principles and Methods,
NASA Measurement Quality Assurance Handbook—Annex 3; National Aeronautics and Space Administration:
Washington, DC, USA, 2010.
14. Chen, C. Evaluation of measurement uncertainty for thermometers with calibration equations.
Accredit. Qual. Assur. 2006, 11, 75–82. [CrossRef]
15. Myers, R.H. Classical and Modern Regression with Applications, 2nd ed.; Duxbury Press: Pacific Grove, CA,
USA, 1990.
16. Weisberg, S. Applied Linear Regression, 4th ed.; Wiley: New York, NY, USA, 2013.
17. Lu, H.; Chen, C. Uncertainty evaluation of humidity sensors calibrated by saturated salt solutions.
Measurement 2007, 40, 591–599. [CrossRef]
18. Greenspan, L. Humidity fixed points of binary saturated aqueous solutions. J. Res. Natl. Bur. Stand. 1977,
81A, 89–96. [CrossRef]
17](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-32-320.jpg)
![Sensors 2019, 19, 1213
19. OMIL. The Scale of Relative Humidity of Air Certified Against Saturated Salt Solutions; OMIL R 121; Organization
Internationale De Metrologie Legale: Paris, France, 1996.
20. Wernecke, R.; Wernecke, J. Industrial Moisture and Humidity Measurement: A Practical Guide; Wiley: Hoboken,
NJ, USA, 2014.
21. Wiederhold, P.R. Water Vapor Measurement; Marcel Dekker, Inc.: New York, NY, USA, 1997.
22. Matko, V.; Đonlagić, D. Sensor for high-air-humidity measurement. IEEE Trans. Instrum. Meas. 1996, 4,
561–563. [CrossRef]
23. Matko, V. Next generation AT-cut quartz crystal sensing devices. Sensors 2011, 5, 4474–4482. [CrossRef]
[PubMed]
24. Zheng, X.Y.; Fan, R.R.; Li, C.R.; Yang, X.Y.; Li, H.Z.; Lin, J.D.; Zhou, X.C.; Lv, R.X. A fast-response and highly
linear humidity sensor based on quartz crystal microbalance. Sens. Actuator B Chem. 2019, 283, 659–665.
[CrossRef]
25. Wexler, A.; Hasegawa, S. Relative humidity-temperature relationships of some saturated salt solutions in the
temperature range 0◦ to 50◦ C. J. Res. Natl. Bur. Stand. 1954, 53, 19–26. [CrossRef]
26. Young, J. Humidity control in the laboratory using salt solutions—A review. J. Chem. Technol. Biotechnol.
1967, 17, 241–245. [CrossRef]
27. Lake, B.J.; Sonya, M.N.; Noor, S.M.; Freitag, H.P.; Michael, J.; McPhaden, M.J. Calibration Procedures and
Instrumental Accuracy Estimates of ATLAS Air Temperature and Relative Humidity Measurements; NOAA Pacific
Marine Environmental Laboratory: Seattle, WA, USA, 2003.
28. Wadsö, L.; Anderberg, A.; Åslund, I.; Söderman, O. An improved method to validate the relative humidity
generation in sorption balances. Eur. J. Pharm. Biopharm. 2009, 72, 99–104. [CrossRef]
29. Duvernoy, J.; Gorman, J.; Groselj, D. A First Review of Calibration Devices Acceptable for Metrology
Laboratory. 2015. Available online: https://www.wmo.int/pages/prog/www/IMOP/publications/IOM-
94-TECO2006/4_Duvernoy_France.pdf (accessed on 11 December 2018).
30. Belhadj, O.; Rouchon, V. How to Check/Calibrate Your Hygrometer? J. Paper Conserv. 2015, 16, 40–41.
[CrossRef]
31. Japan Mechanical Society. The Measurement of Moisture and Humidity and Monitoring of Environment; Japan
Mechanical Society: Tokyo, Japan, 2011. (In Japanese)
32. Japan Industrial Standard Committee. Testing Methods of Humidity; JIS Z8866; JISC: Tokyo, Japan, 1998.
33. Centre Microcomputer Application. Relative Humidity Sensor 025I. Available online: http://www.cma-
science.nl/resources/en/sensors_bt/d025i.pdf (accessed on 2 December 2018).
34. Delta Ohm Company. Calibration Instructions of Relative Humidity Sensors. 2012. Available online:
http://www.deltaohm.com/ver2012/download/Humiset_M_uk.pdf (accessed on 10 December 2018).
35. Omega Company. Equilibrium Relative Humidity Saturated Salt Solutions. 2013. Available online: https:
//www.omega.com/temperature/z/pdf/z103.pdf (accessed on 11 December 2018).
36. TA Instruments. Humidity Fixed Points. 2016. Available online: http://www.tainstruments.com/pdf/
literature/TN056.pdf (accessed on 10 December 2018).
37. Vaisala Ltd. Vaisala Humidity Calibrator HMK 15 User’s Guide. 2017. Available online: www.vaisala.com/
sites/default/files/documents/HMK15_User_Guide_in_English.pdf (accessed on 11 December 2018).
38. Ellison, S.; Williams, A. Eurachem/CITAC Guide: Quantifying Uncertainty in Analytical Measurement, 3rd ed.;
Eurachem: Torino, Italy, 2012.
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
18](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-33-320.jpg)
![sensors
Article
Humidity Sensors with Shielding Electrode Under
Interdigitated Electrode
Hong Liu, Qi Wang, Wenjie Sheng, Xubo Wang, Kaidi Zhang, Lin Du and Jia Zhou *
ASIC and System State Key Lab, Department of Microelectronics, Fudan University, Shanghai 200433, China;
16210720074@fudan.edu.cn (H.L.); 18212020034@fudan.edu.cn (Q.W.); wsheng13@fudan.edu.cn (W.S.);
xbwang16@fudan.edu.cn (X.W.); 15110720079@fudan.edu.cn (K.Z.); 17112020015@fudan.edu.cn (L.D.)
* Correspondence: jia.zhou@fudan.edu.cn; Tel.: +86-13818066203
Received: 14 December 2018; Accepted: 31 January 2019; Published: 6 February 2019
Abstract: Recently, humidity sensors have been investigated extensively due to their broad
applications in chip fabrication, health care, agriculture, amongst others. We propose a capacitive
humidity sensor with a shielding electrode under the interdigitated electrode (SIDE) based on
polyimide (PI). Thanks to the shielding electrode, this humidity sensor combines the high sensitivity
of parallel plate capacitive sensors and the fast response of interdigitated electrode capacitive sensors.
We use COMSOL Multiphysics to design and optimize the SIDE structure. The experimental
data show very good agreement with the simulation. The sensitivity of the SIDE sensor is
0.0063% ± 0.0002% RH. Its response/recovery time is 20 s/22 s. The maximum capacitance drift
under different relative humidity is 1.28% RH.
Keywords: humidity sensor; capacitive; PI; SIDE; IDE
1. Introduction
In addition to daily applications, such as air conditioners and humidifiers, humidity sensors
are widely used in industrial process control, medical science, food production, agriculture,
and meteorological monitoring [1–9]. In industry, the many manufacturing processes, such as
semiconductor manufacturing and chemical gas purification, rely on precisely controlled humidity
levels. In medical science, environmental humidity needs to be controlled during operations and
pharmaceutical processing. In agriculture, humidity sensors are used for greenhouse air conditioning,
plantation protection (dew prevention), soil moisture monitoring, and grain storage. Furthermore,
in meteorological monitoring, weather bureaus and marine monitoring applications rely on accurate
humidity sensing. For modern agriculture [10] and weather stations [11,12], accurate and fast
measurement of humidity is becoming more and more important. Compared to existing infrared
humidity sensors, electronic humidity sensors are cheaper, lighter, and smaller, which makes them
more suitable for sensor networks to feed weather models. Nonetheless, high-precision fast-response
sensors are important for many fields. For instance, fast and accurate humidity measurement are critical
for eddy covariance systems [13]. Hence, electronic sensors have to become faster and more accurate.
Electronic humidity sensors can be divided into resistive and capacitive [14]. Resistive humidity
sensors tend to have higher gain and are usually cheaper to manufacture than capacitive humidity
sensors. However, these sensors do not respond well when operating at low relative humidity (about
10% RH) because they exhibit very poor conductivity in low relative humidity environments, making
it difficult to measure the output response [15]. In contrast, capacitive humidity sensors have better
linearity, accuracy, and higher thermal stability than resistive humidity sensors [16–19]. A capacitive
humidity sensor responds to changes of humidity by changes of the relative dielectric constant of the
sensing layer, e.g., polymer film, upon water vapor absorption. Therefore, it is possible to directly
Sensors 2019, 19, 659; doi:10.3390/s19030659 www.mdpi.com/journal/sensors
19](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-34-320.jpg)
![Sensors 2019, 19, 659
detect changes in capacitance to monitor changes in humidity. Unlike resistive humidity sensor,
capacitive humidity sensors respond linearly with humidity, which simplifies the sensor readout.
Various materials can be used as humidity sensing materials, such as electrolyte [20],
ceramics [21,22], porous inorganic material [23–26], and polymers [27–30]. In particular, polymers
have been used as sensing materials for capacitive humidity sensors owing to their good dielectric
properties arising from their microporous structure and measurable physical property changes due to
water absorption. PI is among the most commonly used moisture sensing material [31] for its good
mechanical strength, electrochemical stability, and flexibility [32]. It remains stable after long time
exposure to the measurement environment. Furthermore, PI is a microporous material with imide
groups that strongly bond water molecules, which makes the material dielectric constant very sensitive
to humidity. Therefore, we used PI in the proposed capacitive sensor.
Capacitive humidity sensors have two basic structures: parallel plate (PP) capacitance (Figure 1a)
and interdigital electrode (IDE) capacitance (Figure 1b).
Figure 1. Structure diagram of parallel plate (PP) and interdigital electrode (IDE) sensors. (a) PP
sensors composed of a solid substrate, two layers of parallel plate electrode, and a sensing material
between them. (b) IDE sensors composed of an inert substrate, IDEs, and sensing material layer atop
of the IDEs. A partial enlarged detail of IDE is shown on the right.
In PP sensors, the upper plate is perforated by an array of holes or parallel stripes to allow water
molecules from the air to reach the sensing material underneath. Since the sensing area of the PP
capacitor is sandwiched between two parallel plates, the change in the relative dielectric constant
of the sensing material in the PP sensors affects the overall capacitance change. Unlike PP sensors,
IDE sensors usually only affect the change in the upper capacitance of the IDEs, which makes them
less sensitive than PP sensors. However, the exposed sensing area of the PP sensors is smaller than for
IDE sensors, which causes a slower response than for IDEs.
20](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-35-320.jpg)
![Sensors 2019, 19, 659
The IDEs are fabricated on an inert solid or flexible substrate as parallel comb electrodes that
overlap each other [6,33]. IDE sensors are easier to fabricate than PP ones. The sensitive area of the
IDEs is typically a few square millimeters, and the electrode gap is a few microns. The sensitivity
of this type of sensor increases with decreasing pitch [34]. The electric field strength above the IDEs
decreases exponentially away from the electrode surface, and becomes one-thirtieth, or even lower,
of the surface value [35] after a few microns. Therefore, in the case where the gap between the IDEs
is several microns, a sensing layer only a few microns thick is enough. Thanks to this layer being
completely exposed to the measurement environment, the IDE sensors are faster. However, in the
IDEs, only half of the electric field lines pass through the sensing layer, and the other half of the electric
field lines pass through the underlying substrate. Therefore, the IDE sensors will have only half or less
sensitivity (depending on the relative dielectric constant of the substrate) compared to an equivalent
PP sensor [36].
It is clear that there are advantages and disadvantages of these two types of sensors. There has
been a significant effort to improve the sensor structures. For example, Zhao et al. used RIE (Reactive
Ion Etching) and ICP (Inductively Couple Plasma) to etch sensing materials between parallel plates of
the sensors to obtain a larger contact area with the tested environment to reduce response time from
35 s to 25 s [37], but this was still slower compared to typical equivalent IDEs.
Inspired by combining the advantages of PP and IDE structures, this paper proposes a novel
IDE humidity sensor with a shielding electrode under the IDEs, namely, SIDE. On the SIDE,
the capacitance of the lower half of the IDEs is shielded by an additional electrode underneath
the IDEs, which effectively raises the relative capacitance change as it becomes exposed to moisture.
Thus, a SIDE humidity sensor combines the high sensitivity of PP sensors and the fast response (20 s)
as the IDE ones.
In this work, we first verified the feasibility of the SIDE structure in the simulation software.
Secondly, the thickness of the sensing layer with different electrode gaps and the dielectric thickness
between the shielding electrode and the IDEs were optimized regarding the sensitivity and response
speed. The SIDE sensor with optimized parameters was fabricated. The sensitivity, response time,
recovery time, and stability of the sensor were measured.
2. Simulation of SIDE
COMSOL Multiphysics®(Stockholm, Sweden) is applied to simulate the SIDE and IDE structure.
Figure 2a shows the SIDE structure. The size of this sensor is 13 mm × 6 mm with a sensing area
of 1.6 mm × 1 mm. The sensor consists of a 100 nm-thick shielding electrode, a 1 μm-thick silicon
dioxide dielectric layer, a standard 100 nm IDE layer, and a PI film as the sensing layer. The finger
length of the interdigitated electrode is 1 mm, with the width and the gap both being 5 μm. A total
of 80 pairs of IDEs are used. A 5 μm-thick PI layer is utilized as the humidity sensing layer. Since
the PI’s relative dielectric constant increases linearly with humidity [38], we simulate variations of
humidity by directly changing the relative dielectric constant of the PI. An IDE model with the same
structural parameters as the SIDE one is implemented with the only difference being the absence of
the shielding electrode.
Figure 2b shows the simulation results of the capacitance change rate (ΔC/C0) of SIDE and IDE
under different relative dielectric constant of PI representing the humidity conditions. C0 is the total
capacitance when the relative dielectric constant of the sensing layer is 2.9. ΔC is the capacitance
difference between any other relative dielectric constant of PI and 2.9. It can be seen that under the
same conditions, ΔC/C0 of the SIDE structure, is about 4 times bigger than that of the IDE structure,
which implies that the SIDE will have much higher sensitivity than IDE with the same parameters.
The effect of the thickness of the sensing film on ΔCmax/C0 is also simulated by COMSOL
Multiphysics®(Stockholm, Sweden). We define that ΔCmax/C0 equals to ΔC/C0 with the relative
dielectric constant of PI at 2.9 (C0) and 3.7 (Cmax), which indicates the sensitivity of the sensor.
21](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-36-320.jpg)
![Sensors 2019, 19, 659
The response and recovery dynamics are among the most important characteristics for evaluating
the performance of humidity sensors. The response time for RH increase and the recovery time for
RH decrease are usually defined for a sensor as the time taken to reach 90% of its total capacitance
variation. The response and recovery curves are measured by exposing the SIDE sensor to alternate
levels of humidity between 2.0% ± 0.8% and 77.0% ± 0.8% RH.
In order to evaluate the functioning of the humidity sensor over long periods of time, we measured
the sensor’s capacitance over the duration of 20 h at 25 ◦C with relative humidity levels of 25.7% ± 0.8%,
34.4% ± 0.8%, 45.0% ± 0.8%, 57.0% ± 0.8%, and 73.5% ± 0.8% RH.
4. Results and Discussion
A sensitivity test is carried out on the SIDE and IDE structure. Figure 8 shows the capacitance
measured from SIDE and IDE at different levels of humidity, and their linear fits with R2 of 0.996 and
0.991, respectively. The slopes of the line, i.e., S of SIDE and IDE are 0.0063 and 0.001,65, respectively.
Taking the uncertainty of HC2-S into consideration, the S of SIDE and IDE are 0.0063 ± 0.0002 and
0.001,65 ± 0.000,05, respectively. Hence, the sensitivity of the SIDE structure is 3.82 times bigger
than that of the IDE. These results show the significant improvement of sensitivity brought by the
shielding electrode, that minimizes the large constant capacitance of the substrate. Indeed, whatever
substrate the IDE is built on, the relative dielectric constant of the substrate is larger (e.g., Si is 11.9,
glass is 10) or close to (e.g., flexible polymer films) the relative dielectric constant of PI (2.9–3.7).
The experimental result and simulation data verify the effects of the shielding electrode and shows
high agreement as well. It is clear that our proposed SIDE structure can provide an effective way
to measure relative humidity more sensitively and accurately. Another advantage of the shielding
electrode is that it can effectively suppress the external electromagnetic interference and reduce the
noise in the measurement process.
Figure 8. Experimental measurement of sensitivity of SIDE and IDE humidity sensors.
Figure 9 shows the responses of the SIDE sensor. The absorption curve represents the response of
the sensor as a function of time, from an environment with low relative humidity to an environment
with high relative humidity. The desorption curve represents the response of the sensor as a function
of time, from an environment with high relative humidity to an environment with low relative
humidity. The curve can switch to steady states rapidly after the RH level changes. Our sensor’s
response/recovery time is 20 s/22 s, which is comparable to 1 s/15 s for normal IDE reported in
the literature [39], but a little worse. This is because in their work, the thickness of the sensing
25](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-40-320.jpg)
![Sensors 2019, 19, 659
film is only 0.65 μm, while ours is 5 μm. If we scale down our sensors to reduce the IDE gap,
the required sensing film thickness will also decrease, resulting in great improvement in response
speed. Limited to laboratory conditions, we fabricated the sensor with 5 μm gap. However, our
sensor’s response/recovery time is still much better than 122 s for PP sensors [40].
Figure 9. The response and recovery curves are measured by switching the SIDE sensor, alternately,
between 2.0% ± 0.8% and 77.0% ± 0.8% RH. The response/recovery time is 20 s/22 s.
Figure 10 shows the stability characteristic of the SIDE sensor. The sensor is kept in the incubator
for 20 h at 25.7% ± 0.8%, 34.4% ± 0.8%, 45.0% ± 0.8%, 57.0% ± 0.8%, and 73.5% ± 0.8% RH, respectively.
The magnitude of the drift of sensor capacitance is converted into the apparent changes in relative
humidity, D, which is calculated by
D = (Cmax − Cmean)/(C0·S) (3)
where Cmax is the maximum measured capacitance after the sensor is exposed to different RH atmosphere,
and Cmean is the average capacitance of all recorded values at a certain relative humidity, C0 is the
capacitance measured when the RH is 23.7% ± 0.8%. The maximum drift value (D) obtained from
Figure 10 under different relative humidity was 1.28% RH. Thus, our sensor is able to achieve satisfactory
stability from a practical standpoint, which makes it promising as a commercially available sensor.
Figure 10. Stability of SIDE sensor. The sensor is kept in the incubator for 1200 min at 25.7% ± 0.8%,
34.4% ± 0.8%, 45.0% ± 0.8%, 57.0% ± 0.8%, and 73.5% ± 0.8% RH, respectively.
26](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-41-320.jpg)
![Sensors 2019, 19, 659
5. Conclusions
In summary, we propose a novel shielded interdigitated electrode structure for humidity sensing.
We perform a comprehensive simulation of this structure to optimize the parameters for the sensor
fabrication. In simulation and actual testing, we find that the sensitivity of the SIDE structure is
much higher than that of the IDE structure because of the effect of the shielding electrode on the
capacitance change rate. Since the surface structure of the SIDE structure is still the same as IDE,
the SIDE sensor combines the high sensitivity of the parallel plate sensors and fast response of the
IDE sensors. The sensitivity of SIDE is 0.0063% ± 0.0002% RH, and the response/recovery time is
20 s/22 s. The stability of the SIDE sensor was also characterized. The maximum drift value under
different relative humidity is 1.28% RH.
Meanwhile, since the basic operating principle of many capacitive sensors is the same, the SIDE
structure can even be applied to capacitive gas sensors, such as volatile organic compound (VOC)
sensors which are used to monitor toxic gases. This shows that SIDE can replace IDE in various sensors
that are more sensitive to the accuracy and response speed.
Author Contributions: Conceptualization, J.Z. and H.L.; methodology, H.L.; software, K.Z.; validation, Q.W., W.S.
and L.D.; formal analysis, Q.W.; investigation, H.L.; resources, J.Z.; data curation, H.L.; writing—original draft
preparation, H.L.; writing—review and editing, J.Z.; visualization, X.W.; supervision, J.Z.; project administration,
J.Z.; funding acquisition, J.Z.
Acknowledgments: This work was supported by the National Natural Science Foundation of China (Grant No.
61874033), Science Foundation of Shanghai Municipal Government (Grant No.18ZR1402600) and the State Key
Lab of ASIC and System, Fudan University with Grant No.2018MS003.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Tételin, A.; Pellet, C.; Laville, C.; N’Kaoua, G. Fast response humidity sensors for a medical microsystem.
Sensors Actuators B Chem. 2003, 91, 211–218. [CrossRef]
2. Chen, Z.; Lu, C. Humidity Sensors: A Review of Materials and Mechanisms. Sens. Lett. 2005, 3, 274–295.
[CrossRef]
3. Lee, C.W.; Lee, S.J.; Kim, M.; Kyung, Y.; Eom, K. Capacitive Humidity Sensor Tag Smart Refrigerator System
using the Capacitive to Voltage Converter (CVC). Int. J. Adv. Sci. Technol. 2011, 36, 15–26.
4. Kolpakov, S.A.; Gordon, N.T.; Mou, C.; Zhou, K. Toward a new generation of photonic humidity sensors.
Sensors 2014, 14, 3986–4013. [CrossRef] [PubMed]
5. Farahani, H.; Wagiran, R.; Hamidon, M.N. Humidity sensors principle, mechanism, and fabrication
technologies: A comprehensive review. Sensors 2014, 14, 7881–7939. [CrossRef] [PubMed]
6. Pavinatto, F.J.; Paschoal, C.W.A.; Arias, A.C. Printed and flexible biosensor for antioxidants using
interdigitated ink-jetted electrodes and gravure-deposited active layer. Biosens. Bioelectron. 2015, 67, 553–559.
[CrossRef] [PubMed]
7. Lee, C.-Y.; Lee, G.-B. Humidity Sensors: A Review. Sens. Lett. 2005, 3, 1–15. [CrossRef]
8. Rittersma, Z.M. Recent achievements in miniaturised humidity sensors—A review of transduction
techniques. Sensors Actuators A Phys. 2002, 96, 196–210. [CrossRef]
9. Willett, K.M.; Gillett, N.P.; Jones, P.D.; Thorne, P.W. Attribution of observed surface humidity changes to
human influence. Nature 2007, 449, 710–712. [CrossRef] [PubMed]
10. Imam, S.A.; Choudhary, A.; Sachan, V.K. Design issues for wireless sensor networks and smart humidity
sensors for precision agriculture: A review. In Proceedings of the 2015 International Conference on Soft
Computing Techniques and Implementations (ICSCTI), Faridabad, India, 8–10 October 2015; pp. 181–187.
11. Chandana, L.S.; Sekhar, A.J.R. Weather Monitoring Using Wireless Sensor Networks based on IOT. Int. J. Sci.
Res. Sci. Technol. 2018, 4, 525–531.
12. Yawut, C.; Kilaso, S. A Wireless Sensor Network for Weather and Disaster Alarm Systems. Int. Conf. Inf.
Electron. Eng. 2011, 6, 155–159.
13. Baldocchi, D.D. Assessing the eddy covariance technique for evaluating carbon dioxide exchange rates of
ecosystems: Past, present and future. Glob. Chang. Biol. 2003, 9, 479–492. [CrossRef]
27](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-42-320.jpg)
![Sensors 2019, 19, 659
14. Fenner, R.; Zdankiewicz, E. Micromachined Water Vapor Sensors: A Review of Sensing Technologies. IEEE Sens. J.
2001, 1, 309–317. [CrossRef]
15. Blank, T.A.; Eksperiandova, L.P.; Belikov, K.N. Recent trends of ceramic humidity sensors development:
A review. Sensors Actuators B Chem. 2016, 228, 416–442. [CrossRef]
16. Dokmeci, M.; Najafi, K. A high-sensitivity polyimide capacitive relative humidity sensor for monitoring
anodically bonded hermetic micropackages. J. Microelectromech. Syst. 2001, 10, 197–204. [CrossRef]
17. Gu, L.; Huang, Q.A.; Qin, M. A novel capacitive-type humidity sensor using CMOS fabrication technology.
Sensors Actuators B Chem. 2004, 99, 491–498. [CrossRef]
18. Wagner, T.; Krotzky, S.; Weiß, A.; Sauerwald, T.; Kohl, C.D.; Roggenbuck, J.; Tiemann, M. A high temperature
capacitive humidity sensor based on mesoporous silica. Sensors 2011, 11, 3135–3144. [CrossRef]
19. Lee, H.; Lee, S.; Jung, S.; Lee, J. Nano-grass polyimide-based humidity sensors. Sensors Actuators B Chem.
2011, 154, 2–8. [CrossRef]
20. Yang, M.-R.; Chen, K.-S. Humidity sensors using polyvinyl alcohol mixed with electrolytes. Sensors Actuators
B Chem. 1998, 49, 240–247. [CrossRef]
21. Kim, Y.; Jung, B.; Lee, H.; Kim, H.; Lee, K.; Park, H. Capacitive humidity sensor design based on anodic
aluminum oxide. Sensors Actuators B Chem. 2009, 141, 441–446. [CrossRef]
22. Feng, Z.S.; Chen, X.J.; Chen, J.J.; Hu, J. A novel humidity sensor based on alumina nanowire films. J. Phys. D
Appl. Phys. 2012, 45, 225305. [CrossRef]
23. Tudorache, F.; Petrila, I. Effects of partial replacement of iron with tungsten on microstructure, electrical,
magnetic and humidity properties of copper-zinc ferrite material. J. Electron. Mater. 2014, 43, 3522–3526.
[CrossRef]
24. Tudorache, F.; Petrila, I.; Popa, K.; Catargiu, A.M. Electrical properties and humidity sensor characteristics of
lead hydroxyapatite material. Appl. Surf. Sci. 2014, 303, 175–179. [CrossRef]
25. Tudorache, F.; Petrila, I.; Condurache-Bota, S.; Constantinescu, C.; Praisler, M. Humidity sensors applicative
characteristics of granularized and porous Bi2O3 thin films prepared by oxygen plasma-assisted pulsed laser
deposition. Superlattices Microstruct. 2015, 77, 276–285. [CrossRef]
26. Tudorache, F.; Petrila, I.; Slatineanu, T.; Dumitrescu, A.M.; Iordan, A.R.; Dobromir, M.; Palamaru, M.N.
Humidity sensor characteristics and electrical properties of Ni–Zn–Dy ferrite material prepared using
different chelating-fuel agents. J. Mater. Sci. Mater. Electron. 2016, 27, 272–278. [CrossRef]
27. Suzuki, T.; Tanner, P.; Thiel, D.V. O2 plasma treated polyimide-based humidity sensors. Analyst 2002, 127, 1342–1346.
[CrossRef] [PubMed]
28. Zampetti, E.; Pantalei, S.; Pecora, A.; Valletta, A.; Maiolo, L.; Minotti, A.; Macagnano, A.; Fortunato, G.;
Bearzotti, A. Design and optimization of an ultra thin flexible capacitive humidity sensor. Sensors Actuators
B Chem. 2009, 143, 302–307. [CrossRef]
29. Kim, J.H.; Hong, S.M.; Moon, B.M.; Kim, K. High-performance capacitive humidity sensor with novel
electrode and polyimide layer based on MEMS technology. Microsyst. Technol. 2010, 16, 2017–2021. [CrossRef]
30. Liu, M.Q.; Wang, C.; Kim, N.Y. High-sensitivity and low-hysteresis porous mim-type capacitive humidity
sensor using functional polymer mixed with TiO2 microparticles. Sensors 2017, 17, 284. [CrossRef] [PubMed]
31. Wang, H.; Feng, C.-D.; Sun, S.-L.; Segre, C.U.; Stetter, J.R. Comparison of conductometric humidity-sensing
polymers. Sens. Actuators B Chem. 1997, 40, 211–216. [CrossRef]
32. Fujita, S.; Kamei, Y. Electrical properties of polyimide with water absorption. In Proceedings of the 11th
IEEE International Symposium on Electrets, Melbourne, VIC, Australia, 1–3 October 2002; pp. 275–278.
33. Olthuis, W.; Sprenkels, A.J.; Bomer, J.G.; Bergveld, P. Planar interdigitated electrolyte-conductivity sensors
on an insulating substrate covered with Ta2O5. Sensors Actuators B Chem. 1997, 43, 211–216. [CrossRef]
34. Singh, K.V.; Bhura, D.K.; Nandamuri, G.; Whited, A.M.; Evans, D.; King, J.; Solanki, R.
Nanoparticle-enhanced sensitivity of a nanogap-interdigitated electrode array impedimetric biosensor.
Langmuir 2011, 27, 13931–13939. [CrossRef]
35. Schaur, S.; Jakoby, B. A numerically efficient method of modeling interdigitated electrodes for capacitive
film sensing. Procedia Eng. 2011, 25, 431–434. [CrossRef]
36. Blue, R.; Uttamchandani, D. Chemicapacitors as a versatile platform for miniature gas and vapor sensors.
Meas. Sci. Technol. 2017, 28, 22001–22024. [CrossRef]
28](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-43-320.jpg)
![Sensors 2019, 19, 659
37. Qiang, T.; Wang, C.; Liu, M.Q.; Adhikari, K.K.; Liang, J.G.; Wang, L.; Li, Y.; Wu, Y.M.; Yang, G.H.; Meng, F.Y.;
et al. High-Performance porous MIM-type capacitive humidity sensor realized via inductive coupled plasma and
reactive-Ion etching. Sensors Actuators B Chem. 2018, 258, 704–714. [CrossRef]
38. Schubert, P.J.; Nevin, J.H. A polyimide-based capacitive humidity sensor. IEEE Trans. Electron Devices 1985,
32, 1220–1223. [CrossRef]
39. Laville, C.; Delétage, J.Y.; Pellet, C. Humidity sensors for a pulmonary function diagnostic microsystem.
Sensors Actuators B Chem. 2001, 76, 304–309. [CrossRef]
40. Kim, J.H.; Hong, S.M.; Lee, J.S.; Moon, B.M.; Kim, K. High sensitivity capacitive humidity sensor with a novel
polyimide design fabricated by mems technology. In Proceedings of the 4th IEEE International Conference on
Nano/Micro Engineered and Molecular Systems, NEMS 2009, Shenzhen, China, 5–8 January 2009; pp. 703–706.
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
29](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-44-320.jpg)
![sensors
Article
A Fast Response−Recovery 3D Graphene Foam
Humidity Sensor for User Interaction
Yu Yu 1,2, Yating Zhang 1,2,*, Lufan Jin 1,2, Zhiliang Chen 1,2, Yifan Li 1,2, Qingyan Li 1,2,
Mingxuan Cao 1,2, Yongli Che 1,2, Junbo Yang 3 and Jianquan Yao 1,2
1 Department of Electrical and Electronic Engineering, South University of Science and Technology of China,
Shenzhen 518055, China; yuyu1990@tju.edu.cn (Y.Y.); jlfking@tju.edu.cn (L.J.); chenzl@tju.edu.cn (Z.C.);
yifanli@tju.edu.cn (Y.L.); liqingyan216@163.com (Q.L.); mingxuancao@tju.edu.cn (M.C.);
cheyongli@tju.edu.cn (Y.C.); jqyao@tju.edu.cn (J.Y.)
2 Key Laboratory of Opto-Electronics Information Technology, Ministry of Education, School of Precision
Instruments and Opto-Electronics Engineering, Tianjin University, Tianjin 300072, China
3 Center of Material Science, National University of Defense Technology, Changsha 410073, China;
yangjunbo008@sohu.com
* Correspondence: yating@tju.edu.cn
Received: 9 October 2018; Accepted: 3 December 2018; Published: 8 December 2018
Abstract: Humidity sensors allow electronic devices to convert the water content in the
environment into electronical signals by utilizing material properties and transduction techniques.
