Deep-Ultraviolet Resonance Raman Excitation Proﬁles of ...asher/homepage/spec_pdf/Deep-Ultraviolet Resonance Raman... · Deep-Ultraviolet Resonance Raman Excitation Proﬁles of

Deep-Ultraviolet Resonance Raman Excitation Profiles ofNH4NO3, PETN, TNT, HMX, and RDX

Manash Ghosh, Luling Wang, Sanford A. Asher*

Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260 USA

We measured the dispersion of the absolute-differential Raman cross-

sections of ammonium nitrate (NH4NO3), pentaerythritol tetranitrate

(PETN), trinitrotoluene (TNT), nitroamine (HMX), and cyclotrimeth-

ylene-trinitramine (RDX) in acetonitrile and water solutions between 204

and 257 nm. The ultraviolet (UV) resonance Raman/differential Raman

cross-sections of NH4NO3, PETN, TNT, HMX, and RDX dramatically

increase as the excitation wavelength decreases deep into the UV to 204

nm. NH4NO3, PETN, and RDX are best resonance-enhanced by the 204

nm excitation used here, while the optimum excitation wavelength for

TNT and HMX is ;230 nm. The excitation profile of TNT roughly follows

its absorption band shape. The excitation profiles for the different Raman

bands of each explosive molecule differ, indicating that multiple-excitation

wavelength spectra are not redundant and can offer additional

information on the species present. We see no evidence of any nonlinear

spectral response or sample degradation at the fluences and spectral

accumulation times used here. However, we previously observed such

phenomena at longer spectral accumulation times and higher fluences.

These results are promising for the development of standoff deep-UV

Raman methods for explosive molecule determinations.

Index Headings: Explosives; Resonance Raman; Raman excitation profile;

Energetic materials; Raman cross-section.

INTRODUCTION

The need for methods to rapidly (and remotely) detectexplosives has increased enormously recently due to the use ofimprovised explosive devices (IEDs) by terrorists.1–3 Remoteand standoff detection are generally envisioned as methods thatuse laser sources to remotely interrogate surfaces. Light emittedor scattered from the surface would be collected and analyzedfor information on the molecular species present.

Raman spectroscopy (RS) appears to be the most promisingtechnique for rapidly detecting and analyzing explosivematerials,4 partly because RS identifies the molecule throughits unique vibrational signature. The sensitivity of normal RS islimited because of small RS cross-sections.5,6 Further, similarlysmall RS cross-sections occur for most materials. Thus,congested spectra are likely to occur in situations where theexplosive species are intermixed with environmental interfer-ents. Also spectral signal-to-noise (S/N) ratios are generallylow for normal RS measurements of impure samples excited inthe near-ultraviolet (near-UV), visible, and near-infrared (NIR)due to strong fluorescence from the ubiquitous impurities thatfluoresce in these spectral regions.7

These considerations have motivated interest in using RSwith UV excitation to increase the sensitivity, selectivity, andS/N in remote RS detection.8,9 Excitation in the UV results in

increased sensitivity because of the increased Raman cross-sections due to resonance and the m4 Raman cross-sectionincreases. There can be large selectivity enhancement byexciting at wavelengths selectively in resonance with theanalyte, which can minimize resonance with interferents.Furthermore, fluorescence is completely extinguished whenexciting with deep-UV excitation less than ;260 nm, becausemolecules whose first singlet states reside at wavelengths lessthan 260 nm are flexible and relax nonradiatively.7 Moleculeswith first singlet states that reside at wavelengths greater than260 nm emit wavelengths greater than 260 nm and do notspectrally interfere. Most explosives absorb in the deep-UVregion.8 Therefore, excitation in the deep-UV will optimizestandoff Raman scattering S/N and may enable sensitivestandoff detection out in the field where it is so needed.

Our research group has been developing deep-UV resonanceRaman spectroscopy (DUVRR) for applications in analyticalchemistry, physical chemistry, and biological chemistry.5–12

We8,13–15 and a few others16–20 recently began exploring theuse of DUVRR for explosive detection. In the first stage, it isessential to determine the dependence of the Raman cross-sections on the excitation wavelength in order to quantify andpredict the spectral S/N of the Raman spectral measurements.

Thus, in the work here, we measured the dispersion of theUV differential Raman cross-sections of trinitrotoluene (TNT),pentaerythritol tetranitrate (PETN), cyclotrimethylene-trinitr-amine (RDX), the chemically related nitroamine, HMX, andammonium nitrate (NH4NO3) in solution using excitationbetween 200 nm and 260 nm. For these studies we began ourinvestigations by measuring the UV Raman (UVR) excitationprofiles for these molecules in transparent solution. Thisisolates the Raman response of these chemical species such thatthey are not electronically or vibrationally coupled to adjacentanalyte molecules. This allows us to characterize the intrinsicresponse of these species.

Future work will determine the dispersion of the Ramancross-sections for these explosives in the solid state. We arealso developing methods to predict the UVR spectral intensitiesand the S/N of solid explosives dispersed onto surfaces.

EXPERIMENTAL

Materials. PETN and RDX, received as gifts from theFederal Aviation Administration, were used as received. Wepurchased TNT and HMX from Omni Distribution, Inc.(Marion, AR). TNT was received in the form of yellow flakes,and HMX (PBXN-5, desensitized) was received in the form ofa powder and used as received. Ammonium nitrate (NH4NO3)was purchased from J.T. Baker (Phillipsburg, NJ) and used asreceived. High-performance liquid chromatography (HPLC)-grade acetonitrile (ACN) was purchased from EMD Chemicals,Inc. (Gibbstown, NJ) and used as received.

Received 10 February 2012; accepted 31 May 2012.* Author to whom correspondence should be sent. E-mail: [email protected]: 10.1366/12-06626

Volume 66, Number 9, 2012 APPLIED SPECTROSCOPY 10130003-7028/12/6609-1013$2.00/0

� 2012 Society for Applied Spectroscopy

Absorption Measurements. The UV absorbance spectrawere measured by using a Cary 5000 Varian UV–VIS–NIRspectrophotometer. A 1 mm-path-length fused-silica cuvettewas used for absorption measurements of the 0.1 mg/mLsamples of TNT, PETN, RDX, HMX, and NH4NO3. ACN wasused as the solvent for TNT, PETN, RDX, and HMX, and a23% water/77% ACN solution by volume was used forNH4NO3.

