GC lab report

Schnell
1

Experiment 6
Gas Chromatography
Margaret Schnell
maschnell
00784786
Partner: Sanja Tresnjic
March 22-24, 2016
April 7, 2016
Analytical Chemistry
CHEM:3430:0A01
Prof. Scott Shaw

Schnell
2

Introduction:
Gas chromatography is an analytical separation technique that analyzes compounds
eluted from a column, which carried through the column by a gas mobile phase. The carrier gas
in our experiment was helium. We used helium gas because it has a thermal conductivity that is
higher than any of the analytes we used. A detector is located at the end of the column in gas
chromatography to measure the solutes that are eluted from it. The detector used in our
experiment was a thermal conductivity detector (TCD). TCDs work by passing the eluted analyte
over a heated filament. The thermal conductivity of the gas stream decreases as the temperature
of the filament increases. These changes also change the resistance of the filament; this change
causes a change in the voltage, which is the measured response for this detector. This signal is
proportional to the sample concentrations.
To obtain an ideal chromatogram, there are many parameters that need to be adjusted.
Ideal parameters give chromatograms with the high resolutions in the short amounts of time.
There are many parameters that can affect both the elution time and the resolution, but for this
experiment we focused on the effects of temperature and pressure. In gas chromatography, an
increase in temperature is only favored up to a certain point. The increased temperature
decreases the time of elution for all of the analytes, but after a certain point, it causes a decrease
in the resolution. Pressure also needs to be controlled in order to obtain a favorable
chromatogram. Higher pressures cause backflow which causes elution time to increase. Since the
temperature and pressure are dependent on each other, and it would take too long to figure out
mathematically, these parameters were optimized in our experiment with the Simplex
Optimization software. The Simplex Optimization software can only be used if the
parameters being optimized are dependent on each other. This software used data from test runs

Schnell
3

to calculate what it thought to be a good set of parameters. The results from those parameters
were then reentered into the system to make a better measurement. In our experiment, we
performed as many trials as possible in the time allotted. In the end, we used the parameters that
gave a chromatogram that looked to have the highest resolution in the shortest amount of time.
Experimental Methods:
The CRF values were only recorded for the first three parameters plugged into the
Simplex Optimization. We did this because were not given CRF values for the optimized trials.
We choose the parameters based on looking at the chromatograms from the GC, and not on the
CRF values obtained. Manipulations and calculations were also made to the refractive indices
obtained experimentally. We did this because we needed to correct for the temperature difference
of 23 °C (experimental) and 20 °C (literature values).
The average peak area of ethanol and the internal standard, 1-propanol, were used to
obtain the ratio of ethanol peak area to 1-propanol peak area for each ethanol (v/v) %. A
regression line was included in the calibration graph to show that the results were linear and
reliable to help determine the ethanol (v/v) % in our vodka sample.
Data and Results:
The first part of this experiment was done to determine the optimal conditions for the GC
to run the samples at. These
conditions were determined using
Simplex Optimization. The first
three sets of conditions were set to 30
°C with 30 psi, 50 °C with 20 psi,
and 80 °C with 10 psi. The results from these runs and their respective CRF values are displayed

Table 1 This table shows the results from the Simplex Optimization of the
original three sets of parameters. These were used to give the optimization
program sets of data to start choosing the values that would give the optimal
results.
Original Parameters
Trial
Temperature
(°C)
Pressure
(psi) CRF
1 30 30 1.1273
2 50 20 0.8344
3 80 10 0.9747

Schnell
4

in Table 1. From the table, we can see that the pair in the mid-range had the lowest CRF value.
The points that the Simplex Optimization suggested were closest to the mid-range values for
both the temperature and
pressure. The final set of
parameters we decided to
run the samples at was 64
°C and 15.6 psi. The
entire set of parameters
and their times to elute
are recorded in Appendix
B, Calculation 6. Figure 1
is a graphical
representation of the
progression for the measurements made at the different parameters that the Simplex
Optimization program suggested. The points closer to the center of the graph belong to better
sets of parameters. Trials 4-9 are all optimization parameters chosen by the Simplex program.
Trial 5 was chosen as the set of parameters with the best resolution in the shortest amount of
time.