Three-dimensional graphene foam (3DGF) can be exploited in humidity sensors due to its convenient
features including low-mass density, large specific surface area, and excellent electrical. In this
paper, 3DGF with super permeability to water enables humidity sensors to exhibit a broad relative
humidities (RH) range, from 0% to 85.9%, with a fast response speed (response time: ~89 ms,
recovery time: ~189 ms). To interpret the physical mechanism behind this, we constructed a 3DGF
model decorated with water to calculate the energy structure and we carried out the CASTEP as
implemented in Materials Studio 8.0. This can be ascribed to the donor effect, namely, the electronic
donation of chemically adsorbed water molecules to the 3DGF surface. Furthermore, this device can
be used for user interaction (UI) with unprecedented performance. These high performances support
3DGF as a promising material for humidity sensitive material.
Keywords: three-dimensional graphene foams; humidity sensor; fast response; user interaction
1. Introduction
Humidity sensors have aroused attention in many fields such as industry, agriculture, and
environment [1,2], and medical devices [3]. Generally, they measure humidity through a variety of
transduction techniques, including the use of resistive [4,5], capacitive [6], optical fiber [7], and field
effect transistors [8,9]. There are also some high precision impedance-frequency transducers using
quartz crystals which compensate temperature drift, and have fast response, as investigated by in
Matko et al. [10]. In high air humidity measurement there is a problem with response time of the sensors
in conventional methods. A solution for this problem is sensors for high air humidity measurement
which use open capacitors with very low response time such as is described by Vojko et al. [11].
As an active material for absorbing water molecules, a series of sensing materials
including polymers [12], metal oxides [8], carbon nanotubes [13,14], graphene dioxide [15,16],
and composites [5,17] have been exploited in humidity sensors. For instance, Zhang et al. [12]
described humidity sensors utilizing poly(N-vinyl-2-pyrrolidone) (PVP), poly(vinyl alcohol) (PVA),
and hydroxyethyl cellulose (HEC). In particular, after using PVP, the humidity sensors exhibited
response and recovery times between 11% and 95% relative humidity (RH) were about 37 s and 10 s,
Sensors 2018, 18, 4337; doi:10.3390/s18124337 www.mdpi.com/journal/sensors
30](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-45-320.jpg)
![Sensors 2018, 18, 4337
respectively. Wang et al. [8] applied a single SnO2 nanowire (NW) to fabricate a humidity sensor,
which exhibited a wide sensor RH range (5~85%), and the response and recovery times were 120~170 s
and 20~60 s, respectively. Zhao et al. [15] investigated a humidity sensor based on multi-wall carbon
nanotubes where the sensor testing range was about 11% to 97% RH, the response time was 45 s,
and the recovery time was 15 s. Borini et al. [15] exploited graphene oxide in a humidity sensor and
obtained an unprecedented response speed (~30 ms response and recovery times) in a range of 30% to
70% RH. Zhang et al. [5] utilized a graphene oxide (GO)/poly(diallyldimethylammonium chloride)
(PDDA) nanocomposite film to fabricate a humidity sensor. The humidity sensor exhibited ultrahigh
performance over a wide range of 11~97% RH, and the recovery time is 125 s at 11% RH. Thus, each
sensing material has its own advantages and specific conditions of application. In addition, with large
surface area to volume ratio, nanomaterials are attractive to fabricate humidity sensors with ultrahigh
performance features including high sensitivity and fast response times.
Recently, graphene with three dimensional (3D) architectures, including foams, networks, and gels
have been investigated [18–21]. These 3D graphene-based materials not only have the characteristics
of graphene, but also have high specific surface area, low density, good mechanical strength and
good conductivity [22]. Because of its wide accessibility, easy synthesis and solution processability,
high chemical stability and strong adaptability [23,24], 3D graphene foam (3DGF) has attracted great
interest in various sensing applications. Meanwhile, 3DGFs are efficient materials for biosensors
and gas-sensing devices given their low-mass density, large surface area, good mechanical stability,
and high electrical conductivity. Huang and coworkers [25] synthesised 3DGF/CuO nanoflower
composites as single-chip independent 3D biosensors for the electrochemical detection of ascorbic
acid with outstanding biosensing properties, such as an ultrahigh sensitivity of 2.06 mA mM−1 cm−2
to ascorbic acid at a 3 s response time. Besides that, Yavari et al. [24] used macroscopic 3DGF to
fabricate gas detectors with high sensitivity. Generally, these electrical-type 3DGF sensors exhibit high
sensitivity due to these properties including an ultrahigh surface area, and its electronic properties.
It shows a strong dependence on surface absorbents (including gas molecules), which can change
the carrier density of graphene [24]. Therefore, it is necessary to develop a new type of humidity
sensor based on 3DGF by utilizing the unique structure and chemical characteristics and avoiding
its shortcomings.
In this paper, we fabricate a humidity field effect transistor based on 3DGF and develop test
equipment to measure the properties of the device. It exhibits a high performance over a broad RH
range from 0% to 85.9%, with fast response and recovery times. To interpret the physical mechanism,
we construct the 3DGF model decorated with water and apply CASTEP in the Materials Studio
software to calculate the energy structure. Herrin, we explore the potential of 3D GF for portable,
reliable and low cost humidity sensing applications in the future.
2. Materials and Methods
Utilizing a modified Hummers’ method, [19–21] graphene oxide, denoted as GO, was synthesized
from natural graphite powder by an oxidation reaction. GO ethanol solution (50 mL) with the
concentration of 1 mg mL−1 was sealed in a 100 mL Teflon-lined autoclave which was then heated
up to 180 ◦C and held for 12 h. Then the autoclave was cooled naturally to room temperature.
The prepared ethanol intermediates were carefully removed from the autoclave by a slow and gradual
solvent exchange with water. After the solvent exchange process was completed, the product filled
with water was freeze-dried and then dried at 120 ◦C for 2 h in a vacuum oven. Finally, the sample
was annealed at 450 ◦C in H2/Ar (5/95, v/v) for 6 h. Finally, the sample was treated in a UV ozone
system for 15 min to obtain the final 3DGF. The infrared spectrum of 3DGF was recorded on a Fourier
transform infrared (FTIR) spectrophotometer using potassium bromide (KBr) pellets. Figure 1a shows
the FTIR spectra of three-dimensional graphene foam (3DGF)) with water molecules (black line) and
dry (red line) conditions. It can be seen that a broad peak at 3436 cm−1 corresponds to the vibration
due to the stretching and bending of OH groups present in the water molecules adsorbed by 3DGF.
31](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-46-320.jpg)
![Sensors 2018, 18, 4337
Thus, it was concluded that 3DGF exhibits strong hydrophilicity. Meanwhile, the absorption peaks
at 565, 1163, and 1640 cm−1 correspond to the symmetric and antisymmetric stretching vibrations
of C=O, C–O, and C–C groups for 3DGF, respectively. Figure 1b shows the surface morphologies
of the 3DGF. Field emission scanning electron microscopy (SEM) images show clear, layered and
interconnected three-dimensional uniform graphene sheets. It can be concluded that it forms a spongy
porous network structure. [20]. The samples are cut into rectangular slabs (14 mm × 2 mm), and both
sides are pasted by copper conductive adhesives on silicon substrates with a size of 14 mm × 14 mm
for electrical contact.
Figure 1. (a) FTIR spectra of 3D graphene with or without water molecule. (b) Field emission scanning
electron microscopy (SEM) images of 3DGF.
For humidity sensors, chemical or physical reactions between water molecules and materials
induce changes in channel current. External factors including the water concentration, temperature,
and operating conditions will impact the performance of the device. For accurate measurements,
as shown in Figure 2a, we used a closed box as an experimental chamber to control the humidity.
In detail, the water concentrations were controlled by the ratio of saturated water vapor generated by a
humidifier to high-purity nitrogen. We assure high quality humidity measurement in different ambient
temperature operating conditions in climate chamber as shown in [26]. In order to measure the channel
current flowing into the drain electrode (IDS) [27–29], the source (with ground connection) and drain
electrodes were connected with a Keithley 2400 apparatus (Tektronix China Ltd, Shanghai, China).
The electrical measurements were also performed with this system, and the RH of the environment
was measured by a commercial humidometer. Therefore, as described by Figure 2b, the output
characteristics of the device were measured under dry and humid conditions. It shows that when
the RH level was fixed to 100%, the channel current (IDS) became lower than the conditions under
drying. Meanwhile, the Dirac point shifted towards the positive direction. This donor effect [1] has
been ascribed to the donation of electrons from the chemically adsorbed water molecules to the 3DGF.
It can be concluded that the water molecules decorated in 3DGF will attract electrons and remain as
holes, leading to p-type doping. Furthermore, water molecules decrease the charge mobility of 3D
graphene, leading to lower currents. Through swelling or the 2D capillary effect [7,15,24], the dielectric
constant will increase and the resistance decrease after adsorbing water molecules (confirmed using
FTIR, as shown in Figure 1a). At the same time, the space charge polarization effect can be enhanced by
adsorbing more water molecules, leading to the rapid diffusion of 3DGF and the formation of protons
between hydroxyl groups. [6]. To investigate the mechanism, band energy of graphene decorating with
water molecule was theoretically simulated by density functional theory (DFT) in the Material Studio
8.0 software (Neotrident Technology Ltd. Beijing, China). Simply speaking, graphene is simulated by
plane wave program implemented in CASTEP. Considering the single and double supercells (2 × 1 × 1
allowing edge reconstruction) under GGA-PBE with 9 × 1 × 1 k-points Monkhorst-Pack point grid
32](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-47-320.jpg)
![Sensors 2018, 18, 4337
and 500 eV plane wave base truncation, the graphene is simulated by plane wave program with basis
cutoff of 500 eV. The geometry was optimized until the total energy reached 2 × 10−5 eV/atom and
the maximum force acting on each atom is less than 0.05 eV/Å. For the 3D graphene foam and 3D
graphene foam adding water molecule calculations, the CASTEP plane wave code was used under
GGA-PBE considering a Monkhorst−Pack grid with 9 × 9 × 1 k-points and a plane wave basis cutoff
of 500 eV; optimizing the geometry until the total energy reaches 2 × 10−5 eV/atom and the maximum
force per atom exhibits values less than 0.05 eV/Å [30,31].
Figure 2. (a) Testing equipment used for the electrical characterization of 3DGF humidity sensors.
(b) Output characteristic of the device decorated with or without water molecules.
3. Results and Discussion
Furthermore, the humidity-sensing performance of the 3DGF sensors exposed to different RH
levels (0%, 10.0%, 19.9%, 30.3%, 44.5%, 51.4%, 57.1%, 60.3%, 66.4%, 70.5%, 75.2%, 80.2%, and 85.9% RH)
are presented in Figure 3a. In a closed air-tight box, the humidity sensors were measured by different
RH values ranging from 0 to 85.9%. It can be seen that as the RH level increased, the obtained channel
currents of the sensor reduced monotonically. To consider the real-time response and recovery times of
the devices, the time-dependent response and recovery curves of the device to 85.9% RH are plotted
in Figure 3b. The time taken by a sensor to achieve 85% RH of the total channel current was defined
as the response or recovery time. The response and recovery times of the sensor were approximately
89 ms and 189 ms, respectively. Additionally, our humidity sensors exhibited reproducibility and
long-term stability. Professionally, the hysteresis value is a vital parameter for humidity sensors as
it determines the maximum time lag between the response time (adsorption process) and recovery
time (desorption process). With respect to the water content in the environment, the hysteresis effect
is defined by the difference between the resistances. In particular, for a perfect humidity sensor, the
hysteresis value should be as small as possible or can even be negligible.
(a) (b)
Ⲵはᆀ൘ᧂ⡸䱦⇥䈧൘䘉ਕ䈍ਾ䶒࣐кĀ
Figure 3. (a) Channel current response measurement of the 3DGF humidity sensor with varying
different RH. (b) Response and recovery times of the device at 85% RH and the drain voltage was fixed
at 1 V.
33](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-48-320.jpg)
![Sensors 2018, 18, 4337
Table 1 compares the different characteristics of graphene-type humidity sensors including the
response/recovery time, fabrication method, and sensitivity range. It was observed that the3DGF
sensor exhibited broad sensitivity and rapid response and recovery rates.
Table 1. Comparison of different reported humidity sensors with graphene series materials.
Reference Material Sensing Range Response/Recovery Time
Smith [30] Graphene 1–96% 0.6 s/0.4 s
Ghosh [32] Graphene 4–84% 180 s/180 s
Cai [33] reduced graphene oxide (rGO)/graphene oxide (GO)/rGO 6.3–100% 1.9 s/3.9 s
Zhang [34] Graphene oxide foam 36–92% 2 s/10 s
Trung [35] rGO-polyurethane composites 10–70% 3.5 s/7 s
Leng [36] GO/Nafion composite 11.3–97.3% 100–300 s/not shown
Bi [6] GO 15–95% 10.5 s/41 s
Naik [37] GO 30–95% 100 s/not shown
Yu [38] GO/poly (sodium 4-styrenesulfonate) (PSS) composite 20–80% 60 s/50 s
Zhang [5] rGO/poly(diallylimethyammonium chloride) PDDA composite 11–97% 108 s/94 s
Guo [39] rGO 10–95% 50 s/3 s
This work 3DGF 0–85.9% 89 ms/189 ms
It can be seen that our devices showed good uniformity. Quantitatively, the effect of relative
humidity on the device is depicted in Figure 4. Figure 4a describes the relationship between channel
current and relative humidity. It can be seen that the relationship showed a decreasing trend with the
increase in water humidity. This also showed that the channel current (IDS) decreased more rapidly as
relative humidity increased. To characterize the performance of the humidity sensor, the sensitivity (S)
of the device was defined by Equation (1) [4,5,30,40]:
S =
Iwet − Idry
IdryRH
× 100 (1)
where Iwet and Idry represent the channel current of the device under wet and dry conditions (RH = 0%),
respectively. As shown in Figure 4b, the sensitivity increased rapidly as RH increased. Due to its
perfect performance, including its ultrafast response/recovery rate, our humidity sensors can be used
for breathing monitoring or for developing new user interfaces (UIs). Figure 3b presents the ability
of a 3DGF sensor to monitor human breathing. In particular, during the user’s speech, the ultrafast
humidity sensor allowed the capture of fine features due to moisture modulation. Therefore, the 3DGF
ultra-fast RH sensor can be used to identify different whistles, which can make use of low-cost and
low-power sensors for user authentication.
Figure 4. Relative humidity effect on the device performance. (a) Channel currents (IDS) with the
relationship of RH (b) The variation in sensitivity of the device for different RH values.
34](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-49-320.jpg)
![Sensors 2018, 18, 4337
A schematic model of humidity sensing at a 3DGF film is shown in Figure 5a. To investigate
its mechanism, the band energy of graphene decorated with water molecules was theoretically
simulated by density functional theory (DFT) in the Material Studio 8.0 software. As shown in
Figure 5b, conductivity and valence are at K Brillouin point, which makes the material a direct
bandgap semiconductor. The direct band gap at the K point was ~0.172 eV, as shown in Figure 5c.
This can be ascribed to the donor effect [3] attributed to the donation of electrons from the chemically
adsorbed water molecules to the 3DGF surface. The water molecules decorated in 3DGF will attract
electrons. Simply, water molecules open the band gap of 3DGF. Meanwhile, electron density will
decrease and the conduction level will rise, leading to the formation of band energy.
Figure 5. (a) The bonding mechanism between the graphene and water molecules. (b) The electronic
band structure of graphene decorated with water. (c) The energy gap at the K point location.
4. Conclusions
In summary, a three-dimensional graphene foam (3DGF) exhibiting super permeability to water
was exploited in humidity sensors, enabling a humidity sensor with a broad range of % RH values
and unprecedented response speed (response time: ~89 ms, recovery time: ~189 ms). The ultra-fast
response speed of these sensors enables us to observe the regulation of moisture in a user’s breath.
We constructed the 3DGF model decorated with water molecules theoretically and conducted the
CASTEP as implemented in Materials Studio to calculate the energy structure. This allows sensors to
be used in a variety of applications, such as humidity sensing, which we have experimentally verified
with a cheap and easily available identification system. In addition, for different 3D materials, such
as 3D transition metal dihalogenated hydrocarbons, ultra-thin nanoporous membranes for sensing
applications can be realized in the interaction with different vapors and gases, which can be explored.
Author Contributions: Conceptualization, Y.Y. and Y.Z.; methodology, Y.Z.; formal analysis, Y.Y., Z.C., Y.L., L.J.,
Q.L., Y.C., and M.C; data curation, Y.Y.; writing—original draft preparation, Y.Y.; writing—review and editing, Y.Y.;
supervision, Y.Z. and J.Y. (Junbo Yang); project administration, Y.Z. and J.Y. (Jianquan Yao); funding acquisition,
J.Y. (Jianquan Yao).
Funding: This work was supported by the National Natural Science Foundation of China (Nos. 61675147,
61605141 and 61735010), Basic Research Program of Shenzhen (JCYJ20170412154447469) and Wenzhou City
Governmetal Public Industrial Technology Project (G20160014).
Acknowledgments: We thank Yongshen Chen group in Nankai University, which he provides three dimensional
graphene foam.
Conflicts of Interest: The authors declare no conflict of interest.
35](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-50-320.jpg)
![Sensors 2018, 18, 4337
References
1. Traversa, E. Ceramic sensors for humidity detection: The state-of-the-art and future developments.
Sens. Actuators B Chem. 1995, 23, 135–156. [CrossRef]
2. Chu, J.; Peng, X.; Feng, P.; Sheng, Y.; Zhang, J. Study of humidity sensors based on nanostructured carbon
films produced by physical vapor deposition. Sens. Actuators B Chem. 2013, 178, 508–513. [CrossRef]
3. Chen, Z.; Lu, C. Humidity Sensors: A Review of Materials and Mechanisms. Sens. Lett. 2005, 3, 274–295.
[CrossRef]
4. Mogera, U.; Sagade, A.A.; George, S.J.; Kulkarni, G.U. Ultrafast response humidity sensor using
supramolecular nanofibre and its application in monitoring breath humidity and flow. Sci. Rep. 2014, 4, 4103.
[CrossRef] [PubMed]
5. Zhang, D.; Tong, J.; Xia, B. Humidity-sensing properties of chemically reduced graphene oxide/polymer
nanocomposite film sensor based on layer-by-layer nano self-assembly. Sens. Actuators B Chem 2014, 197,
66–72. [CrossRef]
6. Bi, H.; Yin, K.; Xie, X.; Ji, J.; Wan, S.; Sun, L.; Terrones, M.; Dresselhaus, M.S. Ultrahigh humidity sensitivity
of graphene oxide. Sci. Rep. 2013, 3, 2714. [CrossRef] [PubMed]
7. Wong, W.C.; Chan, C.C.; Chen, L.H.; Li, T.; Lee, K.X.; Leong, K.C. Polyvinyl alcohol coated photonic crystal
optical fiber sensor for humidity measurement. Sens. Actuators B Chem. 2012, 174, 563–569. [CrossRef]
8. Kuang, Q.; Lao, C.; Wang, Z.L.; Xie, Z.; Zheng, L. High-Sensitivity Humidity Sensor Based on a Single SnO2
Nanowire. J. Am. Chem. Soc. 2007, 129, 6070–6071. [CrossRef]
9. Hu, P.; Zhang, J.; Li, L.; Wang, Z.; O’Neill, W.; Estrela, P. Carbon nanostructure-based field-effect transistors
for label-free chemical/biological sensors. Sensors 2010, 10, 5133–5159. [CrossRef]
10. Vojko, M. Next generation AT-cut quartz crystal sensing devices. Sensors 2011, 11, 4474–4482.
11. Matko, V.; Donlagic, D. Sensor for high-air-humidity measurement. Sens. Actuators A Phys. 1997, 61, 331–334.
[CrossRef]
12. Fei, T.; Zhao, H.; Jiang, K.; Zhou, X.; Zhang, T. Polymeric humidity sensors with nonlinear response:
Properties and mechanism investigation. J. Appl. Polym. Sci. 2013, 130, 2056–2061. [CrossRef]
13. Chen, W.P.; Zhao, Z.G.; Liu, X.W.; Zhang, Z.X.; Suo, C.G. A Capacitive Humidity Sensor Based on Multi-Wall
Carbon Nanotubes (MWCNTs). Sensors 2009, 9, 7431–7444. [CrossRef] [PubMed]
14. Han, J.-W.; Kim, B.; Li, J.; Meyyappan, M. A carbon nanotube based ammonia sensor on cellulose paper.
RSC Adv. 2014, 4, 549–553. [CrossRef]
15. Borini, S.; White, R.; Wei, D.; Astley, M.; Haque, S.; Spigone, E.; Harris, N.; Kivioja, J.; Ryhänen, T. Ultrafast
Graphene Oxide Humidity Sensors. ACS Nano 2013, 7, 11166–11173. [CrossRef] [PubMed]
16. Zhao, X.; Long, Y.; Yang, T.; Li, J.; Zhu, H. Simultaneous High Sensitivity Sensing of Temperature and
Humidity with Graphene Woven Fabrics. ACS Appl. Mater. Interfaces 2017, 9, 30171–30176. [CrossRef]
[PubMed]
17. Zhang, D.; Tong, J.; Xia, B.; Xue, Q. Ultrahigh performance humidity sensor based on layer-by-layer
self-assembly of graphene oxide/polyelectrolyte nanocomposite film. Sens. Actuators B Chem. 2014, 203,
263–270. [CrossRef]
18. Ma, Y.; Chen, Y. Three-dimensional graphene networks: Synthesis, properties and applications. Natl. Sci. Rev.
2015, 2, 40–53. [CrossRef]
19. Wu, Y.; Yi, N.; Huang, L.; Zhang, T.; Fang, S.; Chang, H.; Li, N.; Oh, J.; Lee, J.A.; Kozlov, M.; et al.
Three-dimensionally bonded spongy graphene material with super compressive elasticity and near-zero
Poisson’s ratio. Nat. Commun. 2015, 6, 6141. [CrossRef]
20. Zhang, T.; Chang, H.; Wu, Y.; Xiao, P.; Yi, N.; Lu, Y.; Ma, Y.; Huang, Y.; Zhao, K.; Yan, X.-Q.; et al. Macroscopic
and direct light propulsion of bulk graphene material. Nat. Photonics 2015, 9, 471–476. [CrossRef]
21. Chang, H.; Qin, J.; Xiao, P.; Yang, Y.; Zhang, T.; Ma, Y.; Huang, Y.; Chen, Y. Highly Reversible and Recyclable
Absorption under Both Hydrophobic and Hydrophilic Conditions using a Reduced Bulk Graphene Oxide
Material. Adv. Mater. 2016, 28, 3504–3509. [CrossRef]
22. Cao, X.; Yin, Z.; Zhang, H. Three-dimensional graphene materials: Preparation, structures and application in
supercapacitors. Energy Environ. Sci. 2014, 7, 1850–1865. [CrossRef]
36](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-51-320.jpg)
![Sensors 2018, 18, 4337
23. Chabot, V.; Higgins, D.; Yu, A.; Xiao, X.; Chen, Z.; Zhang, J. A review of graphene and graphene oxide
sponge: Material synthesis and applications to energy and the environment. Energy Environ. Sci. 2014, 7,
1564–1596. [CrossRef]
24. Yavari, F.; Chen, Z.; Thomas, A.V.; Ren, W.; Cheng, H.M.; Koratkar, N. High sensitivity gas detection using a
macroscopic three-dimensional graphene foam network. Sci. Rep. 2011, 1, 166. [CrossRef]
25. Ma, Y.; Zhao, M.; Cai, B.; Wang, W.; Ye, Z.; Huang, J. 3D graphene foams decorated by CuO nanoflowers for
ultrasensitive ascorbic acid detection. Biosens. Bioelectron. 2014, 59, 384–388. [CrossRef] [PubMed]
26. Brezovec, B.; Matko, V. Software and Equipment for Remote Testing of Sensors. Sensors 2007, 7, 1306–1316.
[CrossRef]
27. Yu, Y.; Zhang, Y.; Zhang, Z.; Zhang, H.; Song, X.; Cao, M.; Che, Y.; Dai, H.; Yang, J.; Wang, J.; et al.
Broadband Phototransistor Based on CH3NH3PbI3 Perovskite and PbSe Quantum Dot Heterojunction. J.
Phys. Chem. Lett. 2017, 8, 445–451. [CrossRef]
28. Yu, Y.; Zhang, Y.; Song, X.; Zhang, H.; Cao, M.; Che, Y.; Dai, H.; Yang, J.; Zhang, H.; Yao, J. High Performances
for Solution-Pocessed 0D-0D Heterojunction Phototransistors. Adv. Opt. Mater. 2017, 5, 1700565. [CrossRef]
29. Yu, Y.; Zhang, Y.; Song, X.; Zhang, H.; Cao, M.; Che, Y.; Dai, H.; Yang, J.; Zhang, H.; Yao, J. PbS-Decorated
WS2 Phototransistors with Fast Response. ACS Photonics 2017, 4, 950–956. [CrossRef]
30. Smith, A.D.; Elgammal, K.; Niklaus, F.; Delin, A.; Fischer, A.C.; Vaziri, S.; Forsberg, F.; Rasander, M.;
Hugosson, H.; Bergqvist, L.; et al. Resistive graphene humidity sensors with rapid and direct electrical
readout. Nanoscale 2015, 7, 19099–19109. [CrossRef]
31. Gutierrez, H.R.; Perea-Lopez, N.; Elias, A.L.; Berkdemir, A.; Wang, B.; Lv, R.; Lopez-Urias, F.; Crespi, V.H.;
Terrones, H.; Terrones, M. Extraordinary room-temperature photoluminescence in triangular WS2
monolayers. Nano Lett. 2013, 13, 3447–3454. [CrossRef] [PubMed]
32. Ghosh, A.; Late, D.J.; Panchakarla, L.S.; Govindaraj, A.; Rao, C.N.R. NO2 and humidity sensing characteristics
of few-layer graphenes. J. Exp. Nanosci. 2009, 4, 313–322. [CrossRef]
33. Cai, J.; Lv, C.; Aoyagi, E.; Ogawa, S.; Watanabe, A. Laser Direct Writing of a High-Performance All-Graphene
Humidity Sensor Working in a Novel Sensing Mode for Portable Electronics. ACS Appl. Mater. Interfaces
2018, 10, 23987–23996. [CrossRef] [PubMed]
34. Zhang, K.-L.; Hou, Z.-L.; Zhang, B.-X.; Zhao, Q.-L. Highly sensitive humidity sensor based on graphene
oxide foam. Appl. Phys. Lett. 2017, 111, 153101. [CrossRef]
35. Trung, T.Q.; Duy, L.T.; Ramasundaram, S.; Lee, N.-E. Transparent, stretchable, and rapid-response humidity
sensor for body-attachable wearable electronics. Nano Res. 2017, 10, 2021–2033. [CrossRef]
36. Leng, X.; Luo, D.; Xu, Z.; Wang, F. Modified graphene oxide/Nafion composite humidity sensor and its
linear response to the relative humidity. Sens. Actuators B Chem. 2018, 257, 372–381. [CrossRef]
37. Naik, G.; Krishnaswamy, S. Room-Temperature Humidity Sensing Using Graphene Oxide Thin Films.
Graphene 2016, 5, 1–13. [CrossRef]
38. Yu, H.W.; Kim, H.K.; Kim, T.; Bae, K.M.; Seo, S.M.; Kim, J.M.; Kang, T.J.; Kim, Y.H. Self-powered humidity
sensor based on graphene oxide composite film intercalated by poly(sodium 4-styrenesulfonate). ACS Appl.
Mater. Interfaces 2014, 6, 8320–8326. [CrossRef]
39. Guo, L.; Jiang, H.B.; Shao, R.Q.; Zhang, Y.L.; Xie, S.Y.; Wang, J.N.; Li, X.B.; Jiang, F.; Chen, Q.D.; Zhang, T.
Two-beam-laser interference mediated reduction, patterning and nanostructuring of graphene oxide for the
production of a flexible humidity sensing device. Carbon 2012, 50, 1667–1673. [CrossRef]
40. Zhu, Z.-T.; Mason, J.T.; Dieckmann, R.; Malliaras, G.G. Humidity sensors based on pentacene thin-film
transistors. Appl. Phys. Lett. 2002, 81, 4643–4645. [CrossRef]
© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
37](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-52-320.jpg)
![sensors
Article
Design and Implementation of an Infrared Radiant
Source for Humidity Testing
Hong Zhang *, Chuansheng Wang, Xiaorui Li, Boyan Sun and Dong Jiang
School of Computer Science and Technology, Harbin University of Science and Technology, 52 Xuefu Road,
Harbin 150080, China; wangchuansheng994@163.com (C.W.); 13682088813@163.com (X.L.);
15754509280@163.com (B.S.); wdyu2004@163.com (D.J.)
* Correspondence: zhangh@hrbust.edu.cn; Tel.: +86-177-6655-5090 or +86-138-3609-9065
Received: 26 June 2018; Accepted: 11 September 2018; Published: 13 September 2018
Abstract: A novel way to measure humidity through testing the emissivity of an area radiant source
is presented in this paper. The method can be applied in the environment at near room temperature
(5~95 ◦C) across the relative humidity (RH) range of 20~90% RH. The source, with a grooved radiant
surface, works in the far infrared wavelength band of 8~12 μm. The Monte-Carlo model for thermal
radiation was set up to analyze the V-grooved radiant surface. Heat pipe technology is used to
maintain an isothermal radiant surface. The fuzzy-PID control method was adopted to solve the
problems of intense heat inertia and being easily interfered by the environment. This enabled the
system to be used robustly across a large temperature range with high precision. The experimental
results tested with a scanning radiant thermometer showed that the radiant source can provide a
uniform thermal radiation capable of satisfying the requirements of humidity testing. The calibration
method for the radiant source for humidity was explored, which is available for testing humidity.
Keywords: infrared radiant source; Monte Carlo method; emissivity; calibration; humidity
1. Introduction
Compared to traditional humidity measurement methods, innovative electronic testing methods
involving humidity sensors such as hygristors and humicaps are the current research direction.
Even though electric methods have fast responses, they often lack stability and their accuracy is
improved little. Widely available commercial humidity sensors composed of humicaps use embedded
microprocessors, such as the DHT11 with ±5% RH precision, and 1% RH resolution which are
convenient to use. Because humidity is often mingled together with temperature, the precision and
the humidity measurement range are easily affected by the temperature. Humidity, which reflects the
degree of dryness of the atmosphere, is an important variable that is extensively tested in agriculture,
industry, hospital and warehouse. Some sensors with new materials possessing resistive and capacitive
features have been explored, which include a sulfonated polycarbonate resistive humidity sensor [1],
polyimide-based capacitive humidity sensor [2], a high-performance capacitive humidity sensor with
novel electrode and polyimide layer capacitive humidity sensors [3], and some sensors with improved
sensing properties whose response and recovery times are 14.5 s and 34.27 s, respectively, for humidity
levels between 33% RH and 95% RH at 102 Hz. [4]. These kinds of humidity sensors often have long
response times. Most studies focus on material and processing innovations to increase the humidity
testing precision. There have been few breakthroughs in hygristors because these are easily interfered
by the environment. Novel ways using new effects such as optical properties present approaches to
measure humidity [5]. Because humidity is a factor that affects the radiation measurement, humidity
can be measured indirectly by testing radiation changes [6,7].
