J. Phys. Chem. A 2011, 115, 439–442
439
Orientation Dependent Far-Infrared Terahertz Absorptions in Single Crystal Pentaerythritol Tetranitrate (PETN) Using Terahertz Time-Domain Spectroscopy Von H. Whitley,* Daniel E. Hooks, Kyle J. Ramos, and Timothy H. Pierce WX-9, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States
John F. O’Hara, Abul K. Azad, and Antoinette J. Taylor MST-CINT, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States
Jeffrey Barber Battelle, Egg Harbor Township, New Jersey 08234, United States
Richard D. Averitt Department of Physics, Boston UniVersity, Boston, Massachusetts 02215, United States ReceiVed: September 2, 2010; ReVised Manuscript ReceiVed: NoVember 18, 2010
Terahertz time-domain spectroscopy (THZ-TDS) has been used to measure the absorption spectra in the range 7-100 cm-1 (0.2-3 THz) of single crystal pentaerythritol tetranitrate (PETN). Absorption was measured in transmission mode as a function of incident polarization with the incident and transmitted wave vectors oriented along the crystallographic directions [100], 〈10(a/c)2〉, and 〈110〉. Samples were rotated with respect to the incident polarization while absorption was measured at both 300 and 20 K. Comparatively minor differences were observed among the three orientations. Two broad absorptions at 72 and >90 cm-1, and several weaker absorptions at 36, 55, 80, and 82 cm-1, have been observed at cryogenic temperatures. Introduction Pentaerythritol tetranitrate (PETN) is a stable crystalline energetic material with a failure diameter below 1 mm,1 relatively high detonation velocity of 8.16 mm/µs,2 and a higher initiation sensitivity compared to hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) and octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX). Formulations of PETN are frequently used in detonators, thin sheet explosives, detonation cords, and so forth.3 We are interested in characterizing the absorptions in energetic materials below 100 cm-1 to gain insight into shock interactions with explosive molecules. For chemical reactions to occur under shock compression, energy must be transferred from the shock to the individual bonds within the explosive molecules. Although there is debate about the nature of the energy transfer mechanism, the vibrational properties of both the crystal lattice and the individual molecules are recognized to play an important role in this process. These low-frequency phonon modes are thought to be important in transferring energy from the shock front to individual chemical bonds.4-6 Characterizing optical modes in the region 0-3 THz as a function of orientation is crucial to enable the basic physical interpretation of how shock leads to detonation. Despite the potential importance of these modes in the initial stages of detonation, there is little published data characterizing the optical modes in 0-3 THz region. There are even fewer studies on the orientation dependence of these modes. Many of the previous experiments on PETN absorptions in the 7-100 cm-1 region were directed toward detection and identification techniques. PETN is generally found as a white * To whom correspondence should be addressed. E-mail: VWhitley@ lanl.gov.
powdery substance similar in color and texture to sugar or table salt. In the 7-100 cm-1 region of the spectrum, however, sugar and PETN have different and unique absorption signatures, allowing for identification of unknown powders. In addition, electromagnetic waves in the 7-100 cm-1 region can pass through visibly opaque objects like plastic, wood, paper, and cloth. The ability to penetrate objects and positively identify an unknown powder has led to numerous studies on the viability of THz spectroscopy as a means for screening for explosives. A recent review of this work is thoroughly covered by LeahyHoppa et al.7 Single-crystal absorption spectra of HMX, PETN, and RDX have been shown to be distinctly different than powdered absorptions. Spectra of single crystals of RDX, PETN, and HMX along fixed crystallographic orientations bear so little resemblance to their powder counterparts as to be unrecognizable using a powder spectrum, especially in the case of RDX.8 Follow-up work on RDX demonstrated that many of the absorptions were highly dependent upon the orientations of the crystallographic axes with respect to the incident polarization angle.9 An average of the measured absorptions across the various crystal orientations of RDX produced a spectrum similar to the RDX powder spectrum. Powder measurements10-13 of PETN below 100 cm-1 show two broad distinct peaks at 67 and 96 cm-1. The lower frequency peak often shows a shoulder at 72 cm-1. Raman experiments on PETN powder14 and molecular dynamics calculations15-19 suggest a number of additional absorptions should exist in the 30-100 cm-1 region. The orientation dependence of the PETN spectrum, similar to the RDX spectrum discussed previously, has not been published. Here, we present absorption measure-
10.1021/jp108388c 2011 American Chemical Society Published on Web 01/06/2011
440
J. Phys. Chem. A, Vol. 115, No. 4, 2011
Whitley et al.