Raman Instrumentation. The UVRR instrumentation wasdescribed previously.21–23 Continuous wave (CW) excitation at229, 244, and 257 nm was obtained with an Innova 300 FReDargon ion laser (Coherent, Inc.) with appropriate doublingcrystals. Excitation at 204 nm was generated by the fifth anti-Stokes hydrogen Raman shift of the third harmonic of aCoherent Infinity Nd:YAG, 100 Hz, 3 ns pulsed laser. The CWlaser beams were focused into a 25 lm diameter spot. The 204nm beam was focused into a 200 lm diameter spot, onto acirculating open-flow stream21,22 to avoid any spectral impactof photodegradation.24,25 The average laser power on thesample was between 0.3 and 2.0 mW. Each 15 to 20 mLsample was irradiated for a total of 3 to 5 min to collect threeconsecutive spectra to check for degradation. Power-dependentmeasurements were obtained by placing neutral density filtersin the excitation beam.13

The Raman scattered light was collected in a 1508backscattering geometry. For 229, 244, and 257 nm excitation,the scattered light was collected and directed into a SPEX 1877Triplemate spectrometer modified for deep UVRR measure-ments. A subtractive double monochromator, SPEX 1401,modified for deep UVRR measurements, was used for 204 nmexcitation.21

The throughput of the spectrometer varies with wavelength.These differences in throughput were removed by compensat-ing for the spectrometer efficiency.21 The Raman scattered lightwas detected by a liquid-nitrogen-cooled charge-coupleddevice (CCD) (Princeton Instruments, Spec-10:400B). Wins-pec software (Roper Scientific) was used to record the spectrathat were analyzed and deconvoluted using Grams/32 AI 8.0software (Thermo Electron Corporation, Waltham, MA).

Absolute Differential Raman Cross-Sections. The abso-lute differential Raman cross-sections of NH4NO3, PETN,TNT, HMX, and RDX bands were determined by using theintegrated intensities of the 918 cm�1 C–C stretching band ofACN as an internal intensity standard. The absolute differentialRaman cross-sections (rR) of the NH4NO3, PETN, TNT,HMX, and RDX bands corrected for self absorption werecalculated:

rR ¼IbandjðkACNÞCACN

IACNjðkbandÞCsample

� rACN �es þ eex

er þ eex

� �ð1Þ

where Iband and IACN are the observed integrated intensities ofthe explosive molecule and the ACN Raman bands in solution,respectively; j(kband) and j(kACN) are the spectrometerefficiencies at the wavelengths of the explosive and ACNRaman bands; Csample and CACN are the concentrations (M) ofthe explosive and ACN; eex is the extinction coefficient for theexplosive at the excitation wavelength; es is the extinctioncoefficient for the explosive at each Raman band wavelength;and er is the extinction coefficient at the ACN Raman bandwavelength. The expression in brackets corrects for theresonance Raman self-absorption.26

RESULTS AND DISCUSSION

Excitation Profiles of NH4NO3. Figure 1 shows theDUVRR solution spectra of NH4NO3 in 77% ACN/23% waterby volume with the ACN and water Raman contributionsnumerically removed, excited between 204 and 257 nm.Tuschel et al.8 showed that all of the 229 nm excited resonanceRaman bands of NH4NO3 in solution result from NO3

�

vibrations with little contribution from NH4þ vibrations.

Excitation between 204 and 257 nm gives rise to essentiallyidentical NH4NO3 DUVRR spectra. The small differences inthe ;1390 cm�1 region result from over- or under-subtractionof ACN.

The strongest Raman band at 1044 cm�1 is assigned to theNO3

� totally symmetric stretching vibration.27–29 Its intensitymonotonically increases with excitation frequency as shown inFig. 1 and by the excitation profiles in Fig. 2. Theantisymmetric stretching vibration of NO3

� couples to watermotion to give rise to bands between ;1300 and 1420 cm�1

whose intensities also monotonically increase with excitationfrequency. In contrast, the 1662 cm�1 band overtone of the out-of-plane deformation (2 3 m2, Fig. 1) and the 2085 cm�1

overtone of the totally symmetric stretching (2 3 m1, Table I)bands have their largest cross-sections at 229 nm.

The wavelength dependence of the solution-phase DUVRRdifferential cross-sections of the 1044, 1338, and 1403 cm�1

bands are shown in Fig. 2 together with the absorptionspectrum of NH4NO3 in the ACN/water solvent. The NH4NO3

ACN/aqueous solution shows an absorption maximum for thep!p* transition of NO3

� at ; 200 nm with an extinctioncoefficient of 0.96 3 104 L�mol�1�cm�1.30 Table I tabulates theNO3

� band assignments as well as their measured Ramancross-sections.

The Raman cross-sections increase with excitation frequency(Fig. 2). The 1044 cm�1 symmetric stretching band (204 nm)Raman cross-section is ;40 times larger than at 257 nm. The1338 cm�1 and 1403 cm�1 band Raman cross-sections at 204nm are ;70 times larger than at 257 nm, while the 1665 and2085 cm�1 overtones show their highest cross-sections at 229nm.

Figure 2 (inset) overlays the sodium nitrate (NaNO3)

FIG. 1. DUVRR spectra excited between 204 and 257 nm of NH4NO3 inACN/water (77%/23% by volume) with the ACN and water contributionsnumerically removed. The spectra were approximately scaled to the relativecross-sections of the 1044 cm�1 band. The small shift observed for the 1662cm�1 band at 204 nm is due to a small calibration error.

1014 Volume 66, Number 9, 2012

preresonance Raman cross-section data of Dudik et al.31 forexcitation wavelengths between 640 and 280 nm with theresults measured here. NO3

� has a ;30 000-fold larger Ramancross-section at 204 nm than at 640 nm. Further, Dudik’sAlbrecht’s A-term fitting demonstrates that the 1044 cm�1

symmetric stretch band is completely preresonance enhancedby the ;200 nm NO3

� p!p* transition. The A-termextrapolation well into resonance underestimates the DUVRRcross-sections. These results show that detection of NO3

� willbe maximized at ;200 nm where excitation is deepest inresonance.

