Nasa Nnc05 Ca90 C Final Report

"fi, sf" RTA P2004#65 NASA Topic No B3.05 CN: NNC05CA09C Phase II Final Report June 2, 2007

Executive Summary

The overall goal of this program (through Phase III) is the development of an analyzer integrated into the
International Space Station toilets capable of detecting key chemicals in urine to monitor and assess astronaut health.
The analyzer will employ a novel metal-doped sol-gel material to simultaneously extract and quantify these key
chemicals using surface-enhanced Raman spectroscopy (SERS). During the Phase I program feasibility was
successfully demonstrated by chemically extracting 3-methylhistidine, a muscle-loss indicator, and Raloxifene, a
bone-loss inhibitor from reconstituted urine and obtaining their SER spectra in 1-mm glass capillaries. During the
Phase II program the following success were achieved:

1) 50 biochemicals (12 biomarkers, 13 drugs, and 25 potential interfering urine components) were
successfully measured using these SERS-active capillaries.
2) The biomarkers and drugs were successfully detected at physiological concentrations (1 mg/L, and 0.01
mg/L, respectively).
3) Prototype micro-fluidic chips were fabricated that contained separation and the SERS-active materials.
4) 3-methylhistidine and Risedronate were separated from a real urine sample and detected at these
required concentrations using the chip shown below (Figure E.1).
5) A design is offered that can be integrated into Hamilton Sundstrand’s ISS toilet that would allow automated
extraction, measurement and analysis of each astronaut’s urine every day to monitor and assess health.
6) The success of this program has broad commercial value, considering that 55% of the US population over
age 50 suffer from osteoporosis, and the micro-fluidic chips developed here can identify biomarkers
indicative of bone loss as much as 1 year prior to standard X-ray measurements. This would allow
administering bisphosphonate-based drugs to potentially avert fractures, which represent a $17 billion
annual cost to the US healthcare system.

A C
3 2 B
a

b

4 1 c
d d 5

6 Top
Views

D E

350 500 750 1000 1250 1500 1850 350 500 750 1000 1250 1500 1850
Raman Shift, cm-1 Raman Shift, cm-1
Figure E.1. Photographs of A) 1) urine sample, 2) syringe, 3) filter, 4) ion-retardation capillary, 5) transfer vial, 6)
lab-on-a-chip containing a) flow-control syringe, b) ports, c) ion exchange resin loaded channel, and d) SERS-
active sol-gel loaded channels. B) Chip mounted on XY plate reader. C) RTA Raman Analyzer. SERS of D) 1
mg/L 3-methyl histidine and E) 0.01 mg/L Risedronate extracted from actual urine sample using this apparatus
and detected in the d) channels. Complete analysis takes less than 10 minutes. This figure is proprietary.

2


" On-Demand Urine Analyzer "

Part 1: Table of Contents
Project Summary ..........................................................................................................................................................1
Executive Summary.......................................................................................................................................................2
Part 1: Table of Contents ..............................................................................................................................................3
Part 2: Identification and Significance of the Innovation .............................................................................................3
Part 3: Technical Objectives.......................................................................................................................................11
Part 4: Work Plan (Phase II Results) ..........................................................................................................................13
Part 5: Potential Applications .....................................................................................................................................53
Part 6: Contacts...........................................................................................................................................................53
Part 7: Future Technical Activities .............................................................................................................................55
Part 8: Potential Customer and Commercialization Activities ...................................................................................55
Part 9: Resources Status .............................................................................................................................................57
Part 10: References.....................................................................................................................................................58

Part 2: Identification and Significance of the Innovation
This part is reproduced verbatim from the Phase I proposal.
This Small Business Innovation Research program will develop a novel surface-enhanced Raman (SER) sensor that
will perform real-time chemical analysis of urine. It will provide key physiologic information to monitor astronaut
health and indicate appropriate preventative treatment. The Phase I program will demonstrate feasibility by
establishing the ability of sol-gel chemistry to both select key chemicals: amino acids, biomarkers, drugs, and
metabolites, and enhance their Raman signals. The Phase II program will design and build a prototype “On-
Demand Urine Analyzer” for ground-based measurement. This will include interfacing the SER sensor between a
sampling system and a Raman instrument. The Phase II program will also design a low mass, low power version of
this system (Figure 1) to be used on the International Space Station (ISS) and other vehicles employed during
extended space flight missions (e.g. Mars expedition).
Toilet Data & Power Field Bus
Urine Sample

Separator System
Power
Supply
To Solid Molecules Sol-Gel
Waste Computer in Solution Matrix
(circuit boards) Raman
Filter Scattering

Laser
Fiber
Optics Raman Metal
Spectrometer Particle
Sample Laser
Delivery
Pump Adsorbed
Flow Cell with Molecules
One of four
Block Out Chemically Selective Chemically Selective
To Water SER-Active Discs
Reclaimation Valves SER-Active Discs
System

Figure 1. Block diagram of surface-enhanced Raman based urinalysis system. A pump or flush mechanism draws
the sample from the process through a particle filter into a flow cell and back into the process. Four sol-gel coated
discs in the side wall provide polar-positive, polar-negative, weakly polar-positive and weakly polar-negative
chemical selectivity and SER-activity. The proposed system will weigh 7.6 kg, occupy a 20x12x10 cm space (0.1
cubic foot), and require 20 watts (pump not included). Expanded view illustrates surface-enhanced Raman
scattering from one of the sol-gel coated discs. This Illustration is Confidential and Proprietary.

2.1. The Problem or Opportunity - Extended weightlessness causes numerous deleterious changes in human
physiology, including space motion sickness (SMS), cephalad fluid shifts, reduced immune response, and loss of
bone and muscle mass.1-6 The need to monitor and assess these effects is critical to developing exercise programs or
drug regimes that would maintain astronaut health.7 Many of these physiological changes are reflected in the
chemical composition of urine.8-13 For example, 3-methylhistidine can be used to assess loss of muscle mass,
hydroxyproline to assess bone loss, and uric acid to assess renal stone formation,11 while metabolites can be
analyzed to regulate dosage of anti-SMS drugs.14 According to the National Research Council Space Studies Board,

3


"Analysis of in-flight specimens for markers of bone resorption and formation would offer a unique opportunity to
determine relative efficacy of these various exercise programs."15 The Board further states that current exercise
regimes are ineffective and physiological data is inadequate to properly develop methods to offset the maladies of
weightlessness. Furthermore, these physiological changes also influence metabolism of therapeutic drugs used by
astronauts during space flight. Unfortunately, current earth based analytical laboratory methods that employ liquid
or gas chromatography for separation and fluorescence or mass spectrometry for trace detection are labor intensive,
slow, massive, and not cost-effective for operation in space, regardless of the type of bio-fluid sample analyzed.16-20
Therefore, the ability to assess and rapidly diagnose the status of individual bio-fluid samples “on-demand” is a
critical and necessary component for monitoring the health and future well being of the crew members during
extended flight missions in space.

2.2. The Innovation - We at Real-Time Analyzers (RTA) believe a light-weight, real-time analyzer can be
developed based on surface-enhanced Raman spectroscopy (SERS) to provide detection and identification of several
key chemicals (amino acids, biomarkers, drugs, and metabolites) in urine at relevant concentrations (from ng/mL to
microg/mL) in under 5 minutes. This approach is based on the extreme sensitivity of the SERS technique, which
has been demonstrated by detection of single molecules,21,22 and the ability to identify molecular structure of key
physiological chemicals and drugs through the abundant vibration information provided by Raman spectroscopy.

We proposed this concept in 2001 and received a high score, but not an award. Since that time we have made
several significant advances in our SER-active sol-gel technology, two of which address reviewer’s comments.
First, the sol-gel process is highly reproducible, and we guarantee SER-activity at 20% RSD for our Simple SERS
Sample Vials, now sold commercially for 2 years (Figure 2). Second, we have successfully coated glass capillaries
that are capable of measuring analytes in a flowing solution, reversibly (Figure 3) and capable of performing
chemical separations.23 Additional important aspects of the technology include: third, we have patented the sol-gel
process (NASA sponsored) that incorporates silver and/or gold nanoparticles into a stable porous silica matrix.24,25,26
1008 cm-1 band intensity for BA

45
A
40.0-45.0
40 35.0-40.0
35 30.0-35.0
30 25.0-30.0
25
20.0-25.0
B
20
15.0-20.0
15
10.0-15.0
10
5.0-10.0
5
6 0 0.0-5.0
330

9
300
270
240

Height
210

12
180
150

along
120
90
60

15
vial (mm)
30

15o increments around vial
0

Figure 2. Reproducible SER-intensity response for Figure 3. Reversible SER spectra of 30-sec “plug” of
benzoic acid over entire surface of a Simple SERS benzoic acid flowing through a sol-gel coated capillary.
Sample Vial. Average = 29.1± 4.26 (14.6%) for 240 A) Spectra from time 0 to 2.4 min (bottom to top) and
points (10 sec per point). B) plot of corresponding 1000 cm-1 band intensity.
Spectra are each 8-sec, using 100 mW of 785 nm.

Fourth, virtually all solvents can be used, including aqueous solutions ranging in pH from 2 to 11. No special
reagents or conditions are required. Fifth, we have used these metal-doped sol-gels to measure SER spectra of
several hundred chemicals,27,28,29,30,31,32 with typical detection limits of 1 ng/mL using 100 mW of 785 nm and 3-min
acquisition time. Sixth, we have measured creatinine, lactic acid, uric acid, and actual urine specimens by simply
adding the samples to vials and recording the SER spectra (Figure 4).33-35 Furthermore, in preparation of this
proposal, we measured 1 mg of 3-methylhistidine in 1 mL of water and estimate a current detection limit of 60
microg/mL (0.35 mM, 6 ppm, Figure 5). It is worth noting that most drugs contain nitrogen functionalities
indicative of SER activity, and we have been able to measure drugs and their metabolites (Figure 6).36 Although
physiological measurements require detection limits below 1 microg/mL for biomarkers and as low as 10 ng/mL for
drugs and their metabolites, we believe improvements in sol-gel chemistry and instrumentation would allow

4


Confidential Proprietary Information

achieving these detection limits. Seventh, the choice of metal and alkoxide can be used to develop chemically
selective sensors. We have successfully developed sol-gels that select for polar-positive, polar-negative, weakly
polar-positive, and weakly polar-negative chemical species.37 This will allow discriminative enhancement of the
biomarkers or drugs in favor of urine components. Finally, it is worth noting that RTA has developed an extremely
rugged, compact Raman instrument that employs interferometry for absolute wavelength accuracy and an avalanche
Si detector that improves sensitivity by ~100 times.38 And we are currently developing a hand portable, battery
powered version of the system for the Navy, which will weigh 9.8 kg, occupy 13,500 cc (0.5 cubic foot), require
23.5 W, and will be capable of wireless communication.39

N COOH
H
NH 2
N
relative intensity

R

relative intensity
A
A

B
B
uric acid

Wavenumbers (∆cm-1) Wavenumbers (∆cm-1)
Figure 4. SERS of A) female and B) male urine Figure 5. A) Raman and B) SERS of 3-methylhistidine
specimens. Both samples were pH of 6.5 and diluted (R= CH3). Conditions: Raman, pure solid, 500 mW and 3
to 50% with water. Bands attributed to uric acid occur min scan, SER, 6mM, 100 mW and 3 min. Labeled SER
at 502, 650, 815, 1134, and 1616 cm-1. Conditions: bands have been observed for histidine (R = H) by
120 mW of 1064 nm, 50 scans at 8 cm-1. electrolytic SERS. Inset: molecular structure.

2.3. The Proposal - The overall objective of the proposed program (through Phase III) is the development of an
analyzer integrated into the ISS toilets capable of immediate detection of key chemicals in urine to monitor and
assess astronaut health (complete analysis within 5 minutes of flush). The focus of Phase I will be to establish the
ability of the sol-gel chemistry to both select these key chemicals and enhance their Raman signals. These key
chemicals, listed in Table 1, include: urine products, hormones, muscle loss, bone loss and stone forming indicators
(biomarkers), drugs and their metabolites. This will be achieved in three tasks. The first task will employ
combinatorial chemistry to synthesize 36 libraries of metal-doped sol-gel coated micro-plates varying in alkoxide
composition (Si:Si = 1:0 or 1:1 v/v) to be screened for analyte specific SER activity in Task 2. This will be
Urine Products (g/L):* Muscle Loss Indicators (<mg/L):**
creatinine (1.4) 3-methyl histidine A
glucose (0.1)
glutamic acid (0.3) Bone Loss Indicators(<mg/L):
hippuric acid (0.9) hydroxyproline
hydroxyproline (0.9) deoxypyridinoline
lactic acid (0.2) B
nicotinic acid (0.25) Stone Formation Indicators (<mg/L):
PABA (0.2) calcium oxalate
uric acid (0.2) calcium phosphate
thiamine (0.2) uric acid C
histidine (0.2)
pyridoxamine (0.1) Drugs (~microg/L):
Hormones (g/L): alendronate - for Anti bone loss
pregnanediol (0.9) scopolamine - for Anti motion sickness D
pregnanetriol (2.2) D-penicilamine - for Anti stone formation
raloxifene - for Anti bone loss
estradiol lovastatin - for Anti bone loss

Table 1. Partial list of chemicals in urine, muscle loss,
Wavenumbers (∆cm-1)
bone loss, and stone formation indicators, and
Figure 6. SER spectra of A) amobarbital, B) barbital,
administered drugs.
C) phenobarbital, and D) secobarbital. Conditions: 1
*To be measured in Task 3.
**Chemicals in italics to be measured in Task 2. mg/ml (analyte/methanol) in sol-gel coated sample
vials, 80 mW of 1064 nm, 50 averaged scans.

5


accomplished by varying reactant concentrations delivered to 96 well micro-plates. Reactant variables include eight
Si-alkoxide precursors (tetramethylorthosilicate (TMOS), tetraethylorthosilicate (TEOS), methyltrimethoxysilane
(MTMOS), ethyltrimethoxysilane (ETMOS), methyltriethoxysilane (MTEOS), ethyltriethoxysilane (ETEOS),
aminotrimethoxysilane (ATMOS), and aminotriethoxysilane (ATEOS)), and two types of metal particles (silver or
gold). The second task will screen the sol-gel libraries with key physiological chemicals and drugs for SER activity.
This will be accomplished by measuring the SER spectra of library subsets using the following four standard
chemicals: p-aminobenzoic acid (PABA), aniline (AN), benzoic acid (BA), and phenyl acetylene (PA); and twelve
initial target chemicals: 3-methyl histidine, hydroxyproline, deoxypyridinoline, calcium phosphate, uric acid,
alendronate, scopolamine, dextroamphetamine, raloxifene, lovastatin, acetaminophen, and acetylsalicylic acid. The
third task will demonstrate the ability of the proposed sol-gel SERS plates to discriminatively detect and quantify the
key chemicals in a chemical matrix equivalent to a urine specimen. This will be accomplished by analyzing the
chemicals in simulated urine samples prepared according to clinically defined formulations representing an average
human composition. An initial chemometrics urinalysis model will be developed for identifying, quantifying and
correlating key components in urine of physiological interest with the Raman spectra.

2.4. The Probability of Success - Dr. Frank Inscore, as the Principal Investigator, and Dr. Stuart Farquharson, as
Program Manager, and Dr. John Murren of Yale University as a Consultant have the required expertise to perform
the proposed research. The PI has over eight years experience in designing Raman systems to make difficult and
demanding measurements requiring extreme sample preparation protocols and rigorous optical alignment procedures
for collecting reproducible spectral data. This includes the implementation of continuous wave (CW), pulsed, and
solid-state lasers combined with dispersive instrumentation and multichannel detection employing a charge coupled
device (CCD) for acquiring normal Raman and resonance enhanced Raman spectra of various inorganic-organic
models of related metalloprotein enzymatic systems.40,41,42,43 The PI also has extensive experience in designing and
collecting Raman data for various sampling configurations, sample states and conditions (e.g., in vacuo and at
cryogenic temperatures). The PI also has acquired considerable expertise and experience in the analysis and
application of FT-Raman, SERS and metal-doped sol-gel chemistry at RTA that is relevant to the successful
completion of the proposed project. The PM has the experience and expertise in the analysis and application of FT-
Raman.25,44,45,47 The PM also has considerable experience and expertise in designing new Raman analyzers for
many applications including sensor design for Raman and surface-enhanced Raman spectroscopy applications.24-,46-
52
This includes numerous sampling systems, and several Raman systems, including a state-of-the-art fiber optic FT-
Raman spectrometer.52,53 He has designed, patented, and implemented several fiber optic probes for in-situ remote
monitoring in harsh physical and chemical environments.54 Dr. Farquharson also has extensive experience in
performing and managing large interdisciplinary experimental research projects. He has been the Principal
Investigator or Manager on contracts from DOD, DOE, NASA, NIH and NSF. Dr. John Murren is a Professor at
Yale University School of Medicine (Medical Oncology) and Director of the Lung Cancer Treatment Unit. Dr.
Murren is very active in the evaluation of new chemotherapy drugs and drug combinations used for cancer
treatment. He will provide guidance in our urinalysis experiments (e.g. likely interferents), and understanding of
drug metabolic pathways.55 Finally, we presented a conceptual Urine Analyzer design to Hamilton Sundstrand,
which was very well received. Based on their design review, and our previous working relationship, Hamilton
Sundstrand will support our Phase II research and Phase III commercialization efforts (see support letter).

2.5. Background and Technical Approach - We at Real-Time Analyzers believe that a method based on surface-
enhanced Raman spectroscopy can be developed to provide real-time detection and quantification of several key
chemicals, biochemicals, and metabolites in urine to monitor astronaut health and indicate appropriate preventive
treatment. This is based on our SERS detection of several chemicals, such as creatinine, lactic acid, and uric acid in
urine specimens, the DNA bases, amino acids, including 3-methylhistidine, and numerous drugs and their
metabolites (see Figures 4 – 6).24 A background to this approach follows.

Microgravity and Human Physiology - Extended weightlessness causes numerous deleterious changes in human
physiology, including space motion sickness (SMS), cephalad fluid shifts, reduced immune response, and loss of
bone and muscle mass. The signs of SMS (nausea, dizziness) and fluid shifts (headaches, increased heart rate) are
easily detected, while changes in hormone and bone metabolism are not. Consequently, a more detailed analysis of
astronaut physiology is required to assess these effects. Many of these physiological changes are reflected in the
chemical composition of urine. For example, 3-methylhistidine is a known product of muscle protein breakdown
and is quantitatively excreted in urine,8-10 while the urinary concentration of hydroxyproline and deoxypyridinoline
released during collagen breakdown show promise as indicators of bone turnover.11,12 Renal stone formation,

6


associated with bone loss, can be evaluated by analyzing for calcium oxalate, calcium phosphate, uric acid, citrate,
magnesium ammonium phosphate, and other stone-forming salts. Furthermore, the metabolic products of drugs
administered to relieve SMS (e.g. promethazine and scopolamine),56-58 reduce muscle loss (amino acid infusion59) or
bone loss (alendronate, lovastatin,60 raloxifene61), or renal stone formation (e.g. D-penicilamine) can be analyzed to
regulate dosage and adjust diet. Unfortunately, the approach used in previous missions (shuttle and Mir), astronauts
logging their diet, collecting and sending urine samples back to earth for analysis, would not allow timely preventive
measures. This wait may further jeopardize the health of astronauts performing physical labor associated with the
construction of the International Space Station. In particular, a normal dose of an anti-SMS drug for one astronaut
may metabolize as a high dose for another, causing drowsiness and making tasks difficult to perform and potentially
dangerous.7,62,63,64

Earth based urinalysis employs multiple steps to separate the chemicals and perform the required detection. This
typically includes particulate filtration, pH adjustment, and chromatographic separation (usually high performance
liquid chromatographic, HPLC), prior to introduction into a mass spectrometer (MS). Inclusion of standards
throughout this process is required to ensure measurement accuracy. These methods are labor intensive and time
consuming, and the massive instruments (e.g. MS) are inappropriate for the ISS. It should be noted, however, that
an effort to make this traditional approach to urinalysis practical for the ISS has been undertaken by a research team
at Johns Hopkins University headed by Dr. Potember.18,65 They have developed a rugged time-of-flight mass
spectrometer (TOFMS) to measure biomarkers. Virtues of the TOFMS technologies are that it is small (less than one
cubic ft), lightweight (less than 5 kg), and requires low power (less than 50 watts). To introduce quantitative
samples into the TOFMS, without time consuming chromatographic columns (>30 minutes analysis times), the
Johns Hopkins team has been investigating matrix-assisted laser desorption ionization (MALDI) sampling. With
MALDI sampling, a matrix standard must be used (and supplied), and high-powered ultraviolet pulse lasers are
required. Unfortunately, these lasers are inefficient and typical power requirements are near 100 W.

The ability to monitor and assess the effectiveness of therapeutic agents used by astronauts during space flight is
also problematic. Evidence exists that suggest the therapeutic effectiveness of some drugs, such as
scopolamine/dextroamphetamine (a drug combination used to prevent motion sickness) and acetaminophen (a drug
used frequently for pain relief by astronauts) may change in space. This is reflected by the fact that concentration
levels of such drugs (and hence their pharmacokinetic behavior) measured in bio-fluid samples (blood-plasma,
urine, and saliva) during the course of pre- and post-flight time by typical chromatographic, mass spectrometric and
immuno-assay techniques on earth are not invariant, and that these subsequent changes ultimately depend on
mission length and individual physiological responses to space flight.

According to the National Research Council Space Studies Board "Analysis of in-flight specimens for markers of
bone resorption and formation would offer a unique opportunity to determine relative efficacy of these various
exercise programs."15 The Board further states that the current exercise regimes are ineffective and physiological
data is inadequate to properly develop methods to offset the maladies of weightlessness. The proposed system will
allow pre- and post-flight ground analysis at the end of Phase II and on-station analysis shortly thereafter (through
Phase II production by Hamilton Sundstrand).

Raman Spectroscopy - Similar to an infrared spectrum, a Raman spectrum consists of a wavelength distribution of
bands specific to molecular vibrations corresponding to the sample being analyzed, which allows confident
identification of chemicals and biochemicals (selectivity). For example, Figure 5 shows the Raman spectrum of 3-
methyl histidine, which is slightly different from that of histidine. In practice, a laser is focused on the sample, the
inelastically scattered radiation (Raman) is optically collected, and directed into a spectrometer, which provides
wavelength dispersion, and a detector converts photon energy to electrical signal intensity. Historically, the very
low conversion of incident radiation to inelastic scattered radiation limited Raman spectroscopy to applications,
which were difficult to perform by infrared spectroscopy, in particular, aqueous solutions. In addition to sensitivity,
Raman spectroscopy has been limited by long-term instrument stability, fluorescence interference, and wavelength
reproducibility. These four limitations have been largely overcome in the past decade by several technological
advances, principally: air cooled stable diode lasers, notch filters, full spectrum detectors (i.e. no scanning), high
quantum efficiency detectors, and associated fast electronics, data acquisition and analysis, which have made Raman
spectroscopy standard equipment in analytical laboratories,66 and allowed the development of portable systems. An
attractive advantage to this technique is that in many cases samples do not have to be extracted or prepared, and a
fiber optic probe can simply be aimed at a sample to perform chemical analysis. In this regard, Raman spectroscopy

7


has been used to identify various chemicals (within glass cylinders). However, these measurements are best
performed for pure or at least highly concentrated samples. Further improvements can be realized by using a
Fourier transform Raman spectrometer.67 These systems employ diode pumped Nd:YAG lasers that provide
excitation in the near infrared, virtually eliminating fluorescence interference associated with visible laser excitation.
Another advantage of FT systems is high wavelength accuracy (Connes advantage)68afforded by the HeNe
wavelength reference laser. This allows reliable spectral subtraction and library matching,69 as well as continuous or
"on-demand" monitoring. Spectral subtractions can be used to isolate contributions of trace chemicals in the
presence of much more concentrated interferent chemicals,69,69 and library matching can provide rapid chemical
identification. The former may be very important, in that uric acid produces a significant SER spectrum that might
need to be subtracted to observe other key chemicals. It also allows employing other urine components as internal
concentration standards, such as creatinine to quantify 3-methyl histidine. Here, however, spectral analysis will be
augmented by the chemical selectivity of the sol-gels to be developed (see below). Nevertheless, even with these
improvements, the very low conversion of incident radiation to inelastic scattered radiation limits the sensitivity of
Raman spectroscopy and provides only moderate detection limits for normal Raman scattering. Relatively high
laser powers and long acquisition times are required to obtain a spectrum with a reasonable S/N, which in general
still allow only modest detection limits to be estimated for normal Raman. For example, Figure 5 shows Raman
spectra of pure 3-methyl histidine. The detection limit for 3-methyl histidine in water or lactic acid is ~1% (Figure 5
is a pure solid sample). Thus, normal Raman scattering would be capable of only moderate detection limits, such as
100 mg/mL.