Figure 1. Absorption spectrum of (a) [001], (b) 〈10(a/c)2〉, and (c) 〈110〉 oriented PETN at 20 K as a function incident polarization angle. The in-plane crystallographic axis is shown on the [001] oriented sample. Purple represents transparent regions, red represents strong absorption regions. The dark circle in the center covers the 0-7 cm-1 spectral range. The dotted circles move outward in 20 cm-1 increments to a maximum of 100 cm-1.
ments on oriented single crystal PETN measured in transmission along the [001], 〈10(a/c)2〉, and 〈110〉 crystallographic directions and the dependence of these absorptions on the incident polarization angle. Experimental Section For these experiments, PETN samples were prepared with parallel faces normal to [001], 〈10(a/c)2〉, and 〈110〉. The samples were cut from large single crystals, nominally 4-6 cm on a side. PETN powder used for crystal growth was obtained from the Dupont Company, dissolved in ethyl acetate, filtered, recrystallized, and dried. The crystals were grown by controlled evaporation of saturated solutions of PETN dissolved in ethyl acetate. Contact goniometry was used to locate planes in PETN by redundant reference to crystal facets. The samples were cut using a low-speed diamond impregnated wire saw using a solution of Alconox in deionized water as a lubricant. Sample spectra were measured in the as-cut configuration; no further polishing was performed. The final prepared sample thickness for the Terahertz time-domain spectroscopy (THZ-TDS) measurements were 1 mm thick and approximately 1 cm × 1 cm wide. The samples’ normal and in-plane crystallographic directions were determined by transmission Laue X-ray diffraction. The uncertainty in crystallographic directions is within 2° of those reported. Crystallographic directions are given with respect to the P4j21c space group of PETN.20 Additional details on cryostat and THz-TDS system have been described previously.9 Spectra were recorded at two different temperatures, 20 and 300 K. To cool the samples to 20 K without cracking them, we used a cooling rate of 0.025 K/s. To produce the crystal orientation dependence on the input polarization direction, the sample was rotated about its normal, which was either the [100], 〈10(a/c)2〉, or 〈110〉 crystallographic direction, in the cryostat instead of rotating the beam polarization. The sample was rotated 9° after every time-domain spectrum at 20 K, and 22.5° at 300 K. Reference data with the sample removed from the beam path were taken at the beginning and end of each sample data set. The ratio of the Fouriertransformed THz-TDS sample data to the reference data set yielded the complex transmission function. The combination of decreasing emission amplitude above 100 cm-1 and strong sample absorption at frequencies greater than 95 cm-1 limits the reported spectra to 105 cm-1, which is our detection limit. The 〈110〉 orientation shows a shoulder on the low energy side at ∼65 cm-1, suggesting multiple components to this absorption. The strong absorption at >90 cm-1 is present
Absorptions in Single Crystal PETN Using THZ-TDS
J. Phys. Chem. A, Vol. 115, No. 4, 2011 441
Figure 4. Comparison of single crystal spectra at 20 K (solid) and 300 K (dotted) averaged over all orientations (solid) with the powder spectrum from Fan et al. taken at 4 K (dashed).