Our excitation profiles of NH4NO3 differ significantly fromthose of potassium nitrate (KNO3) in water solution studied byMyers et al.28 In contrast to their monotonically increasingRaman cross-sections for excitation between 246 nm and 204nm, we observe that the 1662 cm�1 (2 3 m2) overtone and the2085 cm�1 totally symmetric stretching overtone show maximaat 229 nm excitation. More importantly, our measured 204 nmDUVRR cross-sections are a factor of ;4 smaller than those ofMyers et al.28 Our 229 nm results are similar, but our longerexcitation wavelength cross-sections are smaller. We have noready explanation for the differences.

The peculiar subsidiary maxima of the overtone bands

suggest underlying electronic transitions within the broad;200 nm NO3

� absorption band. We will examine this issue inthe future.

Excitation Profiles of PETN. PETN contains four nitrate(NO3) groups connected to methylene groups. Figure 3compares the DUVRR spectra of PETN in ACN excitedbetween 204 and 257 nm. These spectra are qualitativelyidentical and very similar to the normal Raman spectrum ofPETN in acetone (d6 at 488 nm excitation).32 The strongestband at 1294 cm�1, along with a shoulder at 1279 cm�1, ismainly due to the –NO2 symmetric stretching with a minorcontribution from CH bending and CH2 wagging. –NO2

antisymmetric stretching contributes to the 1658 cm�1

vibration. The medium intensity broad ;865 cm�1 band isassigned to O–N stretching with some contribution from C–Cstretching. The medium weak ;1510 cm�1 band is assigned toCH2 scissoring. Detailed normal mode assignments are foundin Gruzdkov et al. and Perger et al.33,34

Figure 4 shows that the Raman cross-sections of the 865,1294, and 1658 cm�1 bands monotonically increase as theexcitation frequency increases. Figure 4 also shows the PETNabsorption spectrum.

According to Mullen and Orloff,35 the PETN absorptionmaxima at ;194 nm results from a p!p* transition localizedon the NO2 groups with an additional large contribution froman intramolecular charge transfer transition that involveselectron density transfer from the C2H5O groups to the –NO2

groups. There are two additional very weak absorption bands

TABLE I. Absolute differential Raman cross-sections of NH4NO3 at different excitation wavelengths.

Ramanbands (cm�1) Assignment

204 nm absoluteRaman cross-sections(10�26 cm2/molc�sr)



257 nm absoluteRaman cross-sections(10�26 cm2/molc�sr )

1044 NO3– symmetric stretching 13.6 7.3 0.40 0.36

1338 NO3– antisymmetric stretching

coupled to water motion16.1 3.3 0.10 0.23

1403 NO3� antisymmetric stretching

coupled to water motion14.5 2.4 0.10 0.21

1662 Overtone of the out-of-planedeformation, (2 3 m2, A 002)

0.46 1.3 0.04 0.11

2085 2 3 m1s 0.3 3.0 0.03 —

FIG. 3. DUVRR spectra of PETN excited between 204 and 257 nm in ACNwith the ACN contribution numerically removed. The spectra wereapproximately scaled to the relative cross-sections of the 1294 cm�1 band.

FIG. 2. Absolute differential Raman cross-sections of 1044, 1338, and 1403cm�1 bands of NH4NO3 between 204 and 257 nm excitation and the molarabsorptivity absorption spectrum of NH4NO3 in ACN/water (77%/23% byvolume). Inset: overlay of the sodium nitrate (NaNO3) preresonance differentialRaman cross-section data of Dudik et al.31 for the 1044 cm�1 band for excitationwavelengths between 640 and 280 nm with the results measured here.

APPLIED SPECTROSCOPY 1015

of PETN at ;260 and ;290 nm that are assigned to n!p*transitions of the –NO2 groups.36 Table II indicates the bandassignments and the Raman cross-sections at the differentexcitation wavelengths.

Characteristic –NO2 symmetric and antisymmetric stretchingmodes have a ;115-fold increased cross-section at 204 nmcompared to 257 nm excitation. Figure 4 (inset) shows that theRaman cross-section excitation profile of the PETN 1294 cm�1

Raman band is fit well by an Albrecht A-term37 yielding aresonance frequency of 55 400 cm�1 (;180.5 nm), which isclose to the 187 nm value that Mullen and Orloff35 predictedon the basis of molecular orbital (MO) calculations. Thesimilarity of the resonance Raman (RR) spectra to thosemeasured in the visible spectral region strongly suggests thatthe intensities observed with visible and near-UV excitationderive from preresonance enhancement by this ;194 nm deep-UV electronic transition.

The 230 nm value that Nagli et al.17 found from their solidPETN modified A-term fit is spurious due to their use of anexternal standard method to measure Raman intensities. As faras we can determine, our results are qualitatively consistentwith deep-UV Raman measurements of PETN by Grun et al.16

The relative intensities of the PETN Raman bands changesignificantly with excitation wavelength, especially on theshoulder of the UV absorption. This demonstrates thatadditional confirmation can be obtained by multiple-excitationwavelength measurements where the relative band intensities

rapidly change. In addition, it is clear that the highestintensities will derive from excitation deepest in resonancewith PETN at 200 nm or less.

Excitation Profiles of TNT. Figure 5 shows Raman spectraof TNT in ACN solution excited between 204 and 257 nm. Thestrongest TNT Raman band at 1354 cm�1 is assigned to NO2

symmetric stretching coupled to C–N stretching. The mediumstrong band at 1616 cm�1 is due to NO2 antisymmetricstretching coupled to aromatic ring stretching. The 826 cm�1

medium weak band corresponds to NO2 scissoring.38 Table IIIlists the band assignments and the Raman cross-sectionsmeasured at the different excitation wavelengths. Our solutionTNT DUVRR spectra are qualitatively similar to those of Grunet al.16 as well as the solid-state TNT DUVRR spectra of Nagliet al.17 The similarities of the DUVRR spectra to thepreresonance spectra indicate that the ;210 to 270 nmelectronic transitions dominate the preresonance Ramanintensities.