After we chose our optimal set of parameters, we
ran known standard solutions of pure ethanol, 1-propanol,
and water in the GC. The results obtained are tabulated in
Table 2. These were run to be sure we knew which peaks Table 2 The table shows times at which the
three standards of the compounds we are
testing are eluted. These are controls so we
know which peak corresponds to which
compound in our mixture samples.
Known Standards
Standard
Retention Time
(min)
Pure EtOH 1.658
1-propanol 2.248
Water 2.536
10
14
18
22
26
30
30
40
50
60
70
80

Pressure(psi)
Temperature
(Celsius)

Simplex Optimization Trials Trial
1

Trial
2

Trial
3

Trial
4

Trial
5

Trial
6

Trial
7

Trial
8

Trial
9

Figure 1 This graph follows original parameters run then the parameters chosen by the
Simplex Optimization program. The points closer to the center of the graph are the better
parameter settings. The parameters shown by the marker for Trial 5 was selected as the best,
and the parameters were used for the remainder of the solutions to be tested.

Schnell
5

in the chromatograms corresponded to which compounds in the mixture. It is important to note
that these are just estimates and the actual retention times vary with each solution and each
individual run.
Next, we ran the four mixtures of water, ethanol, and 1-propanol and the vodka sample in
a refractometer. This was done to be sure we obtained the exact ethanol (v/v) % of the solutions.
The data was collected and manipulated, and the results are recorded in Table 3. The refractive
indices were corrected from 23 °C to 20 °C. The corrected ethanol (v/v) % allowed us to make a
more precise and accurate calibration curve for the known samples.
After the refractive indices were
measured, the samples were ready to be run
at the optimal parameters in the GC.
Mixtures of 0.6 uL of each ethanol (v/v) %
standard with 0.6 uL of the 40% (v/v) 1-
propanol standard were tested to make a
calibration curve. The curve obtained is
Sample
Refractive
Index
Measured
Refractive
Index
Corrected
Mass (g)/g
Solution
Density
Density
Pure EtOH
Actual
EtOH
(v/v)%
20% (v/v)
EtOH
1.3428 1.3440 0.150 0.977 0.7893 18.6%
30% (v/v)
EtOH
1.3481 1.3493 0.238 0.963 0.7893 29.0%
40% (v/v)
EtOH
1.3531 1.3543 0.320 0.950 0.7893 38.5%
50% (v/v)
EtOH
1.3569 1.3581 0.383 0.940 0.7893 45.6%
Table 3 The column of refractive indices measured is the data that was obtained from the refractometer. The measurements
were made at 23 °C, and so had to be corrected for a temperature of 20 °C. Once this was done, a series of calculations was
made to determine the actual ethanol (v/v) %.
y = 0.0252x + 0.0542
R² = 0.98554
0.5
0.75
1
1.25
1.5
15 25 35 45 55
PeakAreaEtOH:PeakArea1-
propanol
EtOH (v/v) %
Peak Area EtOH:Peak Area 1-
propanol vs. EtOH (v/v)%
Figure 2 (Left) This is a graph of the calibration curve of the
solutions with known EtOH (v/v) %. The R2
value indicates that the
curve is a reliable measurement, and that it is safe to determine the
EtOH (v/v) % of the vodka with this curve.

Schnell
6

shown in Figure 2, and was used to determine the exact ethanol (v/v) % in our vodka solution.
The graph was corrected for the new ethanol (v/v) % that were determined by the corrected
refractive indices. The slope of the graph is 0.0252 and an R2
value of 0.98554 was obtained.
The ethanol (v/v) % in the vodka sample was determined to be 42.0% from the calibration curve
with an error of 6%.
Discussion:
The temperature and pressure are dependent on each other and both affect the time of
elution of the sample. The Simplex Optimization program was run to help us in determining the
optimal parameters to get the best resolution between components in the fastest time. The
original parameters (Table 1) were run to give the program starting points to determine optimal
parameters. The lowest CRF value is generally correlated to the optimal conditions, but this is
not always true. The middle to high temperature and the middle to low pressures of the original
parameters gave the lowest CRF values. Because of this, we assume that the optimal parameters
will fall around and between these sets of parameters.
After plugging in the original values, the program gave us the first set of parameters to
use on the GC. After we ran the GC with the suggested set of parameters we reported the values
obtained back into the Simplex Optimization program and it would give us a new set of
parameters, and the process was repeated again as many times as possible (in our case, six
times). We looked at each chromatogram to find which we thought had the best resolution (and
most narrow bandwidths) in the shortest amount of time. We decided to run the rest of our
samples at 64 °C and pressure of 15.6 psi. These parameters gave us a resolution of 3.06 in 2.83
minutes. We choose the parameters based on how the chromatograms looked and not actually
calculating the resolutions for the sake of time. Due to poor injections, some of the peaks were

