J . Phys. Chem. 1990, 94, 3344-3347
3344
Kinetics of CO and H Atom Production from the Decomposition of HNCO in Shock Waves C.H.Wu, H . 4 . Wang, Chemistry Division, Code 6180, Naval Research Laboratory, Washington, D.C. 20375-SO00
M. C . Lin,* Department of Chemistry, Emory University, Atlanta, Georgia 30322
and R. A. Fifer Ballistic Research Laboratory, Aberdeen Proving Ground, Maryland 21 005-SO66 (Received: October 18, 1989; In Final Form: January 23, 1990)
The production of CO and H atoms from the thermal decomposition of HNCO in shock waves at temperatures between 2120 and 2570 K has been measured by resonance absorption. Kinetic modeling of these product formation rates using a recently established mechanism yielded the second-order rate constants for the primary decomposition processes HNCO + Ar NH + CO + Ar (1) and HNCO + Ar H + NCO + Ar (2): k l = 10’s.41M.16 exp(-39800 f 7 0 0 / T ) cm3/(mol.s) and k , = 1017.0 exp(-56400/T) cm3/(mol.s). The values of k , determined in this work agree quantitatively with those reported recently by Hanson, Bowman, and co-workers. Combination of these two sets of data gives rise to the Arrhenius expression k l = 10’5~925f0~085 exp(-42640 i 450/T) cm3/(mol.s) for 1830 K IT 5 3340 K.
-
-
Introduction
lsocyanic acid (HNCO) has been shown to be an early decomposition product of cyclic nitramines (such as RDX and HMX).I In these systems it is a likely product of HCN oxidation by OHZduring the early stages of the decomposition processes. Since HCN is also an important product of N-containing fuels derived from shale oil or coal, high-temperature HNCO chemistry thus becomes quite important with respect to the formation of NO, in these combustion systems.2 Recently, HNCO has also been proposed by Perry and Siebers3 as a potential agent for “rapid reduction of nitrogen oxides (RAPRENO,) in exhaust gas streams“. The kinetics and mechanisms of the high-temperature reactions of HNCO, including the thermal unimolecular decomposition reaction, are therefore very relevant and important to many practical processes. The thermal decomposition of HNCO was first investigated by Back and Childs4 in a static cell at temperatures between 823 and 973 K and pressures from 5 to 20 Torr. The major products observed were C 0 2 , CO, N2, HCN, and HZ. Because of the presence of surface effects and apparent polymerization reactions: no quantitative kinetic data could be obtained from this brief study. In view of the great thermal stability of the HNCO molecule, the low-temperature reaction may be initiated by either bimolecular or surface catalytic processes. The first homogeneous pyrolytic study of HNCO was made by Fueno and co-workers in the temperature range of 2100-2500 K using a shock tube.5 The reaction was monitored by UV absorption at 206 and 336 nm for the disappearance of HNCO and the appearance of NH(3Z-), respectively. Analysis of measured kinetic data, aided by computer-modeling and RRKM calculations, gave rise to the second-order rate constant for the unimolecular dissociation process: HNCO + Ar NH(3Z-) CO Ar (1)
-
+
k , = 1017.23”0.36 exp(-48400 f 2050/7‘)
+
cm3/(mol.s)
( I ) Boggs, T. L. Fundamentals of Solid-Propellant Combustion. f r o g . Aeronaur. Astronaut. 1984. 90, 121. (2) Miller, J. A.; Bowman, C. T. frog. Energy Combust. Sci., in press. (3) Perry, R . A.; Siebers, D. L. Nature 1986, 324, 657. (4) Back, R. A.; Childs, J. Can. J . Chem. 1968, 46, 1023. ( 5 ) Kajimoto. 0.; Kondo, 0.;Okada, K.; Fujikane, J.; Fueno, T. Bull. Chem. Sac. Jpn. 1985, 58, 3469.