Figure 3. Comparison of spectra of (a) [001], (b) 〈10(a/c)2〉, and (c) 〈110〉 PETN at both 20 K (solid line) and 300 K (dotted line). The presented spectra are averages of all in-plane spectra for a given orientation.
in all three samples with both [010] and 〈110〉 orientations exhibiting a shoulder at ∼87 cm-1. Weaker absorptions are found at 80 cm-1 in 〈10(a/c)2〉 and 〈110〉 and 82 cm-1 in [001] and 83 cm-1 in 〈110〉. A few weak absorptions at 36 and 55 cm-1 are primarily associated with the [001] orientation. Figure 3 presents a comparison of the spectra measured at 20 and 300 K for the three orientations of PETN. Similar to the 20 K spectra, the 300 K spectra have no dependence on the in-plane axes, so we only present the average spectral data here for the three measured crystal orientations. All three orientations show phenomenologically the same temperature dependence. As the sample is cooled, the peaks become narrower and shift to higher frequency, similar to the behavior reported by Barber et al. on single crystals8 and by Fan et al. on powders.11 The weak absorptions between 80-83 cm-1 cannot be resolved at 300 K due to the temperature-dependent spectral shift and broadening of features. The weak absorptions at 36 and 55 cm-1 were not resolved at 300 K either. Otherwise, the 300 K spectrum is a good representation of the spectrum measured at 20 K. Finally we present a comparison of our normalized single crystal spectra taken at 20 and 300 K with normalized powder data taken at 4 K in Figure 4. To produce this figure, we assume that the powder spectrum represents equal contributions from all possible crystal orientations. If the measured spectra from the [001], 〈10(a/c)2〉, and 〈110〉 orientations are a suitable representation for “all possible orientations”, then an average
of all the orientation-dependent spectra should reasonably match the powder spectrum. The averaged single-crystal spectrum presented in Figure 4 shows similarities with the powdered spectrum measured at 4 K by Fan et al. Both show strong absorptions in the 65-75 cm-1 range and strong absorption at frequencies greater than 90 cm-1. Unfortunately, we cannot compare the linewidths between the powder spectrum and the single crystal spectrum since we were unable to measure the maximum absorptions in the single crystal data. The powder spectrum does not resolve the absorptions between ∼80-83 cm-1, which we know from Figure 2 to be from at least three distinct contributions. Neither the averaged single crystal spectrum nor the powder spectrum managed to resolve the weak absorptions at 36 and 56 cm-1. There is one noteworthy difference that we cannot adequately explain. The ∼70 cm-1 peak in the powder spectrum is shifted to a lower frequency compared to the 20 K single crystal spectrum. This shift could be caused by inability to cool the powders to the same temperature as the cryostat. The packed powders are poor thermal transmitters compared to a homogeneous crystal, so it is plausible that the powders are not at 4 K. However, Fan’s 4 K peak looks to be centered within our 300 K peak. Furthermore, their data taken at 300 K is shifted another 6 cm-1 toward lower frequencies. The spectral shift between the powder and single crystal data is too large to be explained by temperature differences alone. One potential explanation for these spectral shifts could be related to crystal quality differences between the powders and the single crystals. A large series of experiments undertaken on RDX to understand differences in detonation sensitivity revealed huge variations in crystal quality, morphology, and purity.21 Wilkinson et. al.22 found notable differences in the 0-3.5 THz spectrum of some of these materials, suggesting that material quality affects the measured spectrum in RDX. Conclusions We have measured the orientation dependence of the absorption for transmission along [001], 〈10(a/c)2〉, and 〈110〉 PETN single crystals from 7-100 cm-1 at 20 and 300 K. Two strong absorptions were found in the 72 and >90 cm-1 range for all three crystallographic orientations. Weaker absorptions were found at 80, 82, and 83 cm-1. Very weak absorptions at 36 and 56 cm-1 were present in [001] but were not present in the other orientations. In contrast to RDX, PETN does not show any measurable dependence on the orientation of the polarization
442
J. Phys. Chem. A, Vol. 115, No. 4, 2011
with respect to the crystallographic axes. A comparison of averaged single crystal spectra with the spectrum of powdered PETN shows reasonable agreement between the two, although the locations of some of the peaks in the powder spectra are shifted to lower frequencies compared to the single crystal spectra. Acknowledgment. This work was supported in part by the Office of Naval Research and was performed at Los Alamos National Laboratory, operated by Los Alamos National Security, LLC, for the National Nuclear Security Administration of the U.S. Department of Energy under contract DE-AC52-06NA25396. J.B. was supported in part by the Department of Homeland Security under contract DTFACT-03-C-00042. References and Notes (1) Campbell, A. W.; Engelke, R. The diameter effect in high-density heterogeneous explosives. In Sixth Symposium (International) on Detonation; Edwards, D. J., Ed.; Office of Naval Research: Coronado, CA, 1976; pp 642-652. (2) Gibbs, T. R.; Popolato, A. LASL explosiVe property data; University of California Press: Berkeley, CA, 1980. (3) Meyer, R.; Ko¨hler, J.; Homburg, A. ExplosiVes, 5th ed.; WileyVCH: Weinheim, 2002. (4) Tokmakoff, A.; Fayer, M. D.; Dlott, D. D. J. Phys. Chem. 1993, 97, 1901–1913. (5) Fried, L. E.; Ruggiero, A. J. J. Phys. Chem. 1994, 98, 9786–9791. (6) Tarver, C. M. J. Phys. Chem. A 1997, 101, 4845–4851. (7) Leahy-Hoppa, M. R.; Fitch, M. J.; Osiander, R. Anal. Bioanal. Chem. 2009, 395, 247–257.