As a standard radiant source, a blackbody is usually adopted for calibrating infrared instruments
such as pyrometers and radiant thermometers. Industrial blackbodies ranging from −50~2500 ◦C
Sensors 2018, 18, 3088; doi:10.3390/s18093088 www.mdpi.com/journal/sensors
38](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-53-320.jpg)
![Sensors 2018, 18, 3088
have been developed by the National Research Council (NRC) of Canada in order to calibrate optical
testing devices [8]. A high-spatial-resolution multi-spectral imager (ASTER) on the first platform
(Terra) of NASA’s Earth Observing System requires a blackbody radiant source on a satellite for
calibration purposes [9]. Traditional blackbody cavities evaluated by the Bedford methods [10] usually
possess symmetrical shapes with small apertures, which makes them suitable for the high temperature
range case, but not for the case of near room temperature range measurements. Minimum Resolvable
Temperature Difference (MRTD) sensitivity requires that a radiant source working in the far infrared
scope should be an area source which can provide a stable radiant flux with high uniformity. As to
environmental humidity measurement, according to MRTD, an area radiant source ranging from
5~95 ◦C in the wavelength bands of 8~12 μm, is required. Under a certain temperature, a source
should emit a stable radiation which is monitored by a radiometer. In order to enhance the effective
emissivity of a radiant surface, its surface is often processed into grooves or mini holes. The radiant
surface of the source possesses concentric V-grooves which can increase the effective emissivity.
Among statistical evaluation methods, Monte Carlo methods have been widely applied in optical
radiometry and blackbody cavity analysis [11]. Monte Carlo methods possess advantages which are
greater than exactitude methods in complex radiant characteristic analysis. Therefore, the Monte
Carlo method is adopted to analyze the effective emissivity of the radiant source. After a theoretical
analysis on the distribution of the effective emissivity of the radiant surface, the source structure was
constructed. Heat pipe technology keeps the source isothermal and the temperature control system
ensures that the source is stable at a certain temperature. The radiant source has a broad applications
in various fields, such as infrared imaging, infrared measurement and humidity test.
Although there are tens of ways to measure humidity, among which the most traditional methods
are the dry and wet bulb thermometer whose precision is lower compared with modern electronic
methods, most of these ways are not satisfactory in terms of precision and stability [7]. Capacitor
sensors which possess fast response advantages are employed to test humidity, but their measuring
precision is easily affected by electromagnetic interference. Besides, humidity testing is often affected
by the environmental temperature, which is a factor that makes humidity sensors’ precision not be
high. Humidity testing through radiation possesses advantages of fast response and robustness with
high precision. Our research on an infrared source which has highly sensitivity for humidity may
provide an improved way to measure humidity.
2. Analysis on Characteristics of Radiant Surface
The Monte-Carlo method was utilized for analyzing the radiant surface with concentric V-grooves.
Assuming that the surface is diffuse (Lambertian), the calculation on its luminance follows Lambert’s
cosine Law. The Monte-Carlo Model of thermal radiation was set up. A Monte-Carlo simulation
is implemented through random sampling that is based on probability models whose deduction
methods are based upon actual physical models. Exactitude methods like the Bedford method are
just suitable to calculate simple symmetric cavity shapes. Although Monte-Carlo methods are flexible
enough to be used for complicated cases, they are often regarded as unreliable methods with low
precision. The effective emissivity of a cone was calculated by both the Monte Carlo method and the
Bedford method, respectively. By comparing the results from the both methods, the correctness of the
Monte Carlo method was proven.
2.1. Monte-Carlo Model in Thermal Radiation
The idea of a Monte-Carlo method is to set up a probability model or stochastic process whose
parameters are equal to the solution of the problem to investigate, then to calculate the statistical
characteristics of the required parameters through sampling, and the solution can be solved based
on a vast number of observations. In thermal radiation calculations, local temperature and radiation
fluxes are usually involved. The process of thermal radiation exchange is regarded as the movement of
discrete energy beams. In this way, the local radiation flux can be obtained by calculating the number
39](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-54-320.jpg)
![one-gross-of-gloves long. Buffon would only sit down to write after
taking a bath and donning pure linen with a full frilled bosom.
Haydn[71] declared that he could not compose unless he wore the
large seal-ring which Frederick the Great had given him. He would
sit wrapped in silence for an hour or more, after which he would
seize his pen and write rapidly without touching a musical
instrument; and he rarely altered a line. In early life, poor, freezing
in a miserable garret, he studied the rudiments of his favorite art by
the side of an old broken harpsichord. For a period of six years he
endured a bitter conflict with poverty, being often compelled for the
sake of warmth to lie in bed most of the day as well as the night.
Finally he was relieved from this thraldom by the generosity of his
patron, Prince Esterhazy, a passionate lover of music, who appointed
him his chapel-master, with a salary sufficient to keep him supplied
with the ordinary comforts of life.
Crébillon the elder, a celebrated lyric poet and member of the French
Academy, was enamoured of solitude, and could only write
effectively under such circumstances. His imagination teemed with
romances, and he produced eight or ten dramas which enjoyed
popularity in their day,—about 1776. One day, when he was alone
and in a deep reverie, a friend entered his study hastily. Don't
disturb me, cried the author, I am enjoying a moment of
happiness: I am going to hang a villain of a minister, and banish
another who is an idiot.
We have lately mentioned Dumas. Hans Christian Andersen,
speaking of the various habits of authors, thus refers to the elder
Dumas, with whom he was intimate: I generally found him in bed,
even long after mid-day, where he lay, with pen, ink, and paper by
his side, and wrote his newest drama. On entering his apartment I
found him thus one day; he nodded kindly to me, and said: 'Sit
down a minute. I have just now a visit from my Muse; she will be
going directly.' He wrote on, and after a brief silence shouted 'Vivat'
sprang out of bed, and said, 'The third act is finished!'[72]](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-56-320.jpg)
![written fastest have always pleased the most. I wrote the 'Essay on
Criticism' fast, for I had digested all the matter in prose before I
began it in verse.
I never work better, says Luther, than when I am inspired by
anger: when I am angry, I can write, pray, and preach well; for then
my whole temperament is quickened, my understanding sharpened,
and all mundane vexations and temptations depart. We are
reminded of Burke's remark in this connection: A vigorous mind is
as necessarily accompanied with violent passions as a great fire with
great heat. Luther, however ribald he may have been at times, had
the zeal of honesty. There was not a particle of vanity or self-
sufficiency in the great reformer. Do not call yourselves Lutherans,
he said to his followers; call yourselves Christians. Who and what is
Luther? Has Luther been crucified for the world?
Churchill,[73] the English poet and satirist, was so averse to
correcting and blotting his manuscript that many errors were
unexpunged, and many lines which might easily have been improved
were neglected. When expostulated with upon this subject by his
publisher, he replied that erasures were to him like cutting away so
much of his flesh; thus expressing his utter repugnance to an
author's most urgent duty. Though Macaulay tells us that his vices
were not so great as his virtues, still he was dissipated and
licentious. Cowper was a great admirer of his poetry, and called him
the great Churchill. George Wither,[74] the English poet, satirist,
and political writer, was compelled to watch and fast when he was
called upon to write. He went out of himself, as he said, at such
times, and if he tasted meat or drank one glass of wine he could not
produce a verse or sentence.
Rogers, who wrote purely con amore, took all the time to perfect his
work which his fancy dictated, and certainly over-refined many of his
compositions. The Pleasures of Memory occupied him seven years.
In writing, composing, re-writing, and altering his Columbus and
Human Life, each required just double that period of time before](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-58-320.jpg)
![in full canonicals, a fact which he frankly admitted. Milton would not
attempt to compose except between the vernal and autumnal
equinoxes, at which season his poetry came as if by inspiration, and
with scarcely a mental effort.[75] Thomson, Collins, and Gray
entertained very similar ideas, which when expressed so incensed
Dr. Johnson that he publicly ridiculed them. Crabbe fancied that
there was something in the effect of a sudden fall of snow that in an
extraordinary manner stimulated him to poetic composition; while
Lord Orrery found no stimulant equal to a fit of the gout!—all of
which fancies are but mild forms of monomania. James Hogg (the
Ettrick Shepherd) was only too glad to write without any of these
accessories, when he could get any material to write upon. He used
to employ a bit of slate, for want of the necessary paper and ink.
The son of an humble Scottish farmer, he experienced all sorts of
misfortunes in his endeavors to pursue literature as a calling. He was
both a prose and poetic writer of considerable native genius, and
formed one of the well-drawn characters of Christopher North's
Noctes Ambrosianæ. N. P. Willis in the latter years of his life was
accustomed to ride on horseback before he sat down to write. He
believed there was a certain nervo-vital influence imparted from the
robust health and strength of the animal to the rider, as he once told
the writer of these pages; and, so far as one could judge, the
influence upon himself certainly favored such a conclusion.
Some authors frankly acknowledge that they have not the necessary
degree of patience to apply themselves to the correction of their
manuscripts. Ovid, the popular Roman poet, admitted this. Such
people may compose with pleasure, but there is the end; neither a
sense of responsibility nor a desire for correctness can overcome
their constitutional laziness. Pope, Dryden, Moore, Coleridge, Swift,
—in short, nine-tenths of the popular authors of the past and the
present, all change, correct, amplify, or contract, and interline more
or less every page of manuscript which they produce, and often to
such a degree as greatly to confuse the compositors. Richard
Savage, the unfortunate English poet, could not, or would not, bring
himself to correct his faulty sentences, being greatly indebted to the](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-62-320.jpg)
![intelligence of the proof-reader for the presentable form in which his
writings finally appeared. Julius Scaliger, a celebrated scholar and
critic, was, on the other hand, an example of remarkable
correctness, so that his manuscript and the printer's pages
corresponded exactly, page for page and line for line. Hume,[76] the
historian, was never done with his manifold corrections; his sense of
responsibility was unlimited, and his appreciation of his calling was
grand. Fénelon and Gibbon were absolutely correct in their first
efforts; and so was Adam Smith, though he dictated to an
amanuensis.
We are by no means without sympathy for those writers who dread
and avoid the reperusal and correction of their manuscripts. Only
those who are familiar with the detail of book-making can possibly
realize its trying minutiæ. When one has finished the composition
and writing of a chapter, his work is only begun; it must be read and
re-read with care, to be sure of absolute correctness. When once in
type, it must be again carefully read for the correction of printer's
errors, and again revised by second proof; and finally a third proof is
necessary, to make sure that all errors previously marked have been
corrected. By this time, however satisfactory in composition, the text
becomes more tedious than a twice-told tale. Any author must be
singularly conceited who can, after such experience, take up a
chapter or book of his own production and read it with any great
degree of satisfaction. Godeau, Bishop of Venice, used to say that
to compose is an author's heaven; to correct, an author's
purgatory; but to revise the press, an author's hell!
Guido Reni, whose superb paintings are among the gems of the
Vatican, in the height of his fame would not touch pencil or brush
except in full dress. He ruined himself by gambling and dissolute
habits, and became lost as to all ambition for that art which had
been so grand a mistress to him in the beginning. He finally arrived
at that stage where he lost at the gaming-table and in riotous living
what he earned by contract under one who managed his affairs,
giving him a stipulated sum for just so much daily work in his studio.](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-63-320.jpg)
![Such was the famous author of that splendid example of art, the
Martyrdom of Saint Peter, in the Vatican. Parmigiano, the eminent
painter, was full of the wildness of genius. He became mad after the
philosopher's stone, jilting art as a mistress, though his eager
creditors forced him to set once more to work, though to little effect.
Great painters, like great writers, have had their peculiar modes of
producing their effects. Thus Domenichino was accustomed to
assume and enact before the canvas the passion and character he
intended to depict with the brush. While engaged upon the
Martyrdom of Saint Andrew, Caracci, a brother painter, came into
his studio and found him in a violent passion. When this fit of
abstraction had passed, Caracci embraced him, admitting that
Domenichino had proved himself his master, and that he had learned
from him the true manner of expressing sentiment or passion upon
the canvas.
Richard Wilson, the eminent English landscape-painter, strove in
vain, he said, to paint the motes dancing in the sunshine. A friend
coming into his studio found the artist sitting dejected on the floor,
looking at his last work. The new-comer examined the canvas and
remarked critically that it looked like a broad landscape just after a
shower. Wilson started to his feet in delight, saying, That is the
effect I intended to represent, but thought I had failed. Poor Wilson
possessed undoubted genius, but neglected his art for brandy, and
was himself neglected in turn. He was one of the original members
of the Royal Academy.
Undoubtedly, genius is at times nonplussed and at fault, like plain
humanity, and is helped out of a temporary dilemma by accident,—
as when Poussin the painter, having lost all patience in his fruitless
attempts to produce a certain result with the brush, impatiently
dashed his sponge against the canvas and brought out thereby the
precise effect desired; namely, the foam on a horse's mouth.
Washington Allston[77] is recalled to us in this connection, one of the
most eminent of our American painters, and a poet of no ordinary](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-64-320.jpg)
![pretensions. The Sylphs of the Seasons and other Poems was
published in 1813. He was remarkable for his graphic and animated
conversational powers, and was the warm personal friend of
Coleridge and Washington Irving. Irving says, His memory I hold in
reverence and affection as one of the purest, noblest, and most
intellectual beings that ever honored me with his friendship. While
living in London he was elected associate of the Royal Academy.
Bostonians are familiar with Allston's half-finished picture of
Belshazzar's Feast, upon which he was engaged when death
snatched him from his work.
CHAPTER IV.
It has been said that the first three men in the world were a
gardener, a ploughman, and a grazier; while all political economists
admit that the real wealth and stamina of a nation must be looked
for among the cultivators of the soil. Was it not Swift who declared
that the man who could make two ears of corn or two blades of
grass grow upon a spot of ground where only one grew before,
deserved better of mankind than the whole race of politicians?
Bacon, Cowley, Sir William Temple, Buffon, and Addison were all
attached to horticulture, and more or less time was devoted by them
to the cultivation of trees and plants of various sorts; nor did they
fail to record the refined delight and the profit they derived
therefrom. Daniel Webster was an enthusiastic agriculturist; so were
Washington, Adams, Jefferson, Walter Scott, Horace Greeley,
Gladstone, Evarts,[78] Wilder, Loring, Poore, and a host of other
contemporaneous and noted men. They who labor in the earth,
said Jefferson, are the chosen people of God.](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-65-320.jpg)
![It was asked in Jerrold's club, on a certain occasion, what was the
best definition of dogmatism. There is but one, he instantly
replied,—the maturity of puppyism. A member remarked one day
that the business of a mutual acquaintance was going to the devil.
All right, said Jerrold; then he's sure to get it back again. Another
member who was not very popular with the club, hearing a certain
melody spoken of, said, That always carries me away when I hear
it. Cannot some one whistle it? asked Jerrold. Another member,
who was rather given to boasting, said: Very singular! I dined at the
Marchioness of So-and-so's last week, and we actually had no fish.
Easily explained, said Jerrold; no doubt they had eaten it all
upstairs. When Heraud, a somewhat bombastic versifier, asked him
if he had read his Descent into Hell, Jerrold instantly replied, No; I
had rather see it. Being asked what was the idea of Harriet
Martineau's rather atheistical book, he answered that it was plain
enough,—There is no God, and Harriet is his Prophet. This is even
better than the remark of another wit who, when asked what was
the outcome of a meeting before which three of the ablest and most
dogmatic Positivists in England made speeches, replied that the
result arrived at was this: that there were three persons and no God.
Jerrold could not confine himself to any regular system of work, but
drove the quill at such times and only to such purpose as his erratic
mood indicated, jumping from one subject to another like one
crossing a brook upon stepping-stones. This, however, was a habit
by no means peculiar to Douglas Jerrold. There are some ludicrous
stories told of him; like that of his being pursued by a printer's boy
about the town, from house to club, from club to the theatre, and so
on, and finally of his being overtaken, getting into a corner and
writing an admirable article with pencil and paper on the top of his
hat.
Agassiz,[79] the great Swiss naturalist, who became an adopted and
honored son of this country, was singularly unmethodical in his
habits of professional labor. If he was suddenly seized with an
interest in some scientific inquiry, he would pursue it at once, putting
by all present work, though it might be that he had just got fairly](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-68-320.jpg)
![Coleridge was much addicted to the habit of marginal writing; which,
though sadly wasteful on his own part, was very enriching to those
friends who loaned him from their libraries.[80] Charles Lamb, who
was not inclined to spare book-borrowers as a tribe, had no
reflections to cast upon Coleridge for this habit. The depth, weight,
and originality of his comments as hastily and carelessly penned on
the margins of books were wonderful, and if collected and classified
would form several volumes, not only of captivating interest, but of
rare critical value, as the few which have been brought together
abundantly prove. In one volume which he returned to Lamb is this
memorandum: I shall die soon, my dear Charles Lamb, and then
you will not be vexed that I have be-scribbled your book. S. T. C.,
May 2d, 1811. Elia valued these marginal notes beyond price, and
said that to lose a volume to Coleridge carried some sense and
meaning with it. These critical notes often nearly equalled in
quantity of matter the original text. In his article upon the subject,
Lamb says, I counsel thee, shut not thy heart nor thy library against
S. T. C. As we have already said, while this erratic expenditure of
Coleridge's rare literary taste and judgment enriched others, it in a
degree impoverished himself; for had the same time and thought
been expended upon consecutive literary work, it would have
produced volumes of inestimable value to the world at large, and
have proved monumental to their author.
Byron was addicted to marginalizing; and though he could not equal
Coleridge in the profundity of his criticisms, or impart such charming
interest to them, still he was quite original and often piquant. Burns
contented himself with trifling criticisms of approval or disapproval
pencilled in the margin of books, especially poetical ones, which
were nearly all he was in the habit of reading.
Many famous authors and public men have been extravagantly fond
of the rod and line, disciples of that patient and poetical angler,
Izaak Walton. George Herbert, the English poet; Henry Wotton,
diplomatist and author; Dr. Paley, Archdeacon of Carlisle; John
Dryden, poet and dramatist; Sydney Smith, the witty divine; Sir](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-70-320.jpg)
![Humphry Davy, the eminent chemist,—all were devoted anglers.[81]
This brief list might be largely increased. Bulwer-Lytton says:
Though no participator in the joys of more vehement sport, I have
a pleasure that I cannot reconcile to my abstract notions of the
tenderness due to dumb creatures, in the tranquil cruelty of angling.
I can only palliate the wanton destructiveness of my amusement by
trying to assure myself that my pleasure does not spring from the
success of the treachery I practise towards a poor little fish, but
rather from that innocent revelry in the luxuriance of summer life
which only anglers enjoy to the utmost. Walton puts himself on
record in these words: We may say of angling, as Dr. Boteler said of
strawberries: 'Doubtless God could have made a better berry, but
doubtless God never did;' and so, if I might be judge, God never did
make a more calm, quiet, innocent recreation than angling. Sydney
Smith declared it to be an occupation fit for a bishop, and that it
need in no way interfere with sermon-making.
Perhaps the best thing said or done in angling is an unpublished
anecdote of the great preacher to the seamen,—the late Father
Taylor, of Boston. He was once lured to try his hand at the rod, and
soon brought up a very little fish that had been tempted by his bait.
He took the small creature carefully from the hook, gazed at it a
moment, and then cast it back into the water, with this advice: My
little friend, go and tell your mother that you have seen a ghost!
Dr. Parr, the profound English scholar, was a most inveterate smoker;
so was Charles Lamb,[82] who one day said to his doctor, I have
acquired this habit by toiling over it, as some men toil after virtue.
Robert Hall, the popular English divine, was very much addicted to
tobacco and other stimulants. A friend who found him in his study
blowing forth clouds of smoke from his lips, said, There you are, at
your old idol! Yes, replied the divine, burning it. Napoleon could
never abide smoking tobacco; yet observing how much other men
seemed to enjoy it, he tried to acquire the habit, but finally gave it
up in disgust. He, however, took snuff to excess. Sir Walter Scott](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-71-320.jpg)
![was very fond of smoking. Thackeray, like Burns, loved to get away
by himself and enjoy the flavor of a rank tobacco-pipe. Carlyle, like
Tennyson, did not care for a cigar, but kept a pipe in his mouth most
of his waking hours. Bulwer-Lytton was a ceaseless smoker; and
there are few if any notable Germans who have not been addicted to
the same indulgence. The nicotine produced from tobacco is one of
the most deadly of all poisons, as has been proven by some startling
experiments in the Paris hospitals.[83] Thackeray said there was
good eating in Scott's novels. Extending the remark, it might be
added that there was good drinking in those of Dickens, and good
smoking in those of Thackeray.
Dean Swift relieved his sombre moods by harnessing his servants
with cords and driving them, school-boy fashion, up and down the
stairs and through the garden of the deanery of St. Patrick's
Cathedral, Dublin. Dickens was controlled by a nervous activity
which made him crave physical exercise of some sort, and he daily
found relief in an eight or ten mile walk. Thackeray once told the
author of these pages that he preferred to take his exercise driving
upon very easy roads. When Dickens was in this country he was
frequently accompanied in his long walks by the late James T. Fields,
who was ever ready to sacrifice himself to the pleasure of others. Mr.
Fields was not partial to extreme pedestrian exercise, and the author
of the Pickwick Papers tested his good-nature to the verge of
exhaustion in this respect. Dumas, when not otherwise engaged,
was accustomed to go down into his kitchen, and, deposing the
servants, cook his own dinner; and an excellent cook he must have
been, if one half the stories rife about him be true. Besides, did he
not write an original cook-book, which still stands for good authority
in the cafés of the boulevards?
Dr. Warton, the English critic and author, as represented by
contemporary authority, was noted for a love of vulgar society, which
he daily sought in low tap-rooms and gin-shops, where he joked
away the evening hours. Turner the painter had similar tastes and
habits, though he was of a reserved and unsociable character, and](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-72-320.jpg)
![was a child himself in many respects; in illustration of which the
reader has only to recall the poet's singular amusement of sailing
paper boats whenever he found himself conveniently near a pond.
So long as the paper which he chanced to have about him lasted, he
remained riveted to the spot. First he would use the cover of letters,
next letters of little value; but he could not resist the temptation,
finally, of employing for the purpose the letters of his most valued
correspondents. He always carried a book in his pocket, but the fly-
leaves were all consumed in forming these paper boats and setting
them adrift to constitute a miniature fleet. Once he found himself on
the banks of the Serpentine River without paper of any sort except a
ten-pound note. He refrained for a while; but presently it was rapidly
twisted into a boat by his skilful fingers, and devoted to his boat-
sailing purpose without further delay. Its progress being watched, it
was finally picked up on the opposite shore of the river and returned
to the owner for more legitimate use.
Charles Lamb in his quaint way says: I know that sweet children are
the sweetest things in nature, not even excepting the delicate
creatures which bear them; but the prettier the kind of a thing is,
the more desirable it is that it should be pretty of its kind. One daisy
differs not much from another in glory; but a violet should look and
smell the daintiest.[84]
Good and substantial food is quite as necessary to authors and
public men, as to those who gain their livelihood by laborious
physical employment. Authors are, however, as a rule, rather inclined
to free indulgence at table. There is as much intemperance in eating
as in drinking. Tom Moore, who was the best diner-out of his day,
said, by way of excusing this habit, In grief, I have always found
eating a wonderful relief. N. P. Willis was quite a gourmand. There
are, he once wrote, so few invalids untemptable by those deadly
domestic enemies, sweetmeats, pastry, and gravies, that the usual
civilities at a meal are very like being politely assisted to the grave.
It is certainly better to punish our appetites than to be punished by
them. Dickens and Thackeray were both inclined to free indulgence](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-75-320.jpg)
![at the table, the former being struck with death at a public banquet.
Dean Swift often gave better advice than he was himself inclined to
follow. He says: Temperance, meaning both in eating and drinking,
is a necessary virtue to great men, since it is the parent of the
mind, which philosophy allows to be one of the greatest felicities in
life. Macready, the famous English tragedian, would not touch food
of any kind for some hours before making one of his grand dramatic
efforts, but drank freely of strong tea before appearing in public,—a
subtle stimulant in which the late Rufus Choate freely indulged,
particularly before addressing a jury.
Abstinence in diet was a special virtue with Milton. Shelley utterly
despised the pleasures of the table. Walter Scott was an abstemious
eater. Pope was a great epicure, and so was the poet Gay. Speaking
of appetite, Coleridge tells us of a man he once saw at a dinner-
table, who struck him as remarkable for his dignity and wise face.
The awful charm of his manner was not broken until the muffins
appeared, and then the wise one exclaimed, Them's the jockeys for
me! Dignity is sometimes very rudely unmasked, and an imposing
air is nearly always the cloak of a fool. Newton lived on the simplest
food. If Aristotle could diet on acorns, he said, so can I; and
before sitting down to study he exercised freely and abstained from
food. Dr. George Fordyce, the eminent Scotch physician, ate but one
meal a day, saying that if one meal in twenty-four hours was enough
for a lion, it was sufficient for a man; but in order not to be like the
lion, he drank a bottle of port, half a pint of brandy, and a pitcher of
ale with his one meal. Lamartine used to pass one day in ten fasting,
as he said, to clear both stomach and brain. Aristo, the stoic
philosopher, used to fast for days on acorns. Thomas Byron, a well-
known author, never ate flesh of any sort. Dryden's favorite dish was
a chine of bacon. Charles Lamb was enamoured of roast pig. He
said, You can no more improve sucking pig than you can refine a
violet! Keats was a very fastidious eater, but was fond of the table,
especially where there was good wine,[85] and yet he was not
addicted to its intemperate use. Dr. Johnson was greedy over boiled
mutton; and Dr. Rhondelet, the famous writer on fishes, was so fond](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-76-320.jpg)
![of figs that he died from having at one time eaten immoderately of
them. Barrow, one of the greatest of English theologians and
mathematicians, is said to have died of a surfeit of pears,—a fruit of
which he was extravagantly fond.
Gastronomic appetite and reason have been compared to two
buckets in a well; when one is at the top the other is at the bottom.
Byron nearly starved himself to prevent growing gross and
uninteresting in physical aspect. Addison was addicted to port and
claret, and was accustomed, as already spoken of, while meditating
a moral or political essay, to pace up and down the long gallery of
Holland House.[86] When a humorous suggestion occurred to his
fertile fancy, he solaced himself with claret; or fortified himself with a
glass of port when a moral sentiment required to be enforced by an
impressive close to a beautifully constructed sentence.[87] This was
after his frigid marriage to the Dowager Countess of Warwick. On his
death-bed he is reported to have said to her graceless son, See
how a Christian can die! Probably the profligate youth, spying his
father-in-law as he walked in the gallery, might have irreverently
remarked: See how a Christian can drink! But the truth is that
Addison, judged by the habits of his time, should be considered a
moderate drinker. Poe's nerves were so shattered that a slight
amount of wine would intoxicate him into a frenzy of dissipation; the
same amount swallowed by a regular toper would hardly disturb his
brain at all. While Pitt was quite a young man, he was so weakly that
his physician ordered him to drink freely of port wine, and he thus
contracted the habit of depending upon stimulants, and could not do
without them. Lord Greville tells us he has seen him swallow a bottle
of port wine by tumblerfuls before going to the House. This,
together with the habit of late suppers, helped materially to shorten
his life.[88]
Goldsmith had a queer fancy for sassafras tea, from which he
imagined he derived an excellent tonic effect. Such a relish had
certainly one element to recommend it,—and that was its
harmlessness. Dr. Shaw, the English naturalist, nearly killed himself](https://image.slidesharecdn.com/18186769-250521105949-233153b6/85/Humidity-Sensors-Peter-W-Mccarthy-Editor-Zhuofu-Liu-Editor-77-320.jpg)
Humidity Sensors Peter W Mccarthy Editor Zhuofu Liu Editor
- 1. Humidity Sensors Peter W Mccarthy Editor Zhuofu Liu Editor download https://ebookbell.com/product/humidity-sensors-peter-w-mccarthy- editor-zhuofu-liu-editor-36373538 Explore and download more ebooks at ebookbell.com
- 2. Here are some recommended products that we believe you will be interested in. You can click the link to download. Humidity Sensors Types Nanomaterials And Environmental Monitoring Types Nanomaterials And Environmental Monitoring 1st Edition Christopher T Okada https://ebookbell.com/product/humidity-sensors-types-nanomaterials- and-environmental-monitoring-types-nanomaterials-and-environmental- monitoring-1st-edition-christopher-t-okada-51348842 Humidity And Electronics Corrosion Reliability Issues And Preventive Measures Rajan Ambat https://ebookbell.com/product/humidity-and-electronics-corrosion- reliability-issues-and-preventive-measures-rajan-ambat-47937956 Temperature And Humidity Independent Control Thic Of Airconditioning System 1st Edition Xiaohua Liu https://ebookbell.com/product/temperature-and-humidity-independent- control-thic-of-airconditioning-system-1st-edition-xiaohua-liu-4636056 Handbook Of Humidity Measurement Volume 1 Spectroscopic Methods Of Humidity Measurement First Edition Korotcenkov https://ebookbell.com/product/handbook-of-humidity-measurement- volume-1-spectroscopic-methods-of-humidity-measurement-first-edition- korotcenkov-7033408
- 3. Profiles Of Wind Temperature And Humidity Over The Arabian Sea F I Badgley C A Paulson M Miyake https://ebookbell.com/product/profiles-of-wind-temperature-and- humidity-over-the-arabian-sea-f-i-badgley-c-a-paulson-m- miyake-51896630 Specificity Of Deformation And Strength Behavior Of Massive Elements Of Concrete Structures In A Medium With Low Humidity 1st Edition Koryun Karapetyan https://ebookbell.com/product/specificity-of-deformation-and-strength- behavior-of-massive-elements-of-concrete-structures-in-a-medium-with- low-humidity-1st-edition-koryun-karapetyan-11047186 Photocatalytic No Removal Over Defective Bibiobr Nanoflowers The Inhibition Of Toxic No2 Intermediate Via High Humidity Yue Xin https://ebookbell.com/product/photocatalytic-no-removal-over- defective-bibiobr-nanoflowers-the-inhibition-of-toxic- no2-intermediate-via-high-humidity-yue-xin-59171334 Humility Aladdin A Dark Contemporary Retelling Brooklyn Cross https://ebookbell.com/product/humility-aladdin-a-dark-contemporary- retelling-brooklyn-cross-54778554 Humility The Virtues Jennifer Cole Wright https://ebookbell.com/product/humility-the-virtues-jennifer-cole- wright-55661624
- 5. Humidity Sensors Advances in Reliability, Calibration and Application Peter W. McCarthy, Zhuofu Liu and Vincenzo Cascioli www.mdpi.com/journal/sensors Edited by Printed Edition of the Special Issue Published in Sensors sensors
- 8. Humidity Sensors Advances in Reliability, Calibration and Application Special Issue Editors Peter W. McCarthy Zhuofu Liu Vincenzo Cascioli MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade
- 9. Zhuofu Liu Harbin Univesity of Science and Technology China Special Issue Editors Peter W. McCarthy University of South Wales UK Vincenzo Cascioli Murdoch University Australia Editorial Office MDPI St. Alban-Anlage 66 4052 Basel, Switzerland This is a reprint of articles from the Special Issue published online in the open access journal Sensors (ISSN 1424-8220) from 2018 to 2019 (available at: https://www.mdpi.com/journal/sensors/special issues/humidity sensors) For citation purposes, cite each article independently as indicated on the article page online and as indicated below: LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. Journal Name Year, Article Number, Page Range. ISBN 978-3-03921-122-7 (Pbk) ISBN 978-3-03921-123-4 (PDF) c 2019 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.
