69453

I R N E

U R
L S A
T S P M
RA I E A
AC G C
U R S V A F P M I E A N C N G T

A J S . I . AE C F I ND . I I DL N A F

S 2 E 0 P 0 T 1 E

E N S D V C E I I

B NR L AO A TO

1 , U L IP N
O YS T 1 EN OL O
 DISCLAIMER
This report was prepared as an account of work sponsored by an agency of the
United States Government. Neither the United States Government nor any agency
thereof, nor any employees, nor any of their contractors, subcontractors or their
employees, makes any warranty, express or implied, or assumes any legal liability
or responsibility for the accuracy, completeness, or any third party’s use or the
results of such use of any information, apparatus, product, or process disclosed,
or represents that its use would not infringe privately owned rights. Reference
herein to any specific commercial product, process, or service by trade name, trade-
mark, manufacturer, or otherwise, does not necessarily constitute or imply its en-
dorsement, recommendation, or favoring by the United States Government or any
agency thereof or its contractors or subcontractors. The views and opinions of
authors expressed herein do not necessarily state or reflect those of the United
States Government or any agency thereof.

Available electronically at
http://www.doe.gov/bridge
Available to U.S. Department of Energy and its contractors in paper from-
U.S. Department of Energy Office of Scientific and Technical Information
P.O. Box 62 Oak Ridge, TN 37831
(423) 576-8401
Available to the public from-
U.S. Department of Commerce National Technical Information Service
5285 Port Royal Road Springfield, VA 22131
(703) 487-4650

PRINTED ON RECYCLED PAPER

 BNL-
Informal Report

ULTRAVIOLET RAMAN SPECTRAL SIGNATURE ACQUISITION:

UV RAMAN SPECTRAL FINGERPRINTS

A. J. Sedlacek III and C. Finfrock

Environmental Sciences Department
Brookhaven National Laboratory
Upton, NY 11973-5000

September 2001

Year End Report

submitted to
Edgewood Chemical Biological Center (ECBC)
in support of the Raman spectroscopy research effort

RESEARCH BY BNL INVESTIGATORS WAS PERFORMED UNDER THE AUSPICES OF THE

U.S. DEPARTMENT OF ENERGY UNDER CONTRACT NO. DE-AC02-98CH10886.
Executive Summary:

As a member of the science-support part of the ITT-lead LISA development program, BNL is tasked with
the acquisition of UV Raman spectral fingerprints and associated scattering cross-sections for those
chemicals-of-interest to the program’s sponsor. In support of this role, the present report contains the first
installment of UV Raman spectral fingerprint data on the initial subset of chemicals. Because of the unique
nature associated with the acquisition of spectral fingerprints for use in spectral pattern matching
algorithms (i.e., CLS, PLS, ANN) great care has been undertaken to maximize the signal-to-noise and to
minimize unnecessary spectral subtractions, in an effort to provide the highest quality spectral fingerprints.

This report is divided into 4 sections. The first is an Experimental section that outlines how the Raman
spectra are performed. This is then followed by a section on Sample Handling. Following this, the spectral
fingerprints are presented in the Results section where the data reduction process is outlined. Finally, a
Photographs section is included.

Experimental Section:

Figure 1 is a schematic diagram of the experimental apparatus used to measure Raman cross sections. A
Spectra- Physics MOPO-730 system generates continuously tunable radiation from 410 nm to 2000 nm.
The MOPO-730 is a BBO-based optical parametric oscillator (OPO) pumped by the third harmonic (355
nm) of a 30 Hz Nd:YAG laser. The UV light (210–400 nm) is obtained by doubling either the signal or the
idler beam from the MOPO. The pulse width of the UV light is 3–4 ns, and the linewidth is approximately
0.3 cm-1. Photographs of the Raman system can be found at the end of this report.

Figure 1: Schematic of Experimental Setup Used to Collect UV Raman Signatures

A 254 mm focal length cylindrical lens focuses the excitation light into the sample. The UV energy is 0.5
mJ/pulse or less, and the spot size at the sample is approximately 0.733 mm. The scattered light is collected
at 180° with respect to the incident UV beam by a 50.8 mm diameter spherical lens (f number=1). The
angle between the UV beam and the surface of the sample stream is 45° to reduce back reflections from the
cell walls. Another 50.8 mm diameter spherical lens focuses the scattered light into the prism predisperser.
The f number of the second lens matches the f number of the prism predisperser and spectrometer for

-1-
maximum throughput of the scattered light. The predisperser works as a sharp-cut bandpass filter with high
throughput (40%–50%) and negligible aberration.