Figure 6 shows the Raman cross-section excitation profilesof the 826, 1209, 1354, and 1616 cm�1 TNT bands togetherwith the TNT absorption spectrum. The absorption spectrum ofTNT shows a maximum at ;229 nm with a shoulder at ;260nm. Weak, broad features occur between 280 and 340 nm. Thecharge-transfer electronic transitions that give rise to the strong;230 nm TNT absorption are thought to change the electrondensity in the aromatic ring and the –NO2 groups. Thus, weexpect and observe that the resonance Raman spectra showenhancement of both aromatic ring and –NO2 vibrations.39

The Raman cross-sections roughly follow the absorptionband shape with a global maximum at 229 nm, although a

FIG. 4. Absolute differential Raman cross-sections of 865, 1294, 1510, and1656 cm�1 bands of PETN between 204 and 257 nm excitation and the molarabsorptivity absorption spectrum of PETN in ACN. Inset: Raman cross-sectionexcitation profile Albrecht A-term fit for the PETN 1294 cm�1 band.

TABLE II. Absolute differential Raman cross-sections of PETN at different excitation wavelengths.






865 O–N stretching with somecontribution from C–C stretching

9.3 1.1 0.29 0.21

1294 NO2 symmetric stretching 29.8 1.8 0.30 0.261510 CH2 scissoring 3.7 0.90 0.05 0.041658 NO2 antisymmetric stretching

vibration13.3 1.4 0.12 0.12

FIG. 5. TNT DUVRR spectra excited between 204 and 257 nm in ACN withthe ACN contribution numerically removed. The spectra were approximatelyscaled to the relative cross-sections of the 1354 cm�1 band.


pronounced dip is observed with 244 nm excitation for the1209, 1354, and 1616 cm�1 bands. This dip does not occur forthe 826 cm�1 NO2 scissoring band. The very low cross-sections observed at 204 nm excitation indicate destructiveinterference between the enhancements from the electronictransitions between 210 and 270 nm and deeper UV transitions, 200 nm.40,41

Our results differ from Grun et al,16 who found maximumintensities at ;260 nm for most peaks. They, however, did findthat the 1552 cm�1 band shows a maximum at around 230 nm.

The fact that the excitation profiles differ between thedifferent bands clearly indicates that multiple electronictransitions underlie the broad, antisymmetric 230 nm absorp-tion band, consistent with the results of Quenneville et al.39 Wedo not have a sufficient density of excitation wavelengths toresolve the different transitions. However, the overall shape ofthe 826 cm�1 NO2 scissoring band excitation profile indicatesthat NO2 electronic transitions are involved in both the 229 nmabsorption component as well as in the 260 nm shoulder.

Clearly the optimal excitation wavelength to detect TNT is;230 nm. Deep-UV excitation around 200 nm will give rise toexceptionally weak enhancement due to the destructiveinterference from the higher energy transitions.

Excitation Profiles of HMX. Figure 7 shows that theDUVRR spectra of HMX in ACN solution, excited between204 and 257 nm, contains at least 16 resolved bands. Werecently reviewed8 the HMX Raman band assignmentsdescribed by Brand et al.42 and Zhu et al.43 The HMX Ramanspectra in Fig. 7 are dominated by the 937 cm�1 band (except at204 nm excitation) that involves antisymmetric –C–N–Nstretching with some CH2 rocking motion. The strong 881cm�1 band is assigned to a vibration that includes N–N–C2

symmetric stretching with some ONO bending.As the excitation wavelength decreases all the Raman bands

increase their intensity, showing maxima at 229 nm. Themaximum is especially sharp for the 937 cm�1 antisymmetricC–N–N stretching band, as it should be if its enhancement weredominated by this absorption band; the excitation profile willconsequently show a band shape approximating the square ofthe absorption band.5,9,10 Table IV shows the band assignmentsand the absolute differential Raman cross-sections at thedifferent excitation wavelengths. The intensity decreasesdramatically with 204 nm excitation, indicating destructiveinterference with higher energy electronic transitions.40,41

Figure 8 shows the Raman excitation profiles of the 881,937, 1224, 1260, 1330, and 1520 cm�1 bands of HMX alongwith the HMX absorption spectrum. The HMX absorptionshows a resolved electronic absorption band at ;230 nm andan additional absorption band significantly below 200 nm.

TABLE III. Absolute differential Raman cross-sections of TNT at different excitation wavelengths.






826 NO2 scissoring vibration 1.6 4.0 1.7 1.61170 Combination of C–C ring in-plane

trigonal bending with C–N andpossible C–CH3 stretching

0.31 2.3 1.0 1.2

1209 Symmetric aromatic ‘‘ring-breathing’’ mode

1.5 5.8 1.5 1.9

1354 NO2 symmetric stretching coupledto C–N stretching

9.8 65.4 28.5 41.6

1616 NO2 antisymmetric stretchingcoupled to aromatic ringstretching

4.9 33.0 13.3 21.4

FIG. 6. Absolute differential Raman cross-sections of the 826, 1209, 1354,and 1616 cm�1 bands of TNT between 204 and 257 nm excitation and themolar absorptivity absorption spectrum of TNT in ACN.

FIG. 7. DUVRR spectra excited between 204 and 257 nm of HMX in ACNwith the ACN contribution numerically removed. The spectra wereapproximately scaled to the relative cross-sections of the 937 cm�1 band.


According to Stals et al.,44,45 the transition at 230 nm originates

from the p!p* transition of the N–NO2 group, and the broad

band at ;200 nm derives from an electronic transition between

orbitals involving intimately mixed r, p, r*, p*, and n orbitals

of similar energies.