This result differs somewhat from that obtained recently by Hanson, Bowman, and co-workers6 carried out in a broad temperature range of 1830-3380 K using reflected shocks. The kinetics of HNCO decomposition was followed by its IR emission at 5 pm and NH(32;-) absorption at 336 nm employing a narrow line width CW dye laser. Kinetic modeling of these data yielded
k, = 9.84
X
10l5exp(-43000/T)
cm3/(mol.s)
and several other rate constants important to the high-temperature N H reactiow6 Although the absolute values of k l obtained by Fueno and co-workers do not differ significantly from those given above, the A factors derived by these two groups vary by as much as a factor of 17. In this work, we focus our attention to the production of C O and H atoms by IR and vacuum-UV resonance absorption spectroscopy, respectively. CO, similar to NH, is a primary HNCO decomposition product. Similarly, the H atom may also be formed by the alternative, primary unimolecular decomposition channel HNCO
+M
-
H
+ NCO + M
(2)
as well as by other secondary dissociation processes involving NH2 and NH.5f’ The measured kinetic data are used to test the validity of the mechanism and several key rate constants derived by previous workers through detailed kinetic modeling and sensitivity analysis. These results are presented herein. Experimental Section
Two different shock tube systems located at the Naval Research Laboratory (NRL) and the Ballistic Research Laboratory (BRL) were employed in the present study. The shock tube-CW CO laser resonance absorption apparatus at NRL has been described in detail The 6.3-cm NRL stainless steel shock tube, which could be readily diffusion-pumped down to 10” Torr prior (6) Mertens, J. D.;Chang, A. Y.; Hanson, R. K.; Bowman, C. T. Int. J . Chem. Kine!. 1989, 2 l . 1049.
(7) Hsu, D.S.Y.;Shaub, W. M.;Blackburn, M.; Lin, M. C. Symp. ( I n r . ) Combust. [ f r o c . ] , 19th 1982, 89. (8) Hsu, D. S. Y . ; Lin, M. C. J . Energ. Mater. 1985, 3, 95. Also see: S P I E Proc. 1984, 482, 79.
This article not subject to U S . Copyright. Published 1990 by the American Chemical Society
Kinetics of C O and H Atom Production
The Journal of Physical Chemistry, Vol. 94, No. 8, I990
TABLE I: Initial Conditions of Shocked Mixtures Used To Evaluate Values of k , and k2*
mixtureb
T/K
P/atm
A A A A A A A A A A A
2123 2147 2166 2172 21 79 223 1 2279 2416 2436 2470 2523 2124 2254 231 1 2322 2335 2374 2409 2416 2467 2484 2502 2505 2519 2523 2534 2556 2622 241 7 2486 2568
0.5 14 0.523 0.491 0.352 0.43 1 0.458 0.472 0.394 0.369 0.348 0.354 0.533 0.438 0.430 0.329 0.412 0.405 0.386 0.394 0.357 0.364 0.362 0.365 0.363 0.329 0.365 0.567 0.534 0.390 0.345 0.603
B B B B B B B B B B B B B B B C C D D D
log k, 7.35 7.40 7.45 7.47 7.52 7.65 7.70 8.15 8.30 8.40 8.62 7.35 7.75 7.88 7.92
4
-
1
HNW
- ..
A
I 6
.
8.00 8.1 1 8.25 8.19 8.40 8.51 8.45 8.48 8.45 8.60 8.58
e
4 s
f TIm-
l a 0 PPU HNCO
I
I
7.38 7.72 6.86 7.29 7.46
-
-
(9) Wu,C. H.; Fifer, R. A. To be published. (IO) Kee, R. J.; Miller, J. A.; Jefferson, T. H. CHEMKIN: A GeneralPurpose, Problem-Independent, Transportable, Fortran-Chemical Kinetics Code Package. Report SAND 80-8003; Sandia National Laboratories:
Albuquerque, NM, 1980. ( I I ) Lutt, A. E.; Kee, R. J.; Miller, J. A. SENKIN: A Fortran Program for Redicting Gas Phase Chemical Kinetics with Sensitivity Analysis. Report SAND 87-8248: Sandia National Laboratories: Albuquerque, NM, 1988.
Y
7 6
-10‘
34
to each experimental run, has a relatively low leak and degassing rate of 1 X IO4 Torr/min under normal operation conditions. The system is currently interfaced with two frequency-stabilized CW CO lasers which can be simultaneously employed to monitor CO and H 2 0 (or NO) using 2 1 P( 10) and 8 7 P( 15) (or 7 6 P( 13)) lines, respectively.* In this study, CO and H20were monitored; the latter was not defected in those test runs, suggesting that the isomerization/decompositionprocess HNCO HOCN O H + C N is not important. The BRL shock tube, which has a diameter of 5.3 cm, was also constructed with stainless steel. The system can be readily evacuated down to IO-’ Torr with a turbomolecular pump; it has been interfaced with a CW H atom resonance lamp for H atom concentration measurement using the Lyman-a line at 121.6 nm. A detailed discussion of this system and of the H atom concentration calibration procedure will be made el~ewhere.~ Four different mixtures-0.0010, 0.0025, 0.25, and 0.50% in Ar-were employed in our experiments. The first two were utilized in the H atom production study at BRL, whereas the latter two higher concentration mixtures were used in C O formation experiments at NRL. The initial conditions of shocked mixtures were calculated with the NASA/Lewis equilibrium program,’-* and the kinetics of C O and H atom production were computed with the CHEMKIN’O and SENKIN” programs supplied by Sandia National Laboratory. HNCO used in this work was prepared by the acidification of KOCN with stearic acid at about 390 K, followed by drying with
-
0 26Y
a
HNCO in Ar; B, 0.25%; C, 0.010%;D, 0.025%.
