J . Phys. Chem. 1989, 93, 462-466
462
though the results reported in ref 2 were derived from the effect of pressure on time-to-explosion measurements. On the basis of theoretical electronic orbital calculations a bimolecular process has been proposedZoas follows: H3C-N
7+ \
H
/O
H3C-N
\0
0
- \/ P y H-C-N
**>C-N
'0
H/\H H
\ / H
+
H-C-N=O
-
/o
tOL*'
H
H-0-C-N
r/" I
H
\
(1)
O
The quantum mechanical calculations show this to be a threecenter reaction complex, where, under high pressure, the potential surface contacts result in a simultaneous motion of atoms, as indicated by the arrows, to form the resultant products. The reaction of the nitrosomethane rapidly proceeds to form H C N and finally N H 3 and HCOOH: H
\ / H
H-C-N=O
-
H\
C=N-0-H
/ H
-
HCN
+
HOH
(2)
At elevated pressure and temperature, H C N hydrolyzes very fast: H C N 2HOH NH3 + HCOOH (3)
+
-+
The remaining product from reaction 1 rapidly decomposes (reaction 4) into the following possible gaseous compounds: H 2 0 , N 2 0 , COz, CO, NO, H2, and C(s). Reaction 1 is rate controlling with a calculated activation energy of 32.5 kcal mol-'. Reactions 2-4 are relatively fast compared with reaction 1. Reactions 3 and 4 provide the pathways for the production of NH3, HCOOH, and H 2 0via chemical reaction with HCN, which is known to be a major pyrolysis product of nitromethane.*' Our experiments confirm the presence of NH3, HCOOH, and H20along with volatile gases as the major products of thermal decomposition of nitromethane under high pressures. While it is true that other mechanisms can result in a positive pressure dependence on the rate of a reaction, in the case of
nitromethane decomposition, our measured effects and experimentally identified reaction products provide strong evidence to support the theoretically proposed mechanisrn?O This conclusion, however, is not unequivocal. This proposed reaction scheme for nitromethane decomposition at elevated pressures suggests a strong intermolecular interaction of nitromethane in the condensed phase. If this is indeed the case, then it is important to investigate the effect of pressure on the vibrational bands in nitromethane and to analyze the observed shifts in light of intermolecular coupling between the NO2 and H,C groups of interacting molecules. This is the subject of the succeeding article in this journal issue. Conclusions Nitromethane thermally decomposes under pressure when heated to temperatures above 440 K at pressures greater than 1.6 GPa. Both pressure and temperature accelerate the rate of thermal decomposition. The condensed product residues are mainly ammonium formate and water in addition to nonretrievable volatile gases. More than one decomposition mechanism or catalytic effect appears to be present as indicated by the complex kinetic results. A proposed bimolecular mechanism is presented that can account for the observed positive pressure dependence of the reaction rate and the experimentally detected reaction products. Nitromethane also was observed to undergo a catastrophic reaction under certain dynamic stress conditions in the DAC at room temperature. The sensitivity level to this catastrophic reaction appears to be crystal orientation dependent with respect to the applied stress in the DAC. The resulting solid residue from this reaction is amorphous and carbonaceous in character, stable to temperatures in excess of 573 K. Under the same conditions, deuteriated nitromethane did not undergo the catastrophic reaction.
Acknowledgment. We thank S . Trevino (NBS) for providing the protonated and deuteriated samples of nitromethane, M. J. Welch (NBS) for carrying out the mass spectrometry measurements, and B. Fanconi (NBS) for helpful assistance with the FTIR instrument. We also acknowledge the US.Army Research Office and the Naval Surface Warfare Center, Independent Research Program, for financial support for this project. Registry No. Nitromethane, 75-52-5.