Our DUVRR spectra and the excitation profiles of HMX in

ACN appear to differ significantly from the only previous

HMX solution DUVRR spectra report (also in ACN) by Grun

et al.16 We measured the HMX spectra at 204 nm with different

excitation powers to look for the occurrence of nonlinear

spectral intensities that might result from Raman spectra of

excited-state species or any Raman saturation phenomena that

could decrease the Raman intensities due to ground-state

concentration depletion.25,46–48,49 All intensities were found to

change linearly in proportion to the excitation power.

We note that our HMX absorption peaks at ;230 nm (Fig.

8) instead of at ;220 nm as in Grun et al.’s sample. This may

indicate that the chemical composition differs between oursamples.

Not surprisingly our HMX solution DUVRR also differsignificantly from previously reported HMX solid-state visibleand NIR Raman spectra. These differences are likely to resultfrom the conformational changes between the solid andsolution states and the fact that our spectra are in resonance,in contrast to the solid-state spectra excited in the NIR andvisible spectral regions. Significant differences are observedbetween different solid-state structures of HMX.50–53

Our results indicate that the best excitation wavelength forHMX detection is ;230 nm. Because the excitation profiles ofthe different bands have different bandshapes, the dependenceof the relative intensities on wavelength can be used to confirmdeterminations of HMX in samples.

Excitation Profiles of RDX. Figure 9 shows the Ramanspectra of RDX in ACN solution excited between 204 and 257nm. These spectra and their variations with excitationwavelength parallel Grun et al.’s16 UVRR spectra. Thewavelength dependences in Fig. 9 are qualitatively similar toGrun et al., although there are some significant differences.

TABLE IV. Absolute differential Raman cross-sections of HMX at different excitation wavelengths.






757 NO2 wagging vibration 0.54 1.4 0.33 0.17843 N–C2 symmetric stretching 1.1 5.3 3.6 1.2881 N–N–C2 symmetric stretching with

O–N–O-b3.0 12.1 5.2 2.2

937 C–N–N antisymmetric stretchingwith some CH2 rocking

4.0 33.0 12.7 6.6

1035 N–N–C2 symmetric stretching 1.0 13.0 1.4 1.91080 N–N–C2 symmetric stretching with

O–N–O-b1.0 5.0 0.86 0.56

1186 NC2 antisymmetric stretching – 8.40 1.4 1.21224 NC2 antisymmetric stretching 4.1 10.9 2.8 1.11260 NO2 symmetric stretching 11.4 15.3 1.7 1.41330 NO2 symmetric stretching with

some CH2 twisting about theN–N bond

12.5 17.1 1.9 1.5

1520 NO2 antisymmetric stretching 5.7 6.5 2.4 2.21577 NO2 antisymmetric stretching 6.3 12.6 2.1 0.85

FIG. 8. Absolute differential Raman cross-sections of the 881, 937, 1224,1260, 1330, and 1520 cm�1 bands of HMX between 204 and 257 nmexcitation, and the molar absorptivity absorption spectrum of HMX in ACN.

FIG. 9. DUVRR spectra of RDX in ACN with the ACN contributionnumerically removed, excited between 204 and 257 nm. The spectra wereapproximately scaled to the relative cross-sections of the 1270 cm�1 band.


Our RDX solution DUVRR spectra are not very close topreviously reported RDX solid-state visible Ramanspectra.54–56 This is because of the solid- and solution-stateconformational changes and also because our spectra are inresonance.

We examined the dependence of our 204 nm excited RDX inACN solution spectra on the incident power and found a lineardependence on intensity in our pulse energy regime between 2and 15 lJ into a 200 lm diameter spot. We did not observe anyexcited-state or sample-degradation features. In contrast, the266 nm excited, broad RDX features of Nagli et al.17 resemblethe photochemically degraded spectra we previously observedwith 229 nm excitation.15

The DUVRR spectra of RDX significantly differ from that ofHMX, though the chromophores are identical. This suggeststhat the enhanced normal modes are not localized within eachNNO2 group; there is possibly different coupling between thesefunctional groups in RDX from that in HMX. In addition, the

electronic transitions of these groups can couple, giving rise tomore delocalized electronic transitions.

However, the absorption bands of HMX and RDX appear tobe similar. RDX has at least one strong electronic transition at;230 nm, whose absorptivity is 57% that of HMX, withanother absorption band below 200 nm and weak absorptionfeatures between 280 and 340 nm (Fig. 10). The assignments ofthe RDX and HMX electronic transitions are similar, but theRaman excitation profiles show that the coupling between thetransitions differs.

The assignments of the DUVRR bands between 700 and1600 cm�1 derive from Dreger and Gupta.57 The 229 nm RDXspectrum in Fig. 9 consists of at least 14 resolved bands: 750cm�1 (ring bending with NO2 scissoring), 795 cm�1 (C–Nstretch and NO2 scissoring), 850 cm�1 (N–N stretching andNO2 axial scissoring), 884 cm�1 (mainly C–N stretching), 924cm�1 (CH2 rocking or combination), 950 cm�1 (N–Nstretching), 1024 cm�1 (N–C stretching with some CH2

rocking), 1216 cm�1 (N–C stretching), 1270 cm�1 (N–Nstretching and O–N–O stretching), 1314 cm�1 (N–N stretchingand CH2 twisting), 1341 cm�1 (CH2 wagging or combination),1378 cm�1 (CH2 twisting), 1458 cm�1 (CH2 scissoring), 1524cm�1 (CH2 scissoring or combination), 1559 cm�1 (O–N–Oequatorial stretching), and 1586 cm�1 (O–N–O axial stretch-ing). The 204 nm RDX spectrum is dominated by the 1270 and1314 cm�1 bands that derive from motion of the NNO2

chromophores.Figure 10 shows the DUVRR differential Raman cross-

section excitation profiles of the 1216, 1270, 1314, and 1586cm�1 RDX Raman bands between 204 and 257 nm along withthe absorption spectrum of RDX in ACN. Table V shows theband assignments and the Raman cross-section values for thedifferent excitation wavelengths. The maximum differentialRaman cross-section of RDX is ;12 3 10�26 cm2/molecule-steradian (molc�sr) at 204 nm for the 1270 and 1314 cm�1

bands, while the maximum cross-section of HMX is 33 310�26 cm2/molc�sr at 229 nm for the 937 cm�1 band.