The spectrometer is a 1.26 m single stage monochromator (Spex Model 1269) with an 1800 grooves/mm
grating. The width of the entrance slit in the spectrometer is controlled to reject elastically scattered laser
light, while the resolution of Raman spectra is governed by the slit width of the predisperser. With this
combination, we are able to measure Raman shifts of less than 200 cm-1 in deep UV with negligible
Rayleigh background. A crystalline quartz wedge polarization scrambler in front of the entrance slit of the
predisperser depolarizes the polarized Raman signal before it enters the spectrometer.

The optical throughput of both the predisperser and the spectrometer has been measured at various
wavelengths in the UV using a NIST standard D2 lamp. The spectra of a standard Hg lamp and known
chemicals serve as wavelength calibrations for the predisperser and the monochromator.

The Raman spectra are captured by a gated, intensified CCD detector (Oriel Instaspec V, 1024X256 pixel
array, 25 mm (and 18 mm) window width) mounted at the exit port of the spectrometer. The CCD chip is
cooled to -25 °C (-35 °C) to yield a dark count of less than 2x10-5 counts element-1 s-1. The adjustable gain
of the intensifier is set to ~50 counts/photoelectron (150 counts/photoelectron) for all measurements. The
Raman spectra are normalized to the measured response uniformity of the ICCD array.

Sample Handling

Preparation and handling of samples for analysis by UV Raman spectroscopy was conducted using the
procedures, protocols and methods described below. Chemicals used in the analysis were obtained from
reputable commercial research chemical suppliers. They were obtained in the highest purity commonly
available to the research community. Certificates of analysis were obtained from the vendor for virtually
all of the chemicals used in this project. Chemical purity was verified in-house by gas chromatography
(HP 5890 Series 2 chromatograph outfitted with a 25 m x 0.32 mm x 0.52 µm FFAP capillary column) for
all tested liquids except SF-96 and HPLC grade water. In order to preserve the purity of the chemicals
tested, only certain container and sample handling materials were permitted. Acceptable sample handling
materials were glass, Teflon, PFA (perfluoro (alkoxyalkane) copolymer), PEEK (polyaryl ether ether
ketone) and stainless steel, all of which were subjected to the following cleaning protocol. All materials
were ultrasonically cleaned in warm distilled water and ultrasonic cleaner detergent solution for a minimum
of fifteen minutes. The materials were then rinsed with copious amounts of distilled water and then rinsed
with spectroscopic grade methanol, followed by drying in an 80oC glassware-drying oven. The only
departure from this protocol was the method for cleaning the magnetic drive pumps used with the
windowless flow cell. As these pumps have interior spaces that are not easily accessible for cleaning,
sufficient pumps were obtained to allow compatible chemicals to have dedicated pumps. In the event that a
pump did require cleaning, the following method was applied. The pump was allowed to pump itself dry
into a waste vessel. Several aliquots of appropriate solvents were cycled through the pumps alternated with
drying of the interior with dry filtered compressed air. This was done over the course of several hours.
The final rinse of a pump was with a solvent of high vapor pressure, such as cyclohexane, which was then
easily removed by flowing dry filtered compressed air through the pump for several hours while it was
slowly rotating.

Some samples required dilution prior to analysis. Dilutions, when required, were prepared gravimetrically
using a Mettler AE 100 or a Mettler SB 8001 electronic balance, as dictated by the amount of sample being
prepared. Concentrations are expressed in terms of weight percent, in other words, mass of minor
component divided by total mass, times 100.

-2-
--
In order to address potential reabsorption, UV/VIS absorption spectra of all samples were collected with a
Perkin-Elmer 320 spectrophotometer. In addition to this, NIR Raman spectra were collected using a
Bruker IFS-66 FT spectrometer outfitted with a NIR Raman attachment (FRA-106).

