Theoretical study of the CH3 + NS and related reactions: mechanism

1 downloads 0 Views 276KB Size Report
turns out to be the most stable isomer followed by thionitrone (2), a three-membered ring, thionitrosyl .... singlet nitrene but can better be regarded as a species.
MOLECULAR PHYSICS, 1999, VOL. 96, NO. 12, 1817± 1822

RESEARCH NOTE Theoretical study of the CH3 + NS and related reactions: mechanism of HCN formation TRUNG NGOC LE1,2 , LOC THANH NGUYEN1,3 1 and MINH THO NGUYEN * 1 Department of Chemistry, University of Leuven, Celestijnenlaan 200F, B-3001 Leuven, Belgium 2 Faculty of Chemistry, University of Danang, Vietnam 3 Faculty of Chemical Engineering, HoChiMinh City University of Technology, Vietnam (Received 16 October 1998, revised version accepted 19 February 1999) More than thirty equilibrium and transition structures on the [CH3 NS] potential energy surface have been located using the B3LYP/6-311+ + G(d,p) method. Thioformaldoxime ( 3) turns out to be the most stable isomer followed by thionitrone ( 2), a three-membered ring, thionitrosyl methane ( 1) and thiazyl methane. These isomers are connected to each other by 1,2H and 1,3H shifts and ring± chain rearrangement, but the associated energy barriers are rather high, making most of them stable with respect to unimolecular transformations. Starting from CH3 + NS, a possible initial atmospheric reaction, HCN formation appears to be the most favoured process through a cascade involvement of 1, 2 and 3. The standard heats of formation, H0f,298 , are calculated to be: 3, 149 kJ mol­ 1 ; and 1, 218 kJ mol­ 1 ; using the CCSD(T)/6-311+ + G(3df,2p) method, with an error of 10 kJ mol­ 1 .

The presence of reduced sulphur compounds in the Earth’ s atmosphere has led to considerable interest in the study of their chemistry [1]. However, compared with the atmospheric cycles for C, N and O, the S cycle remains less well understood [2]. While the chemistry of nitric oxide (NO) has been studied extensively both experimentally (see, e.g., [3, 4]) and theoretically (e.g. [5± 8]), little is known about that of its isovalent sulphur analogue. The fact that the NS radical was produced upon the reaction of nitrogen, sulphur and argon vapour in microwave discharge [9] suggests its probable existence in combustion and/or atmospheric media having high concentrations of nitrogen and sulphur species. If it exists, NS could then undergo a variety of fast reactions with small organic species such as hydrocarbon radicals present in those media. In an attempt to probe the largely unknown reactivity of NS, we have carried out quantum chemical calculations on its reaction with a simple radical, namely the methyl. It is now established that, in the analogous CH3 + NO reaction, HCN+ H2 O are the penultimate products * Author for correspondence. chem.kuleuven.ac.be

e-mail:

minh.nguyen@

[10, 11]. Therefore, a question of interest is whether HCN formation remains the most favoured channel of the CH3 + NS reaction. In the present note we have focused mainly on the essential features of the [CH3 NS] potential energy surface. Given the large number of intermediates and channels, we have opted for density functional theory (DFT) with the popular hybrid three-parameter non-local exchange and correlation functions, B3LYP [12± 13]. The one-electron 6-311+ + G(d, p) basis set [14] was used for all calculations. For each stationary point, a vibrational analysis has been performed to determine its nature and to estimate its zero-point energy (ZPE). When necessary, IRC calculations have been performed to ascertain the identity of the transition structure under consideration. All computations were carried out with the aid of the Gaussian 94 set of programs [15]. Throughout this paper, bond lengths are given in A, bond angles in deg, total energies in Eh , and zeropoint vibrational (ZPE) and relative energies, unless 1 otherwise noted, in kJ mol­ . The main structural and energetic results obtained from B3LYP/6-311+ + G(d,p) calculations are summarized in ® gures 1± 5. Figure 1 displays selected geometrical parameters of the equilbrium structures found on the

Molecular Physics ISSN 0026± 8976 print/ISSN 1362± 3028 online Ñ 1999 Taylor & Francis Ltd http://www.tandf.co.uk/JNLS/mph.htm http://www.taylorandfrancis.com/JNLS/mph.htm

1818

Figure 1.