Effects of Pressure on the Vibrational Spectra of Liquid Nitromethane P. J. Miller,* Naval Surface Warfare Center, White Oak, Maryland 2091 0
S . Block, and G. J. Piermarini Institute for Materials Science and Engineering, National Bureau of Standards, Gaithersburg, Maryland 20899 (Received: January 15, 1988; In Final Form: June 22, 1988)
A complete normal-coordinate calculation for nitromethane is given using the vibrational frequencies measured from the three isotopes CH3N02,CD3N02,and CH31SN02.Pressure effects on the vibrational normal modes of the superpressed liquid to 2.0 GPa also were measured. A softening of the frequency of the asymmetric stretching mode of NO2 with increasing pressure indicated a strong intermolecular interaction. By calculating a minimum energy pair configuration and applying two-site exiton theory, we calculated the shifts of the normal modes of the NO2 stretching vibrations as a function of density and pressure and found them to agree qualitatively with the experimentallymeasured shifts. The results support the bimolecular nature of the mechanism for the thermal decomposition of nitromethane under pressure.
Introduction In a related paper on the effects of pressure on the thermal decomposition kinetics and chemical reactivity of nitromethane, we showed that nitromethane thermally decomposed under pressure through a complex bimolecular mechanism.' A de-
composition mechanism was proposed for this reaction that explains both its bimolecular nature and the observed reaction (1) Piermarini, G. J.; Block, S.;Miller, P. J. J . Phys. Chem., preceding paper in this issue.
This article not subject to U S . Copyright. Published 1989 by the American Chemical Society
Vibrational Spectra of Liquid Nitromethane
The Journal of Physical Chemistry, Vol. 93, No. 1, 1989 463
I .o
=
I
RAMAN S P E C T R - 4
____ _ _ _ _
1.0
l
0.0
I
3600
2800
,
2000
I
1200
400
a
-
I
I
b
WAVE NU,MB E R Figure 1. Infrared spectra of liquid nitromethane taken with the sample in a diamond anvil high-pressure cell at ambient temperature with a nominal pressure seal to prevent loss of sample.
,
~
products, solid residual ammonium formate, water, and volatile gases. To further support the bimolecular nature of the thermal decomposition, we report in the present paper the effects of pressure on the normal vibrations of liquid nitromethane superpressed to approximately 2.0 GPa. We then show through valence field normal-coordinate calculations and an intermolecular potential calculation that nitromethane molecules under pressure interact strongly with their nearest-neighbor molecules. This strong intermolecular interac4ion is a precursor to the bimolecular reaction observed for the thermal decomposition of nitromethane under pressure. Nitromethane has been the subject of numerous spectroscopic studies.*-1° Its vibrational spectrum has been extensively investigated by using 15N and '*Oisotopic substitution.11J2 More recently, the infrared spectra of matrix-isolated nitromethane and trideuterionitromethane have also been studied.I3 Although complete vibrational assignments and normal-coordinate analysis of the molecular vibrations of nitromethane have been reported, we have redone an analysis to investigate the effects of pressure on the reported force constants. In the AI representation there is appreciable mixing of the motions of the CH3 deformation, NO2 stretching, C N stretching, and NOz bending modes resulting from the large motion of the hydrogen atoms in the normal mode.z4 The effect of pressure on these interactions may lead to further understanding of the chemistry of nitromethane decomposition. Evidence of dimer formation in the matrix-isolation studies indicated that at pressures of 10 GPa, which are reached in detonation processes, dimers may play an important role in the reacton mechanism.13 Thus, static high-pressure investigations of nitromethane in the diamond cell could, perhaps, shed further light on the initial chemistry involved in the thermal decomposition of nitromethane under dynamic pressures, particularly in view of the possible dimerlike interaction between molecule pairs.
(2) Wells, A. J.; Wilson, E. B. J . Chem. Phys. 1941, 9, 314. (3) Crawford, B. L.; Wilson, E. B. J . Chem. Phys. 1941, 9, 323. (4) Wilson, T. P. J . Chem. Phys. 1943, ZZ, 361. (5) Smith, D. C.; Pan, C. Y.; Nielson, J. R. J. Chem. Phys. 1950, 18,706. (6) Cox, A. P.; Waring, S. J . Chem. Soc., Faraday Trans. 1972,2, 1060. (7) Singh, K. Spectrochim. Acta, Part A 1967, 23, 439. (8) De Maine, P. A. D.; De Maine, M. M.; Gobie, A. G. Trans. Faraday SOC.1957, 53, 427. (9) De Maine, P. A. D. J. Mol. Spectrosc. 1960, 4,407. De Maine, P. A. D.; De Maine, M. M.; Brigs, A. A.: McAlonie, G. E. J. Mol. Spectrosc. 1960, 4, 398. (10) Varsangi, G.; Holly, S.;Imre, L. Spectrochim. Acta, Part A 1967, 23, 120s. (11) Malewski, G.; Pfeiffer, M.; Reich, P. J. Mol. Struct. 1969, 3, 419. (12) Holman, M.; Pinchas, S. J. Chem. SOC.1960, 1246. (13) Verderame, F. D.; Lannon, J. A,; Harris, L. E.; Thomas, W. G.; Lucia, E. A. J. Chem. Phys. 1972, 56, 2638.