- 10. Contents About the Special Issue Editors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Preface to ”Humidity Sensors” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Hsuan-Yu Chen and Chiachung Chen Determination of Optimal Measurement Points for Calibration Equations—Examples by RH Sensors Reprinted from: Sensors 2019, 19, 1213, doi:10.3390/s19051213 . . . . . . . . . . . . . . . . . . . . 1 Hong Liu, Qi Wang, Wenjie Sheng, Xubo Wang, Kaidi Zhang, Lin Du and Jia Zhou Humidity Sensors with Shielding Electrode Under Interdigitated Electrode Reprinted from: Sensors 2019, 19, 659, doi:10.3390/s19030659 . . . . . . . . . . . . . . . . . . . . . 19 Yu Yu, Yating Zhang, Lufan Jin, Zhiliang Chen, Yifan Li, Qingyan Li, Mingxuan Cao, Yongli Che, Junbo Yang and Jianquan Yao A Fast Response−Recovery 3D Graphene Foam Humidity Sensor for User Interaction Reprinted from: Sensors 2018, 18, 4337, doi:10.3390/s18124337 . . . . . . . . . . . . . . . . . . . . 30 Hong Zhang, Chuansheng Wang, Xiaorui Li, Boyan Sun and Dong Jiang Design and Implementation of an Infrared Radiant Source for Humidity Testing Reprinted from: Sensors 2018, 18, 3088, doi:10.3390/s18093088 . . . . . . . . . . . . . . . . . . . . 38 Zhuofu Liu, Jianwei Li, Meimei Liu, Vincenzo Cascioli and Peter W McCarthy In-Depth Investigation into the Transient Humidity Response at the Body-Seat Interface on Initial Contact Using a Dual Temperature and Humidity Sensor Reprinted from: Sensors 2019, 19, 1471, doi:10.3390/s19061471 . . . . . . . . . . . . . . . . . . . . 56 Amir Orangi, Guillermo A. Narsilio and Dongryeol Ryu A Laboratory Study on Non-Invasive Soil Water Content Estimation Using Capacitive Based Sensors Reprinted from: Sensors 2019, 19, 651, doi:10.3390/s19030651 . . . . . . . . . . . . . . . . . . . . . 72 Torgrim Log Consumer Grade Weather Stations for Wooden Structure Fire Risk Assessment Reprinted from: Sensors 2018, 18, 3244, doi:10.3390/s18103244 . . . . . . . . . . . . . . . . . . . . 101 Andreas Lorek and Jacek Majewski Humidity Measurement in Carbon Dioxide with Capacitive Humidity Sensors at Low Temperature and Pressure Reprinted from: Sensors 2018, 18, 2615, doi:10.3390/s18082615 . . . . . . . . . . . . . . . . . . . . 116 Martta-Kaisa Olkkonen Online Moisture Measurement of Bio Fuel at a Paper Mill Employing a Microwave Resonator † Reprinted from: Sensors 2018, 18, 3844, doi:10.3390/s18113844 . . . . . . . . . . . . . . . . . . . . 127 Zbigniew Suchorab, Marcin Konrad Widomski, Grzegorz Łagód, Danuta Barnat-Hunek and Dariusz Majerek A Noninvasive TDR Sensor to Measure the Moisture Content of Rigid Porous Materials Reprinted from: Sensors 2018, 18, 3935, doi:10.3390/s18113935 . . . . . . . . . . . . . . . . . . . . 138 v
- 11. Yusuke Tsukahara, Osamu Hirayama, Nobuo Takeda, Toru Oizumi, Hideyuki Fukushi, Nagisa Sato, Toshihiro Tsuji, Kazushi Yamanaka and Shingo Akao A Novel Method and an Equipment for Generating the Standard Moisture in Gas Flowing through a Pipe Reprinted from: Sensors 2018, 18, 3438, doi:10.3390/s18103438 . . . . . . . . . . . . . . . . . . . . 158 Jia Qi, Zhen Zhou, Chenchen Niu, Chunyu Wang and Juan Wu Reliability Modeling for Humidity Sensors Subject to Multiple Dependent Competing Failure Processes with Self-Recovery Reprinted from: Sensors 2018, 18, 2714, doi:10.3390/s18082714 . . . . . . . . . . . . . . . . . . . . 169 vi
- 12. About the Special Issue Editors Peter W. McCarthy obtained a BSc jt. Hons in Physiology and a PhD in Neurophysiology from the University of Manchester and the University of St Andrews, respectively. He has valuable experience assessing the activity of the body and its component systems. His awareness for measurement accuracy issues in clinical technology was first raised while working on ear thermometry with the UK’s National Physical Laboratory. He was awarded a full professorship of Clinical Technology at the University of Glamorgan in 2008. His current interests surround the use of technology to better understand the role of neurophysiological sensory feedback mechanisms, with the aim to eventually create intelligent replacements for those with sensory deficits. This includes relating perceptions of the person to body-seat interface parameters, assessing and preventing cervical spine dysfunction in elite sports and optimizing brain-computer interfacing. Zhuofu Liu received his Masters and PhD from Harbin Engineering University, Harbin, China, in 2001 and 2004, respectively. In 2005 he served as an associate professor at the School of Underwater Acoustic Engineering, Harbin Engineering University. In 2006 he worked as an academic visitor at the University of Oxford. From 2007 to 2009 he worked as a research associate at the Welsh Institute of Chiropractic, University of Glamorgan (now University of South Wales), Pontypridd, UK. Since 2010 he has been a professor at the School of Measurement Control and Communication Engineering, Harbin University of Science and Technology. His research interests include image processing, biomedical signal acquisition and analysis, and healthcare information technology. Dr. Liu is currently the principal investigator for several projects investigating the body-seat interface microenvironment. Vincenzo Cascioli obtained a Masters in Chiropractic from Durban University of Technology, South Africa and a PhD in Ergonomics from the University of South Wales, UK. His current research interests involve the use of technology to evaluate the factors, such as temperature, humidity and movement, associated with sitting comfort or discomfort. vii
- 14. Preface to ”Humidity Sensors” This Special Issue, “Humidity Sensors: Advances in Reliability, Calibration and Application”, contains a range of articles illustrating the growth in use and form of humidity sensors. It is obvious from the contents of this volume that humidity detection has come a long way since wet bulb psychrometry. The number of electronic sensor-based methods available for detecting and reporting relative humidity appears to have grown exponentially. However, as one moves further away from the physical measurement of a property, issues of reliability and accuracy of calibration become increasingly important. In the case of humidity, the property of a sensor that enables measurements to be made can also be the property that leads to issues with calibration and sensitivity, as well as recovery of the sensor. All of these factors may limit the uptake and application of the sensors. This volume is a window into the recent, rapid growth in research aimed at finding the best method for sensing humidity in fields ranging from biomedicine, agriculture, and pharmacology to semiconductors and food processing. Never has there been a greater need to study and refine these sensors. In our contribution the editors have taken the opportunity to follow up on colleagues’ questions regarding the source of spurious and short lived, but potentially vital, artifacts associated with one potential use of humidity sensors: assessing seating or mattress breathability. For this, we have gone back to basics to illustrate the effects a delay in the equilibration of temperature at the sensor site can have on the sensor’s reporting of relative humidity in the surrounding environment. This relatively minor artifact shows how believing without questioning can mislead and obfuscate, whereas questioning can open new areas for development. We initially considered this a good point in time to bring together available research (potential and actual) and look at the issues surrounding this measurement. This issue shows the breadth of use and hints at the future potential of these sensors. Peter W. McCarthy, Zhuofu Liu, Vincenzo Cascioli Special Issue Editors ix
- 16. sensors Article Determination of Optimal Measurement Points for Calibration Equations—Examples by RH Sensors Hsuan-Yu Chen 1 and Chiachung Chen 2,* 1 Department of Materials Science and Engineering, University of California, San Diego, CA 92093, USA; wakaharu37@gmail.com 2 Department of Bio-Industrial Mechatronics Engineering, National ChungHsing University, Taichung 40227, Taiwan * Correspondence: ccchen@dragon.nchu.edu.tw; Tel.: +886-4-2285-7562 Received: 26 February 2019; Accepted: 6 March 2019; Published: 9 March 2019 Abstract: The calibration points for sensors must be selected carefully. This study uses accuracy and precision as the criteria to evaluate the required numbers of calibration points required. Two types of electric relative humidity (RH) sensors were used to illustrate the method and the standard RH environments were maintained using different saturated salt solutions. The best calibration equation is determined according to the t-value for the highest-order parameter and using the residual plots. Then, the estimated standard errors for the regression equation are used to determine the accuracy of the sensors. The combined uncertainties from the calibration equations for different calibration points for the different saturated salt solutions were then used to evaluate the precision of the sensors. The accuracy of the calibration equations is 0.8% RH for a resistive humidity sensor using 7 calibration points and 0.7% RH for a capacitance humidity sensor using 5 calibration points. The precision is less than 1.0% RH for a resistive sensor and less than 0.9% RH for a capacitive sensor. The method that this study proposed for the selection of calibration points can be applied to other sensors. Keywords: calibration points; saturated salt solutions; humidity sensors; measurement uncertainty 1. Introduction The performance of sensors is key for modern industries. Accuracy and precision are the most important characteristics. Calibration ensures sensors’ performance. When a sensor is calibrated, the reference materials or reference environments must be specified. For a balance calibration, a standard scale is the reference materials. For temperature calibration, the triple point of ice-water or boiling matter is used to maintain the reference environment. The experimental design for calibration must consider the following factors [1–3]. 1. The number and the location of the calibration points. 2. The regression equations (linear, poly-nominal, non-linear). 3. The regression techniques. 4. The standard references and their uncertainties. Betta [1] adopted minimizing the standard deviations for the regression curve coefficients or the standard deviation for the entire calibration curve to design an experiment to determine the number of calibration points, the number of repetitions, and the location of calibration points. Three types of sensor were used to demo the linear, quadratic and cubic calibration equations: a pressure transmitter, a platinum thermometer and E-Type thermocouple wires. The estimated confidence interval values were used to determine the validity of the regression equation. This method was extended to address calibration for complex measurement chains [2]. Sensors 2019, 19, 1213; doi:10.3390/s19051213 www.mdpi.com/journal/sensors 1
- 17. Sensors 2019, 19, 1213 Hajiyev [3] noted the importance of the selection of the calibration points to ensure the accuracy of the calibration and the optimal selection of standard pressure setters and used an example to verify the method. A dispersion matrix, → D of the estimated coefficients was defined and this matrix → D was used as a scale of the error between the sensor and the reference instruments. Two criteria were used to evaluate the performance. The minimized sum of the diagonal elements of the matrix → D is called the A-optimality criterion. The minimized of the generalized of determinant of the matrix → D is called the D-optimality criterion. The optimal measurement points for the calibration of the differential pressure gages were determined using the A-optimality criterion [3] and the D-optimality criterion [4]. Khan et al. [5] used an inverse modeling technique with a critical neural network (ANN) to evaluate the order of the models and the calibration points. The root-mean-square error (RMSE) was used as the criterion. Recently, modern regression has been used as an important role to express the quantitative relationship between independent and response variables for tests on a single regression coefficient [6–9]. This technique used to address calibration equations and the standard deviations of these calibration equations then served as the criteria to determine their accuracy [10,11]. The confidence band for the entire calibration curve or for each experimental point was used to evaluate the fit of calibration equations [1,2]. The concept of measurement uncertainty (MU) is widely used to represent the precision of calibration equations [12–14]. Statistical techniques can be used to evaluate the accuracy and precision of calibration equations that are obtained using different calibration points [15–17]. Humidity sensors that were calibrated using different saturated salt solutions were tested to illustrate the technique for the specification of optimal measurement points [18,19]. Humidity is very important for various industries. Many manufacturing and testing processes, such as those for food, chemicals, fuels and other products, require information about humidity [20]. Relative humidity (RH) is commonly used to express the humidity of moist air [21]. Electric hygrometers are the most commonly used sensors because they allow real-time measurement and are easily operated. The key performance factors for an electrical RH meter are the accuracy, the precision, hysteresis and long-term stability. At high air humidity measurement, there is a problem with response time of the RH sensors in conventional methods. The solution for this problem for high air humidity measurement is to use an open capacitor with very low response time [22–24] and quartz crystals which compensate temperature drift. An environment with a standard humidity is required for calibration. Fixed-point humidity systems that use a number of points with a fixed relative humidity are used as a standard. A humidity environment is maintained using different saturated salt solutions. The points with a fixed relative humidity are certified using various saturated salt solutions [19]. When the air temperature, water temperature and air humidity reach an equilibrium state, constant humidity is maintained in the air space [19]. The RH value that is maintained by the salt solutions is of interest. Wexler and Hasegawa measured the relative humidity that is created by eight saturated salt solutions using the dew point method [25]. Greenspan [18] compiled RH data for 28 saturated salt solutions. The relationship between relative humidity and ambient temperature was expressed as a 3rd or 4th polynomial equation. Young [26] collected RH data for saturated salt solutions between 0 to 80 ◦C and plotted the relationship between relative humidity and temperature. The Organisation Internationale De Metrologies Legale (OIML) [19] determined the effect of temperature on the relative humidity of 11 saturated salt solutions and tabulated the result. Standard conditions, devices and the procedure for using the saturated salt solutions were detailed. The range for the humidity measurement is from about 11% to 98% RH. Studies show that the number of fixed-point humidity references that are required for calibration is inconsistent. Lake et al. [27] used five salt solutions for calibration and found that the residuals for the linear calibration equation were distributed in a fixed pattern. Wadso [28] used four salt solutions to determine the RH that was generated in sorption balances. Duvernoy et al. [29] introduced seven salt 2
- 18. Sensors 2019, 19, 1213 solutions to generate the RH for a metrology laboratory. Bellhadj and Rouchou [30] recommended five salt solutions and two sulfuric acids to create the RH environment to calibrate a hygrometer. There is inconsistency in the salt solutions that are specified by instrumentation companies and standard bodies. The Japanese Mechanical Society (JMS) specifies 9 salt solutions for the standard humidity environment [31]. The Japanese Industrial Standards Committee (JISC) recommends 4 salt solutions to maintain RH environment [32]. The Centre for Microcomputer Applications (CMA) company specifies 11 salt solutions [33]. Delta OHM use only 3 salt solutions [34]. The OMEGA company use 9 salt solutions [35]. TA instruments specifies 9 salt solutions [36] and Vaisala B.V. select 4 salt solutions [37]. These salt solutions are listed in Table 1. Table 1. The selection of saturated salt solutions that are used to calibrate humidity sensors. Salt Solutions OIMI [19] Lake [27] Wadso [28] Duvernoy [29] Belhadj [30] JMS [31] JISC [32] CMA [33] Delta [34] OMEGA [35] TA [36] Vaisala [37] LiBr * LiCl * * * * * * * * * CH3COOK * * * * MgCl2·GH2O * * * * * * * * * * K2CO3 * * * * * * * Mg(NO3)2 * * * * * * * NaBr * * * * KI * * * SrCl2 * NaCl * * * * * * * * * * * * (NH4)2SO4 * KCl * * * * * * * * KNO3 * * * * K2SO4 * * * * * * * * Note: OIML, The Organisation Internationale De Metrologies Legale. Lu and Chen [17] calculated the uncertainty for humidity sensors that were calibrated using 10 saturated salt solutions for two types of humidity sensors. The study showed that a second-order polynomial calibration equation gave better performance than a linear equation. The measurement uncertainty is used as the criterion to determine the precision performance of sensors [38]. The number of standard relative humidity values for fixed-point humidity systems is limited by the number and type of salt solutions. The number of salt solutions that must be used to specify the calibration points for the calibration of RH sensors is a moot point. More salt solutions allow more calibration points for the calibration of RH sensors. However, using more salt solutions is time-consuming. This study determined the effect of the number and type of salt solutions on the calibration equations for two types of humidity sensors. The accuracy and precision were determined in order to verify the method for the choice of the optimal calibration points for sensor calibration. 2. Materials and Methods 2.1. Relative Humidity (RH) and Temperature Sensors Resistive sensor (Shinyei THT-B141 sensor, Shinyei Kaisha Technology, Kobe, Japan) and capacitive sensor (Vaisala HMP-143A sensor, Vaisala Oyj, Helsinki, Finland) were used in this study. The specification of the sensors is listed in Table 2. 3
- 19. Sensors 2019, 19, 1213 Table 2. The specifications of two humidity sensors. Resistive Sensor Capacitive Sensor Model 1 THT-B121 HMP 140A Sensing element Macro-molecule HPR-MQ HUMICAP Operating range 0–60 ◦C 0–50 ◦C Measuring range 10–99% RH 0–100% Nonlinear and repeatability ±0.25% RH ±0.2% RH ResolutionTemperature effect 0.1% RH (relative humidity)none 0.1% RH0.005%/◦C 2.2. Saturated Salt Solutions Eleven saturated salt solutions were used to maintain the relative humidity environment. These salt solutions are listed in Table 3. Table 3. The Calibration points for saturated salt solutions to establish the calibration equations. Salt Solutions (n1 = 11) Case 1 (n2 = 9) Case 2 (n3 = 7) Case 3 (n4 = 5) Case 4 uc LiCl * * * * 0.27 CH3COOK * 0.32 MgCl2 * * * * 0.16 K2CO3 * * * 0.39 Mg(NO3)2 * * 0.22 NaBr * * * * 0.40 KI * * 0.24 NaCl * * * * 0.12 KCl * * * 0.26 KNO3 * 0.55 K2SO4 * * * * 0.45 Note: uc values were obtained from Greenspan [18] and The Organisation Internationale De Metrologies Legale (OIML) R121 [19]. 2.3. Calibration of Sensors The humidity probes for the resistive and capacitive sensors were calibrated using saturated salt solutions. A hydrostatic solution was produced in accordance with OIML R121 [19]. The salt was dissolved in pure water in a ratio such that 40–75% of the weighted sample remained in the solid state. These salt solutions were stored in containers. The containers were placed in a temperature controller at an air temperature of 25 ± 0.2 ◦C. During the calibration process, humidity and temperature probes were placed within the container above the salt solutions. The preliminary study showed that an equilibrium state is established in 12 h so the calibration lasted 12 h to ensure that the humidity of the internal air had reached an equilibrium state. Experiments for each RH environment were repeated three times. The temperature was recorded and the standard humidity of the salt solutions was calculated using Greenspan’s equation [18]. 2.4. Establish and Validate the Calibration Equation The experimental design and flow chart for the data analysis is shown in Figure 1. The relationship between the standard humidity and the sensor reading values was established as the calibration equation. This study used the inverse method. The standard humidity is the dependent (yi) and the sensor reading values are the independent variables (xi) [17]. The form of the linear regression equation is: Y = b0 + b1 X (1) 4
- 20. Sensors 2019, 19, 1213 where b0 and b1 are constants. The form of the higher-order polynomial equation is: Y = c0 + c1X + c2X2 + c3X3 + . . . +ckXk (2) where c0, c1 to ck are constants. 5HVLVWLYHDQG FDSDFLWDQFHKXPLGLW VHQVRUV DOLEUDWLRQ UHSOLFDWHVRIHDFK VDOWVROXWLRQ VDWXUDWHG 6DOWVROXWLRQV 0RGHOLQJGDWD VDOWVROXWLRQV YVUHDGLQJYDOXHV 'LYLGLQJGDWD VDOWVROXWLRQV YV UHDGLQJYDOXHV VDOWVROXWLRQV YV UHDGLQJYDOXHV VDOWVROXWLRQV YV UHDGLQJYDOXHV VDOWVROXWLRQV YV UHDGLQJYDOXHV (VWDEOLVKLQJFDOLEUDWLRQHTXDWLRQV (TXDWLRQ (TXDWLRQ (TXDWLRQ (TXDWLRQ 9DOLGDWLQJGDWD VDOWVROXWLRQV YVUHDGLQJYDOXHV ULWHULD E(T ULWHULD E(T ULWHULD E(T ULWHULD E(T DOFXODWLQJ80YDOXHV 5HFRPPHQG Figure 1. The experimental design and flowchart of data analysis. 2.5. Different Calibration Points To model the calibration equations, the data for four different salt solutions was used, as listed in Table 3. Case 1: The data set is for 11 salt solutions and 11 calibration points Case 2: The data set is for 9 salt solutions and 9 calibration points Case 3: The data set is for 7 salt solutions and 7 calibration points Case 4: The data set is for 5 salt solutions and 5 calibration points For each sensor, four calibration equations were derived using four different calibration points. 2.6. Data Analysis The software, Sigma plot ver.12.2, was used to determine the parameters for the different orders of polynomial equations. 5
- 21. Sensors 2019, 19, 1213 2.6.1. Tests on a Single Regression Coefficient The criteria to assess the fit of the calibration equations are the coefficient of determination R2, the estimated standard error of regression s and the residual plots. The coefficient of determination, R2 is used to evaluate the fit of a calibration equation. However, no standard criterion has been specified [15,16]. The single parameter coefficient was tested using the t-test to evaluate the order of polynomial regression equation. The hypotheses are: H0 : bk = 0 (3) H1 : bk = 0 (4) The t-value is: t = bk/se(bk) (5) where bk is the value of the parameter for the polynomial regression equation of the highest order, and se(bk) is the standard error of bk. 2.6.2. The Estimated Standard Error of Regression The estimated standard error of regression s is calculated as follows: s = ( (ŷ2 − yi)2 n1 − p ) 0.5 (6) where ŷi is the predicted valued of the response, ŷi is the response, n1 is the number of data and p is the number of parameters. The s value is the criterion that is used to determine the accuracy of a calibration equations [38]. It is used to assess the accuracy of two types of RH sensors that are calibrated using different saturated salt solutions. 2.6.3. Residual Plots Residual plots is the quantitative criterion that is used to evaluate the fit of a regression equation. If the regression model is adequate, the data distribution for the residual plot should tend to a horizontal band and is centered at zero. If the regression equation is not accepted, the residual plots exhibit a clear pattern. For the calibration equation, tests on a single regression coefficient and the residual plots are used to determine the suitability of a calibration equation for RH sensors that are calibrated using different saturated salt solutions. The estimated standard error of the regression equations is then used to determine the accuracy of the calibration equations. 2.7. Measurement Uncertainty for Humidity Sensors The measurement uncertainty for RH sensors using different salt solutions was calculated using International Organization for Standardization, Guide to the Expression of Uncertainty in Measurement (ISO, GUM) [12,13,17]. uc 2 = u2 xpred + u2 temp + u2 non + u2 res + u2 sta (7) where uc is the combined standard uncertainty, uxpred is the uncertainty for the calibration equation, utemp is the uncertainty due to temperature variation, unon is the uncertainty due to nonlinearity, ures is the uncertainty due to resolution, and usta is the uncertainty of the reference standard for the saturated salt solution. The uncertainty of xpred is calculated as follows [38]: 6
- 22. Sensors 2019, 19, 1213 uxpred = s 1 + 1 n + (y − y)2 ∑(yi2) − (∑ yi)2 n (8) where y is the average value of the response. The uncertainty in the value of uref for the saturated salt solutions is determined using the reference standard for the salt solution. The scale and the uncertainty of these saturated salt solutions are listed in Table 3 that are taken from Greenspan [18] and the Organisation Internationale De Metrologies Legale (OIML) R121 [19]: uref = ( ∑(uri)2 N2 ) 0.5 (9) where uri is the uncertainty in the humidity for each saturated salt solution and N2 is the number of saturated salt solutions that are used for calibration. The calibration equations use different numbers of saturated salt solutions had its uncertainty. This criterion is used to evaluate the precision of RH sensors. The accuracy and precision of RH sensors that are calibrated using different saturated salt solutions was determined using the s and uc values. By Equations (7)–(9), the contrast between the number of saturated salt solutions is considered. The greater the number of data points that are used, the smaller is the s value that is calculated by Equation (6). However, this requires more experimental time and cost and the value of uref may be increased. The uncertainty of each calibration point is different because different saturated salt solutions are used. The optimal number of calibration points were evaluated by accuracy and precision. 3. Results and Discussion 3.1. The Effect of the Accuracy of Different Calibration Points 3.1.1. THT-B121 Resistive Humidity Sensor Calibration equations for resistive sensors using 11 salt solutions: The distribution of the relative humidity data for the reading values for a resistive sensor is plotted against the standard humidity values that are maintained using 11 saturated salt solutions in Figure 2. Ϭ ϭϬ ϮϬ ϯϬ ϰϬ ϱϬ ϲϬ ϳϬ ϴϬ ϵϬ ϭϬϬ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ZĞĂĚŝŶŐǀĂůƵĞƐ͕й ^ƚĂŶĚĂƌĚǀĂůƵĞƐ͕й Figure 2. The distribution of the relative humidity data for reading values versus the standard humidity values for THT-B121 resistive humidity sensor using 11 saturated salt solutions (LiCl, CH3COOK, MgCl2, K2CO3, Mg(NO3)2, NaBr, KI, NaCl, KCl, KNO3 and K2SO4). 7
- 23. Sensors 2019, 19, 1213 The estimated parameters and the evaluation criteria for regression analysis are listed in Table 4. The residual plots for the calibration equations for different orders of polynomial equations are shown in Figure 3. Table 4. Estimated parameters and evaluation criteria for the linear and several polynomial equations for THT-B121 resistive sensor using 11 salt solutions. Linear 2nd Order 3nd Order 4th Order b0 0.028672 −2.74999 −11.0702 −20.5303 b1 1.008985 1.13766 1.780025 2.805196 b2 −0.0011437 −0.01432 −0.0491534 b3 7.81681 × 10−5 5.39281 × 10−4 b4 −2.07539 × 10−6 R2 0.9967 0.9974 0.9987 0.9993 s 1.6098 1.4612 0.982 0.7719 Residual plots clear pattern clear pattern clear pattern uniform distribution (a) Linear equation (b) 2nd polynomial equation Ͳϰ Ͳϯ ͲϮ Ͳϭ Ϭ ϭ Ϯ ϯ ϰ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ZĞƐŝĚƵĂůƐ͕й WƌĞĚŝĐƚĞĚǀĂůƵĞƐ͕й Ͳϯ ͲϮ Ͳϭ Ϭ ϭ Ϯ ϯ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ZĞƐŝĚƵĂůƐ͕й WƌĞĚŝĐƚĞĚǀĂůƵĞƐ͕й Figure 3. Cont. 8
- 24. Sensors 2019, 19, 1213 (c) 3rd polynomial equation (d) 4th polynomial equation Ͳϯ ͲϮ Ͳϭ Ϭ ϭ Ϯ ϯ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ZĞƐŝĚƵĂůƐ͕й WƌĞĚŝĐƚĞĚǀĂůƵĞƐ͕й Ͳϯ ͲϮ Ͳϭ Ϭ ϭ Ϯ ϯ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ZĞƐŝĚƵĂůƐ͕й WƌĞĚŝĐƚĞĚǀĂůƵĞƐ͕й Figure 3. The residual plots for the calibration equations for different orders of polynomial equations for THT-B121 resistive humidity sensor using 11 saturated salt solutions (LiCl, CH3COOK, MgCl2, K2CO3, Mg(NO3)2, NaBr, KI, NaCl, KCl, KNO3 and K2SO4). The linear (Figure 3a), 2nd (Figure 3b) and 3rd (Figure 3c) order polynomial equations all exhibit a systematic distribution of residuals. These equations were not satisfactory for resistive sensors. The distribution of residual plots for the 4th order polynomial equations exhibit a uniform distribution (Figure 3d). The t-value for the highest-order parameter (b4 = −2.07539 × 10−6) was significantly different to zero, so the 4th order polynomial equation is the only adequate calibration equation. The equation is: y = −20.530298 + 2.805196x − 0.049153x2 + 0.000539x3 − 2.07539 × 10−6x4 (sb = 2.5004 sb = 0.2590 sb = 0.0082 sb = 0.00016 sb = 4.770 × 10−7 t = −8.2107 t = 11.181 t = −6.005 t = −5.0663 t = −4.3514) R2 = 0.992, s = 0.7719 The coefficient of determination, R2, for the linear, 2nd, 3rd and 4th order polynomial calibration equations are 0.9967, 0.9974, 0.9987 0.9993, respectively. High R2 values do not give useful information 9
- 25. Sensors 2019, 19, 1213 for the specification of an appropriate calibration equation. The estimated values of standard deviation, s, is used to define the uncertainty for an inverse calibration equation [35]. The s values for the four calibration equations are 1.6098, 1.4612, 0.9820 and 0.7719, respectively. It is seen that an appropriate calibration equation gives a significant reduction in uncertainty. Calibration equations for resistive sensor using 5 salt solutions: The estimated parameters and the evaluation criteria for the regression analysis for 5 calibration points for a resistive sensor are listed in Table 5. The residual plots for four calibration equations are shown in Supplementary Materials. Similarly to the regression results for 11 salt solutions, the linear, 2nd and 3rd order polynomial equations all employed a systematic distribution in the residuals plots. These equations are clearly not appropriate calibration equations. For a resistive sensor, the residual plots for the 4th order polynomial equations presented a random distribution. Table 5. Estimated parameters and evaluation criteria for the linear and several polynomial equations for THT-B121 resistive sensors using 5 salt solutions. Linear 2nd Order 3nd Order 4th Order b0 −0.970118 −3.1191770 −12.201481 −19.471802 b1 1.0155235 1.12632754 1.8869907 2.743833 b2 −0.001007316 −0.01685101 −0.04766345 b3 9.34623 × 10−5 5.15689 × 10−4 b4 −1.93676 × 10−6 R2 0.9969 0.9974 0.9994 0.9991 s 1.8109 1.7146 0.7984 1.084 Residual plots clear pattern clear pattern clear pattern uniform distribution The R2 values for the linear, 2nd, 3rd and 4th order polynomial calibration equations are 0.9969, 0.9974, 0.9994 and 0.9998, respectively. However, these higher R2 values do not provide relevant information about the calibration equations. The s values represent the uncertainty of calibration equations. For the linear, 2nd, 3rd and 4th order polynomial calibration equations are 1.8109, 1.7146, 0.7954 and 1.084, respectively. The 4th order polynomial equations is: y = −19.471802 + 2.743833x − 0.047663x2 + 0.0005157x3 − 1.93676 × 10−6x4 (sb = 2.2789 sb = 0.25086 sb = 0.00869 sb = 0.000117 sb = 5.360 × 10−7 t = −8.5447 t = 10.9396 t = −5.4849 t = 4.3946 t = −3.6101) R2 = 0.991, s = 1.014 The regression results for the 4th order polynomial equations using different calibration points in different salt solutions are listed in Table 6. The results for 9 and 7 calibration points are similar to those for 11 and 5 calibration points. Table 6. Estimated parameters and evaluation criteria for the 4th order polynomial equations for THT-B121 resistive sensors using four different calibration points. Case 1 (n1 = 11) Case 2 (n2 = 9) Case 3 (n3 = 7) Case 4 (n4 = 5) b0 −20.530297 −23.41845561 −23.904948 −19.4718019 b1 2.8051965 3.5861653 3.243023015 2.743832845 b2 −0.04915334 −0.06230766 −0.06426625 −0.047663446 b3 5.39281 × 10−4 7.0951 × 10−4 7.34202 × 10−4 5.15689 × 10−4 b4 −2.07539 × 10−6 −2.81734 × 10−6 −2.92042 × 10−6 −1.93676 × 10−6 R2 0.9993 0.9994 0.9994 0.9991 s 0.7719 0.6951 0.8039 1.084 10
- 26. Sensors 2019, 19, 1213 The R2 value is used b to evaluate the calibration equations [27,33]. Even the linear calibration equation for this study shows a high R2 value. However, the estimated error was higher than that for other equations. The residual plots all exhibited a clear pattern distribution so the R2 value cannot be used as the sole criterion to assess the calibration equation. Betta and Dell’Isola [1] mention R2, Chi-square and F-test to verify the accuracy of a model. This study used t-value for a parameter was used as the criterion. This method bases on statistical theory. 3.1.2. HMP 140A Capacitive Humidity Sensor Calibration equations for a capacitive sensors using 11 salt solutions The relationship between the reading values for a capacitive sensor and the standard humidity values that are maintained using 11 saturated salt solutions is shown in Figure 4. Ϭ ϭϬ ϮϬ ϯϬ ϰϬ ϱϬ ϲϬ ϳϬ ϴϬ ϵϬ ϭϬϬ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ZĞĂĚŝŶŐǀĂůƵĞƐ͕й ^ƚĂŶĚĂƌĚǀĂůƵĞƐ͕й Figure 4. The distributions of relative humidity data for standard humidity values versus the reading values for HMP 140A capacitance humidity sensors using 11 saturated salt solutions (LiCl, CH3COOK, MgCl2, K2CO3, Mg(NO3)2, NaBr, KI, NaCl, KCl, KNO3 and K2SO4). The estimated parameters and the evaluation criteria for regression analysis are listed in Table 7. Table 7. Estimated parameters and evaluation criteria for the linear and polynomial equations for HMP 140A capacitive sensor using 11 salt solutions. Linear 2nd Order b0 −0.414520 3.479518 b1 1.031003 0.833274 b2 0.00186718 R2 0.9975 0.9994 s 1.4002 0.6837 Residual plots clear pattern Uniform distribution The residual plots for the calibration equations for different orders of polynomial equations are shown in Figure 5. 11