In contrast to HMX, none of the bands show an excitationprofile maximum within the 230 nm absorption feature. Theband intensities continue to increase as the excitationwavelength decreases to 204 nm, whereupon the 1268 and

FIG. 10. Absolute differential Raman cross-sections of the 1216, 1270, 1314,and 1586 cm�1 bands of RDX between 204 and 257 nm excitation and themolar absorptivity absorption spectrum of RDX in ACN.

TABLE V. Absolute differential Raman cross-sections of RDX at different excitation wavelengths.






750 Ring bending with NO2 scissoring 1.3 1.4 0.18 0.13795 C–N stretching and NO2 scissoring 1.5 0.40 0.51 0.90850 N–N stretching and NO2 axial

scissoring1.0 2.2 2.0

884 C–N stretching 1.5 2.5 2.01024 N–C stretching with some CH2

rocking0.90 6.4 3.5 3.6

1216 N–C stretching 3.9 6.0 3.2 2.21270 N–N stretching and O–N–O

stretching with possible CH2

twisting

13.0 6.8 2.7 2.1

1314 N–N stretching and CH2 twisting,NO2 symmetric vibration

13.9 3.0 0.07 1.3

1352 CH2 wagging or combination 2.9 — 0.01 —1524 CH2 scissoring or combination 3.9 2.0 — 1.31586 O–N–O axial stretching, NO2

antisymmetric vibration4.2 5.5 1.7 0.71


1315 cm�1 bands that derive from motion in the NNO2

chromophores completely dominate the DUVRR. This maysignal a dominating, strong electronic transition in RDX whoseanalogous feature in HMX is at a much higher energy.

Thus, the best excitation wavelength for RDX is deep in theUV, probably lower than 200 nm. RDX can clearly bedifferentiated from HMX. The different RDX Raman bandsshow sufficiently distinct excitation profiles so that multipleexcitation wavelength spectra can be used to obtain animproved determination of these explosive molecules.

CONCLUSIONS

The DUVRR differential Raman cross-sections of NH4NO3,PETN, TNT, HMX, and RDX dramatically increase as theexcitation wavelength decreases deep into the UV to 204 nm.NH4NO3, PETN, and RDX are best resonance-enhanced by the204 nm excitation used here, and the optimum excitationwavelength for TNT and HMX is around ;230 nm. Theexcitation profiles of TNT roughly follow its absorption bandshape.

The excitation profiles for the different Raman bands differfor each explosive molecule, indicating that multiple-excitationwavelength spectra are not redundant but offer additionalinformation on the species present in samples.

We see no evidence of sample degradation at the fluencesand spectral accumulation times used here. However, wepreviously observed such phenomena at longer spectralaccumulation times and for higher fluence measurements.These results are promising for development of standoff deep-UV Raman methods for explosive molecule determinations.

We also determined the detection limit for these explosivesdissolved in ACN/water solutions for 229 nm excitation usingour standard Raman spectrometer.8 Most detection limits were, 1 part per million (ppm).

ACKNOWLEDGMENT

This work was funded by the United States Department of HomelandSecurity’s Science and Technology Directorate and with partial funding fromthe University of Pittsburgh.

1. D.S. Moore. ‘‘Instrumentation for Trace Detection of High Explosives’’.Rev. Sci. Instrum. 2004. 75(8): 2499-2512.

2. D.S. Moore. ‘‘Recent Advances in Trace Explosives Detection Instrumen-tation’’. Sens Imaging. 2007. 8: 9-38.

3. S. Wallin, A. Pettersson, H. Ostmark, A. Hobro. ‘‘Laser-Based StandoffDetection of Explosives: A Critical Review’’. Anal. Bioanal. Chem. 2009.395(2): 259-274.

4. D.S. Moore, R.J. Scharff. ‘‘Portable Raman Explosives Detection’’. Anal.Bioanal. Chem. 2009. 393: 1571-1578.

5. S.A. Asher. ‘‘UV Resonance Raman Spectroscopy for Analytical, Physical,and Biophysical Chemistry–Part 1’’. Anal. Chem. 1993. 65(2): 59A-66A.

6. S.A. Asher. ‘‘UV Resonance Raman Spectroscopy for Analytical Physical,and Biophysical Chemistry–Part 2’’. Anal. Chem. 1993. 65(2): 201A-210A.

7. S.A. Asher, C.R. Johnson. ‘‘Raman Spectroscopy of a Coal Liquid Showsthat Fluorescence Interference is Minimized with Ultraviolet Excitation’’.Science. 1984. 225: 311-313.

8. D.D. Tuschel, A.V. Mikhonin, B.E. Lemoff, S.A. Asher. ‘‘Deep UltravioletResonance Raman Excitation Enables Explosives Detection’’. Appl.Spectrosc. 2010. 64(4): 425-432.

9. R.J.H. Clark. ‘‘Resonance Raman Spectra of Inorganic Molecules andIons’’. In: R.J.H. Clerk, R.E. Hester, editors. Advances in Infrared andRaman Spectroscopy. Heyden, London: John Wiley and Sons, 1975. Vol.1, Chap. 4. 143-172.

10. S.A. Asher. ‘‘UV Resonance Raman Studies of Molecular Structure andDynamics: Applications in Physical and Biophysical Chemistry’’. Ann.Rev. Phys. Chem. 1988. 39: 537-588.

11. S.A. Asher. ‘‘Ultraviolet Raman Spectrometry’’. In: J.M. Chalmers, P.R.Griffiths, editors. Handbook of Vibrational Spectroscopy. New York: JohnWiley and Sons, 2002. Vol. 1. 557-571.

12. S.A. Oladepo, K. Xiong, Z. Hong, S.A. Asher. ‘‘Elucidating Peptide andProtein Structure and Dynamics: UV Resonance Raman Spectroscopy’’. J.Phys. Chem. Lett. 2011. 2(4): 334-344.

13. S.A. Asher, D.D. Tuschel, T.A. Vargson, L. Wang, S.J. Geib. ‘‘Solid Stateand Solution Nitrate Photochemistry: Photochemical Evolution of the SolidState Lattice’’. J. Phys. Chem. A. 2011. 115(17): 4279-4287.