Schnell
7

broken up into multiple peaks, and so, did not look as good as they actually were. Our
parameters were good, although they may not have been the best possible, and they still allowed
us to run our samples while collecting good data.
As mentioned in the data and results section, the known standards were then run at the set
parameters. These served as guides for the chromatograms of the mixtures of the compounds for
the remainder of the experiment. The time of elution of each compound in the chromatograms
would change slightly depending on the mixture, but overall, these were good estimates of where
to look for each compound.
The four different ethanol (v/v) % mixtures were first tested in the refractometer. This
step was essential because it enabled us to find the exact ethanol (v/v) % in each of the solutions.
With small changes in temperature, comes a change in the refractive index. Due to these
changes, it is important to correct the refractive index to a temperature of 20 °C, which is the
temperature at which literature values of densities and mass percentages of ethanol are obtained.
All refractive indices were experimentally measured at 23 °C. The corrected refractive indices all
had similar, but slightly lower, ethanol (v/v) % to what we expected them to be.
The utilization of three different calibration curves enabled us to find the exact ethanol
(v/v) % in the vodka sample. The vodka sample could not be measured in the refractometer
because it was unknown if there were any other analytes in the sample that would interfere with
the instrument. The refractometer was used to obtain a refractive index at 23 °C. Because
literature values of ethanol densities are recorded at 20 °C, we needed to correct for the
temperature difference. The exact mass percentages for each corrected refractive index was
found by using the slope equation in Appendix A, Calibration Curve 1. The exact mass
percentages were then converted to exact densities (g/cm) by using the slope equation in

Schnell
8

Appendix A, Calibration Curve 2. The exact densities and mass percentages were then used to
make the calibration curve in Figure 2. These compared the exact mass percentages to the peak
area of ethanol: peak area of 1-propanol. This calibration curve contains a very good R2
value
that was close to one (perfect). Because of the good R2
value, we can use it as a reliable
calibration curve within the calculated error.
The ethanol (v/v) % in the vodka sample was quoted to be 40% on the Hawkeye bottle.
The ethanol (v/v) % we found from our experiments and related calibration curves was 42%
(v/v). The error in our calibration curve was +/- 6%. The value quoted on the bottle is within the
error range, meaning that the value obtained was accurate. The relative standard deviation
between the quoted and the experimental values was 5%. Since the values fall within this range,
they are precise.
Conclusion:
Overall, it is important that parameters that are dependent on each other are optimized to
the best of our ability. Software’s, such as Simplex Optimizations, make finding these
parameters much easier. Finding the corrected refractive index was an important part of this
experiment because, without doing so, our calibration curve in Figure 2 would have been off.
Overall, the final goal was to determine the ethanol (v/v) % in our vodka sample. The value we
obtained experimentally and mathematically was 42%. This value proved to be both accurate and
precise with respect to the errors associated with it.

Schnell
9

References
1. Dorman, F. L., Whiting, J. J., Cochran, J. W., and Gardea-Torresdey, J. (2012) Gas
Chromatography. Analytical Chemistry Anal. Chem. 12, 4775–4785.
2. Mostafa, A., Edwards, M., and Gorecki, T. (2012) Optimization aspects of comprehensive
two-dimensional gas chromatography. Journal of Chromatography A 1255, 38–55.