0.0 8--
3fiOfl
2800
2000
,
~
1200
400
WAVENUMBER
Figure 2. Raman spectra of nitromethane (a) and trideuteriated nitromethane (b) at ambient pressure and temperature. 10.0
8.0
-
6.0
-
NITROMETHANE
h
E0 \
*
W
A
a
,a' 4.0
PRESSURE (GPa) Figure 3. Effect of pressure on the normal modes (symmetric) of liquid nitromethane, where A1 is (AI)CH3 (sym, str), A2 is (Al)N02 (sym str), A3 is ( A I ) N 0 2 (bend), and A4 is (AI)CH3 (bend). The uncertainty associated with the pressure measurement is *0.025 GPa and with frequency, *0.5 em-'. The curves shown are the result of a cubic spline fit of the data points.
TABLE I: Observed Fundamental Vibrations (cm-') of CH3N02 and Its Isotows mode AI 4 C H 3 ) vdNO2) WH3) 4CN) UNO2) B2 h(CH3) WH3) P(CH3) qJ(N02) B1 d C H 3 ) ua(NOA UCH3) P(CH3) P(NOJ A2 .r(torsion) a Reference
CHqN02 2955 1402 1379 917 655 3050 1426 1125 607 3045 1561 1426 1103 480 =60
CHq1'N02 2967 1391 1361 912 655 3065 1426 1125 595 3045 1524 1426 1102 479
CD3N02 2185 1395 1075 898 630 2280 1040 940 545 2280 1545 1040 883 432 =45
2 1.
Experiments and Results
The method used to study the effects of pressure on the vibrational spectrum of nitromethane is described in the companion paper in this issue' and also in earlier paper^.'^,'^
The Journal of Physical Chemistry, Vol. 93, No. 1, 1989
464
i
1
''O n
E0
-
,OB1
NITROMETHANE
,,u
1
TABLE II: Valence Force Constants for Nitromethane (mdyn/A)O
I
7.0
B2
) ,
5.0
\ 4
3.0
E3 E4
h
4 1.0
a *-A~
.
0.0
O
I
~
I
~
I
~
~t
,
~,
I
0.4
, , , ,,
,
2.0
PREO&URt ( G P 8 Figure 4. Effect of pressure on the normal modes (antisymmetric) of liquid nitromethane, where B1 is (BI)N02(rock), B2 is (B1)CH2(rock), B3 is (BI)N02(asym str), and B4 is (B2)N02(rock). the uncertainty in pressure is f0.025 GPa, in frequency, f0.5 cm-I. The curves shown are the result of a cubic spline fit of the data points.