14. L. Wang, D.D. Tuschel, S.A. Asher. ‘‘Templated Photonic CrystalFabrication of Stoichiometrically Complex Nanoparticles for ResonanceRaman Solid Cross Section Determinations’’. J. Phys. Chem. C. 2011.115(32): 15767-15771.

15. L. Wang, D.D. Tuschel, S.A. Asher. ‘‘229 nm UV PhotochemicalDegradation of Energetic Molecules’’‘‘. Proc. SPIE. 2011. Vol. 8018.80181B-1–80181B-6.

16. G. Comanescu, C.K. Manka, J. Grun, S. Nikitin, D. Zabetakis.‘‘Identification of Explosives with Two-Dimensional Ultraviolet Reso-nance Raman Spectroscopy’’. Appl. Spectrosc. 2008. 62(8): 833-839.

17. L. Nagli, M. Graft, Y. Fleger, M. Rosenbluh. ‘‘Absolute Raman Cross-Sections of Some Explosives: Trend to UV’’. Opt. Mater. 2008. 30(11):1747-1754.

18. M. Gaft, L. Nagli. ‘‘UV Gated Raman Spectroscopy for Standoff Detectionof Explosives’’. Opt. Mater. 2008. 30(11): 1739-1746.

19. Y. Fleger, L. Nagli, M. Gaft, M. Rosenbluh. ‘‘Narrow Gated Raman andLuminescence of Explosives’’. J. Lumin. 2009. 129(9): 979-983.

20. L. Nagli, M. Gaft. ‘‘Raman Scattering Spectroscopy for ExplosivesIdentification’’. Proc. SPIE-Int. Soc. Opt. Eng. 2007. 6552: 65520Z-1 –65520Z-10.

21. S. Bykov, I. Lednev, A. Ianoul, A. Mikhonin, C. Munro, S.A. Asher.‘‘Steady-State and Transient Ultraviolet Resonance Raman Spectrometerfor the 193–270 nm Spectral Region’’. Appl. Spectrosc. 2005. 59(12):1541-1552.

22. I.K. Lednev, A.S. Karnoup, M.C. Sparrow, S.A. Asher. ‘‘a-Helix PeptideFolding and Unfolding Activation Barriers: A Nanosecond UV ResonanceRaman Study’’. J. Am. Chem. Soc. 1999. 121(35): 8074-8086.

23. S.A. Asher, R.W. Bormett, X.G. Chen, D.H. Lemmon, N. Cho, P.Peterson, M. Arrigoni, L. Spinelli, J. Cannon. ‘‘UV Resonance RamanSpectroscopy Using a New CW Laser Source: Convenience andExperimental Simplicity’’. Appl. Spectrosc. 1993. 47(5): 628-633.

24. C.R. Johnson, M. Ludwig, S.A. Asher. ‘‘Ultraviolet Resonance RamanCharacterization of Photochemical Transients of Phenol, Tyrosine, andTryptophan’’. J. Am Chem. Soc. 1986. 108(5): 905-912.

25. J. Teraoka, P.A. Harmon, S.A. Asher. ‘‘UV Resonance Raman SaturationSpectroscopy of Tryptophan Derivatives: Photophysical RelaxationMeasurements with Vibrational Band Resolution’’. J. Am. Chem. Soc.1990. 112(8): 2892-2900.

26. D.F. Shriver, J.B.R. Dunn. ‘‘The Backscattering Geometry for RamanSpectroscopy of Colored Materials’’. Appl. Spectrosc. 1974. 28(4): 319-323.

27. A. Ianoul, T. Coleman, S.A. Asher. ‘‘UV Resonance Raman SpectroscopicDetection of Nitrate and Nitrite in Wastewater Treatment Processes’’. Anal.Chem. 2002. 74(6): 1458-1461.

28. M.R. Waterland, A. Myers-Kelley. ‘‘Far-Ultraviolet Resonance RamanSpectroscopy of Nitrate Ion in Solution’’. J. Chem. Phys. 2000. 113(16):6760-6773.

29. A. Tongraar, P. Tangkawanwanit, B.M. Rode. ‘‘A Combined QM/MMMolecular Dynamics Simulations Study of Nitrate Anion (NO3

–) inAqueous Solution’’. J. Phys. Chem. A. 2006. 110(47): 12918-12926.

30. V. Tomisic, V. Butorac, J. Viher, V. Simeon. ‘‘Comparison of theTemperature Effect on the p* n and p* p Electronic Transition Bands ofNO3(aq)’’. J. Sol. Chem. 2005. 34(6): 613-616.

31. J.M. Dudik, C.R. Johnson, S.A. Asher. ‘‘Wavelength Dependence of thePreresonance Raman Cross Sections of CH3CN, SO4

2�, ClO4�, and

NO3�’’. J. Chem. Phys. 1985. 82(4): 1732-1740.

32. Y.A. Gruzdkov, Z.A. Dreger, Y.M. Gupta. ‘‘Experimental and TheoreticalStudy of Pentaerythritol Tetranitrate Conformers’’. J. Phys. Chem. A.2004. 108(29): 6216-6221.

33. Y.A. Gruzdkov, Y.M. Gupta. ‘‘Vibrational Properties and Structure ofPentaerythritol Tetranitrate’’. J. Phys. Chem. A. 2001. 105(25): 6197-6202.

34. W.F. Perger, J. Zhao, J.M. Winey, Y.M. Gupta. ‘‘First-Principles Study ofPentaerythritol Tetranitrate Single Crystals Under High Pressure: Vibra-tional Properties’’. Chem. Phys. Lett. 2006. 428(4-6): 394-399.

35. P.A. Mullen, M.K. Orloff. ‘‘Ultraviolet Absorption Spectrum of Penta-erythritol Tetranitrate’’. J. Phys. Chem. 1973. 77(7): 910-911.

36. H.H. Jaffe, M. Orchin. ‘‘Simple Chromophores and Solvent Effects’’. In:John Wiley and Sons, Inc., editor. Theory and Applications of Ultraviolet


Spectroscopy. New York, NY: Wiley and Sons, 1962. Chap. 9, pp. 182-184.