- 27. Sensors 2019, 19, 1213 (a) linear equation (b) 2nd polynomial equation Ͳϯ ͲϮ Ͳϭ Ϭ ϭ Ϯ ϯ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ZĞƐŝĚƵĂůǀĂůƵĞƐ͕й WƌĞĚŝĐƚĞĚǀĂůƵĞƐ͕й Ͳϯ ͲϮ Ͳϭ Ϭ ϭ Ϯ ϯ Ϭ ϮϬ ϰϬ ϲϬ ϴϬ ϭϬϬ ϭϮϬ ZĞƐŝĚƵĂůǀĂůƵĞƐ͕й WƌĞĚŝĐƚĞĚǀĂůƵĞƐ͕й Figure 5. The residual plots for the calibration equations for different orders of polynomial equations for HMP 140A capacitance humidity sensor using 11 saturated salt solutions (LiCl, CH3COOK, MgCl2, K2CO3, Mg(NO3)2, NaBr, KI, NaCl, KCl, KNO3 and K2SO4). The linear equation (Figure 5a) exhibited a systematic distribution of residuals. The 2nd (Figure 5b) and 3rd (not presented) order polynomial equations both displayed a uniform distribution. The t-value for the 3rd order parameter was not significantly different to zero, so the 2nd order polynomial equation is the appropriate calibration equation and list as follows: y = 3.479518 + 0.833274x + 0.001867x2, R2 = 0.9994, s = 0.6837 (sb = 0.4805 sb = 0.02028 sb = 0.000187 t = 7.2408 t = 41.098 t = 10.004) The coefficient of determination, R2, for the linear and 2nd order polynomial calibration equations are 0.9975 and 0.9994, respectively. The s values for the two calibration equations are 1.4002 and 0.6837, respectively. An appropriate calibration equation gives a significant reduction in the estimated error. 12
- 28. Sensors 2019, 19, 1213 Calibration equations for a capacitive sensor using 5 salt solutions The estimated parameters and the evaluation criteria for the regression analysis for 5 calibration points for a capacitance are listed in Table 8. The residual plots for four calibration equations are shown in Supplementary Materials. Similarly to the regression results for 11 salt solutions, residuals plots for the linear equation exhibit a systematic distribution. Residual plots for the 2nd order polynomial equations presented a random distribution. Table 8. Estimated parameters and evaluation criteria for the linear and polynomial equations for HMP 140A capacitive sensor using 5 salt solutions. Linear 2nd Order b0 0.226512 2.911321 b1 1.023088 0.814217 b2 0.00155423 R2 0.9981 0.9995 s 1.4386 0.7890 Residual plots clear pattern Uniform distribution The R2 values for the linear and 2nd order polynomial calibration equations are 0.9981 and 0.9995, respectively. The s values for the linear and 2nd order polynomial calibration equations are 1.4386 and 0.7890, respectively. The 2nd order polynomial equations give the smallest estimated errors and listed as follows: y = 2.9113205 + 0.864217x + 0.0015542x2, R2 = 0.9995, s = 0.7890 (sb = 0.63806 sb = 0.02925 sb = 0.000278 t = 74.5628 t = 29.543 t = 5.5872) The regression results for the 2nd order polynomial equations using different calibration points in different salt solutions are listed in Table 9. The results of R2 values for 5, 7, 9 and 11 calibration points are similar. However, the calibration equation for 11 calibration points gives the smallest s value. Table 9. Estimated parameters and evaluation criteria for the 2nd order polynomial equations for HMP 140A capacitive sensors using four different calibration points. Case 1 (n1 = 11) Case 2 (n2 = 9) Case 3 (n3 = 7) Case 4 (n4 = 5) b0 3.479580 3.156891 2.871078 2.9113205 b1 0.833274 0.844157 0.862302 0.8142171 b2 0.00186718 0.00176878 0.00161775 0.00155423 R2 0.9975 0.9992 0.9994 0.9995 s 0.6837 0.7127 0.7490 0.7890 3.1.3. Evaluation of Accuracy The distribution between the number of saturated salt solutions and the estimated standard error for the calibration equations of two types of RH sensors is in Figure 6. For a resistance sensor, the s values of 7, 9, 11 calibration points are 0.8% RH. For a capacitance sensor, the s values for four saturated salt solutions are 0.8% RH. The accuracy of these calibration equations is 0.8% for both types of RH sensors. In terms a practical application [20,21], the calibration equation can be established using 7 salt solutions for a resistance sensor and 5 salt solutions for a capacitance sensor. 13
- 29. Sensors 2019, 19, 1213 Ϭ͘Ϯ Ϭ͘ϰ Ϭ͘ϲ Ϭ͘ϴ ϭ ϭ͘Ϯ ϰ ϱ ϲ ϳ ϴ ϵ ϭϬ ϭϭ ϭϮ ^ƚĂŶĚĂƌĚĚĞǀŝĂƚŝŽŶ͕Ɛ EŽ͘ŽĨƐĂůƚƐŽůƵƚŝŽŶƐ ZĞƐŝƐƚĂŶĐĞ ĂƉĂĐŝƚĂŶĐĞ Figure 6. The distribution between numbers of saturated salt solutions and estimated standard errors of calibration equations of two types of RH sensors. 3.2. The Effect of the Precision of Calibration Points 3.2.1. The Measurement Uncertainty for the Two Humidity Sensors The method that is used to calculate the measurement uncertainty is that of Lu and Chen [17]. Two Types “A” and “B” method are used to evaluate the measurement uncertainty. The Type A standard uncertainty is evaluated by statistical analysis of the experimental data. The Type B standard uncertainty is evaluated using other information that is related to the measurement. The Type A standard uncertainty for the two types of humidity sensors used the uncertainty for the predicted values from the calibration equations. The Type B standard uncertainty for humidity sensors uses the reference standard, nonlinear and repeatability, resolution and temperature effect. The results for the Type B uncertainty analysis for resistive and capacitive sensors are respectively listed in Tables 10 and 11. Table 10. The Type B uncertainty analysis for resistive humidity sensor. Description Estimate Value (%) Standard Uncertainty u(x), (%) Reference standard, Uref N1 = 11, uref = 0.3311 N1 = 9, uref = 0.2983 N1 = 7, uref = 0.3151 N1 = 5, uref = 0.3084 Non-linear and repeatability, Unon ±0.3 0.00866 Resolution, Ures 0.1 0.00290 The combined standard uncertainty of Type B = 0.1926 Table 11. The Type B uncertainty analysis for capacitive humidity sensor. Description Estimate Value (%) Standard Uncertainty u(x), (%) Reference standard, Uref N1 = 11, uref = 0.3311 N1 = 9, uref = 0.2983 N1 = 7, uref = 0.3151 N1 = 5, uref = 0.3084 Nonlinear and repeatability, Unon ±0.1 0.0058 Resolution, Ures ±0.1 0.0029 Temperature effect, Utemp ±0.005 0.0043 The combined standard uncertainty of Type B = 0.1924 The Type A standard uncertainty that are calculated using the predicted values for the 4th order polynomial equation for the resistive sensor and the 2nd order polynomial equation for a capacitive 14
- 30. Sensors 2019, 19, 1213 sensor are added to give a combined uncertainty using Equation (7). The combined uncertainty for three RH observations for the two humidity sensors using calibration equations that use different calibration points are in Figures 7 and 8. Ϭ͘ϱ Ϭ͘ϲ Ϭ͘ϳ Ϭ͘ϴ Ϭ͘ϵ ϭ ϭ͘ϭ ϭ͘Ϯ ϭ͘ϯ ϭ͘ϰ ϮϬ ϯϬ ϰϬ ϱϬ ϲϬ ϳϬ ϴϬ ϵϬ ϭϬϬ ŽŵďŝŶĞĚƵŶĐĞƌƚĂŝŶƚLJ͕й WƌĞĚŝĐƚĞĚŚƵŵŝĚŝƚLJ͕й Eϭсϭϭ EϮсϵ Eϯсϳ Eϰсϱ Figure 7. The distribution between numbers of saturated salt solutions and combined uncertainty of resistance RH sensors. Ϭ͘ϱ Ϭ͘ϲ Ϭ͘ϳ Ϭ͘ϴ Ϭ͘ϵ ϭ ϭ͘ϭ ϮϬ ϯϬ ϰϬ ϱϬ ϲϬ ϳϬ ϴϬ ϵϬ ϭϬϬ ŽŵďŝŶĞĚƵŶĐĞƌƚĂŝŶƚLJ͕й WƌĞĚŝĐƚĞĚŚƵŵŝĚŝƚLJ͕й Eϭсϭϭ EϮсϵ Eϯсϳ Eϰсϱ Figure 8. The distribution between numbers of saturated salt solutions and combined uncertainty of capacitance RH sensors. 3.2.2. The Precision of the Two Types of RH Sensors The combined uncertainty is the criterion that is used to determine the precision of the sensors. The values for the combined uncertainty for the resistive sensor at a RH of 30%, 60% and 90% are 0.8618%, 0.8506% and 0.8647% for the calibration equation that uses 11 calibration points, and 1.1155%, 1.1040% and 1.1271% for the calibration equation that uses 5 calibration points. The calibration equation that uses 9 calibration points gives the smallest uc values. The combined uncertainty for 7, 9 and 11 calibration points is 1.0% RH. The values for the combined uncertainty for a capacitive sensor at a RH of 30%, 60% and 90% are 0.7787%, 0.7690% and 0.7813% for the calibration equation that uses 11 calibration points and 0.8803%, 0.8717% and 0.8890% for the calibration equation that uses 5 calibration points. The combined 15
- 31. Sensors 2019, 19, 1213 uncertainty for 5, 7, 9 and 11 calibration points is 0.9% RH. In terms of practical applications, this performance is sufficient for industrial applications [20,21]. The accuracy and precision are 0.80% and 0.90% RH for a resistance RH sensor that uses 7 calibration points and 0.70% and 0.90% RH for a capacitance RH sensors that uses 5 calibration points. 3.3. Discussion The number of calibration points that are required for sensors represents a compromise between the ideal number of calibration points and the time and cost of the calibration. The criterion that Betta [1] used to determine the optimal number of points used the ratio of the standard deviation of the regression coefficients (sbj) to the established standard error of regression (s). Accuracy and precision are the most important criteria for sensors so this study uses both values. Using statistical theory, the best calibration equation is determined using the t-value for the highest-order parameter and the residual plots. The estimated standard errors for the regression equation are then used to determine the accuracy of the sensors. The combined uncertainty considered the uncertainty of reference materials, the uncertainty for the predicted values and other B type sources. The combined uncertainties for the calibration equations for different numbers of calibration points using different saturated salt solutions are the criteria that are used to evaluate the precision of sensors. Two types of electric RH sensors were calibrated in this study. Some calibration works, such as those for temperature and pressure sensors, are calibrated by an equal spacing of calibration points. The RH reference environments are maintained using different saturated salt solutions. It is seen that the optimum number of calibration points that is required to calibrate a resistive humidity sensors involves 7 saturated salt solutions (LiCl, MgCl2, K2CO3, NaBr, NaCl, KCI and K2SO4), so seven points are specified. Five saturated salt solutions (LiCl, MgCl2, NaBr, NaCl and K2SO4) are specified for a capacitive humidity sensor. Considering factors that influence the choice of salts, such as price, toxicity and rules for disposal, the choice of these salt solutions is suitable. The calibration equations key to measurement performance. This study determines that te 4th order polynomial equation is the adequate equation for the resistive humidity sensor and the 2nd order polynomial equation is the optimum equation for the capacitive humidity sensor. The accuracy of the calibration equations is 0.8% RH for a resistive humidity sensor that uses 7 calibration points and 0.7% RH for a capacitance humidity sensor that uses 5 calibration points. The precision is less than 1.0% RH for the resistive sensor and less than 0.9% RH for the capacitive sensor. The method that is used in this study applicable to other sensors. 4. Conclusions In this study, two types of electric RH sensors were used to illustrate the method for the specification of the optimum number of calibration points. The standard RH environments are maintained using different saturated salt solutions. The theory of regression analysis is applied. The best calibration equation is determined in terms of the t-value of the highest-order parameter and the residual plots. The estimated standard errors for the regression equation are the criteria that are used to determine the accuracy of sensors. The combined uncertainty involves the uncertainty for the reference materials, the uncertainty in the predicted values and other B type sources. The combined uncertainties for the calibration equations for different number of calibration points using different saturated salt solutions are the criteria that are used to evaluate the precision of the sensors. The calibration equations are key to good measurement performance. This study determines that the 4th order polynomial equation is the adequate equation for the resistive humidity sensor and the 2nd order polynomial equation is the best equation for the capacitive humidity sensor. The accuracy of the calibration equations is 0.8% RH for a resistive humidity sensor that uses 7 calibration points and 0.7% RH for a capacitance humidity sensor using 5 calibration points. The precision is less than 1.0% RH for the resistive sensor and less than 0.9% RH for the capacitive sensor. 16
- 32. Sensors 2019, 19, 1213 The method to determine the number of the calibration points used in this study is applicable to other sensors. Supplementary Materials: The following are available online at http://www.mdpi.com/1424-8220/19/5/1213/ s1. The residual plots for the calibration equations for different orders of polynomial equations for resistive humidity sensor using 5 saturated salt solutions (LiCl, MgCl2, NaBr, NaCl and K2SO4). The residual plots for the calibration equations for different orders of polynomial equations for capacitance humidity sensor using 5 saturated salt solutions (LiCl, MgCl2, NaBr, NaCl and K2SO4). Author Contributions: H.-Y.C. drafted the proposal, executed the statistical analysis, interpreted the results and revised the manuscript. C.C. reviewed the proposal, performed some experiments, interpreted some results and criticized the manuscript and participated in its revision. All authors have read and approved the final manuscript. Acknowledgments: The authors would like to thank the Ministry of Science and Technology of the Republic of China for financially supporting this research under Contract No. MOST -106-2313-B-005-006. Conflicts of Interest: The authors declare no conflict of interest. References 1. Betta, G.; Dell’Isola, M. Optimum choice of measurement points for sensor calibration. Measurement 1996, 17, 115–125. [CrossRef] 2. Betta, G.; Dell’Isola, M.; Frattolillo, A. Experimental design techniques for optimizing measurement chain calibration. Measurement 2001, 30, 115–127. [CrossRef] 3. Hajiyev, C. Determination of optimum measurement points via A-optimality criterion for the calibration of measurement apparatus. Measurement 2010, 43, 563–569. [CrossRef] 4. Hajiyev, C. Sensor Calibration Design Based on D-Optimality Criterion. Metrol. Meas. Syst. 2016, 23, 413–424. [CrossRef] 5. Khan, S.A.; Shabani, D.T.; Agarwala, A.K. Sensor calibration and compensation using artificial neural network. ISA Trans. 2003, 42, 337–352. [CrossRef] 6. Chen, C. Application of growth models to evaluate the microenvironmental conditions using tissue culture plantlets of Phalaenopsis Sogo Yukidian ‘V3’. Sci. Hortic. 2015, 191, 25–30. [CrossRef] 7. Chen, H.; Chen, C. Use of modern regression analysis in liver volume prediction equation. J. Med. Imaging Health Inform. 2017, 7, 338–349. [CrossRef] 8. Wang, C.; Chen, C. Use of modern regression analysis in plant tissue culture. Propag. Ornam. Plants 2017, 17, 83–94. 9. Chen, C. Relationship between water activity and moisture content in floral honey. Foods 2019, 8, 30. [CrossRef] 10. Chen, C. Evaluation of resistance-temperature calibration equations for NTC thermistors. Measurement 2009, 42, 1103–1111. [CrossRef] 11. Chen, A.; Chen, C. Evaluation of piecewise polynomial equations for two types of thermocouples. Sensors 2013, 13, 17084–17097. [CrossRef] [PubMed] 12. ISO/IEC 98–3. Uncertainty of Measurement—Part 3: Guide to the Expression of Uncertainty in Measurement; ISO: Geneva, Switzerland, 2010. 13. National Aeronautics and Space Administration. Measurement Uncertainty Analysis Principles and Methods, NASA Measurement Quality Assurance Handbook—Annex 3; National Aeronautics and Space Administration: Washington, DC, USA, 2010. 14. Chen, C. Evaluation of measurement uncertainty for thermometers with calibration equations. Accredit. Qual. Assur. 2006, 11, 75–82. [CrossRef] 15. Myers, R.H. Classical and Modern Regression with Applications, 2nd ed.; Duxbury Press: Pacific Grove, CA, USA, 1990. 16. Weisberg, S. Applied Linear Regression, 4th ed.; Wiley: New York, NY, USA, 2013. 17. Lu, H.; Chen, C. Uncertainty evaluation of humidity sensors calibrated by saturated salt solutions. Measurement 2007, 40, 591–599. [CrossRef] 18. Greenspan, L. Humidity fixed points of binary saturated aqueous solutions. J. Res. Natl. Bur. Stand. 1977, 81A, 89–96. [CrossRef] 17
- 33. Sensors 2019, 19, 1213 19. OMIL. The Scale of Relative Humidity of Air Certified Against Saturated Salt Solutions; OMIL R 121; Organization Internationale De Metrologie Legale: Paris, France, 1996. 20. Wernecke, R.; Wernecke, J. Industrial Moisture and Humidity Measurement: A Practical Guide; Wiley: Hoboken, NJ, USA, 2014. 21. Wiederhold, P.R. Water Vapor Measurement; Marcel Dekker, Inc.: New York, NY, USA, 1997. 22. Matko, V.; Đonlagić, D. Sensor for high-air-humidity measurement. IEEE Trans. Instrum. Meas. 1996, 4, 561–563. [CrossRef] 23. Matko, V. Next generation AT-cut quartz crystal sensing devices. Sensors 2011, 5, 4474–4482. [CrossRef] [PubMed] 24. Zheng, X.Y.; Fan, R.R.; Li, C.R.; Yang, X.Y.; Li, H.Z.; Lin, J.D.; Zhou, X.C.; Lv, R.X. A fast-response and highly linear humidity sensor based on quartz crystal microbalance. Sens. Actuator B Chem. 2019, 283, 659–665. [CrossRef] 25. Wexler, A.; Hasegawa, S. Relative humidity-temperature relationships of some saturated salt solutions in the temperature range 0◦ to 50◦ C. J. Res. Natl. Bur. Stand. 1954, 53, 19–26. [CrossRef] 26. Young, J. Humidity control in the laboratory using salt solutions—A review. J. Chem. Technol. Biotechnol. 1967, 17, 241–245. [CrossRef] 27. Lake, B.J.; Sonya, M.N.; Noor, S.M.; Freitag, H.P.; Michael, J.; McPhaden, M.J. Calibration Procedures and Instrumental Accuracy Estimates of ATLAS Air Temperature and Relative Humidity Measurements; NOAA Pacific Marine Environmental Laboratory: Seattle, WA, USA, 2003. 28. Wadsö, L.; Anderberg, A.; Åslund, I.; Söderman, O. An improved method to validate the relative humidity generation in sorption balances. Eur. J. Pharm. Biopharm. 2009, 72, 99–104. [CrossRef] 29. Duvernoy, J.; Gorman, J.; Groselj, D. A First Review of Calibration Devices Acceptable for Metrology Laboratory. 2015. Available online: https://www.wmo.int/pages/prog/www/IMOP/publications/IOM- 94-TECO2006/4_Duvernoy_France.pdf (accessed on 11 December 2018). 30. Belhadj, O.; Rouchon, V. How to Check/Calibrate Your Hygrometer? J. Paper Conserv. 2015, 16, 40–41. [CrossRef] 31. Japan Mechanical Society. The Measurement of Moisture and Humidity and Monitoring of Environment; Japan Mechanical Society: Tokyo, Japan, 2011. (In Japanese) 32. Japan Industrial Standard Committee. Testing Methods of Humidity; JIS Z8866; JISC: Tokyo, Japan, 1998. 33. Centre Microcomputer Application. Relative Humidity Sensor 025I. Available online: http://www.cma- science.nl/resources/en/sensors_bt/d025i.pdf (accessed on 2 December 2018). 34. Delta Ohm Company. Calibration Instructions of Relative Humidity Sensors. 2012. Available online: http://www.deltaohm.com/ver2012/download/Humiset_M_uk.pdf (accessed on 10 December 2018). 35. Omega Company. Equilibrium Relative Humidity Saturated Salt Solutions. 2013. Available online: https: //www.omega.com/temperature/z/pdf/z103.pdf (accessed on 11 December 2018). 36. TA Instruments. Humidity Fixed Points. 2016. Available online: http://www.tainstruments.com/pdf/ literature/TN056.pdf (accessed on 10 December 2018). 37. Vaisala Ltd. Vaisala Humidity Calibrator HMK 15 User’s Guide. 2017. Available online: www.vaisala.com/ sites/default/files/documents/HMK15_User_Guide_in_English.pdf (accessed on 11 December 2018). 38. Ellison, S.; Williams, A. Eurachem/CITAC Guide: Quantifying Uncertainty in Analytical Measurement, 3rd ed.; Eurachem: Torino, Italy, 2012. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). 18
- 34. sensors Article Humidity Sensors with Shielding Electrode Under Interdigitated Electrode Hong Liu, Qi Wang, Wenjie Sheng, Xubo Wang, Kaidi Zhang, Lin Du and Jia Zhou * ASIC and System State Key Lab, Department of Microelectronics, Fudan University, Shanghai 200433, China; 16210720074@fudan.edu.cn (H.L.); 18212020034@fudan.edu.cn (Q.W.); wsheng13@fudan.edu.cn (W.S.); xbwang16@fudan.edu.cn (X.W.); 15110720079@fudan.edu.cn (K.Z.); 17112020015@fudan.edu.cn (L.D.) * Correspondence: jia.zhou@fudan.edu.cn; Tel.: +86-13818066203 Received: 14 December 2018; Accepted: 31 January 2019; Published: 6 February 2019 Abstract: Recently, humidity sensors have been investigated extensively due to their broad applications in chip fabrication, health care, agriculture, amongst others. We propose a capacitive humidity sensor with a shielding electrode under the interdigitated electrode (SIDE) based on polyimide (PI). Thanks to the shielding electrode, this humidity sensor combines the high sensitivity of parallel plate capacitive sensors and the fast response of interdigitated electrode capacitive sensors. We use COMSOL Multiphysics to design and optimize the SIDE structure. The experimental data show very good agreement with the simulation. The sensitivity of the SIDE sensor is 0.0063% ± 0.0002% RH. Its response/recovery time is 20 s/22 s. The maximum capacitance drift under different relative humidity is 1.28% RH. Keywords: humidity sensor; capacitive; PI; SIDE; IDE 1. Introduction In addition to daily applications, such as air conditioners and humidifiers, humidity sensors are widely used in industrial process control, medical science, food production, agriculture, and meteorological monitoring [1–9]. In industry, the many manufacturing processes, such as semiconductor manufacturing and chemical gas purification, rely on precisely controlled humidity levels. In medical science, environmental humidity needs to be controlled during operations and pharmaceutical processing. In agriculture, humidity sensors are used for greenhouse air conditioning, plantation protection (dew prevention), soil moisture monitoring, and grain storage. Furthermore, in meteorological monitoring, weather bureaus and marine monitoring applications rely on accurate humidity sensing. For modern agriculture [10] and weather stations [11,12], accurate and fast measurement of humidity is becoming more and more important. Compared to existing infrared humidity sensors, electronic humidity sensors are cheaper, lighter, and smaller, which makes them more suitable for sensor networks to feed weather models. Nonetheless, high-precision fast-response sensors are important for many fields. For instance, fast and accurate humidity measurement are critical for eddy covariance systems [13]. Hence, electronic sensors have to become faster and more accurate. Electronic humidity sensors can be divided into resistive and capacitive [14]. Resistive humidity sensors tend to have higher gain and are usually cheaper to manufacture than capacitive humidity sensors. However, these sensors do not respond well when operating at low relative humidity (about 10% RH) because they exhibit very poor conductivity in low relative humidity environments, making it difficult to measure the output response [15]. In contrast, capacitive humidity sensors have better linearity, accuracy, and higher thermal stability than resistive humidity sensors [16–19]. A capacitive humidity sensor responds to changes of humidity by changes of the relative dielectric constant of the sensing layer, e.g., polymer film, upon water vapor absorption. Therefore, it is possible to directly Sensors 2019, 19, 659; doi:10.3390/s19030659 www.mdpi.com/journal/sensors 19
- 35. Sensors 2019, 19, 659 detect changes in capacitance to monitor changes in humidity. Unlike resistive humidity sensor, capacitive humidity sensors respond linearly with humidity, which simplifies the sensor readout. Various materials can be used as humidity sensing materials, such as electrolyte [20], ceramics [21,22], porous inorganic material [23–26], and polymers [27–30]. In particular, polymers have been used as sensing materials for capacitive humidity sensors owing to their good dielectric properties arising from their microporous structure and measurable physical property changes due to water absorption. PI is among the most commonly used moisture sensing material [31] for its good mechanical strength, electrochemical stability, and flexibility [32]. It remains stable after long time exposure to the measurement environment. Furthermore, PI is a microporous material with imide groups that strongly bond water molecules, which makes the material dielectric constant very sensitive to humidity. Therefore, we used PI in the proposed capacitive sensor. Capacitive humidity sensors have two basic structures: parallel plate (PP) capacitance (Figure 1a) and interdigital electrode (IDE) capacitance (Figure 1b). Figure 1. Structure diagram of parallel plate (PP) and interdigital electrode (IDE) sensors. (a) PP sensors composed of a solid substrate, two layers of parallel plate electrode, and a sensing material between them. (b) IDE sensors composed of an inert substrate, IDEs, and sensing material layer atop of the IDEs. A partial enlarged detail of IDE is shown on the right. In PP sensors, the upper plate is perforated by an array of holes or parallel stripes to allow water molecules from the air to reach the sensing material underneath. Since the sensing area of the PP capacitor is sandwiched between two parallel plates, the change in the relative dielectric constant of the sensing material in the PP sensors affects the overall capacitance change. Unlike PP sensors, IDE sensors usually only affect the change in the upper capacitance of the IDEs, which makes them less sensitive than PP sensors. However, the exposed sensing area of the PP sensors is smaller than for IDE sensors, which causes a slower response than for IDEs. 20
- 36. Sensors 2019, 19, 659 The IDEs are fabricated on an inert solid or flexible substrate as parallel comb electrodes that overlap each other [6,33]. IDE sensors are easier to fabricate than PP ones. The sensitive area of the IDEs is typically a few square millimeters, and the electrode gap is a few microns. The sensitivity of this type of sensor increases with decreasing pitch [34]. The electric field strength above the IDEs decreases exponentially away from the electrode surface, and becomes one-thirtieth, or even lower, of the surface value [35] after a few microns. Therefore, in the case where the gap between the IDEs is several microns, a sensing layer only a few microns thick is enough. Thanks to this layer being completely exposed to the measurement environment, the IDE sensors are faster. However, in the IDEs, only half of the electric field lines pass through the sensing layer, and the other half of the electric field lines pass through the underlying substrate. Therefore, the IDE sensors will have only half or less sensitivity (depending on the relative dielectric constant of the substrate) compared to an equivalent PP sensor [36]. It is clear that there are advantages and disadvantages of these two types of sensors. There has been a significant effort to improve the sensor structures. For example, Zhao et al. used RIE (Reactive Ion Etching) and ICP (Inductively Couple Plasma) to etch sensing materials between parallel plates of the sensors to obtain a larger contact area with the tested environment to reduce response time from 35 s to 25 s [37], but this was still slower compared to typical equivalent IDEs. Inspired by combining the advantages of PP and IDE structures, this paper proposes a novel IDE humidity sensor with a shielding electrode under the IDEs, namely, SIDE. On the SIDE, the capacitance of the lower half of the IDEs is shielded by an additional electrode underneath the IDEs, which effectively raises the relative capacitance change as it becomes exposed to moisture. Thus, a SIDE humidity sensor combines the high sensitivity of PP sensors and the fast response (20 s) as the IDE ones. In this work, we first verified the feasibility of the SIDE structure in the simulation software. Secondly, the thickness of the sensing layer with different electrode gaps and the dielectric thickness between the shielding electrode and the IDEs were optimized regarding the sensitivity and response speed. The SIDE sensor with optimized parameters was fabricated. The sensitivity, response time, recovery time, and stability of the sensor were measured. 2. Simulation of SIDE COMSOL Multiphysics®(Stockholm, Sweden) is applied to simulate the SIDE and IDE structure. Figure 2a shows the SIDE structure. The size of this sensor is 13 mm × 6 mm with a sensing area of 1.6 mm × 1 mm. The sensor consists of a 100 nm-thick shielding electrode, a 1 μm-thick silicon dioxide dielectric layer, a standard 100 nm IDE layer, and a PI film as the sensing layer. The finger length of the interdigitated electrode is 1 mm, with the width and the gap both being 5 μm. A total of 80 pairs of IDEs are used. A 5 μm-thick PI layer is utilized as the humidity sensing layer. Since the PI’s relative dielectric constant increases linearly with humidity [38], we simulate variations of humidity by directly changing the relative dielectric constant of the PI. An IDE model with the same structural parameters as the SIDE one is implemented with the only difference being the absence of the shielding electrode. Figure 2b shows the simulation results of the capacitance change rate (ΔC/C0) of SIDE and IDE under different relative dielectric constant of PI representing the humidity conditions. C0 is the total capacitance when the relative dielectric constant of the sensing layer is 2.9. ΔC is the capacitance difference between any other relative dielectric constant of PI and 2.9. It can be seen that under the same conditions, ΔC/C0 of the SIDE structure, is about 4 times bigger than that of the IDE structure, which implies that the SIDE will have much higher sensitivity than IDE with the same parameters. The effect of the thickness of the sensing film on ΔCmax/C0 is also simulated by COMSOL Multiphysics®(Stockholm, Sweden). We define that ΔCmax/C0 equals to ΔC/C0 with the relative dielectric constant of PI at 2.9 (C0) and 3.7 (Cmax), which indicates the sensitivity of the sensor. 21
- 37. Sensors 2019, 19, 659 Figure 2. SIDE structure and simulation results. (a) 3D model of SIDE structure; (b) Comparison of the relative changes in capacitance of the SIDE (red line) and IDE (black line) structure according to numerical simulations. Figure 3 shows that ΔCmax/C0 increases as the thickness of the sensing film increases, but flattens at higher thickness. To optimize the sensing film thickness, two facts should be taken into account. On the one hand, it is clear that when the sensing film thickness is equal to the gap between the IDEs (as those dashed lines in Figure 3), ΔCmax/C0 almost reaches saturated values. There is no significant increase of ΔCmax/C0 with thicker sensing film than the gap. On the other hand, the thickness of the sensing film also affects the speed of water molecules diffusing into the sensing film completely, which defines the sensor response and recovery time. Therefore, we select the optimized sensing film thickness as equal to the gap of the IDEs. Considering the laboratory conditions, we set the width and gap of the IDEs to 5 μm. Figure 3. Influence of sensing film’s thickness on sensor sensitivity. The vertical ordinate of the intersection of all the dashed lines and the solid curves represents the sensor’s ΔCmax/C0 when the sensing film thickness is equal to the gap between the IDEs. The effect of the spacing between the shielding electrode and the IDEs, i.e., the thickness of the silicon dioxide under the IDEs on the sensitivity in the SIDE structure is also studied. Figure 4 shows that with the increasing thickness of the silicon dioxide layer, the ΔCmax/C0 increases first and then decreases, with an optimal value of the SiO2 thickness of 1 μm. 22
- 38. Sensors 2019, 19, 659 Figure 4. Influence of silicon dioxide thickness on the sensor sensitivity. For increasing silicon dioxide layer thickness, the full sensitivity increases first and then decreases past an optimal value. There are several parameters of the optimized SIDE structure through the simulation: the gap of IDEs and spin-coated sensing film thickness are both 5 μm, and the thickness of the silicon dioxide layer is 1 μm. These parameters are used in the fabrication of the sensor. 3. Materials and Methods The sensor is fabricated on a 3-inch silicon wafer according to the following steps: (a) A 2.5 μm-thick negative photoresist is patterned. (b) An e-beam-evaporated Ti/Au layer is deposited and selectively removed by a lift-off process to form the bottom shielding electrode. (c) A layer of 1 μm silicon dioxide is deposited by PECVD (Plasma Enhanced Chemical Vapor Deposition). (d) IDEs are fabricated on the silicon dioxide by the same sequence of lithography, e-beam evaporation, and lift-off. (e) A 5 μm-thick PI is spin-coated. Subsequently, the device is baked at 120 ◦C for 1 h, 180 ◦C for 1 h, and 250 ◦C for 6 h to cure the sensing layer. The completed sensor and cross-section of the SIDE structure under scanning electron microscope (SEM) are shown in Figure 5. The same IDE structure fabricated on the glass substrate without the shielding electrode is studied as the control experiment. Figure 5. SIDE sensor picture under microscopy, and its cross-section image under SEM. The setup for the humidity measurement is shown in Figure 6. The test is always carried out in an incubator. We build the simple incubator with heaters and semiconductor coolers inside. Each of them is controlled by an external PID (proportional integral derivative) controller to keep the temperature constant. In the incubator, we place a bottle of saturated salt solution and the sensor. The humidity is also monitored by a commercial humidity meter (Rotronic, HC2-S) at the same time and in the same 23