Research note

Selected B3LYP/6-311+ + G(d,p) geometries of the equilibrium structures considered.

[CH3 NS] potential energy surface, enumerated from 1 to 10. The ® ve fragmentation products labelled F1± F5 are omitted. Figure 2 gives selected geometrical parameters of the transition structures (TS). The adopted convention X/Y corresponds to a TS connecting the two equilibrium structures X and Y. Figure 3 displays the energy pro® les corresponding to the possible reaction channels starting by an initial carbon± nitrogen recombination. Similarly, ® gure 4 shows the energy pathways starting from an alternative initial combination of the reactants. Finally, ® gure 5 illustrates the dominant reaction behav-

Figure 2. Selected B3LYP/6-311+ + G(d, p) geometries of the transition structures X/Y linking di€ erent X and Y equilibrium structures.

iour by including only the energetically lowest lying paths from both initial recombinations. The fragments F1± F5 are de® ned as follows: F1, CH3 + NS; F2, CH2 N+ SH; F3, CH2 + HNS; F4, CH2 + HSN; and F5, HCN+ H2 S. 2 Isovalent with NO, the NS radical exhibits an X p electronic ground state, followed by at least two quartet states a 4p and b 4 ­ [16]. Due to the presence of S, combination of NS with a radical centre R becomes possible at both N and S ends, giving rise as a consequence to two distinct isomers. Thus the following

1819

Research note

Figure 3. Schematic energy pro® les showing the rearrangements starting from an initial carbon± nitrogen recombination of H3 C and NS. Values at B3LYP/6-311+ + G(d,p)+ ZPE. The total energy of CH3 Ð NÐÐ S ( 1) is: 492.840 18 Eh .

-

discussion is divided according to these initial steps. As seen in ® gure 3, a large number for isomers having not only the open CNS skeleton but also three-membered rings, are connected to each other in this carbon± nitrogen combination channel. We have not considered the isomers having the NCS skeleton such as thioformamide (H2 NÐ C(ÐÐ S)H) and derivatives. Recent surveys [17± 20] of the performance of various functionals for the study of sulphur containing compounds showed that the geometries and energies predicted by available DFT methods are not better than the MP2 results. Thus, the bond distances to sulphur derived from B3LYP functionals are signi® cantly and quasi-systematically longer than the MP2 counterparts [17± 20]. In order to have an evaluation of B3LYP relative energies, coupled-cluster calculations, CCSD(T)/6-311+ + G(d,p), have been carried out also for a subset of structures. We note

that, with respect to the CCSD(T) values, the B3LYP method tends to overestimate the energy gaps between equilibrium structures relative to CH4 NÐÐ S ( 1); 1 the deviations amount up to 20 kJ mol­ , in particular for transition structures and fragments. Nevertheless, the energy ordering remains almost unchanged. A similar trend has been observed also for the analogous [NH2 NS] system [20]. In this context, the accuracy of B3LYP values for either geometries or energies is only semi-quantitative. Therefore, we do not put much importance on the quantitative aspect of the energy surface, but rather on a qualitative description of the associated reaction mechanism. Although a detailed examination of the pathways and intermediates shown in ® gures 1 and 2 is beyond the purpose of this note, a few points of general and qualitative character could be made from the energy pro® les depicted in ® gures 3 and 4.

1820

Research note

Figure 4.

Schematic energy pro® les showing the rearrangements starting from an initial carbon± sulphur association of H3 C and NS. Values at B3LYP/6-311+ + G(d,p)+ ZPE.

Figure 5.

Schematic energy pro® les showing the global and lower lying connection between the H3 C and NS radicals. Values at B3LYP/6-311+ + G(d, p)+ ZPE.