-
s
log k,
and k, are in units of cm3/(mol.s) evaluated from kinetic modeling of CO and H atom production profiles, respectively, using the mechanism summarized in Table 11, which was adopted from ref 6. bThe compositions of shocked mixtures are as follows: A, 0.50%
-
/
3345
T-
I
l
/
e
7-4
266
I 25
2.6
3 76
6
T I U E -10‘
Figure 1. CO and H atom production profiles at different temperatures. (A) CO produced from the pyrolysis of the 0.25% HNCO-Ar mixture at 2124, 2335, and 2534 K. (B) H atoms produced from the pyrolysis of the 0.010% HNCO-Ar mixture at 2337 and 2556 K. Points are experimental data, and curves are values calculated by using the mcchanism listed in Table 11.
P205and subsequent vacuum distillation using different slush baths. An FTIR analysis of the purified sample indicated no absorption by impurities such as CO or C02. Ultrapure Ar (99.9995%, Matheson) was used for preparation of different mixtures, usually many hours before shock experiments. For the H atom kinetic experiment carried out a t BRL, the Ar used for mixture preparation had been purified with a molecular sieve trap. Results The initial conditions for four shocked mixtures are summarized in Table I. Typical C O and H atom production profiles are compared with kinetically modeled values in Figure IA,B. In our kinetic modeling, the values of the rate constants for the two initial decomposition processes
+ Ar N H + C O + Ar -,H + NCO + Ar
HNCO
+
(1)
(2)
were varied to fit observed CO and H atom profiles, with others kept unchanged from the adopted values from different sources as referenced in Table 11. Our initial modeling was carried out with the mechanism employed by Hanson and co-workers6 which included 43 elementary reactions. The result of our sensitivity analysis indicated that many of the reactions, such as NCO NCO and N HNCO, are unimportant under our experimental conditions. They are therefore excluded in our reaction scheme presented in Table 11. Since several secondary reactions involving N H and NH2 (e.g., reactions 8, 12, and 17) do contribute significantly to observed H atom signals, particularly at longer reaction times, we only modeled k2 with initial concentrations of H atoms obtained at high temperatures (T> 2400 K). The values
+
+
Wu et al.
3346 The Journal of Physical Chemistry, Vol. 94, No. 8, 1990 TABLE 11: Mechanism and Associated Rate Constants Used in the Kinetic Modeling of CO and H Atom Production from the Thermal Decomposition of HNCO" reaction A n E. I . H N C O + M = N H C O + M 2.63E1Sb 0.0 79.1 2. HNCO M = H + NCO + M 1.01E17 0.0 112.0 2.OE13 3 . NH HNCO = NH2 CO 0.0 23.8 l.lE14 4. H HNCO = NH2 CO 0.0 12.7 5. NH2 HNCO = NH3 + NCO 7.6El2 0.0 7.0 4.OEl3 6. N HNCO = N H NCO 0.0 35.8 8.51El2 7. NCO + H2 = HNCO + H 0.0 9.0 4.47E13 0.0 0.0 8 . 2 N H = N2 + 2H 6.3 1 El 1 9. NH N = N 2 H 0.5 0.0 IO. NH3 + M = NH2 + H + M 2.51816 0.0 96.0 1 1 . H 2 + M = 2H M 0.0 96.0 2.19E14 12. NzH, M = N2H2 H M l.OE16 0.0 49.7 l.OE16 13. N2H3 + M = NH2 + N H + M 0.0 41.7 14. N2H M = N2 H M 0.0 3.0 I.OE14 l.OE16 15. N2H2 + M = NIH + H M 0.0 49.7 2.57El4 16. NH M = N H M 0.0 75.5 3.16E13 17.NH + N H ; , = N , H I + H 0.0 1.0 3.16E13 18. N H 2 + M = N H H M -2.0 91.4 7.24E13 19. N H 2 + N = N, 2H 0.0 0.0 1.00E14 20. NH H = N H2 0.0 0.0 1.91 E13 21. NH2 + H = N H + H2 0.0 0.0 6.31816 22. NCO + M = N CO M -0.5 48.0 2.OE14 23. NCO + N = N2 + CO 0.0 0.0 1.05E14 24. NCO H = NH CO 0.0 2.0
+ + + +
+ + + +
+
.s
1
l A
+
+
+
I
+ +
/
+
+ +
+ + + + + + + + + + + + + +
"All rate constants, except k , and k2, are taken from ref 6. The rate constants, which are represented by the form k = A T exp(-E, RT), are in units of cm3, mol, s with activation energies in kcal/mol. /Read as 2.63 X I O t 5 .