- 39. Sensors 2019, 19, 659 incubator. The uncertainty of HC2-S is ±0.8% RH. The capacitance measurement uses an IC chip (SMARTEC’s UTI03) and additional circuits. The commercial humidity sensor and the capacitance measurement circuit communicate with the computer using serial port simultaneously. The humidity and capacitance are recorded in parallel by the computer for later analysis. Figure 6. Block diagram of the measurement system consisting of an incubator, a measurement circuit and recording software. The capacitance above the shielding electrode Cx can be directly measured using the circuit shown in Figure 7 without mixing the capacitance between the shielding electrode and IDEs Cpn (n = 1, 2). Cx is the sensing capacitance proportional to the humidity. Cp1 and Cp2 are the capacitances between the shielding electrode and the IDEs. Cf is the fixed capacitance of the IC chip. U1 and U2 are the potentials before the humidity sensor and after the IC chip that both can be measured. Therefore, Cx can be calculated using Equation (1). Cx = −U1/U2·Cf (1) Figure 7. The working principle of the humidity capacitance measurement. The key point is to calculate the capacitance of Cx by measuring the induced charge generated at point B. Before the test, each device is placed in an oven at 100 ◦C for 10 min to get rid of the effect of the previous measurement. The sensitivity (S) can be expressed as Equation (2): S = (ΔC/C0)/Δ(% RH) (2) where ΔC = C1 − C0, C0 is the capacitance measured at the RH, which is 23.7% ± 0.8%, and C1 is the capacitance measured when the RH is 73.0% ± 0.8%. Δ(% RH) is the difference between the relative humidity values when measuring C1 and C0. 24
- 40. Sensors 2019, 19, 659 The response and recovery dynamics are among the most important characteristics for evaluating the performance of humidity sensors. The response time for RH increase and the recovery time for RH decrease are usually defined for a sensor as the time taken to reach 90% of its total capacitance variation. The response and recovery curves are measured by exposing the SIDE sensor to alternate levels of humidity between 2.0% ± 0.8% and 77.0% ± 0.8% RH. In order to evaluate the functioning of the humidity sensor over long periods of time, we measured the sensor’s capacitance over the duration of 20 h at 25 ◦C with relative humidity levels of 25.7% ± 0.8%, 34.4% ± 0.8%, 45.0% ± 0.8%, 57.0% ± 0.8%, and 73.5% ± 0.8% RH. 4. Results and Discussion A sensitivity test is carried out on the SIDE and IDE structure. Figure 8 shows the capacitance measured from SIDE and IDE at different levels of humidity, and their linear fits with R2 of 0.996 and 0.991, respectively. The slopes of the line, i.e., S of SIDE and IDE are 0.0063 and 0.001,65, respectively. Taking the uncertainty of HC2-S into consideration, the S of SIDE and IDE are 0.0063 ± 0.0002 and 0.001,65 ± 0.000,05, respectively. Hence, the sensitivity of the SIDE structure is 3.82 times bigger than that of the IDE. These results show the significant improvement of sensitivity brought by the shielding electrode, that minimizes the large constant capacitance of the substrate. Indeed, whatever substrate the IDE is built on, the relative dielectric constant of the substrate is larger (e.g., Si is 11.9, glass is 10) or close to (e.g., flexible polymer films) the relative dielectric constant of PI (2.9–3.7). The experimental result and simulation data verify the effects of the shielding electrode and shows high agreement as well. It is clear that our proposed SIDE structure can provide an effective way to measure relative humidity more sensitively and accurately. Another advantage of the shielding electrode is that it can effectively suppress the external electromagnetic interference and reduce the noise in the measurement process. Figure 8. Experimental measurement of sensitivity of SIDE and IDE humidity sensors. Figure 9 shows the responses of the SIDE sensor. The absorption curve represents the response of the sensor as a function of time, from an environment with low relative humidity to an environment with high relative humidity. The desorption curve represents the response of the sensor as a function of time, from an environment with high relative humidity to an environment with low relative humidity. The curve can switch to steady states rapidly after the RH level changes. Our sensor’s response/recovery time is 20 s/22 s, which is comparable to 1 s/15 s for normal IDE reported in the literature [39], but a little worse. This is because in their work, the thickness of the sensing 25
- 41. Sensors 2019, 19, 659 film is only 0.65 μm, while ours is 5 μm. If we scale down our sensors to reduce the IDE gap, the required sensing film thickness will also decrease, resulting in great improvement in response speed. Limited to laboratory conditions, we fabricated the sensor with 5 μm gap. However, our sensor’s response/recovery time is still much better than 122 s for PP sensors [40]. Figure 9. The response and recovery curves are measured by switching the SIDE sensor, alternately, between 2.0% ± 0.8% and 77.0% ± 0.8% RH. The response/recovery time is 20 s/22 s. Figure 10 shows the stability characteristic of the SIDE sensor. The sensor is kept in the incubator for 20 h at 25.7% ± 0.8%, 34.4% ± 0.8%, 45.0% ± 0.8%, 57.0% ± 0.8%, and 73.5% ± 0.8% RH, respectively. The magnitude of the drift of sensor capacitance is converted into the apparent changes in relative humidity, D, which is calculated by D = (Cmax − Cmean)/(C0·S) (3) where Cmax is the maximum measured capacitance after the sensor is exposed to different RH atmosphere, and Cmean is the average capacitance of all recorded values at a certain relative humidity, C0 is the capacitance measured when the RH is 23.7% ± 0.8%. The maximum drift value (D) obtained from Figure 10 under different relative humidity was 1.28% RH. Thus, our sensor is able to achieve satisfactory stability from a practical standpoint, which makes it promising as a commercially available sensor. Figure 10. Stability of SIDE sensor. The sensor is kept in the incubator for 1200 min at 25.7% ± 0.8%, 34.4% ± 0.8%, 45.0% ± 0.8%, 57.0% ± 0.8%, and 73.5% ± 0.8% RH, respectively. 26
- 42. Sensors 2019, 19, 659 5. Conclusions In summary, we propose a novel shielded interdigitated electrode structure for humidity sensing. We perform a comprehensive simulation of this structure to optimize the parameters for the sensor fabrication. In simulation and actual testing, we find that the sensitivity of the SIDE structure is much higher than that of the IDE structure because of the effect of the shielding electrode on the capacitance change rate. Since the surface structure of the SIDE structure is still the same as IDE, the SIDE sensor combines the high sensitivity of the parallel plate sensors and fast response of the IDE sensors. The sensitivity of SIDE is 0.0063% ± 0.0002% RH, and the response/recovery time is 20 s/22 s. The stability of the SIDE sensor was also characterized. The maximum drift value under different relative humidity is 1.28% RH. Meanwhile, since the basic operating principle of many capacitive sensors is the same, the SIDE structure can even be applied to capacitive gas sensors, such as volatile organic compound (VOC) sensors which are used to monitor toxic gases. This shows that SIDE can replace IDE in various sensors that are more sensitive to the accuracy and response speed. Author Contributions: Conceptualization, J.Z. and H.L.; methodology, H.L.; software, K.Z.; validation, Q.W., W.S. and L.D.; formal analysis, Q.W.; investigation, H.L.; resources, J.Z.; data curation, H.L.; writing—original draft preparation, H.L.; writing—review and editing, J.Z.; visualization, X.W.; supervision, J.Z.; project administration, J.Z.; funding acquisition, J.Z. Acknowledgments: This work was supported by the National Natural Science Foundation of China (Grant No. 61874033), Science Foundation of Shanghai Municipal Government (Grant No.18ZR1402600) and the State Key Lab of ASIC and System, Fudan University with Grant No.2018MS003. Conflicts of Interest: The authors declare no conflict of interest. References 1. Tételin, A.; Pellet, C.; Laville, C.; N’Kaoua, G. Fast response humidity sensors for a medical microsystem. Sensors Actuators B Chem. 2003, 91, 211–218. [CrossRef] 2. Chen, Z.; Lu, C. Humidity Sensors: A Review of Materials and Mechanisms. Sens. Lett. 2005, 3, 274–295. [CrossRef] 3. Lee, C.W.; Lee, S.J.; Kim, M.; Kyung, Y.; Eom, K. Capacitive Humidity Sensor Tag Smart Refrigerator System using the Capacitive to Voltage Converter (CVC). Int. J. Adv. Sci. Technol. 2011, 36, 15–26. 4. Kolpakov, S.A.; Gordon, N.T.; Mou, C.; Zhou, K. Toward a new generation of photonic humidity sensors. Sensors 2014, 14, 3986–4013. [CrossRef] [PubMed] 5. Farahani, H.; Wagiran, R.; Hamidon, M.N. Humidity sensors principle, mechanism, and fabrication technologies: A comprehensive review. Sensors 2014, 14, 7881–7939. [CrossRef] [PubMed] 6. Pavinatto, F.J.; Paschoal, C.W.A.; Arias, A.C. Printed and flexible biosensor for antioxidants using interdigitated ink-jetted electrodes and gravure-deposited active layer. Biosens. Bioelectron. 2015, 67, 553–559. [CrossRef] [PubMed] 7. Lee, C.-Y.; Lee, G.-B. Humidity Sensors: A Review. Sens. Lett. 2005, 3, 1–15. [CrossRef] 8. Rittersma, Z.M. Recent achievements in miniaturised humidity sensors—A review of transduction techniques. Sensors Actuators A Phys. 2002, 96, 196–210. [CrossRef] 9. Willett, K.M.; Gillett, N.P.; Jones, P.D.; Thorne, P.W. Attribution of observed surface humidity changes to human influence. Nature 2007, 449, 710–712. [CrossRef] [PubMed] 10. Imam, S.A.; Choudhary, A.; Sachan, V.K. Design issues for wireless sensor networks and smart humidity sensors for precision agriculture: A review. In Proceedings of the 2015 International Conference on Soft Computing Techniques and Implementations (ICSCTI), Faridabad, India, 8–10 October 2015; pp. 181–187. 11. Chandana, L.S.; Sekhar, A.J.R. Weather Monitoring Using Wireless Sensor Networks based on IOT. Int. J. Sci. Res. Sci. Technol. 2018, 4, 525–531. 12. Yawut, C.; Kilaso, S. A Wireless Sensor Network for Weather and Disaster Alarm Systems. Int. Conf. Inf. Electron. Eng. 2011, 6, 155–159. 13. Baldocchi, D.D. Assessing the eddy covariance technique for evaluating carbon dioxide exchange rates of ecosystems: Past, present and future. Glob. Chang. Biol. 2003, 9, 479–492. [CrossRef] 27
- 43. Sensors 2019, 19, 659 14. Fenner, R.; Zdankiewicz, E. Micromachined Water Vapor Sensors: A Review of Sensing Technologies. IEEE Sens. J. 2001, 1, 309–317. [CrossRef] 15. Blank, T.A.; Eksperiandova, L.P.; Belikov, K.N. Recent trends of ceramic humidity sensors development: A review. Sensors Actuators B Chem. 2016, 228, 416–442. [CrossRef] 16. Dokmeci, M.; Najafi, K. A high-sensitivity polyimide capacitive relative humidity sensor for monitoring anodically bonded hermetic micropackages. J. Microelectromech. Syst. 2001, 10, 197–204. [CrossRef] 17. Gu, L.; Huang, Q.A.; Qin, M. A novel capacitive-type humidity sensor using CMOS fabrication technology. Sensors Actuators B Chem. 2004, 99, 491–498. [CrossRef] 18. Wagner, T.; Krotzky, S.; Weiß, A.; Sauerwald, T.; Kohl, C.D.; Roggenbuck, J.; Tiemann, M. A high temperature capacitive humidity sensor based on mesoporous silica. Sensors 2011, 11, 3135–3144. [CrossRef] 19. Lee, H.; Lee, S.; Jung, S.; Lee, J. Nano-grass polyimide-based humidity sensors. Sensors Actuators B Chem. 2011, 154, 2–8. [CrossRef] 20. Yang, M.-R.; Chen, K.-S. Humidity sensors using polyvinyl alcohol mixed with electrolytes. Sensors Actuators B Chem. 1998, 49, 240–247. [CrossRef] 21. Kim, Y.; Jung, B.; Lee, H.; Kim, H.; Lee, K.; Park, H. Capacitive humidity sensor design based on anodic aluminum oxide. Sensors Actuators B Chem. 2009, 141, 441–446. [CrossRef] 22. Feng, Z.S.; Chen, X.J.; Chen, J.J.; Hu, J. A novel humidity sensor based on alumina nanowire films. J. Phys. D Appl. Phys. 2012, 45, 225305. [CrossRef] 23. Tudorache, F.; Petrila, I. Effects of partial replacement of iron with tungsten on microstructure, electrical, magnetic and humidity properties of copper-zinc ferrite material. J. Electron. Mater. 2014, 43, 3522–3526. [CrossRef] 24. Tudorache, F.; Petrila, I.; Popa, K.; Catargiu, A.M. Electrical properties and humidity sensor characteristics of lead hydroxyapatite material. Appl. Surf. Sci. 2014, 303, 175–179. [CrossRef] 25. Tudorache, F.; Petrila, I.; Condurache-Bota, S.; Constantinescu, C.; Praisler, M. Humidity sensors applicative characteristics of granularized and porous Bi2O3 thin films prepared by oxygen plasma-assisted pulsed laser deposition. Superlattices Microstruct. 2015, 77, 276–285. [CrossRef] 26. Tudorache, F.; Petrila, I.; Slatineanu, T.; Dumitrescu, A.M.; Iordan, A.R.; Dobromir, M.; Palamaru, M.N. Humidity sensor characteristics and electrical properties of Ni–Zn–Dy ferrite material prepared using different chelating-fuel agents. J. Mater. Sci. Mater. Electron. 2016, 27, 272–278. [CrossRef] 27. Suzuki, T.; Tanner, P.; Thiel, D.V. O2 plasma treated polyimide-based humidity sensors. Analyst 2002, 127, 1342–1346. [CrossRef] [PubMed] 28. Zampetti, E.; Pantalei, S.; Pecora, A.; Valletta, A.; Maiolo, L.; Minotti, A.; Macagnano, A.; Fortunato, G.; Bearzotti, A. Design and optimization of an ultra thin flexible capacitive humidity sensor. Sensors Actuators B Chem. 2009, 143, 302–307. [CrossRef] 29. Kim, J.H.; Hong, S.M.; Moon, B.M.; Kim, K. High-performance capacitive humidity sensor with novel electrode and polyimide layer based on MEMS technology. Microsyst. Technol. 2010, 16, 2017–2021. [CrossRef] 30. Liu, M.Q.; Wang, C.; Kim, N.Y. High-sensitivity and low-hysteresis porous mim-type capacitive humidity sensor using functional polymer mixed with TiO2 microparticles. Sensors 2017, 17, 284. [CrossRef] [PubMed] 31. Wang, H.; Feng, C.-D.; Sun, S.-L.; Segre, C.U.; Stetter, J.R. Comparison of conductometric humidity-sensing polymers. Sens. Actuators B Chem. 1997, 40, 211–216. [CrossRef] 32. Fujita, S.; Kamei, Y. Electrical properties of polyimide with water absorption. In Proceedings of the 11th IEEE International Symposium on Electrets, Melbourne, VIC, Australia, 1–3 October 2002; pp. 275–278. 33. Olthuis, W.; Sprenkels, A.J.; Bomer, J.G.; Bergveld, P. Planar interdigitated electrolyte-conductivity sensors on an insulating substrate covered with Ta2O5. Sensors Actuators B Chem. 1997, 43, 211–216. [CrossRef] 34. Singh, K.V.; Bhura, D.K.; Nandamuri, G.; Whited, A.M.; Evans, D.; King, J.; Solanki, R. Nanoparticle-enhanced sensitivity of a nanogap-interdigitated electrode array impedimetric biosensor. Langmuir 2011, 27, 13931–13939. [CrossRef] 35. Schaur, S.; Jakoby, B. A numerically efficient method of modeling interdigitated electrodes for capacitive film sensing. Procedia Eng. 2011, 25, 431–434. [CrossRef] 36. Blue, R.; Uttamchandani, D. Chemicapacitors as a versatile platform for miniature gas and vapor sensors. Meas. Sci. Technol. 2017, 28, 22001–22024. [CrossRef] 28
- 44. Sensors 2019, 19, 659 37. Qiang, T.; Wang, C.; Liu, M.Q.; Adhikari, K.K.; Liang, J.G.; Wang, L.; Li, Y.; Wu, Y.M.; Yang, G.H.; Meng, F.Y.; et al. High-Performance porous MIM-type capacitive humidity sensor realized via inductive coupled plasma and reactive-Ion etching. Sensors Actuators B Chem. 2018, 258, 704–714. [CrossRef] 38. Schubert, P.J.; Nevin, J.H. A polyimide-based capacitive humidity sensor. IEEE Trans. Electron Devices 1985, 32, 1220–1223. [CrossRef] 39. Laville, C.; Delétage, J.Y.; Pellet, C. Humidity sensors for a pulmonary function diagnostic microsystem. Sensors Actuators B Chem. 2001, 76, 304–309. [CrossRef] 40. Kim, J.H.; Hong, S.M.; Lee, J.S.; Moon, B.M.; Kim, K. High sensitivity capacitive humidity sensor with a novel polyimide design fabricated by mems technology. In Proceedings of the 4th IEEE International Conference on Nano/Micro Engineered and Molecular Systems, NEMS 2009, Shenzhen, China, 5–8 January 2009; pp. 703–706. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). 29
- 45. sensors Article A Fast Response−Recovery 3D Graphene Foam Humidity Sensor for User Interaction Yu Yu 1,2, Yating Zhang 1,2,*, Lufan Jin 1,2, Zhiliang Chen 1,2, Yifan Li 1,2, Qingyan Li 1,2, Mingxuan Cao 1,2, Yongli Che 1,2, Junbo Yang 3 and Jianquan Yao 1,2 1 Department of Electrical and Electronic Engineering, South University of Science and Technology of China, Shenzhen 518055, China; yuyu1990@tju.edu.cn (Y.Y.); jlfking@tju.edu.cn (L.J.); chenzl@tju.edu.cn (Z.C.); yifanli@tju.edu.cn (Y.L.); liqingyan216@163.com (Q.L.); mingxuancao@tju.edu.cn (M.C.); cheyongli@tju.edu.cn (Y.C.); jqyao@tju.edu.cn (J.Y.) 2 Key Laboratory of Opto-Electronics Information Technology, Ministry of Education, School of Precision Instruments and Opto-Electronics Engineering, Tianjin University, Tianjin 300072, China 3 Center of Material Science, National University of Defense Technology, Changsha 410073, China; yangjunbo008@sohu.com * Correspondence: yating@tju.edu.cn Received: 9 October 2018; Accepted: 3 December 2018; Published: 8 December 2018 Abstract: Humidity sensors allow electronic devices to convert the water content in the environment into electronical signals by utilizing material properties and transduction techniques. Three-dimensional graphene foam (3DGF) can be exploited in humidity sensors due to its convenient features including low-mass density, large specific surface area, and excellent electrical. In this paper, 3DGF with super permeability to water enables humidity sensors to exhibit a broad relative humidities (RH) range, from 0% to 85.9%, with a fast response speed (response time: ~89 ms, recovery time: ~189 ms). To interpret the physical mechanism behind this, we constructed a 3DGF model decorated with water to calculate the energy structure and we carried out the CASTEP as implemented in Materials Studio 8.0. This can be ascribed to the donor effect, namely, the electronic donation of chemically adsorbed water molecules to the 3DGF surface. Furthermore, this device can be used for user interaction (UI) with unprecedented performance. These high performances support 3DGF as a promising material for humidity sensitive material. Keywords: three-dimensional graphene foams; humidity sensor; fast response; user interaction 1. Introduction Humidity sensors have aroused attention in many fields such as industry, agriculture, and environment [1,2], and medical devices [3]. Generally, they measure humidity through a variety of transduction techniques, including the use of resistive [4,5], capacitive [6], optical fiber [7], and field effect transistors [8,9]. There are also some high precision impedance-frequency transducers using quartz crystals which compensate temperature drift, and have fast response, as investigated by in Matko et al. [10]. In high air humidity measurement there is a problem with response time of the sensors in conventional methods. A solution for this problem is sensors for high air humidity measurement which use open capacitors with very low response time such as is described by Vojko et al. [11]. As an active material for absorbing water molecules, a series of sensing materials including polymers [12], metal oxides [8], carbon nanotubes [13,14], graphene dioxide [15,16], and composites [5,17] have been exploited in humidity sensors. For instance, Zhang et al. [12] described humidity sensors utilizing poly(N-vinyl-2-pyrrolidone) (PVP), poly(vinyl alcohol) (PVA), and hydroxyethyl cellulose (HEC). In particular, after using PVP, the humidity sensors exhibited response and recovery times between 11% and 95% relative humidity (RH) were about 37 s and 10 s, Sensors 2018, 18, 4337; doi:10.3390/s18124337 www.mdpi.com/journal/sensors 30
- 46. Sensors 2018, 18, 4337 respectively. Wang et al. [8] applied a single SnO2 nanowire (NW) to fabricate a humidity sensor, which exhibited a wide sensor RH range (5~85%), and the response and recovery times were 120~170 s and 20~60 s, respectively. Zhao et al. [15] investigated a humidity sensor based on multi-wall carbon nanotubes where the sensor testing range was about 11% to 97% RH, the response time was 45 s, and the recovery time was 15 s. Borini et al. [15] exploited graphene oxide in a humidity sensor and obtained an unprecedented response speed (~30 ms response and recovery times) in a range of 30% to 70% RH. Zhang et al. [5] utilized a graphene oxide (GO)/poly(diallyldimethylammonium chloride) (PDDA) nanocomposite film to fabricate a humidity sensor. The humidity sensor exhibited ultrahigh performance over a wide range of 11~97% RH, and the recovery time is 125 s at 11% RH. Thus, each sensing material has its own advantages and specific conditions of application. In addition, with large surface area to volume ratio, nanomaterials are attractive to fabricate humidity sensors with ultrahigh performance features including high sensitivity and fast response times. Recently, graphene with three dimensional (3D) architectures, including foams, networks, and gels have been investigated [18–21]. These 3D graphene-based materials not only have the characteristics of graphene, but also have high specific surface area, low density, good mechanical strength and good conductivity [22]. Because of its wide accessibility, easy synthesis and solution processability, high chemical stability and strong adaptability [23,24], 3D graphene foam (3DGF) has attracted great interest in various sensing applications. Meanwhile, 3DGFs are efficient materials for biosensors and gas-sensing devices given their low-mass density, large surface area, good mechanical stability, and high electrical conductivity. Huang and coworkers [25] synthesised 3DGF/CuO nanoflower composites as single-chip independent 3D biosensors for the electrochemical detection of ascorbic acid with outstanding biosensing properties, such as an ultrahigh sensitivity of 2.06 mA mM−1 cm−2 to ascorbic acid at a 3 s response time. Besides that, Yavari et al. [24] used macroscopic 3DGF to fabricate gas detectors with high sensitivity. Generally, these electrical-type 3DGF sensors exhibit high sensitivity due to these properties including an ultrahigh surface area, and its electronic properties. It shows a strong dependence on surface absorbents (including gas molecules), which can change the carrier density of graphene [24]. Therefore, it is necessary to develop a new type of humidity sensor based on 3DGF by utilizing the unique structure and chemical characteristics and avoiding its shortcomings. In this paper, we fabricate a humidity field effect transistor based on 3DGF and develop test equipment to measure the properties of the device. It exhibits a high performance over a broad RH range from 0% to 85.9%, with fast response and recovery times. To interpret the physical mechanism, we construct the 3DGF model decorated with water and apply CASTEP in the Materials Studio software to calculate the energy structure. Herrin, we explore the potential of 3D GF for portable, reliable and low cost humidity sensing applications in the future. 2. Materials and Methods Utilizing a modified Hummers’ method, [19–21] graphene oxide, denoted as GO, was synthesized from natural graphite powder by an oxidation reaction. GO ethanol solution (50 mL) with the concentration of 1 mg mL−1 was sealed in a 100 mL Teflon-lined autoclave which was then heated up to 180 ◦C and held for 12 h. Then the autoclave was cooled naturally to room temperature. The prepared ethanol intermediates were carefully removed from the autoclave by a slow and gradual solvent exchange with water. After the solvent exchange process was completed, the product filled with water was freeze-dried and then dried at 120 ◦C for 2 h in a vacuum oven. Finally, the sample was annealed at 450 ◦C in H2/Ar (5/95, v/v) for 6 h. Finally, the sample was treated in a UV ozone system for 15 min to obtain the final 3DGF. The infrared spectrum of 3DGF was recorded on a Fourier transform infrared (FTIR) spectrophotometer using potassium bromide (KBr) pellets. Figure 1a shows the FTIR spectra of three-dimensional graphene foam (3DGF)) with water molecules (black line) and dry (red line) conditions. It can be seen that a broad peak at 3436 cm−1 corresponds to the vibration due to the stretching and bending of OH groups present in the water molecules adsorbed by 3DGF. 31
- 47. Sensors 2018, 18, 4337 Thus, it was concluded that 3DGF exhibits strong hydrophilicity. Meanwhile, the absorption peaks at 565, 1163, and 1640 cm−1 correspond to the symmetric and antisymmetric stretching vibrations of C=O, C–O, and C–C groups for 3DGF, respectively. Figure 1b shows the surface morphologies of the 3DGF. Field emission scanning electron microscopy (SEM) images show clear, layered and interconnected three-dimensional uniform graphene sheets. It can be concluded that it forms a spongy porous network structure. [20]. The samples are cut into rectangular slabs (14 mm × 2 mm), and both sides are pasted by copper conductive adhesives on silicon substrates with a size of 14 mm × 14 mm for electrical contact. Figure 1. (a) FTIR spectra of 3D graphene with or without water molecule. (b) Field emission scanning electron microscopy (SEM) images of 3DGF. For humidity sensors, chemical or physical reactions between water molecules and materials induce changes in channel current. External factors including the water concentration, temperature, and operating conditions will impact the performance of the device. For accurate measurements, as shown in Figure 2a, we used a closed box as an experimental chamber to control the humidity. In detail, the water concentrations were controlled by the ratio of saturated water vapor generated by a humidifier to high-purity nitrogen. We assure high quality humidity measurement in different ambient temperature operating conditions in climate chamber as shown in [26]. In order to measure the channel current flowing into the drain electrode (IDS) [27–29], the source (with ground connection) and drain electrodes were connected with a Keithley 2400 apparatus (Tektronix China Ltd, Shanghai, China). The electrical measurements were also performed with this system, and the RH of the environment was measured by a commercial humidometer. Therefore, as described by Figure 2b, the output characteristics of the device were measured under dry and humid conditions. It shows that when the RH level was fixed to 100%, the channel current (IDS) became lower than the conditions under drying. Meanwhile, the Dirac point shifted towards the positive direction. This donor effect [1] has been ascribed to the donation of electrons from the chemically adsorbed water molecules to the 3DGF. It can be concluded that the water molecules decorated in 3DGF will attract electrons and remain as holes, leading to p-type doping. Furthermore, water molecules decrease the charge mobility of 3D graphene, leading to lower currents. Through swelling or the 2D capillary effect [7,15,24], the dielectric constant will increase and the resistance decrease after adsorbing water molecules (confirmed using FTIR, as shown in Figure 1a). At the same time, the space charge polarization effect can be enhanced by adsorbing more water molecules, leading to the rapid diffusion of 3DGF and the formation of protons between hydroxyl groups. [6]. To investigate the mechanism, band energy of graphene decorating with water molecule was theoretically simulated by density functional theory (DFT) in the Material Studio 8.0 software (Neotrident Technology Ltd. Beijing, China). Simply speaking, graphene is simulated by plane wave program implemented in CASTEP. Considering the single and double supercells (2 × 1 × 1 allowing edge reconstruction) under GGA-PBE with 9 × 1 × 1 k-points Monkhorst-Pack point grid 32
- 48. Sensors 2018, 18, 4337 and 500 eV plane wave base truncation, the graphene is simulated by plane wave program with basis cutoff of 500 eV. The geometry was optimized until the total energy reached 2 × 10−5 eV/atom and the maximum force acting on each atom is less than 0.05 eV/Å. For the 3D graphene foam and 3D graphene foam adding water molecule calculations, the CASTEP plane wave code was used under GGA-PBE considering a Monkhorst−Pack grid with 9 × 9 × 1 k-points and a plane wave basis cutoff of 500 eV; optimizing the geometry until the total energy reaches 2 × 10−5 eV/atom and the maximum force per atom exhibits values less than 0.05 eV/Å [30,31]. Figure 2. (a) Testing equipment used for the electrical characterization of 3DGF humidity sensors. (b) Output characteristic of the device decorated with or without water molecules. 3. Results and Discussion Furthermore, the humidity-sensing performance of the 3DGF sensors exposed to different RH levels (0%, 10.0%, 19.9%, 30.3%, 44.5%, 51.4%, 57.1%, 60.3%, 66.4%, 70.5%, 75.2%, 80.2%, and 85.9% RH) are presented in Figure 3a. In a closed air-tight box, the humidity sensors were measured by different RH values ranging from 0 to 85.9%. It can be seen that as the RH level increased, the obtained channel currents of the sensor reduced monotonically. To consider the real-time response and recovery times of the devices, the time-dependent response and recovery curves of the device to 85.9% RH are plotted in Figure 3b. The time taken by a sensor to achieve 85% RH of the total channel current was defined as the response or recovery time. The response and recovery times of the sensor were approximately 89 ms and 189 ms, respectively. Additionally, our humidity sensors exhibited reproducibility and long-term stability. Professionally, the hysteresis value is a vital parameter for humidity sensors as it determines the maximum time lag between the response time (adsorption process) and recovery time (desorption process). With respect to the water content in the environment, the hysteresis effect is defined by the difference between the resistances. In particular, for a perfect humidity sensor, the hysteresis value should be as small as possible or can even be negligible. (a) (b) Ⲵはᆀ൘ᧂ⡸䱦⇥䈧൘䘉ਕ䈍ਾ䶒࣐кĀ Figure 3. (a) Channel current response measurement of the 3DGF humidity sensor with varying different RH. (b) Response and recovery times of the device at 85% RH and the drain voltage was fixed at 1 V. 33
- 49. Sensors 2018, 18, 4337 Table 1 compares the different characteristics of graphene-type humidity sensors including the response/recovery time, fabrication method, and sensitivity range. It was observed that the3DGF sensor exhibited broad sensitivity and rapid response and recovery rates. Table 1. Comparison of different reported humidity sensors with graphene series materials. Reference Material Sensing Range Response/Recovery Time Smith [30] Graphene 1–96% 0.6 s/0.4 s Ghosh [32] Graphene 4–84% 180 s/180 s Cai [33] reduced graphene oxide (rGO)/graphene oxide (GO)/rGO 6.3–100% 1.9 s/3.9 s Zhang [34] Graphene oxide foam 36–92% 2 s/10 s Trung [35] rGO-polyurethane composites 10–70% 3.5 s/7 s Leng [36] GO/Nafion composite 11.3–97.3% 100–300 s/not shown Bi [6] GO 15–95% 10.5 s/41 s Naik [37] GO 30–95% 100 s/not shown Yu [38] GO/poly (sodium 4-styrenesulfonate) (PSS) composite 20–80% 60 s/50 s Zhang [5] rGO/poly(diallylimethyammonium chloride) PDDA composite 11–97% 108 s/94 s Guo [39] rGO 10–95% 50 s/3 s This work 3DGF 0–85.9% 89 ms/189 ms It can be seen that our devices showed good uniformity. Quantitatively, the effect of relative humidity on the device is depicted in Figure 4. Figure 4a describes the relationship between channel current and relative humidity. It can be seen that the relationship showed a decreasing trend with the increase in water humidity. This also showed that the channel current (IDS) decreased more rapidly as relative humidity increased. To characterize the performance of the humidity sensor, the sensitivity (S) of the device was defined by Equation (1) [4,5,30,40]: S = Iwet − Idry IdryRH × 100 (1) where Iwet and Idry represent the channel current of the device under wet and dry conditions (RH = 0%), respectively. As shown in Figure 4b, the sensitivity increased rapidly as RH increased. Due to its perfect performance, including its ultrafast response/recovery rate, our humidity sensors can be used for breathing monitoring or for developing new user interfaces (UIs). Figure 3b presents the ability of a 3DGF sensor to monitor human breathing. In particular, during the user’s speech, the ultrafast humidity sensor allowed the capture of fine features due to moisture modulation. Therefore, the 3DGF ultra-fast RH sensor can be used to identify different whistles, which can make use of low-cost and low-power sensors for user authentication. Figure 4. Relative humidity effect on the device performance. (a) Channel currents (IDS) with the relationship of RH (b) The variation in sensitivity of the device for different RH values. 34
- 50. Sensors 2018, 18, 4337 A schematic model of humidity sensing at a 3DGF film is shown in Figure 5a. To investigate its mechanism, the band energy of graphene decorated with water molecules was theoretically simulated by density functional theory (DFT) in the Material Studio 8.0 software. As shown in Figure 5b, conductivity and valence are at K Brillouin point, which makes the material a direct bandgap semiconductor. The direct band gap at the K point was ~0.172 eV, as shown in Figure 5c. This can be ascribed to the donor effect [3] attributed to the donation of electrons from the chemically adsorbed water molecules to the 3DGF surface. The water molecules decorated in 3DGF will attract electrons. Simply, water molecules open the band gap of 3DGF. Meanwhile, electron density will decrease and the conduction level will rise, leading to the formation of band energy. Figure 5. (a) The bonding mechanism between the graphene and water molecules. (b) The electronic band structure of graphene decorated with water. (c) The energy gap at the K point location. 4. Conclusions In summary, a three-dimensional graphene foam (3DGF) exhibiting super permeability to water was exploited in humidity sensors, enabling a humidity sensor with a broad range of % RH values and unprecedented response speed (response time: ~89 ms, recovery time: ~189 ms). The ultra-fast response speed of these sensors enables us to observe the regulation of moisture in a user’s breath. We constructed the 3DGF model decorated with water molecules theoretically and conducted the CASTEP as implemented in Materials Studio to calculate the energy structure. This allows sensors to be used in a variety of applications, such as humidity sensing, which we have experimentally verified with a cheap and easily available identification system. In addition, for different 3D materials, such as 3D transition metal dihalogenated hydrocarbons, ultra-thin nanoporous membranes for sensing applications can be realized in the interaction with different vapors and gases, which can be explored. Author Contributions: Conceptualization, Y.Y. and Y.Z.; methodology, Y.Z.; formal analysis, Y.Y., Z.C., Y.L., L.J., Q.L., Y.C., and M.C; data curation, Y.Y.; writing—original draft preparation, Y.Y.; writing—review and editing, Y.Y.; supervision, Y.Z. and J.Y. (Junbo Yang); project administration, Y.Z. and J.Y. (Jianquan Yao); funding acquisition, J.Y. (Jianquan Yao). Funding: This work was supported by the National Natural Science Foundation of China (Nos. 61675147, 61605141 and 61735010), Basic Research Program of Shenzhen (JCYJ20170412154447469) and Wenzhou City Governmetal Public Industrial Technology Project (G20160014). Acknowledgments: We thank Yongshen Chen group in Nankai University, which he provides three dimensional graphene foam. Conflicts of Interest: The authors declare no conflict of interest. 35