Surface Enhanced Raman Spectroscopy - In 1974,70 it was discovered that when a molecule is in close proximity
to a roughened silver electrode, the Raman signal was increased by as much as six orders of magnitude.49 The
mechanism responsible for this large increase in scattering efficiency has been the subject of considerable research.71
Briefly, the incident laser photons generate a surface plasmon field at the metal surface that provides an efficient
pathway to transfer energy to the molecular vibrational modes, and produce Raman photons.71 This is possible only
if: 1) the material is in the form of particles much smaller than the laser incident wavelength (Raleigh regime,
surface imperfections of similar size also work) to couple the energy, 2) the material has the appropriate optical
properties to couple the light (extinction), 3) the available free electrons, when excited, are confined by the particle
size forming surface modes or generating surface plasmons, and 4) the molecule has matching optical properties
(absorption) to couple to the plasmon field.49,72 These very specific conditions, restrict SERS to the coinage metals,
silver, gold, and copper with diameters between 5 and 200 nm.72,73

SERS has been demonstrated for a number of inorganics, organics, and biochemicals, using three primary methods
developed to produce SER active media: activated electrodes in electrolytic cells,48 activated silver and gold colloid
reagents,74,75,76 and metal coated substrates.77,78,79,80,81 Unfortunately, these methods have not been reduced to a
product because it has proven difficult to manufacture a surface by these methods that yields reproducible
enhancements or reversible adsorptions or both. Oxidation-reduction cycles are used to create surface features
(roughness) on electrodes with the appropriate size to generate surface plasmons, but this roughness is difficult to
reproduce from one measurement to the next.49,82 Reducing a metal salt solution can be used to produce a colloid
containing metal particles capable of generating surface plasmons. The resultant particle size and aggregate size are
strongly influenced by the initial chemical concentrations, temperature, pH, and rate of mixing, and consequently are
also difficult to reproduce.75 Depositing one of the SER active metals onto a surface with the appropriate roughness
can also be used to prepare a surface capable of supporting surface plasmons. The largest enhancements are
obtained when the sample is dried onto the surface, in effect concentrating the analyte on the metal. This also
results in measurements that are difficult to reproduce. The relative merits and limitations of these methods have
recently been reviewed.83

In an effort to overcome these limitations, we have been developing metal-doped sol-gels as an active SER medium.
This medium should be capable of providing SER measurements that are reproducible, reversible, and quantitative,
yet are not restricted to specific environments, such as electrolytes, solvents, or evaporated surfaces. 24 The general
concept is shown Figure 1, where nanocomposite material has been coated on the inside walls of glass vials to yield
a general use SER product: Simple SERS Sample Vials. We have 1) measured the SER spectrum of 1
nanogram/milliliter of PABA in methanol (estimated detection limit of 10 pg/mL), and achieved a signal increase of
a factor of 107,25 2) have reproduced measurements from vial-to-vial with a standard deviation less than 10%, 3)
have demonstrated reversibility using a flowing system (within 5 minutes at 1 ml/min flow through a coated NMR
tube,25 and 4) we have successfully measured the amino acids, DNA bases, several drugs and their metabolites in

8


both aqueous and non-aqueous solutions using our Simple SERS Sample Vials. In particular, we measured 1 mg of
3-methyl histidine in 1 mL of water and the signal-to-noise ratio (S/N) of 52 suggests a current detection limit of 60
microg/mL (defined as S/N=3, Figure 5). Measurements at the required detection limits (<1 microg/mL, Table 1) is
anticipated to be straightforward using a new hybrid FT-Raman spectrometer (increased sensitivity with 785 nm
laser and Si-APD) developed under another SBIR program (see Related R&D). However, we recently employed
785 nm laser excitation and found near-equivalent SERS-enhancement compared to previous 1064 nm laser
excitation, even when taking the ν4 wavelength dependency of Raman scattering into account. This suggested that
our particle size, distribution, and/or aggregation were far from optimum. In recent years we have used our Simple
SERS Sample Vials to measure several hundred different chemicals. In general, we calculate enhancement factors
between 104 and 106. (The difference is due to the polarizability of the analyte and/or the extent of interaction with
the metal.) These values may prove insufficient for detecting certain drugs and/or metabolites present at very low
concentration levels in the body fluids (e.g. 1ng/L to 1ng/mL). Except for highly polarizable molecules, we
typically measure LODs of 10 microg/L, suggesting that an improvement of 2 to 4-orders of magnitude is required
to achieve the lower detection limits. Recently, we re-evaluated our metal-doped sol-gels in regards to optimum
particle size and aggregation. TEM measurements show that the silver particles are largely unnaggregated and small
(5-20 nm). We are currently examining methods to increase particle size and aggregation (variations in
concentration and heating). According to theory improvements of 2 to 6 orders of magnitude can be expected.
Such improvements in both the instrumentation and metal-doped sol-gel chemistry will be important for
achieving the sensitivity required in this proposal.

Sol-Gel Chemistry - The sol-gel process is a chemical route for the preparation of metal oxides and other inorganic
materials such as glasses and ceramics.84,85 The sol-gel process involves the preparation of a sol of metal-alkoxide
precursors in a suitable solvent, which undergo a series of reactions including their initial hydrolysis followed by
poly-condensation to form a gel. Expulsion of the solvent from the gel by a drying process, results in a highly
porous xerogel consisting of the metal oxides and any other additives that may have been introduced during the
process. Additional heating (fired) can be used to crystallize and/or densify the material. Typically, the sol-gel
process involves a silicon alkoxide (such as tetramethyl orthosilicate), water and a solvent (methanol or ethanol),
which are mixed thoroughly to achieve homogeneity on a molecular scale. The sol-gel matrixes offer several
additional properties useful to the proposed ISS application: physical rigidity and high abrasion resistance,
negligible swelling in aqueous solutions, chemical inertness, high photochemical and thermal stability, and excellent
optical transparency, and low intrinsic fluorescence.86 We have successfully developed metal-doped sol-gels that
can coat a variety of substrate-surfaces to produce a wide range of sensor designs, including glass vials, multi-well
microplates (glass and polystyrene), and glass capillaries. We have used the latter to detect flowing samples as well
as to perform chemical separations.

The Simple SERS Sample Vials are produced according to the following procedure. First, a silver amine precursor
complex is prepared from a solution of ammonium hydroxide and silver nitrate. Second, the sol-gel solution is
prepared from TMOS and methanol. Third, the amine complex and sol-gel solution are mixed, then the solution is
spin-coated onto the inner walls of a glass vial, and dried. Fourth, the substrate is heated to form the xerogel. This
step defines the porosity (size and distribution) and silver particle size. Fifth, the silver ion is reduced to silver metal
particles (Ago) using dilute sodium borohydride. And sixth, the substrate is washed and dried prior to the addition of
a sample. Previously, we established that a volume ratio of 1:5:4, silver amine complex to TMOS to methanol
heated at 120 oC for 2 hours yielded optimum SERS signals for PABA.25 It is worth noting that gold-doped sol-gel
coated vials have also been prepared in a similar fashion by using an aqueous solution of HAuCl4 (or NaAuCl4),
nitric acid as a gellation catalyst, and a Si-alkoxide precursor (e.g. TMOS). These conditions were optimized using
a simplistic experimental design approach. Knowledge of sol-gel chemistry and surface-enhancement theory was
used to optimize chemistry and physical properties, while performing a minimum number of experiments.
Nevertheless, the use of only PABA to maximize the sensitivity may have reduced the sensitivity to other analytes.
For example, we estimate a surface enhancement of <105 for 3-methyle histidine. Unfortunately, tailoring
sensitivity to every analyte of interest would be time consuming and tedious.

Combinatorial Chemical Synthesis - To alleviate this problem, we are employing combinatorial chemistry to
systematically synthesize large numbers of well-defined sol-gel compositions (libraries) by combining the reactants
in all combinations.87 We are employing 96-well micro-plates to develop the chemistry. Initially two reservoirs (8
and 12-cell) are filled with the two starting solutions (here TMOS and the amine complex) with incrementally
increasing concentrations. Then multi-channel pipettes (8 and 12) are used to deliver microliter samples to each

9


well fairly rapidly (by hand it takes ~4 minutes). Once completed, the plate is placed into an oven to cure the sol-
gel. We are able to prepare as many as 20 plates per day or 1892 sol-gels with different SERS activity. The success
of a library is determined by testing the activity of the sol-gel coated well. Since the number of wells quickly
escalates using this procedure, testing of each well becomes impossible and high-throughput screening techniques
are used. This is accomplished by selecting and testing an ordered subset of the 96 wells. The most active wells can
be used to refine the chemistry and improve SERS activity.

Using this approach, we have modified the alkoxide chemistry to obtain SERS active sol-gels that preferentially
"solvate" polar or non-polar analytes. For example, the alkoxide precursor MTMOS has a higher -CH3
concentration (lower -OH) than TMOS, and consequently a higher affinity for non-polar chemicals
(hydrophobic).86,88 Task I will focus on employing combinatorial chemistry to develop sol-gels that are selective
and active (and stable) towards various key analytes to be screened for in Task II and Task III. During the past year
we successfully developed our first chemically selective SER-active sol-gels, consisting of the four Libraries
outlined in Table 2.

Table 2. Summary of chemically selective surface-enhanced Raman active metal-doped sol-gels.
Library 1 Ag + TMOS Selective for polar-negative species
Library 2 Ag + (TMOS+MTMS) Selective for weakly polar-negative species
Library 3 Au +TMOS Selective for polar-positive species
Library 4 Au + TEOS Selective for weakly polar-positive species

Figure 7 shows surface-enhanced Raman spectra of p-aminobenzoic acid (PABA) using Library 1 and 3, and phenyl
acetylene (PA) using Library 2 and 4. For Libraries 1 and 3, the polar PABA passes through the polar sol-gel and is
enhanced by either the silver or gold particles. For electropositive silver, the PABA anion (pKa = 4.8) interacts
through the carboxylate group and COO- bands appear at 840 and 1405 cm-1. For electronegative gold, PABA
interacts through the amine group and -NH2 bands appear at 1355 and 1585 cm-1. For Libraries 2 and 4, the non-
polar PA passes through the non-polar sol-gel and is enhanced by either the silver or gold particles. For
electropositive silver, PA interacts strongly through the cylindrical π cloud around the carbon-carbon triple bond and
a -C≡C- doublet occurs near 2000 cm-1. For electronegative gold, this interaction is unlikely and only very weak
bands occur near 2000 cm-1. The polar/non-polar selectivity of the polar-negative and weakly polar-negative sol-
gels was tested by adding a 1:1 molar mixture of PABA and PA. The selective enhancement is quite good (Figure
8). The spectrum obtained using the polar sol-gel suggests 78% PABA and 22% PA reached or is active at the metal
surface, while the spectrum obtained using the weakly polar sol-gel suggests a 9% PABA and 91% PA activity. The
band peak intensities at 2000 cm-1 for PA and 1450 cm-1 for PABA were used for these calculations, and are
expanded in Figure 8 for clarity. Silver-doped TMOS favored more rapid transit of the polar PABA than the non-
polar PA.

A B
C CH

H2N COOH
relative intensity

relative intensity

Wavenumber (∆cm-1) Wavenumber (∆cm-1)
Figure 7. SER spectra of A) PABA using Libraries 1 (top) and 3 (polar-negative and polar-positive sol-gels), and B)
PA using Libraries 2 (top) and 4 (weakly polar-negative and weakly polar-positive sol-gels). PABA is 1 mg/mL,
PA is 1% v/v. Spectral conditions: 75 mw 1064 nm, 100 scans (1.5 min), 8 cm-1 resolution. The y-axis for all
spectra represent intensity in arbitrary units.

10


A C
relative intensity

relative intensity
B D

Wavenumber (∆cm-1) Wavenumber (∆cm-1)
Figure 8. SERS of 1:1 M/M of PABA and PA in A) polar-negative and C) weakly polar-negative sol-gels. The
lower traces, compare the pure chemicals; B) 1 mg/ml PABA in polar-negative sol-gel and D) 1% PA in weakly
polar-negative sol-gel, while the insets magnify the minority species for clarity (x5 in A and x10 in B). Spectral
conditions as in Figure 5.

Part 3: Technical Objectives
The overall objective of the proposed program (through Phase III) is the development of an analyzer integrated into
the International Space Station (ISS) toilets capable of immediate detection of key chemicals in urine to monitor and
assess astronaut health.

The following six comprehensive tasks have been designed to develop this proposed analyzer, as well as the
required method of analysis with the following objectives and specific questions to be answered.

Task 1 - Spectral Library Development. The overall objective of this task is to build a SER spectral library to
allow rapid spectral matching (or functional group analysis) for identification of biochemical markers, drugs and
their metabolites present in human urine. This will be accomplished by extending the Phase I measurements to
include an in-depth analysis of 12 primary bio-indicators specific for assessing muscle/bone loss and renal stone
formation, and 12 priority drugs (and their metabolites) that may be used to minimize or counter the adverse
physiological affects associated with changes in the concentration levels of these biomarkers present in urine.

Questions to be answered? Are all of the 24 chemicals SER-active? What is the preferred sol-gel for each bio-
indicator and target drug? Are there potential interferants? How well does the spectral match software identify
each? How well does it identify a chemical not in the library?

Task 2 – Chemical Selectivity Development. The overall objective of this task is to refine the ability of the 4 basic
SER-active sol-gels developed in Phase I to selectively extract biochemicals and drugs present in human urine, and
enhance their Raman spectra. This will be accomplished by measuring reversibility of representative bio-indicators
and drugs drawn through the 4 sol-gels.

Questions to be answered? Which sol-gels provide irreversible adsorption of the bio-indicators and target drugs
and SER-activity? What are the estimated LODs? Can the spectral deconvolution software identify each urine
component (bio-indicators and drugs) on each sol-gel?

Task 3. Component selection and testing. The overall goal of this task is to design a lab-on-a-chip that can be
used to analyze chemical components present in human urine by SERS. This will be accomplished by designing a
chip based on the Phase I results and the background information provided above.

Questions to be answered: Are there components available to perform the desired extractions and separations?
Are they effective in the context of our SER-active sol-gels, separately, and together? What is the best sequence of
sol-gels? How universal is it for various urine components and drugs? Are the mixtures effectively separated and

11


detected? What happens when other chemicals from Task 1 are analyzed? What are the detection limits?

Task 4. Lab-on-a-chip fabrication (with Advanced Fuel Research as subcontractor). The overall goal of this
task is to build many microfluidic chips that can be used to test the preliminary lab-on-a-chip design. This will be
accomplished by fabricating poly(methyl methacrylate) chips. The process of Muck et al., slightly modified, will be
followed.89 The microfluidic chips will be prepared in a class 10 clean room at AFR, and all safety (HF) procedures
will be followed.

Questions to be answered: Can AFR produce the test chips? Does the process need further modification? Does
the interface perform as planned (no leaks)? Can the various channels be loaded, reduced? What size, length is
best? Can separation materials be introduced?

Task 5. Define Analytical Figures of Merit. The aim of this task is to establish performance criteria for the lab-
on-a-chip. This will be accomplished by measuring the analytical figures of merit for the analyzer, as outlined by
the FDA: sensitivity, reproducibility, linearity, accuracy, precision, resolution, and selectivity.

Questions to be answered: What are the LODs for each of the biomarkers, drugs and their metabolites? What is
the reproducibility of the chips? Which channel design provides the best selectivity, reproducibility, and sensitivity?
What is the best standardization method for quantitating target analyte concentrations?

Task 6 - Prototype Design. The overall goal of this task is to design a prototype system to be used and tested by
NASA in Phase III. This will be accomplished by redesigning the lab-on-a-chip for system integration and
autonomous operation. No questions to be answered.

12


Part 4: Work Plan (Phase II Results)

Task 1 - Spectral Library Development. The overall objective of this task was to build a surface-enhanced Raman
spectral library to allow rapid spectral matching (or functional group analysis) for identification of biochemical
markers, drugs and their metabolites present in human urine. This was accomplished by measuring the SERS-
activity of 25 analytes and 25 potential interferents using some 20 different chemically-selective sol-gels
(Libraries).

The proposed 24 analytes (12 biomarkers and 12 drugs) that were the focus of this program are listed in Table T1.1.
Ten of the 12 biomarkers were commercially available (the two collagen bound bone-loss markers were not) and all
were SERS-active. Similarly, 10 of the 12 proposed drugs were commercially available, and all were SERS-active.
Three additional drugs (secondary) were also measured, including the metabolites of acetaminophen and allopurinol.

Table T1.1. List of biomarkers and associated drugs studied during Phase II.
Biomarkers Drugs
Muscle loss indicators: Deoxypyridinoline – Anti-bone loss: Anti-stone formation:
Creatinine (CRE) collagen bound* Etidronate (ETI) Hydrochlorothiazide (HCT)
3-Methylhistidine (3-MeHIS) Clodronate (CLO) Allopurinol (ALLO)
Bone loss indicators: Stone formation indicators: Pamidronate (PAM) Penicillamine (PEN)
Hydroxyproline (HO-PRO) Calcium oxalate (CaOx) Alendronate (ALE) Anti-motion sickness:
Hydroxylysine (HO-LYS) Calcium phosphate (CaP) Risedronate (RIS) Promethazine (PROM)**
Pyridinoline (H-Pyd) Uric acid (UA) Ibandronate (IBA) Scopolamine (SCOP) **
Pyridinoline-collagen bound* Cystine (CYST) Raloxifene (RAL) Anti-inflammatory:
Deoxypyridinoline (H-dPyd) Tiludronate*/Zoldronic acid* Acetaminophen (AM)**
* **
Four analytes not available. Additional drugs measured.

In the Phase I proposal we described 4 chemically-selective sol-gels (Libraries 1-4) to be used for this study. During
the Phase I program we modified the Library 1 and 2 chemistries to produce 4 additional sol-gels. During the Phase
II program, the number of chemistries was further expanded to some 20 available sol-gels that could be used for
screening chemical selectivity (see Quarterly Report 7). By examining many biomarkers and drugs using all 20
libraries it was found that 6 proved most useful. L1, L2, and L4 correspond to the chemistries designated 1, 2, and 4
while L3 corresponds to the chemistry designated 2d in the Phase I Final Report. We also developed and used two
new chemistries, which essentially are chemistry 1 (or L1) modified by the inclusion of polymers, either
poly(ethyleneglycol) or poly(dimethylsiloxane) (PEG and PDMS, respectively), designated L5 and L6. The relative
concentrations of the precursors used to prepare these 6 libraries are summarized in Table T1.2.
Table T1.2. Synthesis summary for SERS-active, chemically-selective, sol-gel chemical libraries (L1-L6).
Sol-Gel Metal Precursor (A) Sol-Gel Precursor (B) A B Selective for:
(µL) (µL)
L1 5/1/10: 1N AgNO3/28%NH3OH/MeOH 1/5: TMOS/MTMS 100 120 mildly polar - negative
L2 5/5/10: 1N AgNO3/28%NH3OH/MeOH MTMS 100 100 non-polar - negative
L3 5/5/10: 1N AgNO3/28%NH3OH/MeOH 1/5/1:TMOS/MTMS/ODS 100 175 very non-polar-negative
L4 4/1: 0.25N HAuCl4(aq)/70% HNO3 TMOS 100 100 very polar - positive
L5 5/1/10: 1N AgNO3/28%NH3OH/MeOH 1/5: TMOS/MTMS 100 120 polar – negative
(+ 10 µL PEG)
L6 5/1/10: 1N AgNO3/28%NH3OH/MeOH 1/5: TMOS/MTMS 100 120 non-polar - negative
(+ 10 µL PDMS)
APTMS: aminopropyltrimethoxysilane, MTMS: methyltrimethoxysilane, PDMS: polydimethylsiloxane, PEG: polyethylene
glycol, ODS: octadecyltrimethoxysilane, TMOS: tetramethylorthosilicate.

During the Phase I program, screening SERS-activity for the analytes using different sol-gels was initially
performed in 96-well microplates. It was found that better results were obtained using glass capillaries (1.1 mm

13


outer diameter, 800 micron inner diameter) filled with metal-doped sol-gels. These capillaries operate in the active
mode, in that the sample is forced to flow through the sol-gel. We therefore used these capillaries to test SER-
activity for the analytes throughout the Phase II program. Furthermore, the capillaries became the basis of the Phase
II micro-chip sampling system. The basic design and use of the SER-active capillaries is shown in Figure T1.1, and
are prepared as follows. The alkoxide and amine precursors are prepared according to Table T1.2, mixed, and then
drawn into the capillary by syringe. Typically 0.4 mL of solution coats a 4 cm length of capillary. The sol solution
gels in 5 minutes, and a more rigid structure is obtained after 24 hours. A solution of 0.1g/100mL NaBH4 is drawn
through the capillary to reduce the metal. This is followed by a 0.035% HNO3 acid wash, and then the capillary is
then ready to be used.

Figure T1.1. Photograph of sol-gel coated melting point capillaries attached to syringes, before (top) and after
reduction with sodium borohydride.

Initial SERS-activity screening on the various sol-gels employed 1 mg samples in 1 mL HPLC water. The samples
were drawn into the capillaries, which were mounted on an XY sample stage above a fiber optic probe coupled to
RTA’s Industrial Raman Analyzer. Spectra were obtained using 80 or 100 mW of 785 nm excitation at the sample
and 1 minute acquisition time. Once the initial screening was performed, the samples were serially diluted over 4
orders of magnitude to 0.01 mg/L to determine sensitivity. The required sensitivity is ~ 1 mg/L for the metabolites
and 10s of microg/L for the drugs. Tables T1.3 and T1.4 summarize the SERS-activity in terms of the lowest
measured concentration (LMC) for the 10 biomarkers and 13 drugs, respectively. The best measurements were
ultimately obtained by concentrating the sample using ion exchange resins (Task 3), and are included in the tables
for comparison.