Schnell
10

Appendix A: Supplemental Information
Calibration Curve 1 was used to determine the actual mass percent with a known corrected value
for the refractive index (or x value to be plugged into the slope equation).
Calibration Curve 2 was used to determine the actual densities of the solutions with mass
percentages found from Calibration Curve 2 (or x value to be plugged into the slope equation).
y = 1651.5x - 2204.6
R² = 0.99135
16
20
24
28
32
36
40
1.344 1.346 1.348 1.35 1.352 1.354 1.356 1.358
Mass%
Refractive Index
Calibration Plot to Determine Mass Percent
y = -0.0016x + 1.001
R² = 0.99454
0.935
0.94
0.945
0.95
0.955
0.96
0.965
0.97
0.975
16 20 24 28 32 36 40
Density(g/cm)
Mass %
Calibration for Determination of Ethanol Solutions

Schnell
11

Appendix B: Calculations
1. The table below shows the measured refractive indices and the corrected refractive indices
based on the temperature they were measured at. These were corrected by using the equation:
[Corrected RI= Measured RI - (0.0004/°C)
x
(temperature
of
measurement
-‐
20°C)]
Refractive
Indices
of
Samples
(@
23
°C)

Sample

Refractive
Index

Measured

Refractive
Index

Corrected

2%
v/v

EtOH
1.3428
1.344

3%
v/v

EtOH
1.3481
1.3493

4%
v/v

EtOH
1.3531
1.3543

5%
v/v

EtOH
1.3569
1.3581

Vodka
1.3536
1.3548

2. and 3. These are both included under the discussion section in Figure 2.
4. The percentage obtained from the calibration curve was:
slope of the calibration= (peak area ratio)/(EtOH (v/v)%)
0.0252= (1.11)/(EtOH (v/v)%)
EtOH (v/v)% = (1.11)/(0.0252) = 44%
The error in this measurement was:

Schnell
12

The RSD% = [(42-40)/40] = 5%
5. Figure 1 under discussion section contains the plot that traces the progress of the optimization.
6.
Simplex Optimization Trials
Trial
Temperature
(°C)
Pressure
(psi)
Elution Time
(min)
1 48 22.5 2.818
2 64 15.6 2.5
3 62 18.1 2.218
4 68 17.2 2.043
5 57 19.5 2.381
6 60 21 2.056

Schnell
13

7. The number of plates and the plate heights calculated for each sample are recorded below.
The length of the column in this experiment was 25 meters, but 25,000 millimeters is used to
find plate height.
The plate number was found with: (16*Retention Time EtOH2
)/(Width of EtOH Peak2
)
The plate height was found with: (Length of Column)/(Plate Number)
Sample
Average
Time
(min)
Average
Width
(min.)
Average
Plate
Number
Average
Plate
Height
(mm)
Vodka 1.623 0.16 1645 15
20 % EtOH
(v/v)
1.614 0.17 1443 17
30 % EtOH
(v/v)
1.614 0.17 1442 17
40% EtOH
(v/v)
1.619 0.18 1369 18
50% EtOH
(v/v)
1.619 0.19 1161 22

Schnell
14

Appendix C: Questions
1. Thermal Conductivity Detectors measure how well a substance can transfer heat from a
hot region to a cold region. Helium is the carrier gas used in our experiment because its
thermal conductivity is very high; this helps to lower the conductivity of the gas stream
of our ethanol. In turn, the helium gives the lowest limit of detection for the TCD. When
our ethanol leaves the column, it flows into the TCD and over the hot tungsten-rhenium
filament. This causes the conductivity of the gas stream to decrease, the filament to get
hotter, the electrical resistance to increase, and this then causes the voltage across the
filament to change. The change in voltage is what the detector measures. Some TCD split
the carrier gas into the analytical column and the reference column. The reference column
minimizes the change in flow as the temperature is changed. The resistance of the
analytical column is then measured with respect to the reference column.
2. An internal standard is usually used in gas chromatographic analysis because the quantity
of sample analyzed and the instrument response varies slightly from run to run. Since gas
chromatography flow rates can change with each injection (if not all injections are
performed exactly the same way), the internal standard helps to correct for this.
3. UV-Vis is another experiment that would benefit from using Simplex Optimization to
help choose parameters. The resolution and signal to noise ratio are two ways to judge
how good the absorption is. These need to be balanced, just like resolution and time in
this experiment, to obtain good data. These two are balanced by adjusting slit width and
radiant power. Radiant power is related to the square of slit width.

GC lab report

More Related Content

What's hot (20)

Similar to GC lab report (20)

GC lab report