Briefly, the method consists of using a diamond anvil highpressure cell mounted on a micrometer-controlled positioning device inside the sample chamber of an FTIR spectrometer to obtain infrared absorption spectra for the spectral region 4000-450 cm-' or on the microscope stage of a Raman spectrometer equipped with an argon ion laser. Pressures were measured by using the ruby fluorescence method.I6 All spectra were obtained under ambient room-temperature conditions (-298 K), and pressures were measured with an accuracy of f0.025 GPa. Three isotopes of spectroscopically pure (>99%) nitromethane ( C H 3 N 0 2 , CD3N02,and CH31sN02)were obtained from commercial sources and were used directly as received without further purification. Vibrational spectra typically obtained in this investigation are shown in Figures 1 and 2. The observed and assigned fundamental vibrational bands in the liquid state are listed in Table I. These results are in complete agreement with previous investigations for the three isotope^.^^,'^ In Figures 3 and 4, new results on the pressure dependence of the internal molecular vibrations are presented for liquid nitromethane to approximately 2.0 GPa. The nonlinear and negative pressure dependence of the NOz stretching vibrational frequency is clearly indicative of intraand/or intermolecular interactions. In fact, at pressures approaching 1.6 GPa with slightly elevated temperatures (4440 K), nitromethane reacts irreversibly through a bimolecular mechanism to form a solid product stable at ambient pressure and temperature as described in ref 1. To interpret these results, we have carried out a normal-coordinate analysis and an intermolecular potential calculation to evaluate the effects of pressure on the observed vibrational shifts as due to the formation of dimerlike interactions. Normal-Coordinate Analysis and the Pressure Dependence of the Normal Modes of Nitromethane. As previously stated, normal-coordinate calculations on nitromethane have been carried out several times, the most recent of which is given in ref 13. With the additional isotopic information obtained in this study, the calculation was repeated. The bond angles and bond lengths were those obtained from microwave data.'* The normal coordinates and force constants were calculated with the secular equation factored according to C2, symmetry. The internal valence co(14) Piermarini, G. J.; Block, s.;Miller, P. J. J. Phys. Chem. 1987, 3872. (15) Miller, P. J.; Piermarini, G. J.; Block, S . Appl. Specfrosc. 1984, 38, 680. (16) Piermarini, G. J.; Block, S.; Barnett, J. D.; Forman, R. S . J . Appl. Phys. 1975, 46, 2714. ( 1 7 ) Cataliotti, R.; Paliani, G. Can. J . Spectrosc. 1979, 24, 23. (18) Tannenbaum, E.; Meyers, R. J.; Gwinn, W. D. J . Chem. Phys. 1956, 52, 42.
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
CN
4.15
NO
9.45 2.70 -0.10 2.70 1.45 0.625 -1.00 0.41 0.25 5.05
N0,NO CN,NO Y YYY
CN,? NO,? d CN,CH CH
12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
0.043 0.66 0.16 0.613 -0.20 0.02 0.18 -0.04 0.31 -0.035
"The angles are defined in ref 19.
-1.0
-
Miller et al.
ordinates used are described in ref 3. The symmetry coordinates adopted for these calculations were similar to those given in ref 19. Trial force constants were estimated from the published data in ref 13 and refined by the use of the Schactschneider and Snyder programz0 on a CDC 6500 computer. A total of 21 force constants, listed in Table 11, were used to fit 42 observed vibrational frequencies for the three isotopes of nitromethane. The FPERT program of Schactschneider and Snyder gave a weighted leastsquares fit to all the observed frequencies, and the calculated frequencies were refined to be within *0.5% of the observed. As expected, these results readily demonstrated that there is appreciable mixing in the A, modes for CH2 deformation, NOz stretching, C N stretching, and NO2 bending motions. Their interaction constants are correspondingly large, and thus the greatest uncertainty for these calculations exists in their evaluated constants. The calculations were repeated for the observed frequencies of nitromethane at pressures up to 2 GPa. Unfortunately, the observed frequency shifts, which are as small as 0.1% in some cases, are not large enough to arrive at definite conclusions as to which force constants are most effected by pressure. Effects at higher pressures were not studied, because the superpressed liquid was found to rapidly crystallize at pressures greater than ~ 2 . 0 GPa. Furthermore, the effects of unit cell or factor group splitting and shifts in the normal modes could further complicate the analysis. The general increase in frequency for most of the normal modes with respect to pressure could be accounted for by slight increases in the diagonal force constants. The main exception to this was the softening of one of the NOz stretching frequencies. This latter calculated frequency shift could be force fit to the observed frequency by increasing the off-diagonal interaction constants for the NO,NO stretching internal coordinates, and N 0 , C H bending internal coordinates. This may be reasonable, in that it has been shown that the barrier to rotation of the methyl and nitro group changes from roughly 6 cal in the vapor phase to -250 cal in the solid phase at 4 K.21 An increase in the barrier height in the liquid