Also, normal Raman spectra (NRS) of the analytes were acquired as neat liquids or pure solids in the same glass
capillaries using 300 mW at 785 nm for 5 minutes. The NRS and SERS are shown for the 10 urinary biomarkers in
Figures T1.2-T1.11. This includes the two muscle loss indicators, CRE and 3-MeHIS, the quantifiable bone loss
indicators H-Pyd and H-dPyd, the non-quantifiable bone loss indicators HO-PRO and HO-LYS, and the stone
formation indicators, UA, CaP, CaOx and CYST. During the Phase I program we focused on hydroxyproline as an
indicator of bone loss, however, research into the chemical and metabolic reaction pathways associated with
osteoporosis indicates that free and bound pyridinoline and deoxypyridinoline, associated with collagen cross-
linking, are more quantitative indicators of bone loss. And they are present in urine. We were able to obtain the free
forms of pyridinoline and deoxypyridinoline from Quidel Corporation, but not their bound form. Analysis of
hydroxyproline is still important, and remained part of this study

Table T1.3. Summary of Biomarker screening results: SERS-response on select chemistries in mg/L.
CRE 3-MeHis HO-PRO HO-LYS H-Pyd H-Dpd CaP CaOx UA CYST
t
L1 1 1 neg 1000 dnt dnt 500 neg 500 1000
L2 1000 10 neg neg neg dnt 500 500 500 neg
L3 1000 1000 1000 neg 1.8 neg 1.8 500 500 500 neg
L4 neg neg 1 f
neg neg dnt neg neg neg neg
L5 neg 1000 neg neg dnt neg 1.8 500 500 500 1000
L6 1000 1000 neg neg dnt dnt 500 neg 500 neg
IEX
Library L1sL3d L5 L1t L3 L5 L5 L2 L1 L1
LMC 0.1 0.001 dnt 0.01 0.018 1.8 1 1 0.05 0.1
Peaks 1421 1563 1534 1397 1393 1372 925 894 633 613
dnt=did not try, f = flow method, t = TMOS only, S = SPE, d = +PDMS

14


Table T1.4. Summary of Drug screening results: SERS-response on select chemistries in mg/L.
CLO ETI PAM ALE IBA RIS RAL HCT ALLO PEN SCOP PROM AM
L1 100 1000 1000 100 10t 1 10 1000 10 1000 neg 1000 1000
L2 1000 1000 1000 100 neg 100 100 neg 0.1 1000 neg 1000 1000
L3 1000 1000 1000 1000 neg 1000 1000 100 1000 1000 neg 1000 1
L4 neg neg neg neg neg neg neg 0.01 f
neg 100 0.01 f
0.01 f
0.01f
L5 neg 100 100 1000 neg 100 1000 1000 500 neg neg dnt dnt
L6 100 1000 1000 1000 neg 1 1000 100 500 1000 neg dnt dnt
IEX
Library L1 L3s L1 L1 L1t L5 L3d L1 L5 L1
LMC 0.01 0.01 0.01 0.01 0.01 0.001 0.01 1 0.001 0.01 dnt dnt dnt
Peaks 674 644 639 647 1010 1034 1593 678 721 1110 999 1025 1578
dnt = did not try, f = flow method, t = TMOS only, S = SPE, d = +PDMS; Note: OxP (ALLO metabolite) LMC
0.01 mg/L on L5; PAM, IBA, ALE LMC 0.01 mg/L on both IEX and SPE with L1 (L1t for IBA)

N

O
A A
N NH2

OH

B N
O B
N NH2

Fig. T1.2. A) NRS and B) SERS of Creatinine on L1. Fig.T1.3. A) NRS and B) SERS of 3-Methylhistidine on L1
Note: 1-methyl histidine was measured in Phase 1, see Fig T5.2)
.

OH A
O
H2N A
HO O
NH
HO OH

NH2

B
B

Fig. T1.4. A) NRS and B) SERS of Hydroxy-proline, 0.01 Fig. T1.5. A) NRS and B) SERS of Hydroxy-lysine on L1t
mg/L on gold L4. (with HCl wash, and 2nd reduction step.

15


A
NH2

HO
NH2

O
O
NH2
HO
OH
HO

OH
O

B

Fig. T1.6. SERS of H-Pyridinoline in 0.2M acetic acid A) 1.8 Fig. T1.7. SERS of H-deoxyPyridinoline 1.8 mg/L on L5
mg/L on L3; and B) L3 (chem2c, reported in Phase I) with (conditions same as in Fig. T1.6).
acetic acid (SERS) subtracted.

-
O
O
P O
Ca+2
-
O
O- Ca+2 O- A O
A
- O-
O P O-

Ca+2 O-
O Ca+2

B B

Fig. T1.8. A) NRS and B) SERS of Calcium phosphate on Fig. T1.9. A) NRS and B) SERS of Calcium oxalate on L2.
L2.

O NH2 A
A S
O

HO S
HO
NH2

OH

H
N
N
B
O B
N
HO N H

Fig. T1.10. A) NRS and B) SERS of Uric acid on L1. Fig. T1.11. A) NRS and B) SERS of L-Cystine on L1.

16


The NRS and SERS are shown for the 13 drugs in Figures T1.12-T1.24. This includes the anti-bone loss drugs
CLO, ETI, PAM, ALE, IBA and RIS), a selective estrogen receptor modulator, RAL, the anti-stone formation drugs
HCT, PEN and ALLO, the anti-motion sickness drugs SCOP and PROM, and the anti-inflammation drug AM.

O Cl

HO
P
A A
Cl

HO
P OH
HO

O

O CH3

B HO P OH
B
HO
P OH
HO

O

Fig. T1.12. A) NRS and B) SERS of Clodronate disodium on Fig. T1.13. A) NRS and B) SERS of Etidronate disodium on
L1. L1.

NH2
O

HO
NH2
A O
A
P OH HO
P OH
HO
OH HO
P OH
HO P
HO
O
O

B B

Fig. T1.14. A) NRS and B) SERS of Pamidronate disodium on Fig. T1.15. A) NRS and B) SERS of Alendronate sodium on
L1. L1.

O

HO
P OH
N

CH3
CH3 A O
N
A
HO P OH
HO
P OH
HO
HO OH
P
HO
O
O

B B

Fig. T1.16. A) NRS and B) SERS of Ibandronate sodium on Fig. 16 T1.17. A) NRS and B) SERS of Risedronate sodium
L1t. on L1.

17


O
N

O A
HO A
S

OH

B
B H2N
O
O O

S S
NH
O

Cl
N
H C

Fig. T1.18. A) NRS and B) SERS of Raloxifene in MeOH on Fig. T1.19. A) NRS and SERS of Hydrochlorothiazide on B)
L2. L3, and C) 0.01 mg/L on gold L4.

OH

A
N A
N

N
N H

NH2
B
B HS
O

H3C
HO
H3C
C

Fig. T1.20. A) NRS and B) SERS of Allopurinol on L2. Fig. T1.21. A) NRS and SERS of Penicillamine on B) L2,
and C) 0.01 mg/L on gold L4.

H3C
N
CH3
A A
CH3 H
O

N
O N O OH

S
B H
B

C C

Fig. T1.22. A) NRS and SERS of Promethazine HCl on B) Fig. T1.23. A) NRS and SERS of Scopolamine HCl on B)
L2, and C) 0.01 mg/L on gold L4. gold L4 (chem3a, from Phase I) and C) 0.01 mg/L on gold
L4.

18


A
A

O
B O O O NH
O

HN OH

O O O
O
B
C

Fig. T1.24. Acetaminophen, A) NRS and SERS B) 0.1 Fig. T1.25. A) NR and B) SERS of AM-G; 0.1 mg/mL on
mg/mL on L2, and C) L4 (gold). L2.

Although the majority of primary target drugs in this study, such as the bis-phosphonate bone-loss drugs, are
generally excreted unchanged in urine, other drugs metabolize, such as acetaminophen and allopurinol. The SERS
(and NRS) of the metabolites of these two drugs, acetaminophen-glucuronide (AM-G) and oxipurinol (OxP), were
also measured (metabolites from other drugs were not readily available). As shown in Figures T1.25 and T1.25,
acetaminophen and its metabolite produce different SER spectra. ALLO is a xanthine oxidase inhibitor that lowers
the level of uric acid in urine. Approximately 90% of ALLO is metabolized to OxP. ALLO is rapidly excreted in
urine (T1/2 = 40 min), while OxP is excreted over a much longer period (T1/2 =14-30 hrs). The NRS and SERS for
OxP are shown in Figure T1.26 (compare to ALLO Fig T1.20).

OH
A OH

N N
B N N

N N
N H HO N H

C Allopurinol Oxypurinol

Fig. T1.26. A) NRS and B) SERS of Allopurinol; C) NRS and
D D) SERS of Oxypurinol; 0.5 mg/mL on L5.

In addition to these 25 measured analytes, 25 additional chemicals that may be present in astronaut urine and could
be potential interferents were measured (Table T1.5). This included five additional drugs, the pain reliever -
acetylsalicylic acid (ASA, aspirin), representative sleeping aids – barbitol and phenobarbitol, and stimulants caffeine
and Adderall (Figures T1.27-T1.30). The latter drug is a single entity amphetamine product combining the neutral
sulfate salts of dextroamphetamine and amphetamine, with the dextro isomer of amphetamine saccharate and d,l-
amphetamine aspartate monohydrate.

Table T1.5. List of drugs, vitamins, and natural products of metabolism as interferents that may appear in urine.
Drugs Vitamins/Supplements Natural metabolites
Acetylsalicylic acid (ASA) Vitamin A Lactic acid Glucose
Barbital Vitamin E Hippuric acid Gluconic acid
Phenobarbital Thiamine Nicotinic acid Cholesterol
Caffeine Pyridoxamine Glutamic acid Estradiol
Allderall Citric Acid Histidine Pregnane-diol
1-methylhistidine Theophyllene
Cystine/cysteine Xanthene/Hypoxanthene

19


2.600

2.400

2.200
A
2.000

1.800
A
1.600

1.400 O OH O

1.200 HN NH

1.000 O O
O O

0.800
B B
0.600

0.400

0.200

0.000
350 500 750 1000 1250 1500 1750 1850

Fig. T1.27. A) NRS and B) SERS of Acetylsalicylic acid. Fig. T1.28. A) NRS of Barbituric acid, and SERS of B)
on L3 (Chem2c). Barbital, and C) Phenobarbital; 0.1 mg/mL on L3.

A

NH2

N N O B
N
N
O

Fig. T1.29. A) NRS and B) SERS of Caffeine on gold L4. Fig. T1.30. A) NRS and B) SERS of Adderall 18 (25 mg;
active ingredient Dextroamphetamine). L2.

In addition to drugs astronauts also take vitamins (and supplements) to help prevent the negative effects of low
gravity. For this reason we measured vitamins A and E, thiamine, pyridoxamine (B6), and citric acid, all of
which are potential interferents in urine (Figures T1.31- T1.38).

A HO
A
OH O

B
B

C

Figure T1.31. A) NRS and SERS of Vitamin-A in MeOH on B) Figure T1.32. A) NRS and B) SERS of Vitamin-E on L5.
gold L4 and C) silver L2.

20


1.000 2.100 NH2
NH2
2.000 OH
0.900 Cl-
1.900
0.800 N N+ A 1.800
OH
A
1.700
0.700
relative intensity

relative intensity
OH
S 1.600
N N
0.600 1.500
1.400
0.500
1.300
0.400 1.200

0.300
B 1.100 B
1.000
0.200 0.900
0.800
0.100
0.700
0.000 0.600
350 500 750 1000 1250 1500 1750 1850 350 500 750 1000 1250 1500 1750 1850

Fig. T1.33. A) NRS and B) SERS of Thiamine on L2 Fig. T1.34. A) NRS and B) SERS of Pyridoxamine on
(Chem2b). L1.

2.400

A 2.200

2.000

1.800
A
1.600

1.400 OH

O 1.200
O OH OH
1.000
HO O
OH
B 0.800
O
B
HO 0.600

0.400

0.200

0.000
350 500 750 1000 1250 1500 1750 1850

Wavenumbers (∆cm-1)
Figure T1.35. A) NRS and B) SERS of Citric acid on L1. Fig. T1.36. A) NRS and B) SERS of Lactic acid on L1.
3.500 4.500
O
3.250
4.000
3.000
NH
2.750 3.500
2.500 HO A A
relative intensity

relative intensity

3.000
2.250 O

2.000 2.500
1.750
2.000 O
1.500
1.250
1.500
1.000
0.750
B 1.000 N OH B
0.500
0.500
0.250
0.000 0.000
350 500 750 1000 1250 1500 1750 1850 350 500 750 1000 1250 1500 1750 1850

Fig. T1.37. A) NRS and B) SERS of Hippuric acid on Fig. T1.38. A) NRS and B) SERS of Nicotinic acid on
L2. L3 (Chem2c).

During the Phase I program we also measured several natural occurring biochemicals that appear in urine as the
result of various metabolic processes. This included lactic acid, hippuric acid, and nicotinic acid. In addition, all of
the amino acids are present in urine to some extent, and we have measured all of their SER spectra. But since these
measurements are the focus of another SBIR program, here we only report glutamic acid since it has a high urine
concentration, cystine and cysteine (Figures T1.39 and T1.40), and histidine (Figure T1.41). Cystine, composed of
two cysteine amino acids, is of interest because the drug, PEN solubilizes cystine by displacing one of the cysteines
as part of the anti-stone forming process and thereby releasing a free cysteine that ends up in urine. Both cystine
and cysteine have unique SERS features (Figures T1.40.B and T1.40.C).

21


7.500
7.000
6.500
6.000 A
5.500
A

relative intensity
relative intensity

5.000
4.500
4.000
O O
B
3.500
3.000
2.500 HO OH
2.000
C
1.500 NH2
B
1.000
0.500
0.000
D
350 500 750 1000 1250 1500 1750 1850

Fig. T1.39. A) NRS and B) SERS of Glutamic acid on L1 Figure T1.40. A) NRS and B) SERS of Cystine on L1, and C)
(Chem1b). NRS and D) SERS of Cysteine on L3 (chem2c).

0.800
O
0.750
H
0.700 N
OH
0.650
NH2 A A
relative intensity

0.600 N

0.550

0.500 OH

0.450 N
O

0.400 NH2
N

0.350
B B
0.300

0.250

0.200
350 500 750 1000 1250 1500 1750 1850

Wavenumbers (∆cm-1) Fig. T1.42. A) NRS and D) SERS of 1-MeHIS (Nτ); on L1t
Fig. T1.41. NRS and B) SERS of Histidine on L3 (Chem2c). with HCl wash.

Histidine is included in this study since it is the base structure for 3-methyl histidine. For the same reason we also
measured 1-methyl histidine because there were some uncertainty reported in literature as to the position of the
methyl group. Nevertheless, as can be seen in Figures T1.41 and T1.42, all three histidines produce unique spectra.

The biochemical interferents studies are important health indicators in their own right, specifically glucose and
cholesterol. Glucose is an important indicator of diabetes (although blood glucose better represents the condition),
but its SER spectrum has been difficult to obtain (as indicated Phase I Final Report Figure 4.17). We made
additional attempts to measure glucose using many of the sol-gel chemistries. A representative SER spectrum and
the NRS of a solution are shown in Figure T1.43. Only three peaks are observed with significant intensity, a weak
peak at 845 cm-1, a strong peak at 895 cm-1, and a broad peak at 1410 cm-1. These peaks are tentatively assigned to a
C-C stretch, O-C-H bend, and a C-C-H bend, respectively, based on the NRS peak assignments. However, the
paucity of peaks suggested that only part of the molecule was being enhanced (due to surface proximity or selective
plasmon field interaction) or it may be a molecular fragment produced by photo-degradation. To test the latter, we
measured the oxidation product of glucose, gluconic acid, which can also be produced by photo-degradation. As
can be seen, the SERS of gluconic acid is extremely similar to glucose (Figure T1.43). The only difference is the
absence of the weak 845 cm-1 peak. To further clarify the glucose SER spectrum, we also measured glucose using
1064 nm laser excitation (as well as using reduced powers). As Figure T1.44 shows, this glucose SERS looks
virtually identical to gluconic acid. However, the fact that the spectrum is present at all powers (down to 20 mW),
suggest that photo-oxidation is not occurring. Further measurements are required to clarify this point.

22


OH
A
A
O OH
OH
895 1410 O
845 B OH OH
HO OH

OH
B
HO OH

OH
C

Figure T1.43. A) NRS and B) and C) SERS of glucose. Figure T1.44. A) NRS and SERS of gluconic acid.
Conditions: A) 1g/mL in HPLC water B) 1 mg/mL in HPLC Conditions: A) 50 wt% in water, 1 mg/mL in HPLC water on
water on L1, 785 nm, and C) 48 mW of 1064 nm, 1-min. L1.

Cholesterol is normally measured in blood since it is involved in plaque build-up on artery and vein walls, but in
certain pathological conditions cholesterol crystals can be found in acidic to neutral urine. The SERS of cholesterol
was obtained on both gold- and silver-doped sol-gels as shown in Figure T1.45. We also measured the hormones
estradiol and pregnane-diol (Figures T1.46 and T1.47). It should be noted that pregnane-triol (as proposed for in
Phase I), testosterone and spermine were also measured, but were found to be inactive on the initial sol-gel
chemistries tried, and measurements in Phase II were not pursued.
2.400

A 2.200 A
2.000
relative intensity

1.800

1.600 OH

1.400

1.200
B H B
1.000
H H
0.800
OH
0.600
C 0.400

0.200
350 500 750 1000 1250 1500 1750 1850

Figure T1.45. A) NRS and SERS of Cholesterol on B) L4 Fig. T1.46. A) NRS and B) SERS of Estradiol on L2
(chem3b), and C) L2. (Chem 2b).

5.000
O
4.500
N
N A
4.000
A
O N N
3.500
relative intensity

3.000 O B
2.500 O
H H
N
2.000 HN
B H H
N C
1.500 N
O
1.000

0.500
D
0.000
350 500 750 1000 1250 1500 1750 1850

Wavenumbers (∆cm-1) Figure T1.48. A) NRS and B) SERS of Xanthine; C) NRS,
Fig. T1.47. A) NRS and B) SERS of Pregnane-diol on and D) SERS of Hypoxanthine; on L2.
L2.

23


The enzyme, xanthine oxidase, catalyzes the oxidation of hypoxanthine to xanthine, which is further converted to
uric acid. Since allopurinol inhibits this process, it also influences the relative amounts of hypoxanthine, xanthine,
and uric acid in urine. As Figures T1.48 and T1.49 shows, the xanthenes, and theophyllene (which is produced via a
similar metabolic process) produce unique SER spectra.
O

N N A
HN O

N

Fig. T1.49. A) NRS and B) SERS of Theophyllene on
L1.

B

As part of this program, we evaluated the ability of four spectral library search routines to identify the SER spectra
of the biomarkers and drugs. The Euclidean Distance Algorithm (EDA), the Absolute Value Algorithm (AVA), the
Least Squares Algorithm (LSA), and the Correlation Algorithm (CA) have been successfully applied to Raman
spectra, and we incorporated them into a LabVIEW program. These four different search algorithms are
summarized in Figure T1.50.

Euclidean Distance Algorithm (EDA) Absolute Value Algorithm (AVA) Least Squares Algorithm (LSA)
n n
∑ (Libi − Unkni)
2

HQI 2⋅ 1 −
Lib⋅ Unkn ∑ Lib − Unkn
i i
i=1 i=1
Lib⋅ Lib⋅ Unkn⋅ Unkn HQI HQI
n n
Correlation Algorithm (CA) n
n

HQI 1−
( Lib ⋅ Unkn
m m ) 2
∑ Lib
i ∑ Unkn
i
( m m )(
Lib ⋅ Lib ⋅ Unkn ⋅ Unkn
m )
m Where: Lib Lib −
i=n Unkn
m
Unkn −
i=n
m n n
Figure T1.50. Equations for the spectral library search algorithms.

In all cases, Lib is the library entry being searched and Unkn is the unknown sample spectrum. Each method scores
the spectral match in terms of a hit quality index (HQI), where the best score or match is 0 (Lib = Unkn) and the
worst score or match is at higher values. Initially, we tested the algorithms using a very limited library composed of
7 biomarkers and 9 drugs. Spectra different from that used in the library were used for these tests (the HQI scores
were presented in Table 3 of Quarterly Report #4). It was found that the best results were obtained by 1) subtracting
the glass background produced by the capillary from both the sample and the spectra in the library, and 2) taking the
first derivative. Smoothing the data improved scores, but not significantly. In all cases the EDA and CA methods
resulted in a positive match (after this pretreatment). However, the CA method gave the best overall results for
identifying each chemical (low HQI score) as well as discriminating against chemicals (high HQI scores).

This result proved correct even when we added all of the chemicals studied in this program (biomarkers, drugs,
metabolites, and potential interferents) plus all of the basic amino and nucleic acids (a total of 102 SER-active
chemicals). As an important test, the spectrum obtained for a 1 microg/L 3-MeHIS sample pre-concentrated using
an ion exchange column was examined using the Correlation Algorithm. As shown in Figure T1.51, this
biomarker was correctly identified with a very low HQI of 0.069. Furthermore, the next closest match (the
LabVIEW program ranks closest 10) is the structurally similar amino acid histidine (HIS), which nevertheless has a

24


much higher HQI score of 0.434. This is not unexpected as distinct differences observed in the SERS (Fig. T1.52).

OH
A
N
O

NH2
N

OH

N
O

NH2
N
B

Figure T1.51. Display of Spectral Library Search and Match Figure T1.52. SERS of A) 1 microg/L of 3-MeHIS as
software. The program ranks the best matches (lowest the unknown sample, and B) 1 mg/mL of HIS from the
scores) and overlays the unknown and matched spectra for spectral library.
visual comparison (SERS reference library = 102 chemicals).

In addition to these measurements, the search routines were challenged by using SERS of analytes at different pH,
on different SERS-active sol-gels, and with or without an acid treatment (i.e. HCl wash). As demonstrated in Phase
I, the spectra of these chemicals are constant between pH 4 and 8, and only if the pH was outside this range was a
sample misidentified (see Report # 5). In general, the alkoxide chemistry of the sol-gels produces insignificant
changes in the SER spectra. In fact, only in a few cases does the acid wash or metal (i.e. silver vs gold) result in
significant spectral differences. For chemicals that had such spectral differences, proper identification occurred if
both spectral versions were in the library.

We also evaluated S-Quant, a classical least squares algorithm, to determine chemical composition of mixtures.
This program first searches the library for the best spectral match, and then this spectrum is subtracted from the
original spectrum to produce a new spectrum, the residual spectrum. The search routine then uses this residual
spectrum as the unknown. The process is repeated until the noise throughout the residual spectrum is uniform (no
more peaks). The program then creates a spectrum composed of each of the matched spectra. The contribution of
each spectrum is varied (weighted) until the composite spectrum matches the measured spectrum, as defined by a
residual consisting of a flat baseline.

We evaluated this program using a simple 50/50 mixture of Allopurinol and it’s metabolite Oxypurinol (both at 0.25
mg/mL, Figure T1.53), since drugs and their metabolites may not be easily separated using ion exchange (or other
chromatography). It must be stated that these two chemicals can be identified by their unique non-overlapping
peaks at 717 and 651 cm-1, respectively. The results of the S-Quant program are shown in Figure 59D. The
unknown spectrum was fit with equal contributions of Allopurinol and Oxypurinol, at 56.0% and 55.8%
respectively. The total is not 100%, because the program is attempting to fit the noise in the spectrum with 1-2% of
other chemicals (some negative) to achieve a flat baseline. In other words, the program is correctly indicating a
50/50 mix of the two chemicals. Since all of the library spectra are 1 mg/mL concentrations, this program only
gives relative concentrations. Actual concentrations would require a calibration curve for each chemical.

25


A
D
B

C

Figure T1.53. SERS of A) 50/50 mixture of Allopurinol and Oxipurinol (0.25 mg/mL each), and pure B) ALLO
and C) OxP. D) S-Quant Results (part of LabView front panel, see Report #8).