- 51. Sensors 2018, 18, 4337 References 1. Traversa, E. Ceramic sensors for humidity detection: The state-of-the-art and future developments. Sens. Actuators B Chem. 1995, 23, 135–156. [CrossRef] 2. Chu, J.; Peng, X.; Feng, P.; Sheng, Y.; Zhang, J. Study of humidity sensors based on nanostructured carbon films produced by physical vapor deposition. Sens. Actuators B Chem. 2013, 178, 508–513. [CrossRef] 3. Chen, Z.; Lu, C. Humidity Sensors: A Review of Materials and Mechanisms. Sens. Lett. 2005, 3, 274–295. [CrossRef] 4. Mogera, U.; Sagade, A.A.; George, S.J.; Kulkarni, G.U. Ultrafast response humidity sensor using supramolecular nanofibre and its application in monitoring breath humidity and flow. Sci. Rep. 2014, 4, 4103. [CrossRef] [PubMed] 5. Zhang, D.; Tong, J.; Xia, B. Humidity-sensing properties of chemically reduced graphene oxide/polymer nanocomposite film sensor based on layer-by-layer nano self-assembly. Sens. Actuators B Chem 2014, 197, 66–72. [CrossRef] 6. Bi, H.; Yin, K.; Xie, X.; Ji, J.; Wan, S.; Sun, L.; Terrones, M.; Dresselhaus, M.S. Ultrahigh humidity sensitivity of graphene oxide. Sci. Rep. 2013, 3, 2714. [CrossRef] [PubMed] 7. Wong, W.C.; Chan, C.C.; Chen, L.H.; Li, T.; Lee, K.X.; Leong, K.C. Polyvinyl alcohol coated photonic crystal optical fiber sensor for humidity measurement. Sens. Actuators B Chem. 2012, 174, 563–569. [CrossRef] 8. Kuang, Q.; Lao, C.; Wang, Z.L.; Xie, Z.; Zheng, L. High-Sensitivity Humidity Sensor Based on a Single SnO2 Nanowire. J. Am. Chem. Soc. 2007, 129, 6070–6071. [CrossRef] 9. Hu, P.; Zhang, J.; Li, L.; Wang, Z.; O’Neill, W.; Estrela, P. Carbon nanostructure-based field-effect transistors for label-free chemical/biological sensors. Sensors 2010, 10, 5133–5159. [CrossRef] 10. Vojko, M. Next generation AT-cut quartz crystal sensing devices. Sensors 2011, 11, 4474–4482. 11. Matko, V.; Donlagic, D. Sensor for high-air-humidity measurement. Sens. Actuators A Phys. 1997, 61, 331–334. [CrossRef] 12. Fei, T.; Zhao, H.; Jiang, K.; Zhou, X.; Zhang, T. Polymeric humidity sensors with nonlinear response: Properties and mechanism investigation. J. Appl. Polym. Sci. 2013, 130, 2056–2061. [CrossRef] 13. Chen, W.P.; Zhao, Z.G.; Liu, X.W.; Zhang, Z.X.; Suo, C.G. A Capacitive Humidity Sensor Based on Multi-Wall Carbon Nanotubes (MWCNTs). Sensors 2009, 9, 7431–7444. [CrossRef] [PubMed] 14. Han, J.-W.; Kim, B.; Li, J.; Meyyappan, M. A carbon nanotube based ammonia sensor on cellulose paper. RSC Adv. 2014, 4, 549–553. [CrossRef] 15. Borini, S.; White, R.; Wei, D.; Astley, M.; Haque, S.; Spigone, E.; Harris, N.; Kivioja, J.; Ryhänen, T. Ultrafast Graphene Oxide Humidity Sensors. ACS Nano 2013, 7, 11166–11173. [CrossRef] [PubMed] 16. Zhao, X.; Long, Y.; Yang, T.; Li, J.; Zhu, H. Simultaneous High Sensitivity Sensing of Temperature and Humidity with Graphene Woven Fabrics. ACS Appl. Mater. Interfaces 2017, 9, 30171–30176. [CrossRef] [PubMed] 17. Zhang, D.; Tong, J.; Xia, B.; Xue, Q. Ultrahigh performance humidity sensor based on layer-by-layer self-assembly of graphene oxide/polyelectrolyte nanocomposite film. Sens. Actuators B Chem. 2014, 203, 263–270. [CrossRef] 18. Ma, Y.; Chen, Y. Three-dimensional graphene networks: Synthesis, properties and applications. Natl. Sci. Rev. 2015, 2, 40–53. [CrossRef] 19. Wu, Y.; Yi, N.; Huang, L.; Zhang, T.; Fang, S.; Chang, H.; Li, N.; Oh, J.; Lee, J.A.; Kozlov, M.; et al. Three-dimensionally bonded spongy graphene material with super compressive elasticity and near-zero Poisson’s ratio. Nat. Commun. 2015, 6, 6141. [CrossRef] 20. Zhang, T.; Chang, H.; Wu, Y.; Xiao, P.; Yi, N.; Lu, Y.; Ma, Y.; Huang, Y.; Zhao, K.; Yan, X.-Q.; et al. Macroscopic and direct light propulsion of bulk graphene material. Nat. Photonics 2015, 9, 471–476. [CrossRef] 21. Chang, H.; Qin, J.; Xiao, P.; Yang, Y.; Zhang, T.; Ma, Y.; Huang, Y.; Chen, Y. Highly Reversible and Recyclable Absorption under Both Hydrophobic and Hydrophilic Conditions using a Reduced Bulk Graphene Oxide Material. Adv. Mater. 2016, 28, 3504–3509. [CrossRef] 22. Cao, X.; Yin, Z.; Zhang, H. Three-dimensional graphene materials: Preparation, structures and application in supercapacitors. Energy Environ. Sci. 2014, 7, 1850–1865. [CrossRef] 36
- 52. Sensors 2018, 18, 4337 23. Chabot, V.; Higgins, D.; Yu, A.; Xiao, X.; Chen, Z.; Zhang, J. A review of graphene and graphene oxide sponge: Material synthesis and applications to energy and the environment. Energy Environ. Sci. 2014, 7, 1564–1596. [CrossRef] 24. Yavari, F.; Chen, Z.; Thomas, A.V.; Ren, W.; Cheng, H.M.; Koratkar, N. High sensitivity gas detection using a macroscopic three-dimensional graphene foam network. Sci. Rep. 2011, 1, 166. [CrossRef] 25. Ma, Y.; Zhao, M.; Cai, B.; Wang, W.; Ye, Z.; Huang, J. 3D graphene foams decorated by CuO nanoflowers for ultrasensitive ascorbic acid detection. Biosens. Bioelectron. 2014, 59, 384–388. [CrossRef] [PubMed] 26. Brezovec, B.; Matko, V. Software and Equipment for Remote Testing of Sensors. Sensors 2007, 7, 1306–1316. [CrossRef] 27. Yu, Y.; Zhang, Y.; Zhang, Z.; Zhang, H.; Song, X.; Cao, M.; Che, Y.; Dai, H.; Yang, J.; Wang, J.; et al. Broadband Phototransistor Based on CH3NH3PbI3 Perovskite and PbSe Quantum Dot Heterojunction. J. Phys. Chem. Lett. 2017, 8, 445–451. [CrossRef] 28. Yu, Y.; Zhang, Y.; Song, X.; Zhang, H.; Cao, M.; Che, Y.; Dai, H.; Yang, J.; Zhang, H.; Yao, J. High Performances for Solution-Pocessed 0D-0D Heterojunction Phototransistors. Adv. Opt. Mater. 2017, 5, 1700565. [CrossRef] 29. Yu, Y.; Zhang, Y.; Song, X.; Zhang, H.; Cao, M.; Che, Y.; Dai, H.; Yang, J.; Zhang, H.; Yao, J. PbS-Decorated WS2 Phototransistors with Fast Response. ACS Photonics 2017, 4, 950–956. [CrossRef] 30. Smith, A.D.; Elgammal, K.; Niklaus, F.; Delin, A.; Fischer, A.C.; Vaziri, S.; Forsberg, F.; Rasander, M.; Hugosson, H.; Bergqvist, L.; et al. Resistive graphene humidity sensors with rapid and direct electrical readout. Nanoscale 2015, 7, 19099–19109. [CrossRef] 31. Gutierrez, H.R.; Perea-Lopez, N.; Elias, A.L.; Berkdemir, A.; Wang, B.; Lv, R.; Lopez-Urias, F.; Crespi, V.H.; Terrones, H.; Terrones, M. Extraordinary room-temperature photoluminescence in triangular WS2 monolayers. Nano Lett. 2013, 13, 3447–3454. [CrossRef] [PubMed] 32. Ghosh, A.; Late, D.J.; Panchakarla, L.S.; Govindaraj, A.; Rao, C.N.R. NO2 and humidity sensing characteristics of few-layer graphenes. J. Exp. Nanosci. 2009, 4, 313–322. [CrossRef] 33. Cai, J.; Lv, C.; Aoyagi, E.; Ogawa, S.; Watanabe, A. Laser Direct Writing of a High-Performance All-Graphene Humidity Sensor Working in a Novel Sensing Mode for Portable Electronics. ACS Appl. Mater. Interfaces 2018, 10, 23987–23996. [CrossRef] [PubMed] 34. Zhang, K.-L.; Hou, Z.-L.; Zhang, B.-X.; Zhao, Q.-L. Highly sensitive humidity sensor based on graphene oxide foam. Appl. Phys. Lett. 2017, 111, 153101. [CrossRef] 35. Trung, T.Q.; Duy, L.T.; Ramasundaram, S.; Lee, N.-E. Transparent, stretchable, and rapid-response humidity sensor for body-attachable wearable electronics. Nano Res. 2017, 10, 2021–2033. [CrossRef] 36. Leng, X.; Luo, D.; Xu, Z.; Wang, F. Modified graphene oxide/Nafion composite humidity sensor and its linear response to the relative humidity. Sens. Actuators B Chem. 2018, 257, 372–381. [CrossRef] 37. Naik, G.; Krishnaswamy, S. Room-Temperature Humidity Sensing Using Graphene Oxide Thin Films. Graphene 2016, 5, 1–13. [CrossRef] 38. Yu, H.W.; Kim, H.K.; Kim, T.; Bae, K.M.; Seo, S.M.; Kim, J.M.; Kang, T.J.; Kim, Y.H. Self-powered humidity sensor based on graphene oxide composite film intercalated by poly(sodium 4-styrenesulfonate). ACS Appl. Mater. Interfaces 2014, 6, 8320–8326. [CrossRef] 39. Guo, L.; Jiang, H.B.; Shao, R.Q.; Zhang, Y.L.; Xie, S.Y.; Wang, J.N.; Li, X.B.; Jiang, F.; Chen, Q.D.; Zhang, T. Two-beam-laser interference mediated reduction, patterning and nanostructuring of graphene oxide for the production of a flexible humidity sensing device. Carbon 2012, 50, 1667–1673. [CrossRef] 40. Zhu, Z.-T.; Mason, J.T.; Dieckmann, R.; Malliaras, G.G. Humidity sensors based on pentacene thin-film transistors. Appl. Phys. Lett. 2002, 81, 4643–4645. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). 37
- 53. sensors Article Design and Implementation of an Infrared Radiant Source for Humidity Testing Hong Zhang *, Chuansheng Wang, Xiaorui Li, Boyan Sun and Dong Jiang School of Computer Science and Technology, Harbin University of Science and Technology, 52 Xuefu Road, Harbin 150080, China; wangchuansheng994@163.com (C.W.); 13682088813@163.com (X.L.); 15754509280@163.com (B.S.); wdyu2004@163.com (D.J.) * Correspondence: zhangh@hrbust.edu.cn; Tel.: +86-177-6655-5090 or +86-138-3609-9065 Received: 26 June 2018; Accepted: 11 September 2018; Published: 13 September 2018 Abstract: A novel way to measure humidity through testing the emissivity of an area radiant source is presented in this paper. The method can be applied in the environment at near room temperature (5~95 ◦C) across the relative humidity (RH) range of 20~90% RH. The source, with a grooved radiant surface, works in the far infrared wavelength band of 8~12 μm. The Monte-Carlo model for thermal radiation was set up to analyze the V-grooved radiant surface. Heat pipe technology is used to maintain an isothermal radiant surface. The fuzzy-PID control method was adopted to solve the problems of intense heat inertia and being easily interfered by the environment. This enabled the system to be used robustly across a large temperature range with high precision. The experimental results tested with a scanning radiant thermometer showed that the radiant source can provide a uniform thermal radiation capable of satisfying the requirements of humidity testing. The calibration method for the radiant source for humidity was explored, which is available for testing humidity. Keywords: infrared radiant source; Monte Carlo method; emissivity; calibration; humidity 1. Introduction Compared to traditional humidity measurement methods, innovative electronic testing methods involving humidity sensors such as hygristors and humicaps are the current research direction. Even though electric methods have fast responses, they often lack stability and their accuracy is improved little. Widely available commercial humidity sensors composed of humicaps use embedded microprocessors, such as the DHT11 with ±5% RH precision, and 1% RH resolution which are convenient to use. Because humidity is often mingled together with temperature, the precision and the humidity measurement range are easily affected by the temperature. Humidity, which reflects the degree of dryness of the atmosphere, is an important variable that is extensively tested in agriculture, industry, hospital and warehouse. Some sensors with new materials possessing resistive and capacitive features have been explored, which include a sulfonated polycarbonate resistive humidity sensor [1], polyimide-based capacitive humidity sensor [2], a high-performance capacitive humidity sensor with novel electrode and polyimide layer capacitive humidity sensors [3], and some sensors with improved sensing properties whose response and recovery times are 14.5 s and 34.27 s, respectively, for humidity levels between 33% RH and 95% RH at 102 Hz. [4]. These kinds of humidity sensors often have long response times. Most studies focus on material and processing innovations to increase the humidity testing precision. There have been few breakthroughs in hygristors because these are easily interfered by the environment. Novel ways using new effects such as optical properties present approaches to measure humidity [5]. Because humidity is a factor that affects the radiation measurement, humidity can be measured indirectly by testing radiation changes [6,7]. As a standard radiant source, a blackbody is usually adopted for calibrating infrared instruments such as pyrometers and radiant thermometers. Industrial blackbodies ranging from −50~2500 ◦C Sensors 2018, 18, 3088; doi:10.3390/s18093088 www.mdpi.com/journal/sensors 38
- 54. Sensors 2018, 18, 3088 have been developed by the National Research Council (NRC) of Canada in order to calibrate optical testing devices [8]. A high-spatial-resolution multi-spectral imager (ASTER) on the first platform (Terra) of NASA’s Earth Observing System requires a blackbody radiant source on a satellite for calibration purposes [9]. Traditional blackbody cavities evaluated by the Bedford methods [10] usually possess symmetrical shapes with small apertures, which makes them suitable for the high temperature range case, but not for the case of near room temperature range measurements. Minimum Resolvable Temperature Difference (MRTD) sensitivity requires that a radiant source working in the far infrared scope should be an area source which can provide a stable radiant flux with high uniformity. As to environmental humidity measurement, according to MRTD, an area radiant source ranging from 5~95 ◦C in the wavelength bands of 8~12 μm, is required. Under a certain temperature, a source should emit a stable radiation which is monitored by a radiometer. In order to enhance the effective emissivity of a radiant surface, its surface is often processed into grooves or mini holes. The radiant surface of the source possesses concentric V-grooves which can increase the effective emissivity. Among statistical evaluation methods, Monte Carlo methods have been widely applied in optical radiometry and blackbody cavity analysis [11]. Monte Carlo methods possess advantages which are greater than exactitude methods in complex radiant characteristic analysis. Therefore, the Monte Carlo method is adopted to analyze the effective emissivity of the radiant source. After a theoretical analysis on the distribution of the effective emissivity of the radiant surface, the source structure was constructed. Heat pipe technology keeps the source isothermal and the temperature control system ensures that the source is stable at a certain temperature. The radiant source has a broad applications in various fields, such as infrared imaging, infrared measurement and humidity test. Although there are tens of ways to measure humidity, among which the most traditional methods are the dry and wet bulb thermometer whose precision is lower compared with modern electronic methods, most of these ways are not satisfactory in terms of precision and stability [7]. Capacitor sensors which possess fast response advantages are employed to test humidity, but their measuring precision is easily affected by electromagnetic interference. Besides, humidity testing is often affected by the environmental temperature, which is a factor that makes humidity sensors’ precision not be high. Humidity testing through radiation possesses advantages of fast response and robustness with high precision. Our research on an infrared source which has highly sensitivity for humidity may provide an improved way to measure humidity. 2. Analysis on Characteristics of Radiant Surface The Monte-Carlo method was utilized for analyzing the radiant surface with concentric V-grooves. Assuming that the surface is diffuse (Lambertian), the calculation on its luminance follows Lambert’s cosine Law. The Monte-Carlo Model of thermal radiation was set up. A Monte-Carlo simulation is implemented through random sampling that is based on probability models whose deduction methods are based upon actual physical models. Exactitude methods like the Bedford method are just suitable to calculate simple symmetric cavity shapes. Although Monte-Carlo methods are flexible enough to be used for complicated cases, they are often regarded as unreliable methods with low precision. The effective emissivity of a cone was calculated by both the Monte Carlo method and the Bedford method, respectively. By comparing the results from the both methods, the correctness of the Monte Carlo method was proven. 2.1. Monte-Carlo Model in Thermal Radiation The idea of a Monte-Carlo method is to set up a probability model or stochastic process whose parameters are equal to the solution of the problem to investigate, then to calculate the statistical characteristics of the required parameters through sampling, and the solution can be solved based on a vast number of observations. In thermal radiation calculations, local temperature and radiation fluxes are usually involved. The process of thermal radiation exchange is regarded as the movement of discrete energy beams. In this way, the local radiation flux can be obtained by calculating the number 39
- 55. Another Random Scribd Document with Unrelated Content
- 56. one-gross-of-gloves long. Buffon would only sit down to write after taking a bath and donning pure linen with a full frilled bosom. Haydn[71] declared that he could not compose unless he wore the large seal-ring which Frederick the Great had given him. He would sit wrapped in silence for an hour or more, after which he would seize his pen and write rapidly without touching a musical instrument; and he rarely altered a line. In early life, poor, freezing in a miserable garret, he studied the rudiments of his favorite art by the side of an old broken harpsichord. For a period of six years he endured a bitter conflict with poverty, being often compelled for the sake of warmth to lie in bed most of the day as well as the night. Finally he was relieved from this thraldom by the generosity of his patron, Prince Esterhazy, a passionate lover of music, who appointed him his chapel-master, with a salary sufficient to keep him supplied with the ordinary comforts of life. Crébillon the elder, a celebrated lyric poet and member of the French Academy, was enamoured of solitude, and could only write effectively under such circumstances. His imagination teemed with romances, and he produced eight or ten dramas which enjoyed popularity in their day,—about 1776. One day, when he was alone and in a deep reverie, a friend entered his study hastily. Don't disturb me, cried the author, I am enjoying a moment of happiness: I am going to hang a villain of a minister, and banish another who is an idiot. We have lately mentioned Dumas. Hans Christian Andersen, speaking of the various habits of authors, thus refers to the elder Dumas, with whom he was intimate: I generally found him in bed, even long after mid-day, where he lay, with pen, ink, and paper by his side, and wrote his newest drama. On entering his apartment I found him thus one day; he nodded kindly to me, and said: 'Sit down a minute. I have just now a visit from my Muse; she will be going directly.' He wrote on, and after a brief silence shouted 'Vivat' sprang out of bed, and said, 'The third act is finished!'[72]
- 57. Lamartine was peculiar in his mode of composition, and never saw his productions, after the first draft, until they were printed, bound, and issued to the public. He was accustomed to walk forth in his park during the after part of the day, or of a moonlit evening, with pencil and pieces of paper, and whatever ideas struck him he recorded. That was the end of the matter so far as he was concerned. These pieces of paper he threw into a special box, without a number or title upon them. His literary secretary with much patient ability assorted these papers, arranged them as he thought best, and sold them to the publishers at a royal price. We know of no similar instance where authorship and recklessness combined have produced creditable results. Certainly such indifference argued only the presence of weakness and irresponsibility, which were indeed prominent characteristics of Lamartine. The remarkable facility with which Goethe's poems were produced is said to have resembled improvisation, an inspiration almost independent of his own purposes. I had come, he says, to regard the poetic talent dwelling in me entirely as nature; the rather that I was directed to look upon external nature as its proper subject. The exercise of this poetic gift might be stimulated and determined by occasion, but it flowed forth more joyfully and richly, when it came involuntarily, or even against my will. Addison, whose style is perhaps the nearest to perfection in ancient or modern literature, did not reach that standard without much patient labor. Pope tells us that he would show his verses to several friends, and would alter nearly everything that any of them hinted was wrong. He seemed to be distrustful of himself, and too much concerned about his character as a poet, or, as he expressed it, 'too solicitous for that kind of praise which God knows is a very little matter after all.' Pope himself published nothing until it had been a twelvemonth on hand, and even then the printer's proofs were full of alterations. On one occasion this was carried so far that Dodsley, his publisher, thought it better to have the whole recomposed than to attempt to make the necessary alterations. Yet Pope admits that the things that I have
- 58. written fastest have always pleased the most. I wrote the 'Essay on Criticism' fast, for I had digested all the matter in prose before I began it in verse. I never work better, says Luther, than when I am inspired by anger: when I am angry, I can write, pray, and preach well; for then my whole temperament is quickened, my understanding sharpened, and all mundane vexations and temptations depart. We are reminded of Burke's remark in this connection: A vigorous mind is as necessarily accompanied with violent passions as a great fire with great heat. Luther, however ribald he may have been at times, had the zeal of honesty. There was not a particle of vanity or self- sufficiency in the great reformer. Do not call yourselves Lutherans, he said to his followers; call yourselves Christians. Who and what is Luther? Has Luther been crucified for the world? Churchill,[73] the English poet and satirist, was so averse to correcting and blotting his manuscript that many errors were unexpunged, and many lines which might easily have been improved were neglected. When expostulated with upon this subject by his publisher, he replied that erasures were to him like cutting away so much of his flesh; thus expressing his utter repugnance to an author's most urgent duty. Though Macaulay tells us that his vices were not so great as his virtues, still he was dissipated and licentious. Cowper was a great admirer of his poetry, and called him the great Churchill. George Wither,[74] the English poet, satirist, and political writer, was compelled to watch and fast when he was called upon to write. He went out of himself, as he said, at such times, and if he tasted meat or drank one glass of wine he could not produce a verse or sentence. Rogers, who wrote purely con amore, took all the time to perfect his work which his fancy dictated, and certainly over-refined many of his compositions. The Pleasures of Memory occupied him seven years. In writing, composing, re-writing, and altering his Columbus and Human Life, each required just double that period of time before
- 59. the fastidious author felt satisfied to call it finished. Besides this, the second edition of each went through another series of emendations. The observant reader will find that Rogers has often weakened his first and best thoughts by this elaboration. The expression of true genius oftenest comes, like the lightning, in its full power and effect at the first flash. Every event that a man would master, says Holmes, must be mounted on the run, and no man ever caught the reins of a thought except as it galloped by him. One who has had years of active editorial experience on the daily press can hardly conceive of such fastidious slowness of composition as characterizes some authors. Sir Joshua Reynolds, in speaking of Rogers, Rochefoucauld, Cowper, and others, and their dilatory habits of composition, says, that although men of ordinary talents may be highly satisfied with their productions, men of genius never are,—an assumption which is not borne out by facts, as we shall have occasion to show in these chapters. Modesty is not always the characteristic of genius; and very few popular writers are without a due share of vanity in their natures. Voltaire somewhere says that an author should write with the rapidity which genius inspires, but should correct with care and deliberation; which doubtless expresses the process adopted by this unscrupulous but versatile writer, of whom Carlyle said: With the single exception of Luther, there is perhaps, in these modern ages, no other man of a merely intellectual character, whose influence and reputation have become so entirely European as that of Voltaire. Sydney Smith was so rapid a producer that he had not patience even to read over his compositions when finished. He would throw down his manuscript and say: There, it is done; now, Kate, do look it over, and put dots to the i's and strokes to the t's. He was once advised by a fashionable publisher to attempt a three-volume novel. Well, said he, after some seeming consideration, if I do so, I must have an archdeacon for my hero, to fall in love with the pew-opener, with the clerk for a confidant; tyrannical interference of the church- wardens; clandestine correspondence concealed under the hassock; appeal to the parishioners, etc. He was overflowing with humor to
- 60. the very close of life. He wrote to Lady Carlisle during his last illness, saying, If you hear of sixteen or eighteen pounds of human flesh, they belong to me. I look as if a curate had been taken out of me. Buffon caused his Époques de la Nature to be copied eighteen times, so many corrections and changes were made. As he was then (1778) over seventy years of age, one would think this an evidence that his mind was failing him. Pope covered with memoranda every scrap of clear paper which came in his way. Some of his most elaborate literary work was begun and finished on the backs of old letters and bits of yellow wrappers. We do not wonder that such fragmentary manuscript always suggested the idea of revision and correction. It is difficult to understand why Pope should have assumed this small virtue of economy and yet often have been lavish in other directions; indeed, it may be questioned whether it was intended to be an act of economy. Such petty parsimony is inexplicable, but certainly it grew into a fixed habit with him. We believe it was Swift who first called him paper-saving Pope; but Swift was nearly as eccentric a paper-saver as Pope. He wrote to Dr. Sheridan: Keep very regular accounts, in large books and a fair hand; not like me, who, to save paper, confuse everything! Miss Mitford had the same habit of writing upon waste scraps of paper, fly-leaves of books, envelopes, and odd rejected bits, all in so small a hand as to be nearly illegible. William Hazlitt was also remarkable for the same practice, and we are told that he even made the first outline of some of his essays on the walls of his chamber, much to the annoyance of his landlady. Some idea of the rapidity with which Byron wrote may be inferred from the fact that the Prisoner of Chillon was written in two days and sent away complete to the printer. The traveller in Switzerland does not fail to visit the house—once a wayside inn, at Merges, on the Lake of Geneva—where Byron wrote this poem while detained by a rainstorm, in 1816. On the heights close at hand is the Castle of Wuffens, dating back to the tenth century. Morges is a couple of leagues from Lausanne, and the spot where Gibbon finished his
- 61. Rise and Fall of the Roman Empire, in 1787. Colton, the philosophical but erratic author of Lacon, wrote that entire volume upon covers of letters and such small scraps of paper as happened to be at hand when a happy thought inspired him. Having completed a sentence, and rounded it to suit his fancy, he threw it into a pile with hundreds of others, which were finally turned over to the printer in a cloth bag. No classification or system of arrangement was observed. Colton exhibited all the singularities that only too often characterize genius, especially as regards improvidence and recklessness of habit. He lived unattended, in a single room in Princes Street, Soho, London, in a neglected apartment containing scarcely any furniture. He wrote very illegibly upon a rough deal table with a stumpy pen. He was finally so pressed with debts that he absconded to avoid his London creditors, though he held the very comfortable vicarage of Kew, in Surrey. Montaigne, the French philosopher and essayist, whose writings have been translated into every modern tongue, like the musician Sacchini was marvellously fond of cats, and would not sit down to write without his favorite by his side. Thomas Moore required complete isolation when he did literary work, and shut himself up, as did Charles Dickens. He was a very slow and painstaking producer. Some friend having congratulated him upon the seeming facility and appropriateness with which a certain line was introduced into a poem he had just published, Moore replied, Facility! that line cost me hours of patient labor to achieve. His verses, which read so smoothly, and which appear to have glided so easily from his pen, were the result of infinite labor and patience. His manuscript, like Tennyson's, was written, amended, rewritten, and written again, until it was finally satisfactory to his critical ear and fancy. Easy writing, said Sheridan, is commonly damned hard reading. Bishop Warburton tells us that he could only write in a hand-to- mouth style unless he had all his books about him; and that the blowing of an east wind, or a fit of the spleen, incapacitated him for literary work; and still another English bishop could write only when
- 62. in full canonicals, a fact which he frankly admitted. Milton would not attempt to compose except between the vernal and autumnal equinoxes, at which season his poetry came as if by inspiration, and with scarcely a mental effort.[75] Thomson, Collins, and Gray entertained very similar ideas, which when expressed so incensed Dr. Johnson that he publicly ridiculed them. Crabbe fancied that there was something in the effect of a sudden fall of snow that in an extraordinary manner stimulated him to poetic composition; while Lord Orrery found no stimulant equal to a fit of the gout!—all of which fancies are but mild forms of monomania. James Hogg (the Ettrick Shepherd) was only too glad to write without any of these accessories, when he could get any material to write upon. He used to employ a bit of slate, for want of the necessary paper and ink. The son of an humble Scottish farmer, he experienced all sorts of misfortunes in his endeavors to pursue literature as a calling. He was both a prose and poetic writer of considerable native genius, and formed one of the well-drawn characters of Christopher North's Noctes Ambrosianæ. N. P. Willis in the latter years of his life was accustomed to ride on horseback before he sat down to write. He believed there was a certain nervo-vital influence imparted from the robust health and strength of the animal to the rider, as he once told the writer of these pages; and, so far as one could judge, the influence upon himself certainly favored such a conclusion. Some authors frankly acknowledge that they have not the necessary degree of patience to apply themselves to the correction of their manuscripts. Ovid, the popular Roman poet, admitted this. Such people may compose with pleasure, but there is the end; neither a sense of responsibility nor a desire for correctness can overcome their constitutional laziness. Pope, Dryden, Moore, Coleridge, Swift, —in short, nine-tenths of the popular authors of the past and the present, all change, correct, amplify, or contract, and interline more or less every page of manuscript which they produce, and often to such a degree as greatly to confuse the compositors. Richard Savage, the unfortunate English poet, could not, or would not, bring himself to correct his faulty sentences, being greatly indebted to the
- 63. intelligence of the proof-reader for the presentable form in which his writings finally appeared. Julius Scaliger, a celebrated scholar and critic, was, on the other hand, an example of remarkable correctness, so that his manuscript and the printer's pages corresponded exactly, page for page and line for line. Hume,[76] the historian, was never done with his manifold corrections; his sense of responsibility was unlimited, and his appreciation of his calling was grand. Fénelon and Gibbon were absolutely correct in their first efforts; and so was Adam Smith, though he dictated to an amanuensis. We are by no means without sympathy for those writers who dread and avoid the reperusal and correction of their manuscripts. Only those who are familiar with the detail of book-making can possibly realize its trying minutiæ. When one has finished the composition and writing of a chapter, his work is only begun; it must be read and re-read with care, to be sure of absolute correctness. When once in type, it must be again carefully read for the correction of printer's errors, and again revised by second proof; and finally a third proof is necessary, to make sure that all errors previously marked have been corrected. By this time, however satisfactory in composition, the text becomes more tedious than a twice-told tale. Any author must be singularly conceited who can, after such experience, take up a chapter or book of his own production and read it with any great degree of satisfaction. Godeau, Bishop of Venice, used to say that to compose is an author's heaven; to correct, an author's purgatory; but to revise the press, an author's hell! Guido Reni, whose superb paintings are among the gems of the Vatican, in the height of his fame would not touch pencil or brush except in full dress. He ruined himself by gambling and dissolute habits, and became lost as to all ambition for that art which had been so grand a mistress to him in the beginning. He finally arrived at that stage where he lost at the gaming-table and in riotous living what he earned by contract under one who managed his affairs, giving him a stipulated sum for just so much daily work in his studio.