phase over a 2-GPa pressure range would not be unexpected and thus would account for the observed anomalous softening in the NO2 frequency shift. However, the observed shifts of the vibrational modes between the room-temperature vapor phase and the 4 K solidified phase do not correspond to the magnitude of the shifts observed for the pressure effects. This probably indicates that the increase in the barrier height for the torsional mode associated with an intermolecular coupling of N - 0 and H-C bonds are responsible for the observed pressure effects. Future experiments in this laboratory using far-infrared and low-frequency Raman scattering techniques will attempt to determine the pressure effects on the rotational barrier by directly observing the torsional-mode pressure dependence. On the other hand, the frequency shifts of the NO2stretching modes can be more easily accounted for, within the limits of the sensitivity of the calculation, by slightly increasing the force constant of one NO bond and decreasing the force constant of the other. This would result from a strong attractive molecular interaction between the NO2 group (19) Fleming, J. W.; Banwell, C. N. J. Mol. Spectrosc. 1969, 31, 378. (20) Schachtschneider, J. H.; Snyder, R. G. Spectrochim. Acta 1963, 19, 117. (21) Trevino, S. F.; Rymes, W. H.J . Chem. Phys. 1980, 73, 3001
Vibrational Spectra of Liquid Nitromethane and the CH3 group of different molecules that increases with increasing pressure. Matrix-isolation spectra studies13have shown the possible existence of dimers of nitromethane and have suggested that they could play an important role in the high-pressure regimes that occur in detonations and thermal explosions. This kind of interaction could easily explain the observed coupling effects that are responsible for the pressure effects on the NOz stretching modes in which the normal-coordinate calculations seem to indicate that one N-O bond strengthens and the other weakens. Intermolecular Potential Considerations and the Effect of Pressure on the Frequency Shifts. For interpretation of the frequency shifts of nitromethane with respect to pressure as being due to the interaction of neighboring molecules, it is necessary to determine the Cartesian displacements, dx/dQ, of the normal modes of vibration and also the most likely relative orientation of nitromethane molecules in a pair for which the intermolecular potential energy is lowest. Fortunately, a complete description of this calculation, including the parameters used in the potential function, has been reported in the 1 i t e r a t ~ r e . lThus, ~ only the results of the analysis are given here. By application of two-site, molecular excitation theory, the shift due to intermolecular interaction and coupling of vibrations can be estimated. The intermolecular potential energy equation used is
where rij is the distance between atoms i a n d j on two different molecules. The three terms represent repulsive, dispersive, and electrostatic energies, respectively. The constants A and C are atom-atom contact constants, e is the electrostatic charge, and qi and q, are charges on the atoms estimated from molecular orbital theory. From the magnitude of the intermolecular potential energy the likelihood of existence of a given pair configuration can be estimated. This was carried out by finding those configurations for a pair of nitromethane molecules that possess the lowest intermolecular potential energy which would be the most stable pair configuration possible and also fit the observed shifts and splittings assigned to the vibrational modes of the dimer as determined in the matrix isolation spectra. The shifts due to intermolecular coupling are given by
where PI and Ql, are the unperturbed frequency and normal coordinate of the Ith normal mode, rij is the interatomic distance, and the other term is
the values of which depend upon the relative orientation and the Cartesian coordinate displacements of the normal modes as determined by the normal-coordinate analysis. Finally, if we define a term R,, the distance between the centers of the molecules in the most likely pair configuration determined in ref 13, and let this vary as a function of pressure, the pressure shift, dui/dP, for each normal mode may be estimated and compared to the observed shifts. With use of the reported data and configurational analysis, the pressure effects on the intermolecular potential, aV,,/aR, were qualitatively evaluated. It was found that as the value of Ri. decreased, the asymmetric NO stretching frequency 1561 cm- I also decreased in frequency, as was observed. The B1 C H stretching frequency behaved similarly, but it was not studied experimentally. Furthermore, the results qualitatively show that there is a minimum distance, Mi,,or pressure where these two modes would start to increase in frequency as do the other normal modes, this being due to the repulsive term dominating in the potential. This effect is observed experimentally in the 0.4-0.5GPa range. The calculated magnitudes of the shifts do not, however, agree with those observed, but in view of the approximations this is not unexpected. The calculations were made for a molecule pair in free space, where the most stable molecular
The Journal of Physical Chemistry, Vol. 93, No. I, 1989 465 1.00 n
\
NITROMETHANE
i
0 0
%
-0.90
5l
20 3.