Questions to be answered? Are all of the 24 chemicals SERS-active? All of the chemicals that we were able to
obtain were SERS-active. What is the preferred sol-gel for each bio-indicator and target drug? Six sol-gel libraries
allowed measuring all of the target analytes as summarized in Tables 4.3 and 4.4. Are there potential interferants?
Yes, such as histidine for 3-methyl histidine. How well does the spectral match software identify each? It has
proven exceptional, correctly identifying any of the target analytes in a library of over 100 chemicals that could
potentially be in urine. How well does it identify a chemical not in the library? The software ranks all of the
chemicals based on how well their spectra match the measured sample. If there is a close match (i.e. and HQI score
less than 0.5) then the unknown is probably very similar in structure. If no score is below 0.75, then the chemical is
not in the library and likely rather different than any of the library chemicals.

Task 2 – Chemical Selectivity Development. The overall objective of this task was to refine the ability of the 4
basic SER-active sol-gels developed in Phase I to selectively extract biochemicals and drugs present in human urine,
and enhance their Raman spectra to improve sensitivity. This was accomplished by measuring SERS-activity as a
function of flow time for the 25 analytes using the 6 chemically selective sol-gel libraries selected in Task 1.

As stated, the Task 1 results indicated that 6 chemically selective sol-gels allowed measuring the 25 target analytes
at 1 mg/mL in static measurements. During the Phase I program we determined that flowing the analytes through
the sol-gel capillaries improved sensitivity if the sol-gel extracted the analyte, effectively increasing the
concentration. In this task, the biomarkers were prepared at 1 mg/L, while the drugs were prepared at 10 microg/L
to determine if this approach could achieve the required concentration ranges for analysis.

The ability of the sol-gels to extract the target analytes was investigated by measuring the SERS signal as the sample
flowed through a capillary as a function of time. The apparatus is shown in Figure T2.1. In one configuration
Figure T2.1A), a peristaltic pump (VWR model 54856-070, West Chester, PA) was used to cycle the sample (20
mL) through the capillary at a rate of 1 to 2.5 mL/min until a signal was observed, then a 3-way valve was used to
switch the stream to a solvent, either water or methanol to flush the analyte out of the sol-gel. SER spectral
collection was initiated when the sample solution entered the capillary and spectra were collected continuously (20
sec/spectrum) for ~5-30 minutes. Early on it was found that the sol-gels that yielded the most sensitive signals were
those in which the analyte was bound irreversibly. Consequently, the solvent flush was essentially unnecessary, and
instead, a syringe pump (Sage model 341B, Thermo Electron, Waltham, MA, Figure T2.1B) was used to monitor the
time required to generate a signal. Typically, a 50 mL syringe filled with a 20-50 mL sample was passed through
the sol-gel filled capillary at a rate of 1 to 10 mL/min (typically 2 mL/min). It is worth noting that sol-gel plugs of
the MTMS based chemistry (L2) could handle flow rates greater than 10 mL/min, while those using the TMOS and
ODS based chemistries (L1 and L3 respectively) became detached at rates greater than 2.5 mL/min. This limitation
was overcome to some degree by adding the polymer component (PEG or PDMS) to the L1 sol-gel to produce L5
and L6. It also improved sensitivity.

26


A B
Sample Solvent

3-way
valve
XY
Peristaltic Stage
Pump

Syringe Pump
Laser Spot on Capillary
Fig. T2.1. Photographs of apparatus used for flow measurements. A) Peristaltic pump with 3-way valve to flow
sample until detection, then solvent to remove (20 mL vial reservoirs). B) Syringe pump for continuous flow to
monitor signal increase due to chemical extraction by the sol-gel (10-50 mL sample in 50 mL syringe).

The first representative biomarker tested (during Phase I) was 1-MeHis (mislabeled as 3-MeHis by the manufacture,
see Fig T5.2). A 1 mg/L solution was prepared and flowed at 2 mL/min through a capillary containing a 2 cm plug
of L1 (polar sol-gel). As can be seen in Figures T2.2 and T2.3, the spectral intensity (based on the 1564 cm-1 peak
height) reaches 80% of the maximum in the first 20 seconds. A 10 microg/L sample was also measured, and
although a signal was perceptible after just 5 minutes, it was somewhat unstable. Similar experiments were
performed using 1 mg/L samples on the non-polar sol-gels L2 and L3, and in both cases, the 1564 cm-1 peak was
detected, but at a lower intensity.
0.07
A
0.06
1564 cm-1 band intensity

0.05
B
0.04

0.03

0.02
C

0.01

0.00 D
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00

Time (min)
Fig. T2.2. Concentrating 1-MeHIS on L1 as a function of
flow. Conditions: initial concentration was 0.001mg/mL, Fig. T2.3. SERS of 1-MeHIS on L1 after flowing A) 100-
and flow rate was 2 mL/min, spectra: 100 mW of 785 nm, sec (pt.5), B) 80-sec (pt.4), C) 60-sec (pt.3), D) 40-sec (pt.2),
20-sec each. as in Fig.4.61

CRE was also measured as it was flowed through an L1-filled capillary. The signal for a 1 mg/L samples slowly
increased and became significant after 12 minutes (Figures T2.4 and T2.5). Next, the bone-loss indicator, HO-PRO
was tested (we did not have sufficient quantities to test H-Pyd). Initially, this biomarker, active on L4-filled
capillary (gold-doped) was difficult to measure. We found that a second reduction step dramatically improved the
overall sensitivity, typically by a factor of 100. This double reduction allowed measuring 1 mg/L HO-PRO. (It
also allowed detecting many of the drugs at 0.01 mg/L (10 microg/L) for HCT, PEN, SCOP, PROM and AM that
were nominally active on gold.) Here a 1 mg/L HO-PRO sample was flowed through a doubly reduced L4
capillary. The sample signal increased rapidly within 1 minute, as shown in Figure T2.6.

27


As a final example, uric acid was measured during flow through an L1 filled capillary. Here the signal increased
slowly, but after 7 minutes, the laser spot on the capillary was moved and a significant signal appeared, indicating
that the laser was degrading the sol-gel (Figure T2.7)
4.0

3.5

3.0
643 cm-1 Peak Area

2.5

2.0

1.5

1.0

0.5

0.0
0 2 4 6 8 10 12 14
Time (min)
Fig. T2.4. Concentrating CRE as a function of flow on L1 Fig. T2.5. SERS of CRE (pt. 29) after 10 minutes of flow.
with HCl wash. Conditions: initial concentration was 0.001 Conditions as in Fig. 4.63.
mg/mL in HPLC water and flow rate was 1 mL/min,
spectra: 100 mW of 785 nm, 20-sec each.

350 500
1000 1250 1500750 1850
Raman Shift, cm-1
Fig. T2.6 SERS of HO-PRO on L4 A) after 4-min flow B) 20- Fig. T2.7. SERS of UA on L1 after 7 minutes of flow.
sec flow. Conditions: 1 mg/L in HPLC water, flow rate 1.5 Conditions: 1 mg/L in HPLC water, flow rate 1 mL/min, 100
mL/min; spectra: 80 mW of 785 nm, 60-sec. mW of 785 nm, 20-sec.

The first drug tested using flow to concentrate the analyte was raloxifene (Phase I). Initially, a 0.001 mg/mL
sample was flowed at 2 ml/min through an L1 capillary, and the SERS appeared after 1 minute, reaching a
maximum in just 2 minutes, as shown for the1166 cm-1 peak Figure T2.8 (baseline at 1200 cm-1 subtracted). Since
the desired sensitivity for raloxifene is considerably lower (~ 10 microg/L) the flow experiments were repeated
using 100 microg/L. Again, the SER spectrum is detected, but after longer flow times (Figure T2.9).
0.18
A

0.16
0.14
0.12
0.10 B
0.08
0.06
0.04
0.02
0.00
0 1 2 3 4 5 6
Time (min)
Fig. T2.8. Concentrating RAL on L1 as a function of Fig.T2.9. SERS of RAL on L1 A) 1 mg/L after 2-min (see
flow. Initial concentration was 1 mg/L, 100 mW of 785 Fig. T2.8), and B) 0.1 mg/L after 5-min.
nm, 20-sec each.

28


The ability to detect the other drugs at microg/L concentrations using the flow technique was also examined. The
following drugs represent the variety of results. The SERS were measured for a 5 microg/L ALLO sample as it
flowed through an L2 filled capillary. At this low concentration, more time (over 12-min) was required to obtain a
detectable signal (721 cm-1 peak, Figures T2.10 and T2.11). In the case of RIS, 20-min were required for a 10
microg/L sample flowed through an L1 capillary (Figures T2.12 and T2.13). It was also noted, that moving the laser
position along the capillary dramatically increased the signal intensity, suggesting that either a relatively inactive
spot was originally chosen, or the laser caused sol-gel degradation. In the case of ETI, again ~12 minutes were
required to detect the SERS, but shortly thereafter the signal began to decrease (Figures T2.14 and T2.15). This also
suggests that the sol-gel or the analyte may be degrading. In all cases efforts continued to improve the uniformity
and stability of the sol-gels, principally by the incorporation of the polymers into the alkoxides.

0.10
0.09

0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00
Time (min)
Fig. T2.10 Concentrating ALLO as a function of flow on Fig. T2.11. SERS of ALLO (pt. 34) after 12 minutes of flow
L2 Conditions: initial concentration was 0.000005 mg/mL in Conditions: 0.000005 mg/mL on L2 in HPLC water, flow
HPLC water and flow rate was 2.5 mL/min, spectra: 100 rate of 2.5 mL/min, 100 mW at 785 nm, 20-sec.
mW of 785 nm, 20-sec each.
1.60

1.40
A
1.20
1.00
0.80
0.60
B
0.40
0.20
0.00
0.00 5.00 10.00 15.00 20.00 25.00

Time (min)
Fig. T2.12. Concentrating RIS as a function of flow on L1. Fig. T2.13. SERS of RIS A) (pt. 53) after 20 minutes of
Conditions: initial concentration was 0.00001 mg/mL in flow, and B) (pt. 63) on new spot. Conditions: 0.00001
HPLC water and flow rate was 2.5 mL/min, spectra: 100 mg/mL on L1 in HPLC water, flow rate of 2.5 mL/min, 100
mW of 785 nm, 20-sec each. mW at 785 nm, 20-sec.

29


0.18

0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00
Time (min)
Fig. T2.14. Concentrating ETI as a function of flow on L1.
Conditions: initial concentration was 0.00001 mg/mL in Fig. T2.15. SERS of ETI (pt. 32) after 10 minutes of flow
HPLC water and flow rate was 2.5 mL/min, spectra: 100 Conditions: 0.00001 mg/mL on L1 in HPLC water, flow rate
mW of 785 nm, 20-sec each. of 2.5 mL/min, 100 mW at 785 nm, 20-sec.

As described in the proposal, flow could be mimicked
by “pistoning” the sample back and forth through the
sol-gel to concentrate the analyte. This would allow the
use of a syringe to manually manipulate the sample, so
that significant urine samples were not required. To
perform these experiments, a syringe containing 100-
500 microL of sample was pushed back and forth
through the sol-gel 5 times (10 passes) and a spectrum
recorded. A typical example is shown for RIS in Figure
T2.16. As can be seen, the LMC was improved by a
factor of 10 compared to static measurements. In
general, the pistoning approach allowed measuring
samples to 1 mg/L, suggesting that it is a viable Fig. T2.16. SERS of RIS in an L1 capillary before and
approach for biomarkers, but NOT for the drugs. after pistoning (10 passes) at 0.1 mg/L. Conditions: 100 mW
at 785 nm, 20-sec.
The results for this task are summarized in Tables T2.1 and T2.2. Note that most of the biomarkers could be
detected at physiological concentrations by pistoning at 1 mg/L, while most of the drugs could be detected by
flowing at 10 microg/L (indicated in red).

Table T2.1. Summary of Biomarker flow and piston results (vs static): LMC in mg/L on sol-gel libraries.
CRE 3-MeHis HO-PRO HO-LYS H-Pyd H-Dpd CaP CaOx UA CYST
*
Static 1.0 1.0 100 1000 18 dnt 500 500 500 1000
L1 L1 L4 L3 L3 L2 L2 L1
Piston 1.0 1.0 10.0 neg 1 1.8 dnt 1.0 5.0 5.0 neg 1
L1 L1 L4 L1 L3 L3 L6** L2 L1
Flow 1.0 0.01 1.0 neg 1 0.9 dnt 1.0 1.0 1.0 1.0
L1 L1 L4 L1 L3 L3 L6** L2 L1
* = sol-gel chemistry with TMOS only, ** = sol-gel chemistry 2a with PDMS
Table T2.2. Summary of Drug flow and piston results (vs static): LMC in mg/L on sol-gel libraries.
CLO ETI PAM ALE IBA RIS RAL HCT ALLO PEN SCOP PROM AM
Static 100 100 100 100 10* 1 10 100 0.1 100 1.0 1.0 1.0
L6 L5 L5 L1 L1 L1 L6 L2 L4 L4 L4 L3
Piston neg1 neg1 neg1 1.0 neg1 0.1 neg1 neg1 0.01 neg1 1.0 1.0 1.0
L1 L1 L1 L6 L1 L1 L1 L3 L2 L2 L4 L4 L3
Flow 1.0 0.01 neg1 0.01 neg1 0.01 0.01 0.01 0.005 0.01 0.01 0.01 0.01
L2 L1 L1 L6 L1 L1 L1 L4 L2 L4 L4 L4 L4
* = sol-gel chemistry with TMOS only

30


Questions to be answered? Which sol-gels provide irreversible adsorption of the bio-indicators and target drugs
and SER-activity? This data is summarized in Tables T2.1 and T2.2. What are the estimated LODs? Instead of
calculating LODs, we improved sensitivity sufficiently to measure the analytes at the required physiological
concentrations (again see the same tables). Can the spectral deconvolution software identify each urine component
(bio-markers and drugs) on each sol-gel? Yes, see Task 1.

Task 3. Component selection and testing. The overall goal of this task was to design a lab-on-a-chip that can be
used to analyze chemical components present in human urine by SERS. This was accomplished by performing a
series of tests to determine the components and functionality required for the lab-on-a-chip.

A preliminary design illustrating the required functionality that such a chip should have was presented and discussed
in the Phase II proposal. That design and discussion are repeated here verbatim, followed by the experiments
performed to select the components.

1. A sample compartment will be used to accept a liquid sample 100ml Tank
(see components in Fig. T3.1).
2. The urine sample will be drawn through a micron filter to
remove any entrained particles. Filter
3. The sample will continue through an inorganic extraction
material (ion retardation resin) that will remove inorganic salts. Ion Retardant
4. The aqueous sample will be extracted with a polar solvent
(dichloromethane). Neutral organic species (drugs) will be
extracted into the organic phase. Note that this step will require Extract Aq/Org
modifications for application in a microgravity environment.

rg
5. The organic phase will be directed into a series of SER-active

Aq
O
sol-gels designed to selectively extract the target bio-indicators
pH NaOH
and drugs by type (Set 1). This is NOT a chromatographic style
separation. A) A very-weakly-polar (essentially non-polar)

a

b
silver-doped sol-gel will extract non-polar-negative drugs/ Cation Exchange
biomarkers (NP-neg), but pass the other chemicals. B) A

a

b
weakly-polar gold-doped sol-gel will extract weakly-polar-
positive species (WP-pos), but pass other chemicals. C) A Set Set Set
weakly-polar silver-doped sol-gel will extract weakly-polar- One Two Three
negative species (WP-neg), but pass other chemicals. D) A
negative-polar silver-doped sol-gel will extract polar-negative
species (P-neg). As an example of selectivity, we have seen that
Waste
scopolamine and hydroxyproline are most active on
electronegative gold, and that alendronate and 3-methyl-histidine
are most active on electropositive silver (see Fig. T3.2). Fig. T3.1. Flow diagram used to define
separation components.

A D
B C

B
C
A
D

Figure T3.2. SERS 1mg/ml 100 mW, 1min, 785 nm A) Figure T3.3. SERS TMOS/MTMS 100 mW, 1-min, 785 nm A)
ALE, Ag-TMOS/MTMS , B) 3-MeHIS, Ag-MTMS, C) Reconstituted urine, B) reconstituted urine doped with 0.001
HO-PRO, Au-TMOS, D) SCOP, Au-TMOS/MTMS. mg/mL of 3-MeHIS and RAL in a 50:50 mixture, C) RAL
extracted from organic phase (dichloromethane) component 5, D)
3-MeHIS extracted off cationic exchange column (component 9).

31


6. The aqueous phase will be directed into a chamber, where the pH will be adjusted with 1M HCl to a final pH of
1.0 to ensure complete protonation.
7. This pH adjusted aqueous sample will enter into a cation exchange column. The protonated species will
strongly adsorb onto the material, while the basic/neutral species will quickly pass through.
8. These basic/neutral species will enter into a series of SER-active sol-gels designed to selectively extract the
target chemicals by type (Set 2). Set 2 follows a similar sequential sol-gel configuration as described for Set 1
above.
9. A NaOH gradient mobile phase from 0.1 M to 1.0 M will then be used to elute off the remaining species that are
adsorbed on the column. Species elute off the column in order of increasing PKa/pI values (i.e. as pH increases
due to increasing NaOH concentration, the protonated species will elute when the pH equals/exceeds their
respective pKa/pI values. Each fraction is subsequently directed into the appropriate sol-gel (Set 3).
10. The feasibility of this extraction, separation, and detection method to be incorporated into our proposed lab-on-
a-chip is demonstrated in Figure T3.3, where a lyophilized male urine sample (Sigma-Aldrich) was
reconstituted with distilled water and doped with 3-MeHIS and RAL.
11. To optimize the sensitivity, the solution will be “pistoned” in and out of the sol-gel region numerous times. The
Phase I data suggest that sensitivity can be improved by up to 1000 times. This coupled with 532 nm laser
excitation suggest that sensitivity at the part-per-trillion level can be achieved (see below).

The following basic experiments were performed to evaluate the use of a filter and the value of an organic solvent in
the above flow diagram. Since urine contains many large MW molecules and numerous salts, sometimes as
particulate mater, it is necessary to include a filter to remove these substances. A saturated solution of uric acid was
prepared and measured in a SERS-active capillary before and after filtration. A 0.2 µm pre-cut 13 mm Nylon 66
membrane filter (in a filter holder) was used to remove the particulate matter (Figure T3.4). As can be seen the
filtrated sample yields a better quality spectrum (more pronounced peaks and less noise). It was also found that all
of the samples could be passed through this filter without diminishing signal, including urine samples at various pH,
as well as temperature (samples became turbid at room temperature).

A B

Before After

C

Nylon Filter

Fig. T3.4. A) Photograph of syringe and in-line Nylon 66 membrane particle filter and saturated uric acid solutions
before and after filtration. SERS of UA B) before and C) after filtration in L2-filled capillaries.

The role of the organic phase was to extract non-polar drugs and possibly biomarkers into this phase for SERS
measurements, with the goal of reducing the complexity of the spectra analysis. After reviewing the solubility
properties of the 25 target analytes and performing a few experiments, it was realized that only Raloxifene fell into
this category (as shown in Figure T3.3). Since this drug also dissolves in water, the organic solvent extraction step
is not necessary and it was eliminated.

The next series of experiments examined the chemical composition of urine (simulated, reconstituted, and real) in
terms of the Raman and SER spectra, how it and its major components change with pH, and how the proposed ion
retardation and ion exchange filters can be used to remove unwanted urine components and separate the target
analytes. The major components of urine: urea, creatinine, uric acid and lactic acid, and lyophilized urine were
purchased from Sigma-Aldrich. Each of these four urine components as pure solids produce distinctive Raman
spectra (Figure T3.5), but only the urea produces a perceptible Raman signal in either simulated (urea 20 mg/mL,
creatinine 1.4 mg/mL, uric acid 0.15 mg/mL and lactic acid 0.2 mg/mL) or lyophilized urine (reconstituted as 1g/30

32


mL water, pH = 5.67, Figure T3.6). In contrast, only uric acid and creatinine produce SER spectra at neutral pH
(Figure T3.7). Urea does not appear to produce a SER spectrum under any condition, while lactic acid requires a
very acid pH (see below). Furthermore, it is found that the SER spectra of simulated, reconstituted, and real urine
samples are dominated by uric acid with some contribution from creatinine (Figure T3.8).

A A

B
B

C
C
D

Fig.T3.5. NRS of Urine components A) Urea, B) Creatinine, Fig. T3.6. NRS A) Urea (20 mg/mL), B) Simulated urine (see
C) Uric acid, and D) Lactic acid. Conditions: pure solids, 290 text), and C) Reconstituted urine (10,000 mwt cutoff),
mW at 785 nm, 10-min. Conditions: 290 mW at 785 nm, 10-min.

O
A
N
A
H2N NH2
O
N NH2

B B
OH

H
N
N
O

N
OH HO N H

C C
OH

D D
O

Fig. T3.7. SERS of Urine components. A) Urea on L3, B) Fig. T3.8. SERS of A) Simulated urine on L1, B)
Creatinine on L1, C) Uric acid on L1, and D) Lactic acid on Reconstituted urine on L3, C) Real male urine sample (from
L1, Conditions: 1 mg/mL (UA 0.5 mg/mL). Note: all appear to volunteer at RTA) on L1, and D) Reconstituted urine on gold
be inactive on gold L4 (not shown). L4. Conditions: as prepared in Fig. T3.6.

As part of the Phase I study, the pH dependence of creatinine, uric acid and lactic acid were studied in detail.
In each case, stock solutions of the analyte were prepared at 1 mg/mL and then pH adjusted using HNO3 or NaOH
(no buffers, verified by pH electrode), and then for each pH a sample was drawn into a separate SER-active capillary
for measurement. For creatinine, samples were prepared from pH 11 to 3, and their spectra measured (10.8, 7.5, 6.9,
5.9, and 4.5 shown in Fig. T3.9A). The SER signal intensity was good at neutral pHs, but degraded at 4.5 and 10.8.
It was noted that the peak at 614 cm-1 decreased with increasing pH value, while peaks at 675 and 1740 cm-1
increased, and bands at 850, 930, and 1420 cm-1 stayed relatively constant. A plot of the 614, 675 and 1740 cm-1
peak intensities divided by the 1420 cm-1 peak intensity as a function of pH is shown in Fig T3.9B. A plot of the pH
dependence for each of the ion concentrations based on the pKa’s shows that the 614 cm-1 peak corresponds to the
cation species, and the 675 and 1740 cm-1 peaks correspond to the neutral species. It is important to note that in the
pH range of 5-9, which encompasses the urine pH range (5.5 to 7.5), the creatinine SERS spectrum does not change.

33


0.01
Creatinine++ Creatinine+ Creatinine
10.8 1740 0.009
0.008
7.5

Concentration [M]
0.007
0.006
6.9 A B
0.005
pK1 =4.83 pK2 =9.20
614 5.9 0.004
0.003
4.5 0.002
0.001
0
0 2 4 6 8 10 12 14
Wavenumbers (∆cm-1) pH
Figure T3.9. A) SER spectra of 1 mg/mL CRE at pHs indicated (100 mW 785 nm, 1-min, L3 (chem2c)). B) Plot
of 614 ( ), 675 ( ), and 1740 cm-1 (■), normalized band intensities as a function of pH representing CRE++, CRE+
and CRE, respectively. Concentrations of CRE++, CRE+ and CRE based on pKa’s as a function of pH are

Uric acid showed little pH dependence as shown for spectra collected at pHs of 9.10, 7.41, 6.05, 5.25, and 4.08
(Fig. T3.10). The only indication of a change in molecular species as the pH transitions the pKa of 5.4 is a slight
change in the peak at 595 cm-1, which appears to be absent at pH of 4.08. It was noted that at basic pHs (near and
above the other pKa of 10.3) the solubility of uric acid in water was higher.