- 64. Such was the famous author of that splendid example of art, the Martyrdom of Saint Peter, in the Vatican. Parmigiano, the eminent painter, was full of the wildness of genius. He became mad after the philosopher's stone, jilting art as a mistress, though his eager creditors forced him to set once more to work, though to little effect. Great painters, like great writers, have had their peculiar modes of producing their effects. Thus Domenichino was accustomed to assume and enact before the canvas the passion and character he intended to depict with the brush. While engaged upon the Martyrdom of Saint Andrew, Caracci, a brother painter, came into his studio and found him in a violent passion. When this fit of abstraction had passed, Caracci embraced him, admitting that Domenichino had proved himself his master, and that he had learned from him the true manner of expressing sentiment or passion upon the canvas. Richard Wilson, the eminent English landscape-painter, strove in vain, he said, to paint the motes dancing in the sunshine. A friend coming into his studio found the artist sitting dejected on the floor, looking at his last work. The new-comer examined the canvas and remarked critically that it looked like a broad landscape just after a shower. Wilson started to his feet in delight, saying, That is the effect I intended to represent, but thought I had failed. Poor Wilson possessed undoubted genius, but neglected his art for brandy, and was himself neglected in turn. He was one of the original members of the Royal Academy. Undoubtedly, genius is at times nonplussed and at fault, like plain humanity, and is helped out of a temporary dilemma by accident,— as when Poussin the painter, having lost all patience in his fruitless attempts to produce a certain result with the brush, impatiently dashed his sponge against the canvas and brought out thereby the precise effect desired; namely, the foam on a horse's mouth. Washington Allston[77] is recalled to us in this connection, one of the most eminent of our American painters, and a poet of no ordinary
- 65. pretensions. The Sylphs of the Seasons and other Poems was published in 1813. He was remarkable for his graphic and animated conversational powers, and was the warm personal friend of Coleridge and Washington Irving. Irving says, His memory I hold in reverence and affection as one of the purest, noblest, and most intellectual beings that ever honored me with his friendship. While living in London he was elected associate of the Royal Academy. Bostonians are familiar with Allston's half-finished picture of Belshazzar's Feast, upon which he was engaged when death snatched him from his work. CHAPTER IV. It has been said that the first three men in the world were a gardener, a ploughman, and a grazier; while all political economists admit that the real wealth and stamina of a nation must be looked for among the cultivators of the soil. Was it not Swift who declared that the man who could make two ears of corn or two blades of grass grow upon a spot of ground where only one grew before, deserved better of mankind than the whole race of politicians? Bacon, Cowley, Sir William Temple, Buffon, and Addison were all attached to horticulture, and more or less time was devoted by them to the cultivation of trees and plants of various sorts; nor did they fail to record the refined delight and the profit they derived therefrom. Daniel Webster was an enthusiastic agriculturist; so were Washington, Adams, Jefferson, Walter Scott, Horace Greeley, Gladstone, Evarts,[78] Wilder, Loring, Poore, and a host of other contemporaneous and noted men. They who labor in the earth, said Jefferson, are the chosen people of God.
- 66. But the habits and mode of composition adopted by literary men still crowd upon the memory. Hobbes, the famous English philosopher, author of a Treatise on Human Nature, a political work entitled the Leviathan, etc., was accustomed to compose in the open air. The top of his walking-stick was supplied with pen and inkhorn, and he would pause anywhere to record his thoughts in the note-book always carried in his pocket. Virgil rose early in the morning and wrote at a furious rate innumerable verses, which he afterwards pruned and altered and polished, as he said, after the manner of a bear licking her cubs into shape. The Earl of Roscommon, in his Essay on Translated Verse, declared this to be the duty of the poet, — To write with fury and correct with phlegm. Dr. Darwin, the ingenious English poet, wrote his works, like some others of whom we have spoken, on scraps of paper with a pencil while travelling. His old-fashioned sulky was so full of books as to give barely room for him to sit and to carry a well-stored hamper of fruits and sweetmeats, of which he was immoderately fond. Rousseau tells us that he composed in bed at night, or else out of doors while walking, carefully recording his ideas in his brain, arranging and turning them many times until they satisfied him, and then he committed them to paper perfected. He said it was in vain for him to attempt to compose at a table surrounded by books and all the usual accessories of an author. Irving wrote most of the Stout Gentleman mounted on a stile at Stratford-on-Avon, while his friend Leslie, the painter, was engaged in taking sketches of the interesting locality. Jane Taylor, the English poetess and prose writer, began to produce creditable work at a very early age, and used at first to compose tales and dramas while whipping a top, committing them to paper at the close of that somewhat trivial exercise. As she grew older she said that she could find mental inspiration only from outdoor exercise.
- 67. Petavius, the learned Jesuit, when composing his Theologica Dogmata and other works, would leave his table and pen at the end of every other hour to twirl his chair, first with one hand, then with the other, for ten minutes, by way of exercise. Cardinal Richelieu resorted to jumping in his garden, and in bad weather leaped over the chairs and tables indoors,—an exercise which seemed to have a special charm for him. Samuel Clark, the English philosopher and mathematician, adopted Richelieu's plan of exercise when tired of continuous writing. Pope says, with regard to exercise, I, like a poor squirrel, am continually in motion, indeed, but it is only a cage of three feet: my little excursions are like those of a shopkeeper, who walks every day a mile or two before his own door, but minds his business all the while. We are told that Douglas Jerrold, when engaged in preparing literary matter, used to walk back and forth before his desk, talking wildly to himself, occasionally stopping to note down his thoughts. Sometimes he would burst forth in boisterous laughter when he hit upon a droll idea. He was always extremely restless, would pass out of the house into the garden and stroll about, carelessly picking leaves from the trees and chewing them; then suddenly hastening back to his desk, he recorded any thoughts or sentences which had formed themselves in his mind. Jerrold wrote so fine a hand, forming his letters so minutely, that his manuscript was hardly legible to those not accustomed to it. He was very fastidious about his writing-desk, permitting nothing upon it except pen, ink, and paper. Like most persons who habitually resort to stimulants, he could not be content with a single glass of spirits or wine, but consumed many, until he was only too often unfitted for mental labor. Jerrold's wit was of a coarser texture than that of Sheridan, but, unlike his, it came with spontaneous force; it was always ready, though it had not the polish which premeditation is able to impart. Oftentimes his wit was severely sarcastic, but as a rule it was only genial and mirth- provoking.
- 68. It was asked in Jerrold's club, on a certain occasion, what was the best definition of dogmatism. There is but one, he instantly replied,—the maturity of puppyism. A member remarked one day that the business of a mutual acquaintance was going to the devil. All right, said Jerrold; then he's sure to get it back again. Another member who was not very popular with the club, hearing a certain melody spoken of, said, That always carries me away when I hear it. Cannot some one whistle it? asked Jerrold. Another member, who was rather given to boasting, said: Very singular! I dined at the Marchioness of So-and-so's last week, and we actually had no fish. Easily explained, said Jerrold; no doubt they had eaten it all upstairs. When Heraud, a somewhat bombastic versifier, asked him if he had read his Descent into Hell, Jerrold instantly replied, No; I had rather see it. Being asked what was the idea of Harriet Martineau's rather atheistical book, he answered that it was plain enough,—There is no God, and Harriet is his Prophet. This is even better than the remark of another wit who, when asked what was the outcome of a meeting before which three of the ablest and most dogmatic Positivists in England made speeches, replied that the result arrived at was this: that there were three persons and no God. Jerrold could not confine himself to any regular system of work, but drove the quill at such times and only to such purpose as his erratic mood indicated, jumping from one subject to another like one crossing a brook upon stepping-stones. This, however, was a habit by no means peculiar to Douglas Jerrold. There are some ludicrous stories told of him; like that of his being pursued by a printer's boy about the town, from house to club, from club to the theatre, and so on, and finally of his being overtaken, getting into a corner and writing an admirable article with pencil and paper on the top of his hat. Agassiz,[79] the great Swiss naturalist, who became an adopted and honored son of this country, was singularly unmethodical in his habits of professional labor. If he was suddenly seized with an interest in some scientific inquiry, he would pursue it at once, putting by all present work, though it might be that he had just got fairly
- 69. started in another direction. I always like to take advantage, he would say, of my productive moods. The rule that we must finish one thing before we begin another, had no force with him. An individual connected with the lyceum of a neighboring city called upon Agassiz to induce him to lecture on a certain occasion, but was courteously informed by the scientist that he could not comply with the request. It will be a great disappointment to our citizens, suggested the caller. I am sorry for that, replied Agassiz. We will cheerfully give you double the usual price, added the agent, if you will accommodate us. Ah, my dear sir, replied the scientist, with that earnest but genial expression so natural to his manly features, I cannot afford to waste time in making money. A very similar habit of composition or study possessed Goldsmith, Coleridge, Wordsworth, Pope, and some others of the poets, who not infrequently laid by a half-constructed composition for two or three years, then finally took up the neglected theme, finished and published it. This unmethodical style of doing things is but one of the many eccentricities of genius. Scott said he never knew a man of much ability who could be perfectly regular in his habits, while he had known many a blockhead who could. Southey and Coleridge were at complete antipodes in regard to regularity of habits and punctuality: the former did everything by rule, the latter nothing. Charles Lamb said of Coleridge, He left forty thousand treatises on metaphysics and divinity, not one of them complete. Neither Agassiz, Coleridge, nor any of similar irregularity in work, is to be imitated in those respects. Had it not been for Agassiz's far-seeing and vigorous powers,—in short, for his great genius, he could never have accomplished his remarkable mission. The deduction which we naturally draw is, that method is a good servant but a bad master. If genius were to be trammelled by system and order, it would suffocate. Perhaps Montaigne was nearly right when he thought that individuals ought sometimes to cross the line of fixed rules, in order to awaken their vigor and keep them from growing musty.
- 70. Coleridge was much addicted to the habit of marginal writing; which, though sadly wasteful on his own part, was very enriching to those friends who loaned him from their libraries.[80] Charles Lamb, who was not inclined to spare book-borrowers as a tribe, had no reflections to cast upon Coleridge for this habit. The depth, weight, and originality of his comments as hastily and carelessly penned on the margins of books were wonderful, and if collected and classified would form several volumes, not only of captivating interest, but of rare critical value, as the few which have been brought together abundantly prove. In one volume which he returned to Lamb is this memorandum: I shall die soon, my dear Charles Lamb, and then you will not be vexed that I have be-scribbled your book. S. T. C., May 2d, 1811. Elia valued these marginal notes beyond price, and said that to lose a volume to Coleridge carried some sense and meaning with it. These critical notes often nearly equalled in quantity of matter the original text. In his article upon the subject, Lamb says, I counsel thee, shut not thy heart nor thy library against S. T. C. As we have already said, while this erratic expenditure of Coleridge's rare literary taste and judgment enriched others, it in a degree impoverished himself; for had the same time and thought been expended upon consecutive literary work, it would have produced volumes of inestimable value to the world at large, and have proved monumental to their author. Byron was addicted to marginalizing; and though he could not equal Coleridge in the profundity of his criticisms, or impart such charming interest to them, still he was quite original and often piquant. Burns contented himself with trifling criticisms of approval or disapproval pencilled in the margin of books, especially poetical ones, which were nearly all he was in the habit of reading. Many famous authors and public men have been extravagantly fond of the rod and line, disciples of that patient and poetical angler, Izaak Walton. George Herbert, the English poet; Henry Wotton, diplomatist and author; Dr. Paley, Archdeacon of Carlisle; John Dryden, poet and dramatist; Sydney Smith, the witty divine; Sir
- 71. Humphry Davy, the eminent chemist,—all were devoted anglers.[81] This brief list might be largely increased. Bulwer-Lytton says: Though no participator in the joys of more vehement sport, I have a pleasure that I cannot reconcile to my abstract notions of the tenderness due to dumb creatures, in the tranquil cruelty of angling. I can only palliate the wanton destructiveness of my amusement by trying to assure myself that my pleasure does not spring from the success of the treachery I practise towards a poor little fish, but rather from that innocent revelry in the luxuriance of summer life which only anglers enjoy to the utmost. Walton puts himself on record in these words: We may say of angling, as Dr. Boteler said of strawberries: 'Doubtless God could have made a better berry, but doubtless God never did;' and so, if I might be judge, God never did make a more calm, quiet, innocent recreation than angling. Sydney Smith declared it to be an occupation fit for a bishop, and that it need in no way interfere with sermon-making. Perhaps the best thing said or done in angling is an unpublished anecdote of the great preacher to the seamen,—the late Father Taylor, of Boston. He was once lured to try his hand at the rod, and soon brought up a very little fish that had been tempted by his bait. He took the small creature carefully from the hook, gazed at it a moment, and then cast it back into the water, with this advice: My little friend, go and tell your mother that you have seen a ghost! Dr. Parr, the profound English scholar, was a most inveterate smoker; so was Charles Lamb,[82] who one day said to his doctor, I have acquired this habit by toiling over it, as some men toil after virtue. Robert Hall, the popular English divine, was very much addicted to tobacco and other stimulants. A friend who found him in his study blowing forth clouds of smoke from his lips, said, There you are, at your old idol! Yes, replied the divine, burning it. Napoleon could never abide smoking tobacco; yet observing how much other men seemed to enjoy it, he tried to acquire the habit, but finally gave it up in disgust. He, however, took snuff to excess. Sir Walter Scott
- 72. was very fond of smoking. Thackeray, like Burns, loved to get away by himself and enjoy the flavor of a rank tobacco-pipe. Carlyle, like Tennyson, did not care for a cigar, but kept a pipe in his mouth most of his waking hours. Bulwer-Lytton was a ceaseless smoker; and there are few if any notable Germans who have not been addicted to the same indulgence. The nicotine produced from tobacco is one of the most deadly of all poisons, as has been proven by some startling experiments in the Paris hospitals.[83] Thackeray said there was good eating in Scott's novels. Extending the remark, it might be added that there was good drinking in those of Dickens, and good smoking in those of Thackeray. Dean Swift relieved his sombre moods by harnessing his servants with cords and driving them, school-boy fashion, up and down the stairs and through the garden of the deanery of St. Patrick's Cathedral, Dublin. Dickens was controlled by a nervous activity which made him crave physical exercise of some sort, and he daily found relief in an eight or ten mile walk. Thackeray once told the author of these pages that he preferred to take his exercise driving upon very easy roads. When Dickens was in this country he was frequently accompanied in his long walks by the late James T. Fields, who was ever ready to sacrifice himself to the pleasure of others. Mr. Fields was not partial to extreme pedestrian exercise, and the author of the Pickwick Papers tested his good-nature to the verge of exhaustion in this respect. Dumas, when not otherwise engaged, was accustomed to go down into his kitchen, and, deposing the servants, cook his own dinner; and an excellent cook he must have been, if one half the stories rife about him be true. Besides, did he not write an original cook-book, which still stands for good authority in the cafés of the boulevards? Dr. Warton, the English critic and author, as represented by contemporary authority, was noted for a love of vulgar society, which he daily sought in low tap-rooms and gin-shops, where he joked away the evening hours. Turner the painter had similar tastes and habits, though he was of a reserved and unsociable character, and
- 73. noted for his parsimony. Shelley, Goldsmith, and Macaulay delighted in the company of young children. They are so near to God, said Shelley. Intercourse with them freshens and rejuvenates one's soul, wrote Macaulay. I love these little people; and it is not a small thing when they, who are so fresh from God, love us, said Dickens. Children always had a most tender and humanizing effect upon Douglas Jerrold, no matter what was his mood. He writes: A creature undefiled by the taint of the world, unvexed by its injustice, unwearied by its hollow pleasures; a being fresh from the source of light, with something of its universal lustre in it. If childhood be this, how holy the duty to see that in its onward growth it shall be no other! History tells us that Henry of Navarre, who was every inch a king, was often seen upon his palace floor with two of his children upon his back, playing elephant and rider. What a peep into the king's heart we get by this little picture of his domestic life! Where was all the monarch's pride of State, his kingly dignity? How hard it is to hide the sparks of nature! It is related of Epictetus that he would steal away from his philosophical associates to pass an hour romping with a group of children,—to prattle, to creep, and to play with them. Charles Robert Maturin, the poet, author of the tragedy of Bertram, and other successful dramas, could not endure to have children near him during his hours of literary composition. At such times he was particularly sensitive, and pasted a wafer on his forehead as a token to the members of his family that he was not to be interrupted. He said if he lost the thread of his ideas even for a moment, they were gone from him altogether. Sir Walter Scott, on the contrary, was ever ready to lay down his pen at any moment, to exchange pleasant words with child or adult, friend or stranger; and it was notorious that children could always interrupt him with impunity. He declared that their childish accents made his heart dance with glee. He could not check their confidence and simplicity, though pressed upon him when his thoughts were soaring in poetic flights or describing vivid scenes of warfare and carnage. Scott preserved considerable system, nevertheless, in his composition and
- 74. labor. He lay awake, he tells us, for a brief period in the quiet of the early morning, and arranged carefully in his mind the work of the coming day. He laid out systematically the subject upon which he was writing, and resolved in what manner he would treat it. Thus it was that he could lay down his pen at any moment without deranging the purpose of the work. He had one axiom to which he tenaciously adhered, and was often heard to repeat it to his dependants and friends: Do whatever is to be done, at once; take the hours of reflection or recreation after business, and never before it. Schiller said that children made him half glad and half sorry,—always inclined to moralize. Happy child, he exclaims, the cradle is still to thee a vast space: become a man, and the boundless world will be too small for thee. Goethe was ever watchful, loving, and tender with the young. Children, he says, like dogs, have so sharp and fine a scent, that they detect and hunt out everything. He thought their innocent delusions should be held sacred. Elihu Burritt, the Learned Blacksmith, says that he once congratulated an humble farmer upon having a fine group of sons. Yes, they are good boys, was the father's answer. I talk to them often, but I do not beat my children,—the world will beat them by and by, if they live. A fine thought, rudely expressed. Shelley's interest in children was connected with his half belief in the Platonic doctrine of pre-existence. As he was passing over one of the great London bridges, meditating on the mystery, he saw a poor working-woman with a child a few months old in her arms. Here was an opportunity to bring the theory to a decisive test: and in his impulsive way he took the infant from its astonished mother, and in his shrill voice began to ask it questions as to the world from which it had so recently come. The child screamed, the indignant parent called for the police to rescue her baby from the philosophical kidnapper; and as Shelley reluctantly delivered the infant to its mother's arms, he muttered, as he passed on, How strange it is that these little creatures should be so provokingly reticent! Shelley
- 75. was a child himself in many respects; in illustration of which the reader has only to recall the poet's singular amusement of sailing paper boats whenever he found himself conveniently near a pond. So long as the paper which he chanced to have about him lasted, he remained riveted to the spot. First he would use the cover of letters, next letters of little value; but he could not resist the temptation, finally, of employing for the purpose the letters of his most valued correspondents. He always carried a book in his pocket, but the fly- leaves were all consumed in forming these paper boats and setting them adrift to constitute a miniature fleet. Once he found himself on the banks of the Serpentine River without paper of any sort except a ten-pound note. He refrained for a while; but presently it was rapidly twisted into a boat by his skilful fingers, and devoted to his boat- sailing purpose without further delay. Its progress being watched, it was finally picked up on the opposite shore of the river and returned to the owner for more legitimate use. Charles Lamb in his quaint way says: I know that sweet children are the sweetest things in nature, not even excepting the delicate creatures which bear them; but the prettier the kind of a thing is, the more desirable it is that it should be pretty of its kind. One daisy differs not much from another in glory; but a violet should look and smell the daintiest.[84] Good and substantial food is quite as necessary to authors and public men, as to those who gain their livelihood by laborious physical employment. Authors are, however, as a rule, rather inclined to free indulgence at table. There is as much intemperance in eating as in drinking. Tom Moore, who was the best diner-out of his day, said, by way of excusing this habit, In grief, I have always found eating a wonderful relief. N. P. Willis was quite a gourmand. There are, he once wrote, so few invalids untemptable by those deadly domestic enemies, sweetmeats, pastry, and gravies, that the usual civilities at a meal are very like being politely assisted to the grave. It is certainly better to punish our appetites than to be punished by them. Dickens and Thackeray were both inclined to free indulgence
- 76. at the table, the former being struck with death at a public banquet. Dean Swift often gave better advice than he was himself inclined to follow. He says: Temperance, meaning both in eating and drinking, is a necessary virtue to great men, since it is the parent of the mind, which philosophy allows to be one of the greatest felicities in life. Macready, the famous English tragedian, would not touch food of any kind for some hours before making one of his grand dramatic efforts, but drank freely of strong tea before appearing in public,—a subtle stimulant in which the late Rufus Choate freely indulged, particularly before addressing a jury. Abstinence in diet was a special virtue with Milton. Shelley utterly despised the pleasures of the table. Walter Scott was an abstemious eater. Pope was a great epicure, and so was the poet Gay. Speaking of appetite, Coleridge tells us of a man he once saw at a dinner- table, who struck him as remarkable for his dignity and wise face. The awful charm of his manner was not broken until the muffins appeared, and then the wise one exclaimed, Them's the jockeys for me! Dignity is sometimes very rudely unmasked, and an imposing air is nearly always the cloak of a fool. Newton lived on the simplest food. If Aristotle could diet on acorns, he said, so can I; and before sitting down to study he exercised freely and abstained from food. Dr. George Fordyce, the eminent Scotch physician, ate but one meal a day, saying that if one meal in twenty-four hours was enough for a lion, it was sufficient for a man; but in order not to be like the lion, he drank a bottle of port, half a pint of brandy, and a pitcher of ale with his one meal. Lamartine used to pass one day in ten fasting, as he said, to clear both stomach and brain. Aristo, the stoic philosopher, used to fast for days on acorns. Thomas Byron, a well- known author, never ate flesh of any sort. Dryden's favorite dish was a chine of bacon. Charles Lamb was enamoured of roast pig. He said, You can no more improve sucking pig than you can refine a violet! Keats was a very fastidious eater, but was fond of the table, especially where there was good wine,[85] and yet he was not addicted to its intemperate use. Dr. Johnson was greedy over boiled mutton; and Dr. Rhondelet, the famous writer on fishes, was so fond
- 77. of figs that he died from having at one time eaten immoderately of them. Barrow, one of the greatest of English theologians and mathematicians, is said to have died of a surfeit of pears,—a fruit of which he was extravagantly fond. Gastronomic appetite and reason have been compared to two buckets in a well; when one is at the top the other is at the bottom. Byron nearly starved himself to prevent growing gross and uninteresting in physical aspect. Addison was addicted to port and claret, and was accustomed, as already spoken of, while meditating a moral or political essay, to pace up and down the long gallery of Holland House.[86] When a humorous suggestion occurred to his fertile fancy, he solaced himself with claret; or fortified himself with a glass of port when a moral sentiment required to be enforced by an impressive close to a beautifully constructed sentence.[87] This was after his frigid marriage to the Dowager Countess of Warwick. On his death-bed he is reported to have said to her graceless son, See how a Christian can die! Probably the profligate youth, spying his father-in-law as he walked in the gallery, might have irreverently remarked: See how a Christian can drink! But the truth is that Addison, judged by the habits of his time, should be considered a moderate drinker. Poe's nerves were so shattered that a slight amount of wine would intoxicate him into a frenzy of dissipation; the same amount swallowed by a regular toper would hardly disturb his brain at all. While Pitt was quite a young man, he was so weakly that his physician ordered him to drink freely of port wine, and he thus contracted the habit of depending upon stimulants, and could not do without them. Lord Greville tells us he has seen him swallow a bottle of port wine by tumblerfuls before going to the House. This, together with the habit of late suppers, helped materially to shorten his life.[88] Goldsmith had a queer fancy for sassafras tea, from which he imagined he derived an excellent tonic effect. Such a relish had certainly one element to recommend it,—and that was its harmlessness. Dr. Shaw, the English naturalist, nearly killed himself
- 78. by drinking green tea to excess. Haydn partook immoderately of strong coffee, and kept it brewing by his side while he composed. Burns lived on whiskey for weeks together, supplemented by tobacco, which caused Byron to say that he was a strange compound of dirt and deity. Aristippus of old lived up to his own motto; namely, Good cheer is no hindrance to a good life. Few men reason about their appetites, but they give way to them until disease reminds them they are made of mortal stuff. Even Plutarch used to indulge at times in riotous living, saying, You cannot reason with the belly; it has no ears. Addison has pithily recorded his own ideas of this matter. When I behold a fashionable table set out in all its magnificence, he says, I fancy that I see gouts and dropsies, fevers and lethargies, with other innumerable distempers, lying in ambuscade among the dishes. Nature delights in the most plain and simple diet. Every animal but man keeps to one dish. Herbs are the food of this species, fish of that, and flesh of a third. Man falls upon everything that comes in his way; not the smallest fruit or excrescence of the earth, scarce a berry or a mushroom, can escape him. It is among the easiest of all things to outsit both our health and our pleasure at the table. The pleasures of the palate, said shrewd old Seneca, deal with us like Egyptian thieves, who strangle those whom they embrace. Thackeray said towards the close of his life, that his physicians warned him habitually not to do what he habitually did. They tell me that I should not drink wine, and somehow I drink wine; that I should not eat this or that, and, guided by my appetite for this or that, I disregard the warning. Eminent men are not unlike the rest of humanity in a desire for some sort of recreation, and each one finds it after his own natural bent or fancy. Literature is capable of affording the most rational and lasting enjoyment to cultured minds, but physical exercise has also its reasonable demands. The late Victor Emmanuel found recreation only in hunting, having a number of lodges devoted to this purpose in different parts of Italy. McMahon, late President of France, was
- 79. Welcome to our website – the perfect destination for book lovers and knowledge seekers. We believe that every book holds a new world, offering opportunities for learning, discovery, and personal growth. That’s why we are dedicated to bringing you a diverse collection of books, ranging from classic literature and specialized publications to self-development guides and children's books. More than just a book-buying platform, we strive to be a bridge connecting you with timeless cultural and intellectual values. With an elegant, user-friendly interface and a smart search system, you can quickly find the books that best suit your interests. Additionally, our special promotions and home delivery services help you save time and fully enjoy the joy of reading. Join us on a journey of knowledge exploration, passion nurturing, and personal growth every day! ebookbell.com