Two sample-handling systems were developed to facilitate liquid sample exposure to the UV Raman
spectroscopy system. A closed system using quartz flow cells and a syringe pump was developed, as well
as a “windowless” system consisting of an open flow cell and small magnetic drive chemical pumps. The
closed system consisted of a Harvard Apparatus “33” syringe pump, an array of ten milliliter glass
syringes, a set of four way control valves, a collection of quartz flow cells of varying path length, a set of
glass sample reservoirs, required connecting tubing, and a positioning system to locate the cells in the beam
path. The syringe pump was programmed to alternately draw and push the sample between the reservoir
and syringe body via the connecting tubing and the quartz cell. This system provided a completely closed
environment for the sample with good control of the flow rate and path length. Evaporation of samples
with high vapor pressure was well suppressed in this system. A disadvantage of this system was the
spectroscopic contribution of the quartz cells containing the samples. The “windowless” sample
introduction system circumvented this problem. The “windowless” system consisted of an array of
magnetic drive chemical pumps from Micropump Corp., an array of glass sample reservoirs, appropriate
connecting tubing, and a specially fabricated Teflon and glass cell. The upper member of the cell is
machined from Teflon and is a hollow cylinder with a narrow slot milled along the diameter of the end. At
each end of the slot, small glass rods are inserted into drilled holes that terminate the ends of the slot. This
is mounted in a fixture so that the slot and glass rods point vertically downward. Immediately below the
glass rods is a Teflon well that has a PFA compression fitting below for attachment of a glass sample
reservoir. The pump is connected to draw sample from the bottom of the reservoir and deliver it to the
interior of the upper Teflon member. The sample flows out of the slot and wets the glass rods, forming a
falling film or sheet of sample, which is collected back into the reservoir by the well. This assembly may
be mounted to the same positioning system as the quartz cells described above, and is typically positioned
so that the beam passes through the falling film formed between the glass rods. This system has the
advantage of eliminating any contribution to the spectrum by cell materials in the beam path, but the flow
rate and sample film characteristics are somewhat dictated by the sample’s surface tension and wetting
characteristics.

Solid samples were fixed to the positioning system used for the cells above by appropriate clamps, and
positioned in the beam manually.

Results:

As an illustrative example of the data reduction process, we will use cyclohexane. Shown in Figure 2 are
the 4 raw spectral regions collected using the system described above. These Spectra are then corrected for
ICCD pixel-to-pixel variation, the instrumentation response function and ICCD dark current. Following
these corrections the individual spectra are spliced together to form the desired spectral fingerprint. Once
completed, the spectral features’ (in the present case the cyclohexane modes) pixel locations are recorded
as a function of Raman shift. Figure 3 shows the calibration for our example. After application of this
calibration curve, the resulting spectrum is then corrected for instrumentation throughput and, if necessary,
signal reabsorption as a function of Raman shift. Shown in Figure 4 is the final UV Raman spectrum for
cyclohexane. Also shown in Figure 4 is the Raman spectrum collected using the Bruker FTS-66 NIR
Raman spectrometer. As can be clearly seen, the UV spectrum correlates very well with the spectrum
collected in the NIR. It should also be pointed out that use of the windowless flow cell enables acquisition
of Raman spectral features well below 500 cm-1, as evidenced by the clear observation of the 382.7 and
425.2 cm-1 modes of cyclohexane.

-3-
--
 3.5 1.6
SPEX: 254 nm SPEX: 258.5 nm
3.0 1.4

1.2
2.5
1.0
2.0
6

6
x10

x10
0.8
1.5
0.6
1.0
0.4
0.5 0.2
0.0
0 200 400 600 800 1000 0 200 400 600 800 1000
A) ICCD Pixel Number B) ICCD Pixel Number

1.6 SPEX: 260.5 nm 40 SPEX: 265 nm

1.4
1.2 30

1.0
6

6
x10

x10

0.8 20
Atmos. N2
0.6
Atmos. O2
0.4 10

0.2
0
0 200 400 600 800 1000 0 200 400 600 800 1000
C) ICCD Pixel Number D) ICCD Pixel Number

Figure 2: Uncorrected Individual Spectral Slices for Cyclohexane.

Cyclohexane Wavenumber
2500 Best fit to 3-order polynomial
Wavenumbers

2000

1500

1000

500

200 400 600 800 1000 1200 1400

ICCD Pixel Number
Figure 3: Wavenumber to ICCD Pixel Number Calibration
-4-
--
 NIR-FT-Raman: Cyclohexane

UV-Raman: Cyclohexane

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )
Figure 4: UV Raman spectrum of neat cyclohexane. Upper trace is the NIR Raman spectrum of
cyclohexane for comparison purposes.