37. A.C. Albrecht, M.C. Hutley. ‘‘On the Dependence of Vibrational RamanIntensity on the Wavelength of Incident Light’’. J. Chem. Phys. 1971.55(9): 4438-4443.

38. J. Clarkson, W.E. Smith, D.N. Batchelder, D.A. Smith, A.M. Coats. ‘‘ATheoretical Study of the Structure and Vibrations of 2,4,6-Trinitrotolune’’.J. Mol. Struct. 2003. 648(3): 203-214.

39. J. Quenneville, M. Greenfield, D.S. Moore, S.D. McGrane, R.J. Scharff.‘‘Quantum Chemistry Studies of Electronically Excited Nitrobenzene,TNA, and TNT’’. J. Phys. Chem. A. 2011. 115(44): 12286-12297.

40. J. Friedman, R.M. Hochstrasser. ‘‘Interference Effects in ResonanceRaman Spectroscopy’’. Chem. Phys. Lett. 1975. 32(3): 414-419.

41. P. Stein, V. Miskowski, W.H. Woodruff, J.P. Griffin, K.G. Werner, B.P.Gaber, T.G. Spiro. ‘‘Raman Antiresonance: De-enhancement of RamanIntensity by Forbidden Electronic Transitions’’. J. Chem. Phys. 1976.64(5): 2159-2167.

42. H.V. Brand, R.L. Rabie, D.J. Funk, I. Diaz-Acosta, P. Pulay, T.K. Lippert.‘‘Theoretical and Experimental Study of the Vibrational Spectra of the a, b,and d Phases of Octahydro-1, 3, 5, 7-tetranitro-1, 3, 5, 7-tetrazocine(HMX)’’. J. Phys. Chem. B. 2002. 106(41): 10594-10604.

43. W. Zhu, J. Xiao, G. Li, F. Zhao, H. Ziao. ‘‘First-Principles Study of theFour Polymorphs of Crystalline Octahydro-1,3,5,7-tetranitro-1,3,5,7-tetra-zocine’’. J. Phys. Chem. B. 2007. 111(44): 12715-12722.

44. J. Stals. ‘‘The Chemistry of Aliphatic Unconjugated Nitramines III.*Thermal, Electrophilic, and Nucleophilic Reactions’’. Aust. J. Chem. 1969.22(12): 2515-2526.

45. J. Stals. ‘‘Chemistry of Aliphatic Unconjugated Nitramines Part 4.-Photophysical Processes of Secondary Nitramines’’. Trans. Faraday Soc.1971. 67: 1739-1748.

46. C.M. Jones, V.L. Devito, P.A. Harmon, S.A. Asher. ‘‘High-Repetition-Rate Excimer-Based UV Laser Excitation Source Avoids Saturation inResonance Raman Measurements of Tyrosinate and Pyrene’’. Appl.Spectrosc. 1987. 41(8): 1268-1275.

47. N. Cho, S. Song, S.A. Asher. ‘‘UV Resonance Raman and Excited-StateRelaxation Rate Studies of Hemoglobin’’. Biochemistry. 1994. 33(19):5932-5941.

48. P.A. Harmon, J. Teraoka, S.A. Asher. ‘‘UV Resonance Raman SaturationSpectroscopy Measures Protein Aromatic Amino Acid Excited StateRelaxation Rates’’. J. Am. Chem. Soc. 1990. 112(24): 8789-8799.

49. M. Ludwig, S.A. Asher. ‘‘Ultraviolet Resonance Raman Excitation Profilesof Tyrosine: Dependence of Raman Cross Sections on Excited-StateIntermediates’’. J. Am. Chem. Soc. 1988. 110(4): 1005-1011.

50. D.S. Moore, K.-Y. Lee, S.I. Hagelberg. ‘‘Submicron-Sized Gamma-HMX:II. Effect of Pressing on Phase Transition’’. J. Energ. Mater. 2008. 26(1):70-78.

51. F. Goetz, T.B. Brill. ‘‘Laser Raman Spectra of a-, b-, c-, and d-Octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine and Their Temperature Dependence’’.J. Phys. Chem. 1979. 83(3): 340-345.

52. F. Goetz, T.B. Brill, J.R. Ferraro. ‘‘Pressure Dependence of the Raman andInfrared Spectra of a-, b-, c-, and d-Octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine’’. J. Phys. Chem. 1978. 82(17): 1912-1917.

53. Z. Iqbal, S. Bulusu, J.R. Autera. ‘‘Vibrational Spectra of p-Cyclotetra-methylene Tetranitramine and Some of Its Isotopic Isomers’’. J. Chem.Phys. 1974. 60(1): 221-230.

54. P. Torres, L. Mercado, I. Cotte, S.P. Hernandez, N. Mina, A. Santana, R.T.Chamberlain, R. Lareau, M.E. Castro. ‘‘Vibrational Spectroscopy Study ofb and a RDX Deposits’’. J. Phys Chem. B. 2004. 108(26): 8799-8805.

55. J.C. Carter, S.M. Angel, M. Lawrence-Snyder, J. Scaffidi, R.E. Whipple,J.G. Reynolds. ‘‘Standoff Detection of High Explosive Materials at 50Meters in Ambient Light Conditions Using a Small Raman Instrument’’.Appl. Spectrosc. 2005. 59(6): 769-775.

56. N. Gupta, R. Dahmani. ‘‘AOTF Raman Spectrometer for Remote Detectionof Explosives’’. Spectrochim. Acta, Part A. 2000. 56(8): 1453-1456.

57. Z.A. Dreger, Y.M. Gupta. ‘‘High Pressure Raman Spectroscopy of SingleCrystals of Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX)’’. J. Phys. Chem.B. 2007. 111(15): 3893-3903.


Documents

Deep-Ultraviolet Resonance Raman Excitation Proﬁles of ...asher/homepage/spec_pdf/Deep-Ultraviolet Resonance Raman... · Deep-Ultraviolet Resonance Raman Excitation Proﬁles of