0.80
2
Ef: u 0.70 w a
rn 0 60
1
0.2
1
I
0.4
'
I
0.8
'
I
0.8
'
I
1.0
I
1.'2 1.'4
1
PRESSURE (GPa) Figure 5. Pressure dependence of the specific volume of nitromethane at 298 and 373 K. The uncertainty associated with the specific volume measurement is 10.015 cm3/g and with pressure is f0.025 GPa. The curves shown were drawn through the center of each data point. pair was found to be with the second molecule's center of mass displaced from the first by X = 2.8 A, Y = 0 A, and Z = -0.5 A and rotated about the Z axis by 180". In this case the Z axis is parallel to the C-N bond and the X axis is perpendicular to the NO2 plane, whereas the experimental frequencies were obtained from compressed liquids, where the interactions are multiple and three dimensional. However, because the major dimerlike interaction occurs only between two adjacent molecules, the gross changes in the observed vibrational spectrum, we assume, can be approximated by this simple pair calculation. Also, the calculations assume a fixed configuration in space that does not vary with compression except for the separation in the center of masses. Recent similar calculations2zindicate that the configuration is pressure dependent and becomes nearly linear, head-to-tail at high pressures. However, the model definitely predicts the observed early trends in dv/dP at pressures below 1.0 GPa. For comparison of the observed pressure effects on the frequencies of the NOz stretching modes to the calculated frequency shifts due to intermolecular coupling, it is necessary to estimate the change in the intermolecular distance with pressure between interacting pairs of molecules. Because nitromethane is a nonspherical molecule, if dimerlike pairing occurs, this would be a nontrivial problem. However, if the minimum-energy pair configuration of ref 13 is used as an estimated reference separation and if the average separation of the center of masses (assuming spherical geometry) for nitromethane at 1 atm and 298 K is used as another reference, then if we assume that this distance changes linearly with pressure, the value of dP/dRi, can be estimated. Furthermore, the calculated frequency shifts can also be estimated. Figure 5 shows the effect of pressure on the specific volume of nitromethane at 298 and 373 K. The method used for determining these experimental curves has been previously reportedz3 and will not be discussed in any detail here. The specific volume of the compressed liquid in a diamond anvil cell is determined from both optical observations (planimeter measurements on photomicrographs to determine changes in sample area) and resistivity measurements (strain gage transducer measurements to determine changes in sample thickness). The reported error in the specific volume from this kind of measurement is estimated to be f0.015 cm3/g.z3 The specific volume of nitromethane at 298 K appears to change linearly with pressure to about 0.4 GPa. At higher pressures up to about 0.9 GPa the specific volume change with pressure decreases rapidly with the curve becoming very flat, no doubt due to the increasing repulsive forces at those (22) Bardo, R. D., private communication. (23) Brasch, J. W. Rev. Sci. Instrum. 1980, 51, 1358.
J. Phys. Chem. 1989, 93,466-468
466 8.0
1
I
NITROMETHANE 6,0
1
h
E 0
j
4.0
where it becomes sharply flat also could have been chosen. The main point to be made here is that the calculated pressure effects on the NO2 normal modes of a molecular pair show the same trends as the observed pressure effects in the liquid and that the calculated formation of dimerlike intermolecular coupling can account for these effects.
1 -I
\
Summary
4
e
2'
a
-4.0
-
0.0
I
I
'
0.4
PREO&URI?
2.0
(~pA.7
Figure 6. Calculated and observed pressure shifts of the NO2 stretching modes, where a is (AI)N02(sym str) and b is (B,)N02 (asyrn, str).