Whitley et al. (8) Barber, J.; Hooks, D. E.; Averitt, R. D.; Taylor, A. J.; Babikov, D. J. Phys. Chem. A 2005, 109, 3501–3505. (9) Whitley, V. H.; Hooks, D. E.; Ramos, K. J.; O’Hara, J. F.; Azad, A. K.; Taylor, A. J.; Barber, J.; Averitt, R. D. Anal. Bioanal. Chem. 2009, 395, 315–322. (10) Lo, T.; Gregory, I. S.; Baker, C.; Taday, P. F.; Tribe, W. R.; Kemp, M. C. Vib. Spectrosc. 2006, 42, 243–248. (11) Fan, W. H.; Burnett, A.; Upadhya, P. C.; Cunningham, J.; Linfield, E. H.; Davies, A. G. Appl. Spectrosc. 2007, 61 (6), 638–643. (12) Leahy-Hoppa, M. R.; Fitch, M. J.; Zheng, X.; Hayden, L. M.; Osiander, R. Chem. Phys. Lett. 2007, 434, 227–230. (13) Konek, C.; Wilkinson, J.; Esenturk, O.; Heilweil, E.; Kemp, M. In Proceedings of the SPIE, Terahertz Physics, Devices and Systems III: Advanced Applications in Industry and Defense; Anwar, M., Dhar, N. K., Crowe, T. W., Eds.; SPIE: Bellingham, WA, 2009; p 73110K. (14) McGrane, S. D.; Barber, J.; Quenneville, J. J. Phys. Chem. A 2005, 109, 9917–9927. (15) Allis, D. G.; Korter, T. M. ChemPhysChem 2006, 7 (11), 2398– 2408. (16) Piryatinski, A.; Tretiak, S.; Sewell, T. D.; McGrane, S. D. Phys. ReV. B 2007, 75, 214306. (17) Velizhanin, K. A.; Kilina, S.; Sewell, T. D.; Piryatinski, A. J. Phys. Chem. B 2008, 112, 13252. (18) Burnett, A. D.; Kendrick, J.; Cunningham, J. E.; Hargreaves, M. D.; Munshi, T.; Edwards, H. G. M.; Linfield, E. H.; Davies, A. G. ChemPhysChem 2010, 11, 368–378. (19) Pereverzev, A.; Sewell, T. D. 2010, submitted for publication. (20) Cady, H. H.; Larson, A. C. Acta Crystallogr. 1975, B31, 1864– 1869. (21) Doherty, R. M.; Watt, D. S. Propellants, ExplosiVes, Pyrotechnics 2008, 33 (1), 4–13. (22) Wilkinson, J.; Caulder, S. M.; Portieri, A. Manufacturing process effects on the terahertz spectra of RDX. In Proceedings of the SPIE; Jensen, J. O.; Cui, H.-L.; Woolard, D. L.; Hwu, R. J., Eds.; SPIE: Orlando, FL, 2008; Vol. 6949, p 694904.
JP108388C