Lactic Acid samples were prepared from pH 11 to 3, and their spectra measured (9.7, 5.6, 4.4, and 2.9 shown in Fig.
T3.11). Lactic acid is only SER-active in the neutral form below the pKa of 3.08 and consequently will not be
observed in urine at relevant physiological pH’s.
9.10
9.7
7.41
relative intensity

5.6
relative intensity

6.05
4.4
5.25

4.08 2.9

Fig. T3.10. UA SERS pH dependence on L3 Figure T3.11. SERS of 1 mg/mL LA on L3 (chem2c) at
(chem2c), 0.5 mg/mL. 100 mW, 785 nm, 1-min. pH’s indicated; 100 mW, 785 nm, 1-min.
This data suggests that lactic acid will not interfere with measurements of biomarkers or drugs at physiological pHs,
but creatinine and uric acid will. Although it may be possible to remove their spectra contributions, this will be
insufficient, since their nominally high concentrations in urine are likely to block the target analytes from reaching
the SER-active metal. It is clear that these urine components must be removed from the sample in order to perform
analysis of the target analytes.

As proposed, an ion retardation resin was examined to determine if it would remove creatinine and uric acid. A
sample of reconstituted urine at pH 5.7 was pH adjusted to 4.6, and 8.7 to produce 3 samples covering the normal
pH range of urine. The SER spectra for the 3 samples were measured before and after passing through a 5 mL
disposable glass pipette packed with ~1 mL volume of a ion retardation resin (AG 11A8 by BioRad, MA). At pH
4.6 the SER signal is diminished, while at pH 5.7, uric acid dominates the spectrum. At pH 8.7 the uric acid signal
is slightly diminished with a trace of creatinine present. However, in all 3 cases, the ion retardation resin does not
appear to change the spectra, and certainly does NOT remove the uric acid.

34


A A

B B

C C

Fig.T3.12. SERS of Reconstituted urine on L2 before Fig. T3.13. SERS of Reconstituted urine on L2 after ion-
filtration at pH A) 5.7, B) 4.6 and C) 8.7. Note: A) on L1 retardation step, for pH A) 5.7, B) 4.6, and C) 8.7. But here
A) on L2.

Although the focus of the last experiment was to determine if the ion exchange resin could remove creatinine and
uric acid, it also was included as a necessary component to remove dissolved salts that the filter might pass,
particularly NaCl and KCl. Such salts at high concentration can potentially alter the SERS response through
surface-deactivation or particle aggregation. To test the effectiveness of the ion retardation resin to eliminate such
salts, a 1 ml solution containing 0.5 mg/mL of SCOP and KSCN in water was measured on gold L4 (Figure
T3.14B). KSCN was chosen since it produces a very intense SER peak at 2100 cm-1, while the other salts only
produce a Ag-Cl peak at ~ 230 cm-1, which is not observed due to a low-wavenumber cut-off filter in the Raman
analyzer. SCOP was included as a control, since it should appear in the sample spectrum even when passed through
the resin. As shown (Figure T3.14A), the spectrum of the sample passed through the resin is dominated by SCOP
and there is no evidence of SCN-, indicating that this ion had been effectively trapped. It is worth noting that
measurements using a silver-doped sol-gel yielded similar results.

As an initial example of the value of the ion retardation resin, a 0.5 mg/mL sample of allopurinol in reconstituted
urine was measure before and after it was passed through the resin (Figure T3.15). As shown, the removal of salts
improves the spectrum substantially. Although the ion retardation resin did not remove the creatinine or uric acid it
is still an essential component for the proposed lab-on-a-chip, especially since high salt levels diminish the
separation capabilities of the ion-exchange resins (see below).

A
A

B
B
C

Fig.T3.14. SERS of a 1:1 mixture of SCOP and SCN on L4. Fig. T3.15 SERS of 0.25 mg/mL ALLO A) in reconstituted
A) After and B) before passing through ion-retardation urine and B) after filtration and ion-retardation (pH 6.8), and C)
column. On separate L4 capillaries. in pure water for comparison. All on L1.

The next component tested was the ion exchange resin. Again, the first experiment examined the SER spectra
obtained from urine at various pH to determine if uric acid could be removed as a function of pH. As above, a glass
pipette was filled with a strong cation-exchange resin (model AG50W, Bio-Rad). A 1g lyophilized urine in 30 mL
water sample was prepared and pH adjusted to 1.0 using 1M HCl. 1 mL of this sample was passed through the resin
to “load” the column. Next, NaOH was passed through the column, 1 mL at a time, each increasing in strength from

35


0.1M to 1M. Samples were collected in vials then pH 1
transferred to SER-active capillaries and measured
(Figure T3.15, not a 96 well plate as stated in Quarterly 2
Report #6). As the pH increased the SERS of uric acid 4
became apparent from 4-6, while creatinine dominated
at pH 8. 5
6
Next, the 25 target analytes were each separately loaded
on this ion exchange resin and measured as a function of 8
pH, by eluting samples using increasing 1 mL NaOH
12
concentrations and measuring the SERS-activity. In
time it was found that the biomarkers could easily be
measured at the require 1 mg/L, and most of the Fig. T3.15. SERS of reconstituted urine collected from an ion
drugs could be measured at the required 10 exchange column collected as pH fractions of ~1, 2, 4, 5, 6, 8,
microg/L. For example, 3-MeHis, RIS, CaP, Allo, and 12 measured on L1 capillaries, eluted off with NaOH. Note
and OXP were measured at 1 microg/mL (Figures UA and CRE contribution at pH 2-6, and 8 respectively.
T3.16 and T3.17), HO-LYS and IBA at 10
microg/mL (Figures T3.18), and even H-Pyd at 18 microg/mL (Figures T3.19). Note that oxypurinol was captured
on an anion exchange resin and eluted off using an acid (Figure T3.17C).

A A

B

B
C

Fig.T3.16. SERS of 0.001 mg/L A) 3-MeHIS and B) RIS Fig.T3.17. SERS of A) 1 mg/L CaP and B) 0.001 mg/L ALLO
on L5; pre-concentrated with cation exchange resin (pH on L5; pre-concentrated with cation exchange resin (pH ~2)
~2), and eluted off with 1M NaOH. and eluted off with 1M NaOH; C) OXP 0.001 mg/L on L3 pre-
concentrated with anionic exchange resin (pH~10), and eluted
off with 1M HCl.

A

B

Fig.T3.18. SERS of 0.018 mg/L H-Pyd on L3 pre-concentrated Fig.T3.19. SERS of A) 0.01 mg/L HO-LYS and B) 0.01 mg/L
with cation exchange resin (pH ~2), and eluted off with NaOH IBA, both on L1t (HCl washed), pre-concentrated with cation
base gradient. exchange resin (pH ~2), and eluted off with 1M NaOH.

Towards the development of the lab-on-a-chip, methods were developed to load the ion retardation and exchange

36


resins into the 1-mm glass capillaries (Figure T3.20A). This was accomplished by first loading a ~0. 5 mm plug of
MTMS sol-gel (no metal doping) into the capillary and letting it gel to form a porous frit to hold the resins. Then
the resins were dissolved in water and ~ 2 cm segments were loaded. Several experiments were performed to
demonstrate that the resins within these capillaries function as designed. These capillaries, the SERS-active
capillaries, the filter and a syringe as the method of delivering and flowing sample formed the basis of the lab-on-a-
chip components (Figure T3.20) and the first experiment to evaluate the design.

B: ion-retardation resin

D: filter E: syringe
B
A: MTMS Frit F:
C
urine
sample
C: ion-exchange resin G: SERS-active capillary

Fig.T3.20. Photographs of A) MTMS frit, B) ion retardation resin, C) ion-exchange resin, D) filter in filter holder,
E: syringe, F: urine sample, G: SERS-active capillary (L1).

The ideal test sample would contain a biomarker and a drug associated with bone loss or muscle loss (stone
formation is less important). Unfortunately, such a pair could not be measured (the bone loss biomarkers H-Pyd and
H-dPyd available only in acetic acid, cost too much to perform but a few measurements, and there are no accepted
preventive muscle loss drugs). Consequently, the combination of 3-methylhistidine and Risedronate became the
best test case for a biomarker and drug that would realistically be found in an astronaut urine sample.

An artificially doped urine sample was prepared by adding 1 mL of a 1/1 volume mixture of 3-methylhistidine and
Risedronate, both at 1 mg/mL in HPLC water, to a 1 mL reconstituted urine sample (0.5 g in 15 mL HPLC water).
This produced a urine sample containing 0.25 mg/mL of each analyte. This sample solution was vortexed and
allowed to equilibrate at room temperature for 5 minutes. Two 10 microL samples were drawn into SERS-active
capillaries (L1) by syringe and measured, one before and one after passing the 2 mL urine sample through the nylon
filter. In this case and all others, Tygon tubing connected the syringe to the capillary. There was no noticeable
difference between the spectra (Fig.T3.21A and B). The entire urine sample was then drawn through a capillary
filled with ion retardation resin, and again ~10 microL was drawn into a SERS-active capillary and measured.
Peaks associated with both 3-methylhistidine and Risedronate are readily apparent, but so are peaks associated with
uric acid (Fig.T3.21C). Nevertheless, these spectra again show that the ion-retardation resin substantially improves
the SERS-response, presumably by removing dissolved salt ions.

RIS E
UA 3-MeHIS

C

B D

A

Fig.T3.21. SERS of reconstituted urine sample doped with 3-methylhistidine and Risedronate, A) before and B)
after filtration, and C) after passing through an ion retardation (IR) capillary, and after passing D) 0.1 M KOH (3-
MeHIS) and E) 0.5 M KOH (RIS) through the ion exchange (IEX) capillary. Note that the analytes can be identified
in C), but are unmistakable in D) and E).

37


Next, 1 mL of the artificially doped urine sample was pH adjusted to 2 by adding 1M HCl. Approximately 25
microL of this sample was loaded onto an ion exchange resin (cation). Next, three 1 mL solutions of 0.1, 0.5, and
1.0 M KOH were passed through the resin, while collecting the eluted sample, 2 drops each, into the wells of a 96-
well micro-plate. Approximately 1 drop was drawn from each of the 30 filled wells into SERS-active capillaries and
measured (somewhat time consuming). It was found that 3-methylhistidine was carried off the column by the 0.1 M
KOH (wells 1-10, and Risedronate by the 0.5 M KOH (wells 12-20). This is consistent with the pKA’s for these
chemicals, 5.9 and ~8, respectively, since these base concentrations result in pHs of ~6 and 8, respectively.

Although reasonable SER spectra were obtained, it is clear that uric acid is still present in the sample containing 3-
methylhistidine. For this reason, a urine sample doped only with 3-methylhistidine was prepared and measured
using various sol-gel libraries after following the above procedure. It was found that sol-gel chemistry L3 yielded a
quality SER spectrum of this analyte with little contribution from uric acid (Figure T3.22).

A
A

B

C
B

D

Fig.T3.22. SERS of 1 mg/L 3-MeHIS separated from Fig.T3.23. SERS of A) 1 mg/L 3-MeHIS and B) 0.01 mg/L
reconstituted urine (see text) on A) L3 (chem2c), B) L2, C) L1 RIS separated from reconstituted urine (see text). Both on
and D) L5. Note that uric acid is absent in A-C), but not D). L3 (chem2c). Note: separate urine sample for each analyte.

Next a series of experiments were performed using urine samples artificially doped with 3-methylhistidine and
Risedronate at physiologically relevant concentrations to determine the volume of sample required, the amount of
ion retardation and exchange resins required, the volume of acid to prepare the ion exchange column, and the
volume of base to elute the analytes. It was found that the following conditions yielded very good results:
a sample volume of 1 mL, 250 mg of each resin loaded into capillaries, 25 microL of 1M HCl to acidify the ion
exchange resin, 0.5 mL of 0.1M NaOH for 3-methylhistidine or 0.5M NaOH for Risedronate, and collecting the first
100 microL for SERS analysis. These conditions were used to successfully measure 1 mg/L 3-methylhistidine and
0.01 mg/L Risedronate artificially added to reconstituted urine (separately, Figure T3.23)

Finally, based on all of these experiments, a lab-on-a-chip would consist of the components as shown in Figure
T3.24. Note that the use of a continuous base gradient and capillaries (or channels) would be used to measure
additional biomarkers and drugs.

Figure T3.24. Illustration of the
lab-on-a-chip components that
will effectively separate 1 mg/L 3-
methylhistidine and 0.01 mg/L
Risedronate from urine. The
analysis should take less than 10
minutes. Compare to Fig. T3.1.

38


Questions to be answered: Are there components available to perform the desired extractions and separations?
Yes, the evaluation results are summarized in Figure T3.24. Are they effective in the context of our SER-active sol-
gels, separately, and together? Yes. What is the best sequence of sol-gels? L3 for both biomarkers and drugs. How
universal is it for various urine components and drugs? L5 is best for UA, while L3 (chem2c) excludes UA. Are the
mixtures effectively separated and detected? Yes, see Fig. T3.21. What happens when other chemicals from Task 1
are analyzed? Method is amenable to all analytes; spectral deconvolution software provides confirmation. What are
the detection limits? See lowest measured concentrations in Tables T1.3 and T1.4, and T2.1 and T2.2., in general 1
mg/L for biomarkers and 10 microg/L for drugs achieved.

Task 4. Lab-on-a-chip fabrication (with Advanced Fuel Research as subcontractor). The overall goal of this
task was to build many microfluidic chips that can be used to test the preliminary lab-on-a-chip design. This was
accomplished by fabricating two successful polymethylmethacrylate (PMMA) chip designs after making
considerable modifications to the process described by Muck et al.

This task consisted of several parts, 1) preparing a silicon master chip (AFR subcontract), 2) producing PMMA
chips, 3) incorporating SERS-active sol-gels and separation materials into the chip channels, and 4) performing
SERS measurements using the chips. This task was substantially delayed as the initial design mask was incorrect,
and the photo-resist identified by Muck could not be used as described (it is our opinion that much of the chemistry
described by Muck was intentionally miss-stated). Together, these circumstances caused more a 5 month delay in
delivery of the first chips (May, 2006), and consequently we requested a 6 month extension.

While we awaited the chips we focused on part 3) learning to load smaller and smaller channels with our SERS-
active sol-gels. The initial test chip included a variety of channel widths for testing, including ranging from 100 to
500 microns (see below). Consequently, we ordered glass and plastic capillary tubing ranging from 200 to 600
microns. Initial measurements were performed on 600 ID glass capillaries, as these were not much smaller than the
800 micron ID of our standard 1-mm capillaries, and the volume would be close to a 500 micron wide channel that
is 250 microns deep.

Initial measurements were performed by loading the standard silver-doped sol-gel chemistries (L1 - L3) into the
capillaries. It was found that long plugs of 2 cm or more were very difficult to reduce, while short plugs of 0.5 cm
detached during reduction or sample loading. This was even true for L3, which incorporates octadecyltrimethoxy
silane, which produces relatively open structures due to the long organic chain. For these reason, several
modifications to the sol-gel chemistry were investigated. This included the incorporation examining internally
coated, instead of filled capillaries, incorporation of different alkoxides, such as PDMS, and modification of the sol-
gel curing process.

One way to overcome the restricted flow caused by the sol-gel plugs in the capillaries is to open the center of the
capillary so that flow would be unrestricted. This was accomplished by filling a capillary in the usual way (mixed
precursors drawn into the capillary by syringe), but then forcing the sol solution back out after 20 seconds, before it
gels. This process was repeated 6 times to form a thin film of sol-gel on the inner wall of the capillary (Figure
T4.1). This multi-layered approach also gives us flexibility in terms of coating thickness.

These open tubular capillaries were prepared using a modified chemistry L1 (TMOS only). First two 1 mg/mL
samples, pyridine and acetylsalicylic acid, were measured in internally coated capillaries of standard diameter, and
then a 1 mg/mL 3-methyl histidine was measured in a internally coated 600 micron ID capillary (Figure T4.2).

Not surprising, it was found that these coated capillaries were somewhat less sensitive, as the analyte has to diffuse
through the sol-gel to the SERS-active metal, as opposed to the standard capillaries, in which the analyte is forced to
the metal surface as it is introduced into the capillary.

39


A A

B

B

C

Figure T4.1. Microscope photograph of a 600 micron Figure T4.2. SERS of A) PYR, B) ASA, C) 3-MeHIS
ID capillary internally coated with a reduced silver- Conditions: A) and B) 800 micron ID, and C) 600 micron
doped sol-gel. ID capillaries internally coated with L1. Conditions: 1
mg/mL in HPLC water, 100 mw of 785 nm, 1-min.

To aid the development of methods to loading smaller and
smaller diameter capillaries (and ultimately chip channels), Gas Line
an apparatus used to coat gas chromatography columns
was purchased (Rinse Kit FOR GC columns, Part No.
23626, Sigma Aldrich, Figure T4.3). It consists of a Connection
reservoir for the sol-gel solution, a side arm for applying to capillary or
gas pressure, a small slit to fit a capillary into the reservoir chip channel
opening and a side support arm. To load capillaries, first
the reservoir is filled with a sol solution, and one end of the
capillary is inserted into the reservoir opening and sealed
with a graphite ferrule. Gas pressure (40 psi) applied
through a side arm forces the sol solution from the
reservoir through the column and into the capillaries.
Reservoir
Fig. T4.3. Apparatus for loading small capillaries and chip
channels.
The next method investigated to improve the flow through the sol-gels was the incorporation of polydimethyl
siloxane (PDMS) in the alkoxide precursor. As with ODS this siloxane should yield more porous sol-gels. As
mentioned previously, it also appears to adhere to glass and plastic capillary walls very well. 600 micron capillaries
were filled with modified L1 and L3 in which PDMS was added (see L6 chemistry in Table T1.2, Figure T4.4). The
PDMS-L1 modified chemistry (designated L6) in a 600 micron ID capillary was used to measure 1 mg/mL
Risedronate (Figure T4.5).

A

B
A

B C
C

Fig. T4.4. 600 micron ID capillaries; A) L6, B) L1, C) L3 Fig. T4.5. SERS of RIS A) 1mg/mL, L1, B) 1mg/L L6, and C)
(+PDMS). Sealed with rubber caps, cured 24 hours and 1mg/mL L6 600 micron ID capillary; 90 mW, 785 nm, 1-min.
reduced by standard method.

40


Another important requirement is that the sol-gels must bond to the PMMA channels of the chips. To test this idea,
a length of 500 micron ID PTFE (poly(tetrafluoro ethylene)) tubing was filled with L2. Upon curing (~2 hrs), the
sol-gel filled tubing was cut into segments (~5 cm each), which were reduced following the standard method. The
sol-gel did not detach during curing or upon flowing the reducing agent. In fact, one of these segments was reduced
5 days later and as shown in Figure T4.6, it not only remained attached but exhibited good SERS activity for ALLO.
It is anticipated that adherence will be even better in the PMMA channels due to the ability to form chemical bonds
directly with the silanol groups of the sol-gel.

A A

B B

Figure T4.6. A) SERS of 0.5 mg/mL Allopurinol in L2 sol-gel Figure T4.7. SERS of A) 1-MeHIS (Nt-MeHIS), 50 microg/L
loaded poly(tetrafluoro ethylene) tubing, and B) NRS of the urine and B) OxP, 5 microg/L water. Note: samples passed
PTFE tubing, 300 mW at 785 nm, 5-min. Insets, PTFE tubing through sol-gel based ion exchange resin.
with and without sol-gel.

Another requirement for developing the SERS-active microchips is the ability to load the separation materials into
the chip channels. Although we know that we can use MTMS sol-gel frits to hold these materials, we also
investigated the ability of designing ion exchange capability using our sol-gel chemistry. A sol-gel anion-exchange
material was created by using N-octadecyldimethyl[3-(trimethoxysily)propyl]ammonium chloride (C-18 TMOS,
100 microL) with TMOS (100 microL), and 95% triflouroacetic acid (200 microL) as a catalyst. Due to the
presence of a positively charged quaternary ammonium group in the C-18 TMOS, the resulting sol-gel coating
carries a positive charge.

This sol-gel anion-exchange resin was successfully incorporated into a 250 micron capillary. Two samples were
prepared to test the IEX capillary, 1-MeHIS (50 microg/L in urine) and OxP (5 microg/L in water.) First, a 100
microL sample of 1-MeHIS at 50 microg/L doped in reconstituted urine was pH adjusted (15 microL of 1M KOH to
impart a negative charge) and passed through this capillary. The sample was successfully extracted and pre-
concentrated. The sol-gel was washed with 500 microL of HPLC water to remove the sample matrix, following
which the extracted sample was eluted off the sol-gel coating by a 1M HCl solution (100 microL). The eluent was
then drawn into a standard SERS-active capillary (L1) and measured. The same procedure was followed for the
OxP. The SER spectrum for both analytes were weak, with 1-MeHIS dominated by uric acid peaks. Clearly the
extraction efficiency of this ion exchange material requires further development. Since the MTMS frits work,
attempt to develop a sol-gel based ion exchange resin was discontinued.

As the chemical-selectivity and SER-activity for the standard sol-gel libraries in terms of the target analytes was
being developed, it became clear that these libraries yielded the desired results, and that the sol-gel process, not the
chemistry, should be investigated. To this end, the relative precursor volumes and gelling conditions were
investigated (temperature, gel time, amount of oxygen, etc.). It was found that by maintaining a constant
temperature, the sol-gel cure process was more consistent and more uniform activity was obtained. This also
allowed reducing the sol-gel earlier during the gelation with significantly less flow restriction. Consequently,
measurements using smaller diameter glass capillaries filled with sol-gel chemistries L1 and L2 were repeated.
Ultimately, it was found that L2 could readily be incorporated into capillaries with IDs of 250 microns, as well as a
200 micron ID channel etched into a Si-wafer micro-chip (kindly supplied by Dr. Eric Wong, Jet Propulsion
Laboratory). Figures T4.8 and T4.9 show the SER spectra obtained for 10 microg/L samples of etidronate (pre-
concentrated on separation materials) obtained in these devices, respectively.

41


Figure T4.8. SERS of 10 microg/L etidronate Figure T4.9. SERS of 10 microg/L etidronate obtained in
obtained in a sol-gel filled (L2) 250 micron ID glass a sol-gel filled (L2) 200 micron ID U-channel in a Si-chip.
capillary. Conditions: 40 mW of 785 nm, 20-sec. Conditions: 60 mW of 785 nm, 20-sec. Inset: Photo of
Inset: glass capillary, the dark color is the protective chip, actual size.
polymer coating.

In a kick-off meeting between RTA and AFR, the development of the micro-fluids master was discussed, including
the design. Although a 14-channel chip was proposed containing 2- and 4-SERS-active sol-gel segments in
sequence, it was clear from the measurements described above that flow would be very difficult, and so two simpler
designs were developed. They contain most of the important features to develop the methods required to load the
SERS-active sol-gels into the channels. The designs are for 2 inch diameter silicon wafers. The first design is
simply 4-channels (each 250 micron width) suitable for practicing loading sol-gels and obtaining SERS. It also will
serve to produce the 20 deliverable chips for Phase III measurements (each with a different sol-gel chemistry).
The second design contains 6 straight channels and 2 T-channels with widths of width 100, 200, 300, 400, and 500
microns. The mechanical drawings provided to Microtronics to fabricate the masks are shown in Figure T4.10.

A B

Fig.T4.10. Drawing of initial microchip designs. The masks were developed on 4-inch x 4-inch x 0.09-inch soda
lime glass plate (grey regions represent anti-reflective chrome coat). The following specifications were provided:
A) all circular end holes 2.0 mm diameter on 0.25 mm horizontal strips; B) 1) all circular end holes on vertical and
horizontal strips 2.0 mm diameter, 2) vertical strips spaced 3.62 mm apart on centers, 3) width of vertical strips as
follows: a) 0.5, b) 0.4, c) 0.3, d) 0.4 mm including above horizontal connecting strip, e) 0.1 mm including above
horizontal connecting strip, f) 0.2, g) 0.1, h) 0.05 mm.