D
w
0
I
I
I
I
100
E00
300
480
500
400
50-
T I N E Curet>
1 5 -
6
6
1 -
RN I
2 Y
t n
Y
5 2 9
0 5-
0
0 -
I 0
100
200
300
TIME Cuiro3
Figure 3. Sensitivity coefficients for CO and H atom production illustrating the effects various key processes. Mixture: 0.010% HNCO in Ar. T = 2350 K; P = 0.40 atm.
Discussion As alluded to in the preceding section, the measured CO and H atom production profiles can be quantitatively accounted for with the mechanism and associated rate constants presented in Table 11, by only adjusting k , and k2, respectively. All other rate constants were adopted from the recent work of Hanson, Bowman, and co-workers6 without further adjustment. The values of the two rate constants can be effectively represented by the Arrhenius equations k , = 1015,41 exp(-39800/T)
cm3/(mol.s)
(1)
k 2 zz 101'.Oexp(-56400/T)
cm3/(mol-s)
(11)
5~
3 5
3 9
4 3 1/T
4 7
x
5 1
5 5
I0E4
Figure 2. Arrhenius plots of k , and k2: squares, k , determined in this work; triangles, k , determined by Hanson and co-workers (ref 6); dotted line, k l reported by Fueno and co-workers (ref 5); diamonds, k , determined in this work. Solid lines represent least-squares values.
of k , and k2 thus obtained are summarized in Table I as well as in Figure 2. Least-squares analyses of these values lead to k, =
1015.4f0.16
k2 z
exp(-39800 f 700/7') cm3/(mol.s) ( I ) exp(-56400/T)
cm3/(mol.s)
(11)
In Figure 2, we have also compared our k , data with those reported by F ~ e n o Hanson? ,~ and co-workers. Discussion of these results will be made in the following section.
As shown in Figure 2, the absolute values of our kl obtained by kinetic modeling of CO agree quantitatively with those of Hanson and co-workers? although the Arrhenius parameters presented above differ somewhat. Because of the narrower temperature range employed in the present study, we have combined our data with those of Hanson et al.? covering the temperature range of 1830-3340 K. A least-squares analysis of the combined data leads to k I -- 1015.925f0.085 exp(-42640 f 450/T) cm3/(mol-s) (111)
This expression is recommended for future kinetic modelings. There has been no prior determination for kt. The approximate expression for the rate constant given by eq I1 appears to be reasonable and consistent with the large bond energy, D(H-NCO) = 113 k~al/mol.~ Because of the high bond strength and the rapid appearance of secondary H atoms from N H and NH2 reactions
3347
J . Phys. Chem. 1990, 94, 3347-3352 (see Table I]), an accurate determination of k2 becomes difficult. The values of k2 presented in Table I were obtained by modeling the profiles of H atom production during the early stage of HNCO decomposition at higher temperatures at which [HI depends most strongly on k2. The results of a sensitivity analysis carried out for 100 ppm HNCO shocked at T = 2350 K and P = 0.40 atm reveal, for example, the sensitivity coefficient for the H atom ] ) rapidly from 0.96 at t = defined by ( a [ H ] / a k 2 ) ( k 2 / [ Hdrops 1 ps to 0.34 at t = 200 ps (see Figure 3A). On the other hand, for CO, its sensitivity coefficient (t3[CO]/dkl)(kl/[CO]) remains essentially constant at 1.0-0.92, suggesting that reaction 1 is the predominant source of CO. No other reactions were found to affect C O production rate significantly (see Figure 3B). Both reactions I and 2 are assumed to be effectively in the second-order region. For reaction 1, the result of Hanson et aL,6 covering a broader range of pressure (0.33-2.2 atm), indicates that the second-order rate constant k l is essentially pressure independent. A brief RRKM calculation carried out by assuming a tight activation complex for reacton 1, which involves a singlet-triplet transition with a barrier of about 85 kcal/mol (comparing with D(HN-CO) = 8 1 kcal/mo15), suggests that the decomposition would be fully second order if the transition coefficient is near unity. If the transmission coefficient is as small as that of the isoelectronic reaction C 0 2 O('P) + CO, K IC3,then reaction 1 would be about 60-70% into the falloff region. Our present results and particularly those of Hanson and co-workers6 suggest that the transmission coefficient for reaction 1 is probably not
-
-
much less than unity. A similar calculation for reaction 2, assuming a semirigid transition state with a high-pressure A factor of about IOI5 s-I, indicates that the decomposition reaction is fully in the second-order region despite the much larger reaction barrier.