Summary of Chemical Properties and Absorbance Cross-Sections

Chemical Species M.W. Density (g/cc) UV Absorbance Cross-section
at 248 nm (cm2/molecule)†
acetonitrile 41.05 0.786 NA§
cyclohexane 84.16 0.779 NA
DeM 160.17 1.055 1.25±0.04 x 10-20
MeS 152.1 1.174 7.75±0.11 x 10-18
n-hexane 86.18 0.6603 NA
PIBM 300,000 n/a 4.89± 0.17x 10-16
SF96 13,700 0.973 1.24±0.01 x 10-20
water 18.01 1 NA
§
NA:no absorbance
†
Quoted cross-section error bars are 2-σ

-5-
--
The spectra below are in alphabetical order. Quoted spectral locations were generated by curve fitting (a
series of Voigt lineshapes) to a given Raman spectrum. Data reduction was accomplished using Igor Pro.

Acetonitrile

NIR-FT-Raman: Acetonitrile

UV-Raman: Acetonitrile

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Positions (cm-1)

377.407
917.766
1373.94
1437.72
2251.55
2292.25
2731.98
2939.96
3000.16

-6-
--
Concrete

NIR-FT-Raman: Concrete

UV-Raman: Concrete

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

1052.82
1391.64

Featureless.

-7-
--
Cyclohexane

NIR-FT-Raman: Cyclohexane

UV-Raman: Cyclohexane

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

382.675 2695
425.219 2797.68
799.615 2850.44
1026.5 2886.06
1156.14 2921.58
1265.31 2935.09
1345.23
1465.62
1442.69
2631.3
2662.33

-8-
--
Diethyl Malonate (DEM)

NIR-FT-Raman: Neat DEM

UV-Raman: Neat DEM

* Location of new Resonance

Enhanced - based peak

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

340.368 946.646 1416.34
376.239 1031.13 1454.71
568.057 1098.84 1629.71
595.222 1114.06 1746.35
678.746 1151.89 2725.11
785.627 1271.23 2773.66
845.176 1293.22 2873.33
863.538 1368.27 2898.98
931.958 1395.83 2934.98

The UV Raman spectrum of DeM is unique is that a new peak is readily observable (1629.7 cm-1) under
248 nm illumination. Such features can be observed in RR depending upon the interactions between the
ground state and an upper electronic state.

-9-
--
Methyl Salicylate (MeS) in Acetonitrile: 0.15% (w/w)

NIR-FT-Raman: 0.151% (w/w) MeS in Acetonitrile

*
*

*
*
* Mes Res. Raman peak locations

***
* * * *
*

UV-Raman: 0.151% (w/w) MeS in Acetonitrile

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

377.675 1409.02
451.145 1412.23
560.164 1443.49
806.998 1480.41
849.67 1602.8
918.706 1615.4
1033.09 1679.74
1193.59 2252.91
1250.4 2293.8
1333.92 2941.26
1374.45 3002.06

As evidence in the graph above, MeS is undergoing a very large resonance-enhancement at 248 nm.

- 10 -
--
Methyl Salicylate (MeS) in Water: 0.054% (w/w). Comparison with 0.15% MeS in
Acetonitrile

*
*
* * UV-Raman:
0.15% (w/w) MeS in Acetonitrile
*
**
* *
* *

*
*
* * UV-Raman:
0.054% (w/w) MeS in water
**
* * * *

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

814.772
855.631
1040.86
1200.7
1235.71
1252.22
1339.94
1383.51
1446.75
1487.44
1615.71
1674.15

Comparison to the MeS locations in the acetonitrile reveals that some shifts are taking place. See page 11
and page 13 for MeS in acetonitrile and neat MeS (as measured in with the FT-Raman spectrometer).

- 11 -
--
Neat Methyl Salicylate (MeS): NIR-FT-Raman

NIR-FT-Raman: Neat MeS

0 500 1000 1500 2000

-1 2500 3000 3500
Raman Shift (cm )

Raman Shift (cm-1)

185.89 1134.7 3092.31
264.957 1155.91
334.381 1250.4
355.595 1327.54
438.519 1335.26
511.8 1441.32
561.941 1464.46
666.078 1583.97
758.644 1613.93
808.784 1677.05
847.354 2852.93
964.99 2955.34
1032.49 3024.06
1061.41 3054.11
1090.34 3071.39

The very large absorption cross-section for MeS precluded acquisition of a UV-Raman spectra from a neat
sample.