As previously stated, the lab-on-a chip masters were prepared following the method of Muck by Advanced Fuel

42


Research (under subcontract). AFR employed a class 10 clean room to prepare the master and followed all
necessary safety (HF) procedures. The overall process is shown in Figure T4.11 and involves two basic phases: 1)
standard photolithography for patterning the wafer with a protective photoresist coating and 2) wet chemical etching
to thin the silicon wafer and produce the raised ridges.

100 µm

Figure T4.11. Illustration of chip fabrication process.89 Figure T4.12. Photograph of Master Chip (design
#2) and microscope image (x110) of e channel.

Two-inch diameter silicon (100) wafers (p-type, 1-10 ohm-cm, 500 µm thick) with a 1000 nm thick thermal oxide
coating were used. First the wafer was cleaned by soaking in an 80oC bath of NH4OH:H2O2:H2O (1:1:5) for 10
minutes, and then rinsed with de-ionized water. Second, the wafer was spin-coated with a negative tone photoresist
(SU-8 2, Microchem Corp.) at ~ 2000 rpm for 30 seconds to produce a ~ 2 µm coating (the photoresist indicated by
Muck could not be used, and consequently all of the following parameters were refined by AFR). Third, the wafer
was soft-baked on a programmable hotplate by ramping the temperature from 25 to 95oC and holding for two
minutes. The wafer was then allowed to cool to room temperature. Fourth, the mask was manually aligned on the
wafer using the major flat for guidance (A in Figure T4.11). Fifth, the wafer was exposed to 365 nm UV light (B in
Figure T4.11). Sixth, a post-exposure bake was performed to complete the cross-linking of the exposed resist
regions, using conditions identical to the preliminary soft-bake. Seventh, the unexposed resist was then removed by
submersion in a stirred bath of SU-8 developer for two minutes followed by a 90 second rinse in isopropanol, and a
spin-drying stage (C in Figure T4.11). Eighth, the wafer was hard-baked at ~ 165oC for 15 minutes and allowed to
cool on the hotplate, completing the patterning phase of the processing. Ninth, the wafer was submersed in a room
temperature bath of a buffered oxide etchant consisting of a blend of 40% ammonium fluoride and 49% hydrofluoric
acid in a 7:1 volume ratio until the wafer is hydrophobic (~12 minutes, D in Figure T4.11). After a one minute rinse
with de-ionized water, the wafer is immediately submersed in a 30% (wt/wt) solution of aqueous KOH at 80oC with
vigorous stirring (E in Figure T4.11).

A photograph of the second chip design master is shown in Figure T4.12, along with a microscope image of the
channel f T-section. The latter shows well-defined corners, and reveals the tapered shape of the ridges owing to the
anisotropic nature of the Si (100 and111 planes). The top of the ridge was measured at 99.6 µm, which compares
favorably to the target value of 100 µm. Based on the width of the ridge base and assuming that the sidewall angle
with respect to the Si surface is 54.7 degrees [1], the height of the ridge is calculated to be ~ 83 µm. This
corresponds to a KOH etch rate of ~ 0.92 µm/min for the Si(100), which is in fair agreement with the predicted etch
rate of ~1.3 µm/min for 30% KOH at a temperature of 80oC [2]. Thus far, the “tallest” micro-machined ridges we
have achieved on Si were measured to be ~ 138 µm (determined in an optical microscope) for a KOH etch period of
150 minutes.

43


Part 3 of making the micro-fluidic chips was performed at RTA. Following the procedure of Muck, a mold was
prepared by gluing the chip Masters to 4x4” glass supports using epoxy. A Teflon circle was cut and placed around
each Master, and then filled with the PMMA monomers (F in Figure T4.11). Unfortunately, the prescribed PMMA
solution and curing conditions did not work. After investigating these conditions, it was found that a 10% benzoin
methyl ether concentration was required, the 20/80 w/v ratio of PMMA/MMA had to be procured under nitrogen (2-
min at 60 °C), and UV cure required 5 hours (a UV lamp was purchased for this part; Dymax Blue Wave 200 model
38905). Unfortunately, this cured acrylic bonds to the master chip, and it could not be separated without destroying
the master. Eventually, the following procedure was developed. A 4 gm/2 drop solution of PMMA and benzoyl
peroxide (Electron Microscopy Sciences, Hatfield, PA)was mixed and allowed to pre-cure for 10 minutes. At the
same time a thin coat of Teflon spray (DuPont, Delaware) was applied to the master chip. The acrylic monomer
solution was then poured onto the master contained in the Teflon ring. Cure was complete in 24 hours at 25 oC or 2
hours at 75 oC. The acrylic chip separated from the master with a little amount of prying (Figure T4.13B).

Once these chips were formed the ports at the end of the channels were drilled through using a 1 mm bit. Next,
several methods were explored to bond the chips to glass substrates (2”x3” microscope slides were used), as
opposed to another sheet of acrylic described by Muck (since this would produce a strong Raman background).
Ultimately, it was found that coating the microscope slide with a thin layer of the same PMMA/ solution, clamping
it to the chip, and allowing it to cure at room temperature for 4-5 hours resulted in a strong bond with good seals.
Nevertheless, as can be seen in Figure T4.13C, we still have difficulty producing flat chips, and consequently only
preliminary measurements were performed.

A B C

Fig. T4.13. Photograph of A) Chip Design 1 Master, B) PMMA chip produced from master, and C) Chip Design 2
attached to 2”x3” glass slide.

While the above micro fluidic chip development was underway, alternative methods to fabricate credit card sized
lab-on-chips were explored. The focus was to develop a working chip that could be used to separate 3-methyl
histidine and Risedronate from a urine sample and measure them at physiological concentrations. The following
basic design was developed (Figure T4.14). It consists of a central channel in which the ion exchange resin could be
loaded and two side channels containing SERS-active sol-gels to measure the two analytes. The chip was
manufactured as follows. A 1.75” by 2.75” rectangle was cut out of a ¼” sheet of acrylic. Three 1 mm wide
channels, 1 mm deep were milled into the sheet in a trident pattern. Four 1 mm holes were drilled through the sheet
to allow introducing the resin, sol-gels, reducing agent, sample, acid and bases. As described above, the ¼” acrylic
chip was bonded to a 2”x3” glass microscope slide using the PMMA/peroxide monomer solution, clamped and
allowed to cure at room temperature. After 5 hours, a syringe and #22 needle were used to load the outside channels
with L2 sol-gel and a plug of MTMS in the central channel. The sol-gels were allowed to cure for 12 hours. Next
the outside channels were reduced with sodium borohydride, again using a syringe with a #22 needle. The surface
of the chip was then cleaned with isopropyl alcohol and the luer port connectors were attached (NanoPort
Assemblies, Part No N-333, Upchurch Scientific, Oak Harbor, WA). The adhesive gasket is cured at 120 oC in 1
hour (or 60 oC in 12 hours). The luer ports were not attached until after the channels are loaded and reduced, since
they are prone to getting clogged by the sol-gel and could retain some of the reducing agent. These ports provided

44


the lab-on-a-chip to real world fluidic interface (as opposed to the Teflon block interface proposed). To obtain
SERS from these chips, a 4”x 6.5” Plexiglas plate was machined to fit the standard 96-well plate reader and hold the
chips (see Figure E.1).

A B C

Figure T4.14. Photograph of of A) machined Trident chip, B) chip attached to 2”x3” glass slide, and C) Trident
Chip with luer ports.

Questions to be answered: Can AFR produce the test chips? AFR was subcontracted to make the chip masters,
and had a number of set-backs, but eventually developed the process. Does the process need further modification?
Thicker substrates are required. Does the interface perform as planned (no leaks)? Yes, commercial luer locks were
used that simplified the real-world interface. Can the various channels be loaded, reduced? Yes, see above. What
size, length is best? We were able to load 200 micron channels AND flow analyte. Can separation materials be
introduced? Yes.

Task 5. Define Analytical Figures of Merit. The aim of this task was to establish performance criteria for the lab-
on-a-chip. This was accomplished by measuring the analytical figures of merit for the analyzer, as outlined by the
FDA: sensitivity, reproducibility, linearity, accuracy, precision, and selectivity.

Although the focus of this task was to determine the analytical figures of merit for the micro-fluidic chips, the
difficulties encountered in manufacturing these chips, forced us to perform these tests on the SERS-active
capillaries. We believe that the chip performance would have the same linearity, accuracy, precision, and
selectivity, and similar sensitivity, and reproducibility. Sensitivity may be reduced due to a smaller channel size as
compared to the laser focal spot (~365 micron diameter, i.e. fewer analyte molecules in the beam), while the
reproducibility will depend on manufacturing. In the case of the former, changing optical components in the fiber
optic sample probe can address this difference. In the case of the latter, improved replication of the chips using a
photolithographic master, such as shown in Figs.T4.13-14, will result in reproducibility equivalent to the capillaries.

We proposed to improve sensitivity by developing chemically-selective sol-gels, evaluating sample flow and
pistioning, and exploring the use of shorter laser excitation wavelengths. The goal was limits-of-detection (LODs
based on a signal-to-noise ratio of 3) of 1 mg/L for the biomarkers, and 0.01 mg/L for the drugs. Based on Phase I
results, LODs of 10 to 100 microg/L were predicted. In fact, as described in Task 2, not only were these LODs
achieved by using an ion exchange resin, but the analytes were actually measured at these target concentrations (see
lowest measured concentration, LMC). Nevertheless, measurements at even lower concentrations are always
desirable.

The use of shorter laser excitation wavelengths can further improve sensitivity by more closely matching the
plasmon absorption maximum. As described in the proposal, the plasmon maximum for 15 nm silver particles is ~
450 nm, while the maximum SERS response occurs at ~500 nm (Figure T5.1).90 Our silver particles are 30-80 nm in
diameter, which shifts both of these maxima by ~100 nm, but also reduces the magnitude of the plasmon absorption
by ~ 10 (blue line in Figure T5.1). Nevertheless, shifting to laser excitation at 532 or 633 nm should increase the

45


SER signal by at least 20 times, and as much as 100 times. Attempts to obtain SER spectra using 20 mW at 532 nm
laser (1064 nm frequency doubled using a KDP crystal) and a recently purchased 17 mW at 633 nm HeNe laser
(Model MRP170, THOR Labs) failed. This was even true for attempts to obtain normal Raman spectra of pure
chemicals at the shorter wavelength. Closer examination of our Raman analyzer revealed that the spectral response
of the system is essentially non-existent below 650 nm (red line in Figure T5.1). This is not due to the Si detector,
but the inefficiency of the beam splitter, which was not taken into account when proposed. However, there was
sufficient response at wavelengths longer than 725 nm when using 633 nm laser excitation (above 2000 cm-1 in the
Raman spectrum). Unfortunately, our analytes do not produce any significant peaks above ~1700 cm-1. One of the
few chemicals that produce a SERS peak above this wavenumber is cyanide (CN-) with a peak at 2140 cm-1.
6
10 100

15 nm Ag A 90
B
15 nm silver
Enhancement Factor

5 50 nm Ag particles 80
10 Experiment (RTA) 70

60

4
10 50

40
532 633 785
Theory 30
C
103
20

10
2
10 0
800
600 650 700 750 800
400 500 600 700
Wavelength (nm)
Figure T5.1. A) Theoretical SERS enhancement for RTA 50 nm silver particles (blue line) and Raman analyzer
response (red line). SERS of B) 1 and C) 0.1 mg/L using 785 and 633 nm laser excitation, respectively.

We measured 100 microg/L cyanide in a SERS-active vial using the 633 nm laser and a 1 mg/L cyanide sample
using the 785 nm laser. The calculated enhancement was 34 times when including the sample concentration,
signal intensity, laser power, laser wavelength (4th power dependence), and spectral response efficiency ((1mg/L/0.1
mg/L)x (0.07/0.4)x(84 mW/12 mW)x(785 nm/633 nm)4(100%/15%)). Although our Raman system is clearly not
optimized for operation at this wavelength, significant improvements in sensitivity can be expected using a more
efficient beam splitter.

Measuring the SER signal intensity as a function of concentration allows determining linearity, multiple
measurements allows determining precision, measuring a mixture allows determining accuracy, and measuring
several of the samples on multiple capillaries allows determining repeatability. During the Phase I program, 1-
methyl histidine (mislabeled 3-methyl histidine by the supplier, Figure T5.2) was measured over a concentration
range of 0.001 to 1 g/L (pH adjusted to 7.5). For each concentration, 50 microL of the 1-MeHis solution was drawn
into a SERS capillary (chem2c), and five spectra were collected at random spots along the capillary. The
concentration was related to the height of the 1565 cm-1 band in the 1-MeHis SER spectrum. In each case, the
largest 2 outliers were discarded (modified T-test). The results of the concentration curve are shown in Fig. T5.3.
This procedure provided better than normal reproducibility for the SERS capillaries with relative standard deviation
(RSD) of 10-15%. As can be seen, the concentration curve initially increases linearly from zero, but then “rolls-
over” as the available silver surface area decreases and monolayer coverage is reached (at ~0.005 mg/mL). This
SER response has been previously described in the literature and ascribed to a standard Langmuir-Blodgett isotherm
equation.

46


0.9
OH

N
0.8
O
0.7
NH2
N
0.6

Intensity
0.5 mg/mL Int.1565
OH 1 0.672
0.4 0.5 0.617
N
0.25 0.392
O 0.3
0.1 0.329
NH2
N 0.2 0.05 0.278
0.025 0.234
0.1 0.005 0.143
0.001 0.001
0
0 0.2 0.4 0.6 0.8 1 1.2
Concentration (mg/mL)
Fig.T5.2. SERS of A) 1-methyl histidine and
B) 3-methyl histidine. Conditions: 100 mW 785, 1- Fig. T5.3. Plot of the 1565 cm-1 peak height of 1-MeHis
min 1 mg/mL on L1. as a function of concentration (20% error bars included).
Conditions: 100 mW 785, 1-min, on L3 (chem2c).
Although the Phase I measurement was a good start, it did not cover the relevant concentration range and it did not
account for the use of an ion exchange resin to concentrate the sample. A stock solution (3-MeHIS) was prepared in
HPLC water, and serially diluted to generate samples having concentrations of 5000, 1000, 500, 100, 50, and 10
microg/L. Each sample was pH adjusted to 1 using 1M HCl, passed through an ion retardation column, pre-
concentrated on an ion exchange column, eluted off with 0.1M NaOH, and the first 100 microL was loaded into a
SERS-active capillary (L3, chem2c) and measured. Similar to the Phase I measurements, spectra were recorded at
10 positions along the capillary (1.5 mm apart), but none were removed. The ten spots were averaged to produce a
single spectrum at each concentration and the 1364 cm-1 peak height (baseline at 1433 cm-1 subtracted) was used to
prepare a calibration curve (Figure T5.4). The ten points were used to determine the RSD for each concentration,
and it was found the RSD = ~20% for 5, 1, 0.5 microg/L, and 40% for the lower concentrations.

A B
0.08

0.08
0.055
0.055

0.03 0.03

reversed 0.005
0 200 400 600 800 1000

0.005
350 500 750 1000 1250 1500 1850 0 1000 2000 3000 4000 5000
Raman Shift, cm-1 Concentration (microg/L)
Figure T5.4. A) SERS of 3-methyl histidine at 1000, 500, 100, 50, and 10 microg/L. Conditions: pH=1, ion
retardation and ion exchange, measured on L3 capillary, 100 mW of 785 nm, ten 1-min scans averaged. B) Plot of
1364 cm-1 peak height (baseline at 1430 cm-1 subtracted). Inset of higher concentration. Red dot = 25 microg/L.

Since it is clear that the capillaries do not produce the same signal intensity from spot to spot, a program was written
to scan a 15 mm length and continuously collect spectra. In 1 minute it makes 10 passes (5 each way), and
effectively averaged the spot-to-spot intensities. The success of this “Raster” approach is shown in Figure T5.5,
where the identical set of capillaries was measured. Although the S/N was reduced (1 min vs 10 min per spectrum),
the concentration curve becomes linear. Since repeat measurements were not performed, the precision (RSD) was
not calculated for each concentration. However, the improved shape of the concentration curve suggests that the
RSD has improved, and would likely be similar to our SERS-active vials at 15%. Nevertheless, the RSD for all of
the concentrations combined allows calculating the reproducibility of the capillaries, which is 27%. If 10-min
Raster scans were used, this value would also likely improve.

47


0.125
A B
0.105

0.085

0.065

]
0.045

0.025

0.005
350 500 750 1000 1250 1500 1850 0 200 400 600 800 1000
Figure T5.4. A) SERS of 3-methyl histidine at 1000, 500, 100, 50, and 10 microg/L. Conditions: pH=1, ion
retardation and ion exchange, measured on L3 (chem2c) capillary, 100 mW of 785 nm, 1-min Raster scan. B) Plot
of 1364 cm-1 peak height (baseline at 1430 cm-1 subtracted). Inset of higher concentration. Red dot = 25 microg/L.

To test the accuracy of this method, a 25 microg/L 3-methylhistidine sample was independently prepared and
measured, and the peak height plotted in the concentration curve (circle). Based on the concentration curve, the
measured peak height (0.036±0.01) corresponded to 150 microg/L, but is within the RSD.

Risedronate concentration samples were also prepared, but over a lower concentration range (1000 to 1 microg/L).
Again each sample was pH adjusted to 1, passed through an ion retardation column, pre-concentrated on an ion
exchange column, eluted off with 0.8M NaOH, and the first 100 microL was loaded into a SERS-active capillary
(L3, chem2c) and measured. Spectra collected using the Raster approach are plotted in Figure T5.5 along with the
measure peak height at 1032 cm-1 (baseline at 989 cm-1 subtracted). This time the data conformed more closely to
the Langmuir-Blodgett isotherm. Again, the entire data set was used to calculate a capillary-to-capillary RSD of
~32%. Also, a 250 microg/L Risedronate sample was independently prepared, extracted following the procedure
described above, and measured to test the accuracy of the concentration curve. This time the peak height (0.022,
green square) placed the concentration at ~300 microg/L, well within experimental error. As a further test of the
ability to use this calibration curve, a 10 microg/L Risedronate sample was prepared in reconstituted urine,
extracted as above, and measured. The peak height (0.016, red dot) was very close to that measured for the pristine
10 microg/L sample (0.017 peak height).
0.03
A B
0.025

0.02

0.015

0.01

0.005
350 600 800 1000 1250 0 200 400 600 800 1000
Figure T5.5. A) SERS of Risedronate at 1000, 500, 100, 50, 10 and 1 microg/L. Conditions: pH=1, ion retardation
and ion exchange, measured on L3 (chem2c) capillary, 100 mW of 785 nm, 1-min Raster scan. B) Plot of 1032 cm-
1
peak height (baseline at 989 cm-1 subtracted). Inset of higher concentration. Green square = 250 microg/L, red dot
= 10 microg/L in urine.

As proposed, concentration curves for several other biomarkers and drugs were prepared and measured. It was
found that analytes that could be detected at very low concentrations (1 microg/L), followed the Langmuir-Blodgett

48


at these lower concentrations and leveled off within 2 orders of magnitude, such as Risedronate. Conversely,
analytes that could be detected only as low as 50 microg/L followed this curve at higher concentrations, leveling off
at correspondingly higher concentrations. We attribute this response to the chemical-selectivity of the sol-gels and
the analyte-to-metal interactions, i.e. the greater the interaction, the greater the sensitivity, and the greater the
selectivity the lower the surface saturation point.

Questions to be answered: What are the LODs for each of the biomarkers, drugs and their metabolites? See
Tables T1.3 and T1.4, T2.1 and T2.2. What is the reproducibility of the chips? This was not determined, but is
likely to be the same as the glass capillaries, at about 30%. Which channel design provides the best selectivity,
reproducibility, and sensitivity? Since the chip manufacturing was substantially delayed, it was only determined that
we could load and perform measurements in both glass and plastic channels at 800, 600, 500, 250, and 200 microns.
What is the best standardization method for quantifying target analyte concentrations? It appears that the best
method would be to sue creatinine and uric acid as internal intensity references. Creatinine is a known constant for
urine chemistry. Once the relative uric acid concentration is determined, all other chemicals could be referenced to
it, since it appears in virtually all samples due to its 7 ionizable protons (7 pKas.)

Task 6 - Prototype Design. The overall goal of this task was to design a prototype system to be used and tested by
NASA in Phase III. This has been accomplished by redesigning the proposed compact disc (CD) style lab-on-a-
chip based on the Task 3 and 4 data.

The proposed analysis of urine from astronauts on long term space missions in a microgravity environment will
employ a multi-component analyzer that will be integrated into either Node 3 on the International Space Station
(ISS) or on future manned flights. This ultra compact system will be placed into an integrated rack, and will be
connected to a water reclamation system currently being developed by Hamilton-Sundstrand. A 10 mL sample
container designated for each astronaut will collect urine samples over the course of 24 hrs. After 24 hrs a 1 mL
sample is withdrawn for analysis, and the container subsequently flushed with water for the next days use. This
analyzer will consist of three major components, the sampling system, the lab-on-a-chip, and a Raman instrument.
The final lab-on-a-chip design will depend upon the other two components.

Employing a sequence similar to that of the operation outlined in Figure T3.24, the following very simple
design is proposed (Figure T6.1). The system consists of four components, A) an injection cartridge, B) the lab-on-
a-chip, C) pneumatic valve control manifold, and D) a fiber optic linked Raman probe. In many respects it will
operate like a compact disc player. The injection cartridge will contain up to five needles to deliver 1) a vacuum for
micro-fluidics control, 2) the sample from the individual astronaut sample container (which will be pH adjusted to
1 using acid from a reservoir), 3) a base gradient prepared by increasing the amount of NaOH mixed in with water
(note reservoirs), 4) air from a bladder to allow fluid flow control when combined with the vacuum line, and 5)
waste removal line. Again, like a CD player, six chips will be used, one per astronaut (based on expected ISS
capacity). Each lab-on-a-chip will consist of a 12 cm diameter disc (CD-size) divided into 30 pie-shaped “slices”,
12-degrees wide (12 mm across outer edge), one slice per sample. Thus, each lab-on-a-chip will provide 30
measurements; 1 per day. (A practical first version using 36-degee slices can be made with the technology
developed in this program, see below). Each sample segment will contain five ports (6-10), each 1.5 mm diameter,
2 mm deep to match the cartridge delivery needles. A circular layer of butadiene (septa material, 11) will be bonded
to the disc to seal the ports prior to use. The sample acceptance port (7) will also contain a layer of Nylon 66 (or
equivalent particle filter material, 12) melted into place. Valves (13) integrated into a second, lower layer of the lab-
on-a-chip will be computer controlled through a pneumatic or magnetic manifold below the chip. Recently, a plug-
and-play micro-fluidic device has been built that employs computer controlled diaphragm pumps to control flow
through desired channels.91 The choice of valve control will be based on available technology at the time of
implementing this design. The valves are placed to control flow through the various separating materials and direct
the analytes to the appropriate SER-active sol-gel set. An ion retardation channel (14) will contain the required
separation materials to extract salts, while an ion exchange (cationic) column (15) will adsorb the protonated
analytes. The remainder of the sample will be removed from the chip through port 10 to waste (back to water
reclamation). Then the NaOH gradient will be delivered through port 8 through the ion exchange resin and
sequentially to 1 of 12 SER-active sol-gels (16). The first 3 sol-gels are for biomarkers (e.g. 3-methyl histidine), the
next 8 for drugs, and the last for creatinine. The sequence is based on their relative pKa’s (e.g. 3-MeHis = 5.9,
Risendronate ~8.5). These channels will lead back toward the center of the disc where a central vacuum or piston
will be used to move the 3 µL sample through the sol-gels for analysis. A fiber optic linked Raman probe will be

49


fixed above the SER-active channels, such that the 100 micron illumination spot matches the channel.