Conclusion We have investigated the kinetics of the thermal decomposition of HNCO in shock waves at temperatures between 21 20 and 2570 K by monitoring the production of C O and H atoms. Kinetic modeling of measured product formation profiles provided the second-order rate constants for the two decomposition processes HNCO
-.
+ Ar N H + CO + Ar H + NCO + Ar
+
(1)
(2)
The modeled values of k l , which are much greater than those of k2, agree quantitatively with the ones reported by Hanson, Bowman, and co-workers6 using entirely different product diagnostics. Combination of both sets of data for k l gave rise to the expression
k, = 1015~925*0.085 exp(-42640 f 450/T)
cm3/(mol.s)
covering the temperature range of 1830-3340 K. Acknowledgment. C. H. Wu is grateful to the Office of Naval Technology for an O N T postdoctoral fellowship. M . C. Lin acknowledges the Office of Naval Research for the support under Contract No. NO001 4-89-5- 1949. Registry No. HNCO, 420-05-3.
High-Temperature Fast-Flow Reactor Kinetics Studies of the Reactions of Ai0 with C12 and HCI over Wide Temperature Ranges Aleksandar G. Slavejkov, Clyde T. Stanton,+and Arthur Fontijn* High- Temperature Reaction Kinetics Laboratory, Department of Chemical Engineering, Rensselaer Polytechnic Institute, Troy, New York 12180-3590 (Received: October 6, 1989: In Final Form: January 3, 1990}
The kinetics of the title reactions have been studied in a high-temperature fast-flow reactor (HTFFR). The relative concentrations of AI0 were monitored by laser-inducedfluorescence at the B2Z-X2L: and C2Z-X2Z transitions. The following k( T)expressions in cm3 molecule-' s-' are obtained: AI0 + CI2 ( I ) , k , ( T ) = 3.0 X exp(-1250 K I T ) between 460 and 1160 K; A10 + HCI (2), k2(T) = 5.6 X IO-" exp(-139 K/T) between 440 and 1590 K. Confidence limits are given in the text. No fluorescence from a potential four-center product AlCl could be detected. From this it is estimated that less than 5% of the AI0 reacted produced AICI, which indicates that abstraction reactions dominate, Le., OAlCl + CI for reaction 1 and OAlCl + H and/or AlOH + CI for reaction 2.
1. Introduction Reactions of A1 species play a role in several high-temperature environments ranging from rocket exhausts' to dust explosions2 and circumstellar envelopes? To provide a data base for modeling such environments and to guide the development of theory for metallic species reactions, we are engaged in an extensive survey of oxidation reactions of ground-state AI atoms, AlCl (XlZ), and A I 0 (X2Z).4-10 Here we report on the first measurements of reactions between A I 0 (X2Z) and non-oxygen oxidizers. The reactions studied and their thermochemically" accessible pathways are A I 0 + CI2 OAICI + C1 AH0298~= -294 f 28 kJ mol-' ( l a )
-
-
AlCl
+ OCI
AH'298~ = -17 f 16 kJ mol-' (Ib)
'Present address: Naval Research Laboratory, Code 61 10, Washington,
DC 20375-5000.
AI0
+ HC1-
-
-
OAlCl
+ C1 AlCl + O H AlOH
+H
AH0298K= -105 f 2 8 kJ mol-' (2a)
AH0298~ = -33 f 21 kJ mol-' (2b) A H o 2 g 8=~ +I3 f 16 kJ mol-' (2c)
2. Technique The basic HTFFR technique has previously been extensively (1) Park, C. Atmos. Environ. 1976, 10,693. ( 2 ) Ogle, R. A.; Beddow, J. K.;Chen, L. D.;Butler, P. B. Combusr. Sci. Technol. 1988, 61, 15. ( 3 ) Cernichard, J.; Guelin, M. Asrron. Asrrophys. 1987, 183, L10. (4) Rogowski, D.F.; Marshall, P.; Fontijn, A. J. Phys. Chem. 1989, 93. 1 1 18. ( 5 ) Fontijn, A. Specrrochim. Acta 1988, 438, 1075. (6)Fontijn, A. Combust. Sci. Technol. 1986, 50, 151. (7)Rogowski. D.F.; Fontijn, A. Symp. (lnr.) Combust. [Proc.] 2lsr 1988,
943.
0022-3654/90/2094-3347$02.50/0
0 1990 American Chemical Society