densities. By choosing 0.4-0.5 GPa as the pressure at which the calculated dimer separation occurs in the liquid and assuming that distance changes linearly with pressure down to 1 atm as previously described, we can estimate the value of dP/dRij and duNo,/dP. These results are shown in Figure 6 for the symmetric and asymmetric NO2 stretching modes. There is no valid reason for choosing 0.4 GPa as the reference pressure other than that it corresponds to the observed minimum in the pressure effect on the NO2asymetric stretching mode. It is interesting to note that this pressure also corresponds to the equilibrium crystallization pressure for liquid nitromethane at 298 K, the point above which the liquid becomes superpressed. The 0.9-GPa region of the curve
The observed pressure effects on the vibrational normal modes of liquid nitromethane have indicated that there is a strong intermolecular interaction between neighboring molecules. A normal-coordinate analysis of the vibrational force constants of nitromethane as a function of pressure shows that one of the N-O bond force constants increases while the other N-0 bond force constant decreases, indicative of a strong intermolecular interaction. By application of two-site molecular exciton theory, the shifts due to pressure and density of the normal modes of two interacting nitromethane molecules are determined. This calculation for the shifts for the NO stretching modes agrees qualitatively with the observed trends in pressure effects on the normal modes. The results for both experiment and calculation have shown that pressure affects the intermolecular interactions or coupling between normal vibrations in nitromethane and this interaction possibly leads to the observed bimolecular nature of the thermal decomposition of nitromethane under pressure.
Acknowledgment. We express our gratitude to J. W. Brasch for supplying the specific volume data on nitromethane as a function of pressure. We also acknowledge the Naval Surface Warfare Center, Independent Research Program, and the US. Army Research Office for financial support of this project. Registry No. CH3NO2,75-52-5; D2, 7782-39-0; 15N, 14390-96-6.
Pressure Effect on Quenching of Perylene Fluorescence by Halonaphthalenes R. Schmidt, W. Janssen, and H.-D. Brauer* Institut fur Physikalische und Theoretische Chemie, Uniuersitat Frankfurt, Niederurseler Hang, 6000 FrankfurtIMain, West Germany (Received: March 31, 1988; In Final Form: June 27, 1988)
The effect of pressure on the quenching of perylene (P) fluorescence by 1-bromonaphthalene (BN) and 1-iodonaphthalene (IN) has been investigated in toluene at 298 K in the range of 1 bar to 1.5 kbar. The quenching rate of the P-BN system continuously rises with increasing pressure, whereas the quenching rate of the P-IN system initially rises up to a maximum value at about 600 bar and then falls as the pressure is further increased. Treatment of the data in terms of the transition-state theory is consistent with the assumption that the fluorescence quenching occurs via an exciplex. For the quenching of the perylene fluorescence by BN and IN, negative activation volumes, A P , of -8.6 and of -4.6 cm3 mol-', respectively, can be evaluated.
Introduction
The fluorescence of perylene (P) is practically not quenched by molecules containing heavy atoms if the lowest excited singlet and triplet states of these compounds lie much higher than the lowest excited singlet state (S,) of perylene, which is the case for bromobenzene, ethyl bromide, methyl iodide, and n-propyl iodide.1-3 However, one observes a sharp increase in the fluorescence quenching rate for halonaphthalenes (HN) in the series 1(1) Zander, M. Fresenius' 2. Anal. Chem. 1967,229,352; Erdoel Kohle, Erdgas, Petrochem. 1969,22, 81; Fresenius' Z . Anal. Chem. 1973,263, 19. Zander, M. Z . Naturforsch., A: Phys., Phys. Chem., (2) Dreeskamp, H.; Kosmophys. 1973, 28A, 1743. (3) Dreeskamp, H.; Koch, E.; Zander, M . Ber. Bunsen-Ges. Phys. Chem. 1974, 78, 1328.
0022-3654/89/2093-0466$01.50/0
chloronaphthalene (CN), 1-bromonaphthalene (BN), and 1iodonaphthalene (hv)! Since the lowest excited triplet state but not the SI state of these quenchers lies below the SI state of P, it was concluded that the fluorescence quenching was due to heavy-atom-induced intermolecular, "spin forbidden" singlettriplet (ST) energy transfer from perylene to the quencher molec~les.~ Recently direct evidence for the formation of the TI state of P was obtained by TT absorption measurements.5 This result could be explained in terms of the postulated ST process, if a (4) Zander, M.; Dreeskamp, H.; Koch, E. 2.Naturforsch., A: Phys., Phys. Chem., Kosmophys. 1974, 29A, 1518. ( 5 ) Lbhmannsroben, H.-G., University of Braunschweig, FRG, private communication.
0 1989 American Chemical Society