Actual Size
A Base
w Acid
Vie D Water
p
To 1 2 3 4 5

B Side
12 cm
C View

A
11 D

Top View Size x4 16

14 15

12 mm
11
Measurement
Points

13 Side View

Figure T6.1. Illustration of urine analyzer, composed of a A) sample delivery cartridge, B) lab-on-a-chip, C)
pneumatic/magnetic valve control manifold, and D) Raman optic probe. This figure is proprietary.

In operation, the cartridge head (A) will lower, the needles will pierce the seal, and 1 mL of urine, acidified with 25
microL 1M HCl, will be loaded through Port 7, through the particle filter (12), and through both ion resins (14 and
15), then out through the waste port (10). This volume will provide a significant amount of analytes adsorbed onto
the ion exchange resin, and does not influence the size of the channels. One the ion exchange resin is loaded, 120
microL of NaOH gradient will enter through port 8 and flow through the exchange resin into the SERS-active
channels in sequence. Valves would be used to select and deliver 10 microL sample of increasing pH to each
channel (e.g. top to bottom). The air line allows offsetting pressure drops. Based on Task 3, a 200 micron square
channel - 20 cm long will work and we have successfully loaded a chip obtained from NASA (JPL) that has a

50


serpentine channel of these dimensions and obtained a SERS spectrum of Risedronate separated from urine (off-
chip, Figure T6.2). However, this channel dimension would require 36-degree “pie slices” per analysis, limiting
each CD chip to 10 analyses. This should be acceptable for a first version of this analyzer.

At a flow rate of ~ 0.1 mL/min all of the channels would be loaded (we were able to flow 0.5 mL/min through the
800 micron channels). Then the cartridge head separates from the lab-on-a-chip, and the chip rotates in a step-wise
fashion, such that Raman spectra can be collected from each of the SER-active sol-gel segments. Once
accomplished, the chip again rotates to align the cartridge of the ports of the next sample segment and awaits the
next sample. The total analysis for each sample is expected to take no more than 15 minutes. Analysis of the
spectra to identify each biomarker or drug based on spectra matching will be accomplished in less than 1 second.
All spectra could be sent to earth for further analysis, as well as for identifying unknowns. It should be realized that
the pneumatic manifold and Raman probe are fixed, and the only motion is the rotation of the lab-on-a-chip and the
up and down movement of the sample delivery cartridge. This device will require very little power. Furthermore,
only 25 microL acid and 120 microL of base are required per analysis, or ~1.5 mL for a 10 sample CD disc, or 4.5
mL for a 30 sample CD disc! The 1 mL urine sample is not included, since it is provided by the astronaut.

A B C

L3 ( Ag, very non-polar)

L2 (Ag, non-polar)

L5 (Ag, very polar)

L4 (Au, very polar)

Figure T6.2. A) Photograph of lab-on-chip provided by JPL for testing. Note the serpentine channels are 200 x 200
microns, and the serpentine channels are ~ 20 cm long. B) SERS of 1 mg/L Risedronate measured in this chip.
Drug extracted from real urine (off-chip) and loaded into this chip and measured. Conditions: 75 mW of 785 nm, 1-
min. C) Photograph of 4-channel SERS-active test chip (representative of the deliverables).

Mission Tests (Phase III). The overall objective of this task is to test the space-worthiness of the proposed lab-on-a
chip. This will be accomplished by supplying NASA with 20 of SER-active lab-on-a-chips for Phase III testing (a
deliverable).

Twenty 4-channel chips have been manufactured and filled with 4 SERS-active sol-gels (L2, L3, L4 and L5, Figure
T6.2C). These tests will include shelf life (2 years/accelerated tests), temperature cycling between -80 and 40 oC,
vibrational tests, shock, 20-g forces, micro-gravity, radiation resistance (3 krads), cross-contamination-elimination,
etc. This may include shuttle or space station testing as appropriate.

51


Performance Schedule - The Phase II research program was accomplished according to the following schedule,
tasks, and milestones with the work load distributed as SE-150, ENG-200, RA-200 hours:
The program was extended by 6-months.
QUARTERS
1 2 3 4 5 6 7 8
Task 1 - Develop Spectral Library. 1 2 3
Task 2 - Develop Chemical Selectivity. 4 5 6
Task 3 - Select and Test Components. 7 8 9 10 11
Task 4 - Fabricate Lab-on-a-chip 12 13 14 15
Task 5 - Define Analytical Figures of Merit. 16 17 18
Task 6 - Design Prototype. 19 20

Milestones
1. Prepare 96-well SER-active sol-gel plates 11. Re-examine search routine capabilities
2. Measure SERS of 24 chemicals 12. Finalize test chip design
3. Test and refine library search algorithms 13. Fabricate test chips
4. Identify unique bands and determine detection limits 14. Fabricate real-world interface
5. Prepare sol-gel coated capillaries 15. Test fluid deliver, Raman collect, etc.
6. Measure reversibility and prelim LODs 16. Measure sensitivity
7. Purchase components/chromatography supplies 17. Determine reproducibility
8. Prepare 4-zoned capillaries 18. Demonstrate selectivity
9. Test separation methods 19. Perform field test
10. Test interferents 20. Design test chips for NASA

This Final Report was delivered to NASA (June 1, 2007) summarizing the experimental and analytical results of the
project according to contract specifications. In addition, RTA mailed 20 lab-on-chips (see Figure T6.2c above) for
continued testing in Phase III (see Mission Tests above).

52


Part 5: Potential Applications

The overall goal of this program is to develop an analyzer to measure bio-indicators and prescribed drugs and their
metabolites in urine to assess and monitor health of astronauts affected by microgravity conditions during extended
missions in space. The analyzer will employ a lab-on-a-chip to extract and separate out chemical components from
urine samples and surface-enhanced Raman spectroscopy to detect and identify these indicators. The proposed
analyzer is immediately applicable to astronauts working in the International Space Station. For example,
monitoring the concentration of pyridinoline may suggest increasing (or decreasing) the dosage of the anti-bone loss
drug zoledronate. The proposed analyzer will undoubtedly increase our understanding of the adverse effects of
near-zero gravity, allow determining the efficacy of drugs used for treatment, and improve regulating dosage. The
analyzer will have continued value through the life of the ISS and into developing a base on the moon.
Finally, the knowledge gained will undoubtedly be critical to the strategy of traveling to Mars safely.

The proposed analyzer will also have a substantial impact on the health of a significant portion of our population,
the elderly, in its ability to measure bone-loss bio-indicators, as well as the drugs being used to treat osteoporosis.
Osteoporosis, a condition correlated to enhanced bone fragility and increased risk of fracture. As of 2006,
osteoporosis effects 55% of the US population over age 50. More than 50% of healthy American women aged 30-
40 are likely to develop vertebral fractures as they age due to osteoporosis.92 Each year, more than 250,000 hip and
500,000 vertebral fractures occur in postmenopausal women in the USA.93 In 2006, it was estimated that fractures
represent a $17 billion annual cost to the US healthcare system. The World Health Organization has declared
osteoporosis the 2nd largest medical problem, next to cardiovascular diseases.94 Unfortunately, treatment is at best
partially successful once progressive bone weakening has started.

The primary method of detecting osteoporosis is the measurement of mineral density and bone mass using dual-
energy X-ray absorptiometry, which lacks sufficient image quality to adequately detect early stages. It provides a
standard (adult) density (SD) score, SD = 0 = healthy, SD = -2.5 onset of osteoporosis. Recently, studies have
identified pyridinoline and deoxy-pyridinoline as degradation products of type 1 collagen. Not only can these bone-
loss indicators be detected in urine, their concentrations have been found to decrease in response to certain drug
therapies, suggesting their use in monitoring treatment efficacy.93 Two methods are currently being developed to
detect these bio-indicators, high-performance liquid chromatography (HPLC) and immunoassay. HPLC methods
are slow (30-min per analysis), labor intensive, and require daily calibration, while immunoassays are notoriously
inaccurate.

The proposed lab-on-a-chip, coupled to a Raman analyzer, could be used by clinical labs that perform blood and
urine analysis (e.g. Quest Diagostics). The family physician would simply check “test for osteoporosis” during
annual physicals for patients. One milliliter of the urine sample is injected into a disposable lab-on-a-chip ($25), it
is placed on the Raman Analyzer ($75,000), and a “score” is reported (0 to -5 SD). The process takes 5 minutes.
This compares favorably to performing x-rays (machine costs $300,000) and the x-ray cost (>$100 and 30 min
analysis time), which is not covered by most insurance. Furthermore, the patient does not have to make separate
appointment. RTA plans to pursue a marketing agreement with Quest Diagnostics.

Part 6: Contacts

Principle Investigator: Dr. Frank E. Inscore (860-528-9806, x128, inscore@RTA.biz), a Research Scientist and
Manager of Raman Applications at RTA, will be the Principal Investigator of this program. Dr. Inscore received his
Ph.D in May of 2000 under Prof. Martin L. Kirk at the University of New Mexico. His research was directed
towards elucidating the electronic structure of Mo and W complexes employing a combined spectroscopic and
theoretical approach that focused primarily on the use of Raman spectroscopy. The PI has considerable expertise in
CCD dispersive systems, and in designing sampling configurations under a variety of conditions (cryogenic/
anaerobic) employing Raman. The PI was previously the manager and key designer for the Raman Lab of Dr.
Martin Kirk at the University of New Mexico as a graduate student (Research Assistant). As a Research Associate
at the University of Arizona with Prof. John H. Enemark he examined simple and complicated inorganic systems as
structural and electronic models of metalloprotein active sites where he further employed his expertise in Resonance
Raman spectroscopy. The PI has over 12 years experience with vibrational/Raman spectroscopy, and has published
peer-reviewed articles regarding instrumental set-up and analysis of Raman and other spectroscopic data
collected.40,41,42,43 The PI has also presented relevant research at three national ACS meetings, a Gordon Research

53


Conference, two SPIE presentations, and a Pittsburg conference. The PI has also gained extensive experience in sol-
gel chemistry and surface-enhanced Raman spectroscopic applications at RTA. Since joining RTA, Dr. Inscore has
performed critical development work of RTA’s SER-active capillaries. He has successfully used this technology to
measure chemical warfare agents and their hydrolysis products, and trace pesticides residues. He was also a major
contributor for detecting dipicolinic acid extracted from Bacillus anthracis spores on surfaces. This program
represents Dr. Inscore’ first SBIR win. He is on 1 patent and 4 pending.

Principal Investigator: Dr. Frank E. Inscore, Research Scientist, Real-Time Analyzers, Inc.
Education
University of New Mexico - Ph. D. Chemistry - 2000
University of New Mexico - B. S. Chemistry Major/Applied Mathematics Minor - 1993
Austin College - B. A. Biology - 1986
Experience
Real-Time Analyzers, 2003 – present
Manager of Raman applications at RTA: responsible for Raman research and development.
University of Arizona, 2000-2003
Post doctoral Research Associate for J.H. Enemark, University of Arizona.
Managed group research program: responsible for synthesis/purification of 1st generation geometric and
electronic structural models of metalloenzyme active sites, and subsequent characterization via NMR,
FAB/ESI-MS, IR/Raman, XRD, XAS/EXAFS, UV-vis/NIR absorption, gas-phase/solution anionic PES,
CW/pulsed-EPR spectroscopies, and DFT computations. Additional responsibilities included
isolation/purification of various metalloproteins via chromatography, HPLC and FPLC.
University of New Mexico, 1994-2000
Research Assistant for M.L. Kirk University of New Mexico.
Built and managed Raman facilities: responsible for probing contributions to structure/function relationships in
enzyme active sites using resonance Raman, VT-MCD/CD, and UV-vis/NIR absorption spectroscopy.
Honors and Awards
Research Award of Excellence 1998 &1999
DOE GAANN Fellowship 1994 -1996
Dean Uhl Award for Excellence in Chemistry 1994
Inducted Kappa Mu Epsilon National Mathematics Honorary 1993

Program Manager: Dr. Stuart Farquharson (860-528-9806, x127, stu@rta.biz), President at Real-Time Analyzers
will be the Program Manager. He has extensive experience in designing and developing both infrared and Raman
spectrometers for multiple industrial applications.95 While employed by the Dow Chemical Company, he designed
and integrated a Fourier transform infrared spectrometer into a polymer production plant, for continuous process
control. The PM designed and patented the required sample system, designed the optical interface, installed the
instrument, and assisted in designing user-friendly data analysis software to ensure customer satisfaction. The PM
has extensive experience in Raman and surface-enhanced Raman spectroscopic applications, as well as instrument
design, and has published 50 papers in scientific journals and holds 6 patents in the field. He has been a Chair or
Co-chair at 10 SPIE Conferences. He has been an invited speaker at more than 20 conferences. In 2002 this
includes the Pittsburgh Conference (New Orleans, LA, March), CPAC Summer Institute (Seattle, WA, June),
Gordon Research Conference (Newport, RI, July), FACSS (Providence, RI, October), and Eastern Analytical
Conference (Somerset, NJ, November). Dr. Farquharson’s resume is available upon request, and is provided in the
Phase I and II proposals.

Senior Investigator: Mr. Chetan Shende (860-528-9806, x134, chetan@rta.biz), a senior chemist at Real-Time
Analyzers, will be a senior investigator. He recently completed his MS in Analytical Chemistry from the University
of South Florida, under the direction of Dr. Abdul Malik. His main research area was developing column
technology for analytical micro-separations using sol-gel chemistry based systems. His MS thesis work involved
developing Open-Tubular columns for high resolution capillary gas chromatography (CGC), using polyethylene
glycols (PEG) as the stationary phase. Mr. Shende has expertise in sol-gel synthesis and coating capillary columns.
His resume is available on request.

NASA Technical Representative: Robert Hawersaat (216-433-8157, Robert.w.hawersaat@nasa.gov).
Hamilton Sundstrand Contact: Patricia O’Donnell, Program Manager (860-654-5649,
patricia.odonnell@HS.UTC.com)

54


Part 7: Future Technical Activities

The overall objective of the proposed program (through Phase III) is the development of an analyzer integrated into
the International Space Station (ISS) toilets capable of immediate detection of key chemicals in urine to monitor and
assess astronaut health. During Phase I we initiated conversations with Hamilton Sundstrand (HS), and discussed
the design for their toilet and water reclamation systems to be integrated into Node 3. We also proposed to test our
chip performance using HS’s Raman system being developed for a Mars mission. Dr. Farquharson visited HS in
Pomona, CA and found that HS only built a single prototype that was unavailable for tests. However, RTA has
developed a very rugged Raman analyzer, which may be suitable for space operation (~ 1/5th the weight of the HS
prototype), and integration into their toilet system. This is still the desired goal of this program as HS’s Node 3
Water Processor system is finally being integrated into the ISS (Shuttle Mission: 20A/STS-122).

As described in Task 6 above, the analyzer will be comprised of three major components: a urine capture device, a
lab-on-a-chip, and a Raman analyzer. The entire analyzer will be sufficiently small to fit into a standard ISS
integrated rack (IR). In use, 1 ml urine sample will be extracted from water reclamation line exiting HS’ toilet and
collected into one of six sample containers (bladder), one per astronaut (ISS capacity). This will be performed 10
times during a 24 hour period per astronaut for a total of 10 mL each. From these individual bladders, we expect to
draw 1 mL per analysis per day. We will work with HS to develop a rinse cycle to flush the sample containers once
a day after each analysis. The lab-on-a-chip will be designed on a 12-cm disk (1 disk per astronaut, 10 or 30 days of
analysis, versions 1 and 2) that will contain 10 or 30 identical pie “slices”, each section self-contained with full
extraction/separation capabilities and SERS-activity for detection. The disk will rotate once a day for a new urine
analysis. This approach will virtually eliminate the potential for cross-contamination. Multiple disks per astronaut
or higher lab-on-a-chip channel densities can be used for longer missions.

Mission Tests (Phase III). The overall objective of this task is to test the space-worthiness of the proposed lab-on-a
chip. We have supplied NASA with 20 SERS-chips so that some can be 1) tested in space and 2) used to measure
actual astronaut urine brought back to earth. Phase III commercialization tests will begin with comprehensive
analysis of osteoporosis, including analysis of new bis-phosphonate drugs as they become available (e.g. with
Pfizer), and ultimately lead to preliminary clinical tests (possibly at the Nevada Cancer Institute.)

Part 8: Potential Customer and Commercialization Activities

8.1. The Company: Real-Time Analyzers, Inc. (RTA, www.rta.biz) - The mission of Real-Time Analyzers is to
develop, produce and market analyzers that detect, identify and quantify trace chemicals in real-time, in
industrial or field settings, either continuously or on-demand.

Real-Time Analyzers is a spin-off company of Advanced Fuel Research launched September 1, 2001. RTA was
formed to consolidate and commercialize the considerable expertise in Raman and surface-enhanced Raman
spectroscopy developed at AFR into a focused product line. This approach of bringing high-technology products
to market was recently validated when AFR sold its first spin-off company, On-Line Technologies to MKS
Instruments, Inc. for over $20 million.

Dr. Stuart Farquharson has developed a formal Business Plan for RTA focusing on the Chemical Manufacturing
Industry (agricultural chemicals, fine chemicals, petroleum products, polymers, and pharmaceuticals). Dawnbreaker
provided the basic tools for developing this Business Plan (www.dawnbreaker.com through DOE sponsorship).
Foresight is currently performing an osteoporosis market analysis. The Business Plan is a two phase growth
strategy, 1) design, develop, market and sell a portable fiber optic based FT-Raman instrument (Industrial Raman
Analyzer, IRA), and 2) develop a trace chemical analyzer (Surface-enhanced Raman Analyzer, SERA) for
environmental, pharmaceutical, medical applications and homeland security. The IRA is being marketed to 1)
Homeland Security, 2) the DOD, and 3) the Chemical Manufacturing Industry.

RTA completed the first growth Phase as of June 2006 when it sold its first two analyzers! Since that time RTA
has sold 4 more analyzers, penetrating each of these 3 markets!

Initially, Dr. Farquharson was the only full-time employee with 4 part-time employees occupying 600 sq. ft. Early
growth has been focusing on product and application development. In January 2002, two AFR employees, an

55


instrument design engineer and a PhD chemist, transferred to RTA. In June 2002 RTA hired a mechanical engineer,
an optics engineer, and a chemical engineer. In January 2003 RTA added two PhD Raman spectroscopists, an MS
sol-gel chemist, and a full-time accountant. The engineers focus on developing the IRA, while the chemists focus
on developing the SERA and applications. In January 2004 RTA added a product engineer with 12 years experience
producing FT-infrared spectrometers. RTA currently has 7 full-time employees (all with shared and/or overlapping
management, technical and manufacturing responsibilities). RTA moved into new facilities in 2005 (5000 sq. ft. of
office, lab, and manufacturing space). RTA currently employs several consultants and subcontractors to meet their
R&D and manufacturing needs.

RTA’s Revenues 2004-7.
Revenues 2004 2005 2006 2007
Employees 9 8 8 9-12
Analyzer Sales $122,000 $300,000
Service Contracts $28,500 $2,500 $125,000 $400,000
Vial Sales $12,000 $12,000 $12,000 $12,000
SBIR Funding $582,000 $910,000 $717,000 $300,000
% Revenue SBIR 93.4% 98.1% 74.2% 29.6%
Total Revenue $622,500 $927,500 $966,000 $1,012,000

8.2. Markets. This program has provided the foundation for an osteoporosis product, the Urine-analySER. This
product and market are part of RTA’s second growth phase, selling SERA products.

Osteoporosis – When we submitted this proposal in 2005, there were 10 million men and women suffering from
osteoporosis in the USA. As of 2006, there are 44 million US patients! Each year, more than 250,000 hip and
500,000 vertebral fractures occur in postmenopausal women in the USA.93 And more than 50% of healthy
American women aged 30-40 are likely to develop vertebral fractures as they age due to osteoporosis. Metabolic
bone and joint diseases (osteo-arthritis, cancer, etc) account for an additional 12 million cases of accelerated bone
loss per year.92 The World Health Organization has declared osteoporosis the 2nd largest medical problem, next to
cardiovascular diseases.94 The estimated yearly cost associated with acute hospital care and rehabilitation was
estimated at $17 billion for 2006.93 Unfortunately, treatment is only partially successful once progressive bone
weakening has started. Therefore, it is important to identify this condition at an early stage, so that treatment
focuses on prevention instead of fixing fractures and attempting to reverse bone loss. Nevertheless, it should be
noted that drugs developed to prevent, mitigate or even reverse bone loss are showing signs of success.

The primary method of detecting osteoporosis is the measurement of mineral density and bone mass using dual-
energy X-ray absorptiometry, which lacks sufficient image quality to adequately detect early stages. More accurate
analyses, such as bone biopsies, have been developed, but are undesirably invasive. Recently, studies have
identified pyridinoline and deoxypyridinoline as degradation products of type 1 collagen. Not only can these bone-
loss indicators be detected in urine, their concentrations have been found to decrease in response to certain drug
therapies, suggesting their use in monitoring treatment efficacy.92 Two methods are currently being developed to
detect these bio-indicators, high-performance liquid chromatography (HPLC) and immunoassay. HPLC methods
are slow (30-min per analysis), labor intensive, require daily calibration, and high up front and maintenance costs
have limited their market acceptance. Immunoassay kits (colorimetric microassay plate enzyme immunoassay) have
been developed to measure total deoxypyridinoline in urine. However, the analysis requires complicated hydrolysis,
incubation and assay steps totally 48 hrs prior, which must be followed precisely to obtain quantitative results.

The Urine-analySER could be used to rapidly detect and quantify pyridinoline, deoxypyridinoline, and other
biomarkers in urine and assess patient risk, stage, or response to treatment.

8.3. Commercialization Strategy -The overall commercialization strategy is to bring Real-Time Analyzers
products to market through strategic partners with the aid of leveraged investment. This strategic path to
commercialization involves the following specific steps: 1) Form a spin-off company to consolidate expertise and
develop a focused product line (RTA has 4 scientists with over 40 years of Raman experience, and engineers with
over 35 years experience designing spectrometers). 2) Develop a superior produce with substantial patent coverage

56


(the superior performance of the IRA is covered by 4 patents, the SERS technology is covered by 3 patent, and 11
more have been filed). 3) Leverage the protected technology to gain market access through strategic partners (see
below). 4) Grow value of company so that acquisition is desirable.

Strategic Partners - RTA has profiled more than 50 companies in its target markets and has identified several
potential Strategic Partners. Initial meetings have been held to describe RTA's mission, technology, market plans,
and strategic partnering. Relevant to this program, RTA has met with Hamilton Sundstrand and the Nevada Cancer
Institute. RTA plans to initiate discussions with Quest Diagnostics, as well. In each case, RTA plans to sell the
beta-unit (based on successful on-site demonstration) and leverage investment against exclusive-use licenses.

Developments in Strategic Partnerships with Hamilton Sundstrand and Nevada Cancer Institute, are worth
elaborating. In addition, RTA has been actively pursuing matching state funds through Connecticut Innovations.
Hamilton Sundstrand Sensor Systems (HSSS, Pomona, CA) is the Sensor Systems business unit of Orbital
Sciences bought in 2001 by Hamilton Sundstrand, one of United Technologies Corporation six business units. HS,
with annual sales of $3 billion, bought Sensor Systems to complete their portfolio of space-based atmospheric and
space station sensor systems.96 Included in this acquisition was Analect, a company with considerable knowledge of
chemical analyzers, and an understanding of Raman analyzers. Patricia O’Donnell, business development
manager, has been intimately involved in the development of both the HS toilet and water reclamation system.
WE hope to collaborate on the development of the proposed analyzer for ISS Node 3 during Phase III.