- 12 -
--
n-Hexane

NIR-FT-Raman: n-Hexane

UV-Raman: n-Hexane

500 1000 1500 2000 2500 3000 3500

-1
Raman Shift (cm )

Raman Shift (cm-1)

311.93 1165.07
329.545 1356.45
367.367 1451.39
400.116 2654.08
454.314 2729.46
818.937 2855.12
866.99 2873.31
892.993 2894.24
1004.11 2914.42
1038.08 2935.43
1070.38 2960.94
1080.02
1138.56
- 13 -
--
Polyisobutyl Methacrylate: (PIBM);

NIR-FT-Raman: Polyisobutyl Methacrylate (PIBM)

"Thickener"
solid: 300,000 MW

500 1000 1500 2000 2500 3000 3500

-1
Raman Shift (cm )

Raman Shift (cm-1)

261.187 1337.03
384.043 1370.07
430.571 1390.28
544.625 1446.43
599.678 1600.11
796.694 1725.82
829.501 1791.09
852.916 2722.11
879.63 2766.42
1461.52 2871.93
940.425 2888.52
965.565 2916.89
993.406 2935.9
1000.64 2961.51
1123.39 2997.64
1130.81
1178.78
1238.66
1248.86
1299.87
- 14 -
--
SF96 Silicone Oil

NIR-FT-Raman: SF96

UV-Raman: SF96

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

487.506
705.206
789.66
860.241
1260.73
1414.47
2815.66
2905.26
2965.96

This chemical species is highly desirable as a candidate chemical due to it simple Raman spectrum,
minimal absorption and ease of use.

- 15 -
--
Thickened-DeM (TDeM): [2.99% PIBM]

NIR-FT-Raman

*
UV-Raman
Res. Raman signals attributable
* * to Polyisobutyl methacrylate
(PIBM)
*
*

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

331.407 995.339 1747.12
375.283 1026.03 1778.19
425.009 1091.06 2755.81
567.431 1110.14 2872.67
671.519 1176.73 2896.62
780.809 1229.8 2933.89
842.941 1271.38 2974.2
860.374 1451.52 3062.89
927.995 1598.2
951.115 1616.66

The weaker Raman signal (relative to neat DEM) observed for TDEM is attributed to the large backscatter
signal that accompanied the desired Raman signal. The large backscattered signal has been tentatively
ascribed to the polyisobutyl methacrylate. It is estimated that this 300,000 MW “thickener” has a nominal
diameter on the order to microns thereby resulting in Mie scattering contributions. It should also be noted
that PIBM is also undergoing a significant resonance-enhanced as evidenced by new peaks at 995, 1176,
1598, and 1778 cm-1.

- 16 -
--
Teflon

NIR-FT-Raman: Teflon

UV-Raman: Teflon

500 1000 1500 2000 2500 3000

-1
Raman Shift (cm )

Raman Shift (cm-1)

386.717
574.028
733.515
1213.53
1296.33
1377.56
2567.3

- 17 -
--
Water

NIR-FT-Raman of HPLC grade water

UV-Raman of HPLC grade water

500 1000 1500 2000 2500 3000 3500

-1
Raman Shift (cm )

Raman Shift (cm-1)

1633.23
3292.12

- 18 -
--
Photographs of MOPO-based UV Raman Laboratory

Figure 5: Spectra-Physics GCR 290 pumped MOPO 730 laser system

Figure 6: View inside the SP MOPO 730 head. Top is the Master Oscillator and bottom in the Power
Oscillator
- 19 -
--
Figure 7: McPherson Prism Predisperser (blue) used to suppress unwanted stray light. Following the
predisperser the scattered light enters a 1.26 m SPEX single-grating spectrometer and then exits to an array
detector (Andor ICCD housed in an Oriel case).

- 20 -
--
Figure 8: The sample compartment. Four standard quartz flow cells can be readily seen. Each cell is
dedicated to a specific chemical thereby eliminating cross-contamination. To the left of the four quartz
cells is the windowless flow cell. See Figures 9 and 10 for a more detailed picture of this flow cell.

- 21 -
--
Figure 9: BNL designed windowless flow cell. This cell eliminates Raman signal contributions from the
quartz flow cells.

- 22 -
--
Figure 10: Windowless flow cell. Using a red food dye, operation of the windowless flow cell can be
observed. For more viscous materials (TDEM, SF96), less turbulent structure is observed.

- 23 -
--
Figure 11: Bruker FT-Raman system used to collect Raman signatures in the NIR, where contributions due
to resonance-enhancement are virtually non-existent.

- 24 -
--
Figure 12: Perkin-Elmer 320 UV/VIS spectrophotometer (foreground) and an ATi Mattson Genesis Series
FTIR (background).

- 25 -
--
Figure13: HP 5890 Series 2 gas chromatograph used to confirm sample purity upon delivery from a
supplier and to check for sample degradation during laser Raman experiments.

- 26 -
--