Nevada Cancer Institute (Las Vegas, NV) - NCI began operations in 2005 with the goal of becoming a recognized
world-leader performing state-of-the-art research and implementing groundbreaking methods of prevention,
detection and treatment of cancer. According to Director Dr. Vogelzang: "It will expand the tools available for the
treatment of cancer. That will be our first and primary thrust." NCI has agreed to assist RTA in the development of
a chemotherapy drug analyzer based on measurements of saliva using our SERA technology. The importance of
measuring bone-loss in urine due to chemotherapy is a natural extension of this technology, and arrangements
will be made to perform these critical clinical tests with NCI with the goal of obtaining FDA approval.

Quest Diagnostics Incorporated (www.questdiagnostics.com), a $500 million per year revenue company, is the
nation's leading provider of diagnostic testing, information and services that patients and physicians need to make
better healthcare decisions. QD performs diagnostic laboratory tests for approximately 145 million patients each
year, and over 70% of all healthcare treatment decisions. QD has over 2,000 patient service centers. These service
centers each represent one Urine-analySER sale.

Connecticut Innovations (CI) is charged with growing the economy of Connecticut’s entrepreneurial technology
through venture and other investments. RTA has had several meetings with CI over the past 3 years, and they are
currently considering a $500,000 investment.

Part 9: Resources Status

No government equipment was required for the Phase II program. The following resources at RTA are available.

RTA Industrial Raman Analyzer Production (second generation, 3 systems) - RTA latest FT-Raman system is
extremely compact and portable. The entire system fits into a single 19” rack mountable box (19x20x7”). The main
features are: 1) a diode laser (1 W at 785 nm, Process Instruments), 2) an interferometer with a quartz beam splitter
(MKS), 3) a Si avalanche photodiode detector (RTA), 4) numerous fiber optic probes (designed in-house), and 5) a
data station consisting of acquisition and analysis hardware and software (OLT, National Instruments, and a
Gateway Pentium II personal computer). The system measures Raman spectra from 785 to 1080 nm (0 to 3200
∆cm-1).

RTA Industrial Raman Analyzer Prototype (second generation, 1 system) - RTA’s initial compact FT-Raman
system employs a Nd:YAG laser (800 mw at 1064 nm, Keopsis), a compact interferometer (MKS), and an InGaAs
detector (RTA). The system measures Raman spectra from 1064 to 1800 nm (0 to 3500 ∆cm-1). The entire system
also fits into a single 19” box.

RTA Raman Analyzer Research Grade - This system frequency doubles the high powered Spectra Physics laser to

57


generate 20 mW of 532 nm laser excitation. This coupled with the second generation IRA that employs silica
detection will be used to increase the SER scattering efficiency from 50 to 100 times. This will provide important
data regarding potential modifications to the Hamilton Sundstrand Raman system.

General Facilities at RTA - RTA has 15 high-end, interconnected personnel and laptop computers, with Internet
access. The systems are all also tied into a Snap Server for day-to-day back-up. RTA has an extensive wet lab
equipped with two fume hoods for safe preparation of the required samples. A dry box is also available for
preparation of oxygen sensitive chemicals. Also available to RTA are full service machine and electronic shops.

Part 10: References
1
Hughes-Fulford, M., “Review of the Biological Effects of Weightlessness on the Human Endocrine System,”
Receptor, 3:3, 145-54 (Fall 1993).
2
Kohl R.L. and S. MacDonald, “New Pharmacologic Approaches to the Prevention of Space/Motion Sickness,”
J Clin Pharmacol,10, 934-46 (1991).
3
Santy P.,M.Bungo, “Pharmacologic Considerations for Shuttle Astronauts,”J Clin Pharm, 31, 931-3 (1991).
4
LeBlanc, A. and V.Schneider, “Can the Adult Skeleton Recover Lost Bone?” Exp Gerontol, 26, 189 (1991).
5
Lane, H.W., A.D. LeBlanc, L. Putcha, and P.A. Whitson, “Nutrition and Human Physiological Adaptations to
Space Flight,” Am J Clin Nutr, 58:5, 583-6 (Nov 1993).
6
Tipton, C.M., J.E. Greenlead, and C.G.Jackson, “Neuroendocrine and Immune System Responses with
Spaceflights,” Med Sci Sports Exerc, 28:8, 988-98 (Aug 1996).
7
Wood C.D., J.E. Manno, B.R. Manno, H.M. Redetzki, M. Wood, W.A. Vekovius, “Side Effects of Antimotion
Sickness Drugs,” Aviat Space Environ Med, 55:2, 113-6 (1984).
8
Harris, C.I. and G. Milne, "The urinary excretion of N7-methylhistidine by cattle: validation of muscle protein
breakdown," Br. J. Nuutr., 45, 411-422 (1981).
9
Percy, S.D. and M.E. Murphy, "3-methylhistidine excretion as an index of muscle protein breakdown in birds in
different states of malnutrition," Comp. Biochem. Physiol., 116A, 267-272 (1997).
10
Rathmacher J.A. and S.L. Nissen "Development and application of a compartmental model of 3-methylhistidine
metabolism in humans and domestic animals," Adv Exp Med Biol 445, 303-24 (1998).
11
Brazier M, S. Kamel, M. Maamer, F. Agbomson, I. Elesper, M. Garabedian, G. Desmet, J.L. Sebert, "Markers
of bone remodeling in the elderly subject: effects of vitamin D insufficiency and its correction," J Bone Miner
Res,10, 1753-1761(1995).
12
Krall, E.A., B. Dawson-Highes, K. Hirst, J.C. Gallagher, S.S. Sherman, G.J., and G Dalsky, "Bone mineral
density and biochemical markers of bone turnover in healthy elderly men and women," J. Gerontol A Biol Sci
Med Sci, 52, M61-67 (1997).
13
Whitson, P.A., R.A. Pietrzyk, and C.Y. Pak, “Renal Risk Assessment During Space Shuttle Flights,” J Urol,
158:6, 2305-10 (Dec 1997).
14
Taylor G, Houston JB, “Simultaneous Determination of Promethazine and Two of Its Circulating Metabolites
by High-Performance Liquid Chromatography,” J Chromatogr, 230, 194-8 (1998).
15
National Research Council Space Studies Board, A strategy for Research in Space Biology and Medicine in the
New Century, Washington, DC, National Academy Press, 1998.
16
Inoue, H., Y.Date, K.Kohashi, , H.Yoshitomi, , and Y.Tsuruta, “Determination of Total Hydroxyproline and
Proline in Human Serum and Urine by HPLC with Fluorescence Detection,” Biol Pharm Bull, 19, 163-6 (1996).
17
Strickland, P.T., Kang, D, Bowman, E.D., Fitzwilliam, A., Downing, T.E., Rothman, N., Groopman, J.D., and
Weston, A., “Identification of 1-hydroxypyrene Glucuronide as a Major Pyrene Metabolite in Human Urine by
Synchronous Fluorescence Spectroscopy and Gas Chromatrography-Mass Spectrometry,” Carcinogenesis,
15:3, 483-7 (Mar 1994).
18
Potember, R.S., W.A. Bryden, M. Antoine, "Quantitative determination of 3-methylhistidine in urine by matrix-
assisted laser desorption mass spectroscopy," STAIF- 2000, paper 106 (2000).
19
Leelavathi DE, Dressler DE, Soffer EF, Yachetti SD, Knowles JA, “Determination of Promethazine in Human
Plasma by Automated High-Performance Liquid Chromatography with Electrochemical Detection and by Gas
Chromatography-Mass Spectrometry,” J Chromatogr, 339:11, 05-15 (1985).
20
Finnigan product literature (www.finnigan.com), $99,950 for tandem mass spectrometer LCQDUO.
21
Kneipp, K., Wang, Y., Dasari, R. R., and Feld, M. S., “Approach to Single-Molecule Detection Using Surface-
Enhanced Resonance Raman Scattering (SERRS): A Study Using Rhodamine 6G on Colloidal Silver,” Applied

58


Spectroscopy, 49, 780-784 (1995).
22
Nie, S. and Emory, S.R., “Probing Single Molecules and Single Nanoparticles by Surface-Enhanced Raman
Scattering,” Science, 275, 1102 (1997).
23
Farquharson, S. and P. Maksymiuk, “Simultaneous chemical separation and surface-enhancement Raman
spectral detection using silver-doped sol-gels”, Applied Spectroscopy, 57, 479-482 (2003).
24
Lee, Y. H.; W. Smith, S. Farquharson, H.C. Kwon, M.R Shahriari, and P.M. Rainey,"Sol-Gel Chemical Sensors
for Surface-Enhanced Raman Scattering Detection," SPIE, 3537, 252-260 (1998).
25
Lee, Y.-H., S. Farquharson, And P. M. Rainey "Surface-enhanced Raman sensor for trace chemical detection in
water," SPIE, 3857, 76-84 (1999).
26
Farquharson, S., "Miniaturized Fiber Optic Chemical Sensor," STTR Final Phase II Report for NASA Grant
Number NAS9-98024 (submitted February 2000).
27
Farquharson, S., and W. W. Smith, W. H. Nelson and J. F. Sperry, ”Biological agent identification by nucleic
acid base-pair analysis using surface-enhanced Raman spectroscopy,” SPIE, 3533, 207-214 (1998)
28
Farquharson, S., and W. W. Smith, S. Elliott and J. F. Sperry, "Rapid biological agent identification by surface-
enhanced Raman spectroscopy", SPIE, 3855, 110-116 (1999).
29
Lee, Y.H. and S. Farquharson, "Rapid chemical agent identification by surface-enhanced Raman spectroscopy",
SPIE, 4378, 21-26 (2001)
30
Lee, Y.H., and S. Farquharson, "SERS sample vials based on sol-gel process for trace pesticide analysis", SPIE,
4206, 140-146 (2001)
31
Farquharson, S., W.W. Smith, Y.H. Lee, S. Elliott and J. F. Sperry, "Detection of biological signatures: A
comparison of electrolytic and metal-doped sol-gel surface-enhanced Raman media", SPIE, 4575, 62-72 (2002)
32
Farquharson, S. and Lee, Y. H., "High Throughput Identification: Drug Analysis by Surface-Enhanced Raman
Spectroscopy", D&MD Focus Reports, 4-7 (March 2001)
33
Farquharson, S, ”Rapid Drug Assay by Surface-Enhanced Raman Spectroscopy,” SBIR Phase I Final Report for
NIH, Grant No. 1 R43 GM54916-01 (4/14/1998).
34
Farquharson, S., Y.H. Lee, H.C. Kwon, M.R. Shahriari, and P.M. Rainey, "Urinalysis by surface-enhanced
Raman spectroscopy," STAIF- 2000, paper 105 (2000).
35
Farquharson, S., and Y. H. Lee, "Trace drug analysis by surface-enhanced Raman spectroscopy," SPIE, 4200,
89-95 (2000).
36
Wentrup-Byrne, E., Sarinas, S., and Fredericks, P.M., “Analytical Potential of Surface-Enhanced Fourier
Transform Raman Spectroscopy on Silver Colloids,” Applied Spectroscopy, 47, 8, 1192-1197 (1993).
37
Farquharson, S., P. Maksymiuk, K. Ong and S.D. Christesen, "Chemical agent identification by surface-
enhanced Raman spectroscopy", SPIE, 4577, 166-173 (2002).
38
Farquharson, S. PI, "High Sensitivity Raman Spectrometer" NSF CN DMI-0296116.
39
Farquharson, F. PI, "A portable Raman instrument for fuel characterization", Navy, CN M6785-03-M-5043.
40
Inscore, F. E., McNaughton, R., Westcott, B.L., Helton, M.E., Jones, R., Dhawan, I.K., Enemark, J.H., Kirk,
M.L. “Spectroscopic Evidence for a Unique Bonding Interaction in Oxo-Molybdenum Dithiolate Complexes:
Implications for σ Electron Transfer Pathways in the Pyranopterin Dithiolene Centers of Enzymes”, Inorg.
Chem. 38, 1401-1410, (1999).
41
Jones, R.M., Inscore, F.E., Hille, R., Kirk, M.L. “Freeze-Quench Magnetic Circular Dichroism Spectroscopic
Study of the “Very Rapid” Intermediate in Xanthine Oxidase” Inorg. Chem. 38, 4963-4970, (1999).
42
Wang, X.B., Inscore, F.E., Yang, X., Cooney, J.J.A., Enemark, J.H., Wang, L.S. “Probing the Electronic
Structure of [MoOS4]- Centers Using Anionic Photoelectron Spectroscopy”, J. Am. Chem. Soc. 124, 10182-
10191, (2002).
43
Joshi, H.K., Cooney, J.J.A., Inscore, F.E., Gruhn, N.E., Lichtenberger, D.L., Enemark, J.H. “Investigation of
Metal-Dithiolate Fold Angle Effects: Implications for Molybdenum and Tungsten Enzymes”, Proc. Nat. Acad.
Sci., 100, 3719-3724, (2003).
44
Farquharson, S.; Lee, Y. H., and C. Nelson "Material for SERS and SERS sensors and method for preparing the
same", U.S. Provisional Patent Number AFU-17-PROV (1999)
45
Farquharson, S. and S. F. Simpson “Applications of Fiber Optic Raman Spectroscopy to Chemical Processes,”
SPIE, 1796, 272-285 (1992).
46
Woodruff, W.H., Farquharson, S., ”Time-Resolved Resonance Raman Spectroscopy of Hemoglobin
Derivatives: Heme Structure Changes in 7 Nanoseconds,” Science, 201, 831-833 (1978).
47
Farquharson, S., W. W. Smith, W. H. Nelson, and J. F. Sperry, “Biological agent identification by nucleic acid
base-pair analysis using surface-enhanced Raman spectroscopy,” SPIE, 3533-37 (1998)

59


48
Farquharson, S., Weaver, W.J., Lay, P.A., Magnuson, R.H., and Taube, H., “Observation of a Reversible
Adsorbed Redox Couple Using Surface-Enhanced Raman Scattering: Petaammine (pyridine) Osmium (III)/(II)
at Silver Electrodes,” J. Am. Chem. Soc., 105, 3351-3352 (1983).
49
Weaver, M.J., Farquharson, S., Tadayyoni, M.A., “Surface-enhancement factors for Raman scattering at silver
electrodes,” J. Chem. Phys., 82, 4867-4874 (1985).
50
Farquharson, S. and S. F. Simpson, "Applications of Raman spectroscopy to industrial processes," SPIE, 1681,
276-290 (1992).
51
Farquharson, S., S. Bhat, R. Osbaldiston, M. DiTaranto, W. Smith, J. Rose, Y-M. Liu, and M.T. Shaw,
"Characterization of Polymer Composites by Fiber Optic FT-Raman Spectroscopy," SPIE, 3535, (1998).
52
Farquharson, S., W.W. Smith, E. Rigas and D. Granville, "Characterization of polymer composites during
autoclave manufacturing by FT-Raman spectroscopy," SPIE, 4201, 103-111 (2001).
53
Farquharson, S., W.W. Smith, R. C. Carangelo and C. Brouillette “Industrial Raman: providing easy,
immediate, cost effective chemical analysis anywhere”, SPIE, 3859, 14-23 (1999).
54
Farquharson, S. U.S. Patent Number 5,062,713, 1991.
55
Murren JR, Gollerkari A, Anderson S, Lutzker S, Del Prete S, Garrison L, Smith BR: Peripheral blood
progenitor cell cycle kinetics following priming with pIXY321 in patients treated with the 'ICE' regimen. Yale J
Biol Med 1998; 71:355-365.
56
Brown, T. E. and Eckberg, D. L., “Promethazine affects autonomic cardiovascular mechanisms minimally”, J
Pharmacol Exp Ther, 282:2 839-44 (1997).
57
Santy PA, Bungo MW, “Pharmacologic Considerations for Shuttle Astronauts," J Clin Pharmacol, 31:10 931-3
(1991).
58
Dobie TG, May JG, “Cognitive-Behavioral Management of Motion Sickness,” Aviat Space Environ Med, 65:10
Pt 2 C6-7 (1994).
59
Svanberg, E., A.C. Moller-Loswick. D.E. Mathews, U. Koorner, M. Amdersson, and K. Lundham, "The effects
of amino acids to stimulate synthesis and attenuate degradation of skeletal muscle proteins in man," Am. J.
Physiol., 7/6 (1996).
60
Mundy G., Garrett R., Harris S., Chan J., Chen D., Rossini G., Boyce B., Zhao M., Gutierrez G. "Stimulation of
bone formation in vitro and in rodents by statins", Science, 286, 1946-1949 (1999).
61
Ettinger B, Black DM, Mitlak BH, Knickerbocker RK, Nickelsen T, Genant HK, Christiansen C, Delmas PD,
Zanchetta JR, Stakkestad J, Glüer CC, Krueger K, Cohen FJ, Eckert S, Ensrud KE, Avioli LV, Lips P,
Cummings SR "Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with
raloxifene: results from a 3-year randomized clinical trial." JAMA, 282 637-645 (1999).
62
Davis JR, Jennings RT, Beck BG, Bagian JP, “Treatment Efficacy of Intramuscular Promethazine for Space
Motion Sickness”, Aviat Space Environ Med, 64:3 Pt 1 230-3 (1993).
63
Wood CD, Manno JE, Manno BR, Redetzki HM, Wood MJ, Mims ME, “Evaluation of Antimotion Sickness
Drug Side Effects on Performance”, Aviat Space Environ Med, 56:4 310-6 (1985).
64
Collins WE, Schroeder DJ, Elam GW, “A Comparison of Some Effects of Three Antimotion Sickness Drugs on
Nystagmic Responses to Angular Accelerations and to Optokinetic Stimuli,” Aviat Space Environ Med, 53:12
1182-9 (1982).
65
Antoine, M.D., W.A. Bryden, H.W. Ko, P.F. Scholl, R.S. Potember and R.J. Cotter, “Real Time Sampling and
Analysis of Biological Biomarkers by TOF Mass Spectrometry,” Proceeding of the 28th International
Conference on Environmental Systems, (1998)
66
Chase, B. “A New Generation of Raman Instruments,” Appl. Spectrosc. 48, 14A-19A (1994).
67
Chase, D.B. and Rabolt, J.F., Fourier Transform Raman Spectroscopy, Acad. Press, Chapter 1 and p 131
(1994).
68
Connes, J., Rev. Opt. Theor. Instrum., 40, 54 (1961).
69
Vickers, T.J. and Mann, C.K., "On-line monitoring by Raman spectroscopy: instrument control and
calibration," SPIE, 2367, 219-22 (1994).
70
Fleischmann, M., Hendra, P.J., and McQuillan, A.J., Chem. Phys. Lett., 26, 123, (1994).
71
Section Four: Theory in "Surface-Enhanced Raman Scattering," (M. Kerker and B. Thompson Eds.) SPIE, MS
10, also p. 225 (1990).
72
Pettinger, B., "Light Scattering by Adsorbates at Ag Particles; Quantum-Mechanical Approach for Energy
Transfer Induced Interfacial Optical Processes Involving Surface Plasmons, Multipoles, and Electron-hole
Pairs," (M. Kerker and B. Thompson Eds.) SPIE, MS 10, 683-692 (1990).
73
Wang., D.-S., and Kerker, M., "Enhanced Raman Scattering by Molecules Adsorbed at the Surface of Colloidal

60


Spheroids," SPIE, (M. Kerker and B. Thompson Eds.) SPIE, MS 10, 417-429 (1990).
74
Lee, P.C. and D. Meisel. J. Phys. Chem., 86, 3391-3395 (1982).
75
Laserna, J.J., A. Berthod, and J.D. Winefordner, “Evaluation and optimization of experimental conditions for
surface-enhanced Raman detection of analytes in flow injection analysis,” Microchemical J., 38, 125 (1988).
76
Angel, S.M., L.F. Katz, D.D. Archibold, L.T. Lin, D.E. Honigs, Applied Spectroscopy, 42, 1327 (1988).
77
Vo-Dinh, T., Hiromoto, M.Y.K., Begun, G.M., and Moody, R.L., Analytical Chemistry, 56, 1667-1670 (1984).
78
Vo-Dinh, T., Stokes, D.L., Li, Y.S., and Miller, G.H., SPIE 1368, 203-209 (1990).
79
Sutherland, W. S. and Winefordner, W.S., “Preparation of substrates for surface-enhanced Raman microprobe
spectroscopy,” J. Raman Spectrosc., 541-549 (1991).
80
Li, Y.-S., and Wang, Y., Applied Spectroscopy, 46, 142-146, (1992).
81
Maya, L., C.E. Vallet, Y.H. Lee, “Sputtered Gold Films for Surface-Enhanced Raman Scattering,” J. Vac. Sci.
Technol., A 15(2), 238 (1997).
82
Pothier, N.J., and Force, R.K., "Surface-Enhanced Raman Spectroscopy at a Silver Electrode as a Detection
System in Flowing Streams," Analytical Chemistry, 62, 678-680 (1990).
83
Norrod, K.L., L.M. Sudnik, D. Rousell, and K. L. Rowlen in “Quantitative comparison of five SERS substrates:
Sensitivity and detection limit,” Applied Spectroscopy, 51, 994-1001 (1997)
84
Klein, L. C., Sol-Gel Technology for Thin Films, Fibers, Performs, Electronics, and Specialty Shapes, Noyes
Publications, Park Ridge, New Jersey, U.S.A. (1988).
85
Zelinski, B.J., C.J. Brinker, D.C. Clark and D.R. Ulrich, Materials Research Society Symposium Proceedings,
Vol. 121, Better Ceramics Through Chemistry III, Pittsburgh, PA (1988).
86
Lev, O., M., Tsionsky, M., Rabinovich, L., Glezer, V., Sampath, S., Pankratov, I., and Gun, J., “Organically
Modified Sol-Gel Sensors,” Anal. Chem. 67, 22A-30A, (1995).
87
Czarnik, A.W., "Combinatorial: What's in it for chemists?," Anal. Chem., 378A-386A (1998).
88
Lev, O. “Diagnostic applications of organically doped sol-gel porous glass,” Analusis, 20 543-553 (1992).
89
Muck, A., et al., “Fabrication of poly(methyl methacrylate) microfluidic chips by atmospheric molding.”, Anal.
Chem., 76, 2290-2297 (2004).
90
Silman, O, LA Bumm, R Callaghan, CG Blatchford, M Kerker, “SERS by citrate on colloidal silver”, J. Phys.
Chem., 87, 1014-1023 (1983)
91
Fujii, T, Y Sando, K Higashino, Y Fujii, “A plug and play microfliudic device”, Lab Chip, 3, 193-197 (2003).
92
Puri, V., (2003), available at www.bhj.org/journal/2003_4502_april/biochemicalmarketrs_325
93
Weksler, M. E., “Osteoporosis: Up-to-Date Strategies for Prevention and Treatment.” Geriatrics. April 1997
v52 n4 p92(5).
94
Holbrook TL et al. Amer Acad Orthopedic Surg 1984.
95
Farquharson, S. and M. Marcus, "Implementing process instrumentation: worst case scenarios," SPIE, 4201, 1-
11 (2000).
96
See Hamilton Sundstrand Homepage at www.hamiltonsundstrandcorp.com.

61

Nasa Nnc05 Ca90 C Final Report

Nasa Nnc05 Ca90 C Final Report

More Related Content

What's hot (18)

Viewers also liked (7)

Similar to Nasa Nnc05 Ca90 C Final Report (20)

More from inscore (20)

Nasa Nnc05 Ca90 C Final Report