1821

Research note Except for the two fragments F3 and F4, all isomeric equilibrium structures lie lower in energy than, or have at least comparable energetic content with, the starting fragments F1. Among the equilibrium structures 1± 4 having CNS connectivity, thioformaldoxime ( 3) turns out to be the most stable form. Thionitrone ( 2) is even more stable than 1, comparable with the situation in the [CH3 NO] analogues [11]. The fragments F5 containing HCN and H2 S are beyond any doubt the lowest-lying stationary point on the energy surface. As expected, the initial recombination F1 ! 1 is barrier-free. Subsequent conversion of 1 to other isomers could in principle be achieved via direct or multi-step rearrangements. Both one-motion-of-the-nuclei paths, such as 1,2H and 1,3H shif ts, and two-motion paths, such as a 1,2H shift accompanied by a ring-closure, have been identi® ed. Three distinct three-membered rings do exist as local minima among which thia-aziridine ( 5) is quite a stable isomer. Figure 4 indicates that the processes resulting from the alternative carbon± sulphur recombination are much simpler. As a matter of fact, only isomers 8, 9 and 10 bearing the CÐ SÐÐ N skeleton have been located. The thiazyl form 10 having a formal triple SÐÐ N bond is less stable than 8 (having two double SÐÐ N bonds), but far more stable than 9. The latter is not really a singlet nitrene but can better be regarded as a species having hexavalent S atom. A TS for pure ring closure from 9 giving the cycle 6 could not be found. Overall, it seems that only the initial step F1 ! 10 is open in this recombination mode. The global energy pro® le depicted in ® gure 5 suggests two additional important points. (i) There is no other TS formally connecting 1 to 10 than the reactants F1. Therefore, the three-membered ring 5 can be regarded as a possible link between both CNS and CSN skeletons. (ii) Starting from F1, the most favoured channels in which all the intermediate points have lower energy are found to be: F1 ! 1 ! 2 ! 3 ! F5 (HCN+ SH2 ), F1 ! 1 ! 2 ! 5 (ring) and F1 ! 1 ! 2 ! 3 ! F2 (H2 CÐÐ N+ SH). Another important observation is that formation of the ring 5 is associated with low energy barriers but its subsequent unimolecular rearrangements are energetically more demanding. In this sense, the channel going through 5 is not crucial for the disappearance of the reactants F1. Of the two remaining channels, namely formation of hydrogen cyanide F5 and the radical H2 CH F2, both open from thioformaldoxime 3, obviously the former is preferred over the latter. In any case, the radical H2 CN, once formed, is expected rapidly to lose a hydrogen atom leading also to HCN [21].

In view of the scarcity of thermochemical information on the [CH3 NS] isomers, we wish to take this opportunity to evaluate their standard enthalpies of formation 0 ( Hf ,298 ). For this purpose, higher level coupled-cluster CCSD(T) methods of molecular orbital theory were performed in conjunction with the larger 6311+ + G(3df,2p) one-electron basis functions. Geometry and zero-point energies and thermal corrections were obtained from MP2/6-311+ + G(d, p) computa0 tions. To evaluate Hf , the heats of the following working reactions [22] have been computed using CCSD(T) energies: H2 CÐÐ NH + H2 S,

( 1)

H2 CÐÐ NÐ OH + H2 S,

( 2)

H3 CÐÐ NÐ S ( 1) !

HCÐÐ N + H2 S,

H3 CÐ NÐÐ S ( 1) + H2 !

CH4 + HNÐÐ S.

( 3) ( 4)

H2 CÐÐ NÐ SH 3 + H2 ! H2 CÐÐ NÐ SH 3 + H2 O !

The heats of formation of the reference species are taken 1 as follows (values at 298 K, kJ mol­ ): H2 CÐÐ NÐ H, 88.0 [23± 25]; H2 S, 20.5 [26]; HCN, 135.1 [26]; CH4 , 74.9 [26], H2 O: 241.8 [26], HNÐÐ S, 237.5 [27, 28]; and H2 CÐÐ NÐ OH, 14.0 [28, 29]. The heats of formation of H2 CÐÐ NÐ SH ( 3) are cal1 culated to be 148.8 kJ mol­ from equation (1) and ­1 149.1 kJ mol from equation (2). The corresponding 1 values for H3 CÐ NÐÐ S ( 1), are 220.6 kJ mol­ from ­1 equation (3) and 215.7 kJ mol from equation (4). Taking the average of each pair as the best estimate, the values H0f,298 (H2 CÐÐ NÐ SH)= 149 kJ mol­ 1 and 0 1 Hf ,298 ( H3 CÐ NÐÐ S) = 218 kJ mol­ can be proposed. Earlier evaluation of the calculated heats of formation [7, 11, 22, 25, 28] using di€ erent methods indicated that, at the CCSD(T) or a comparable level, in conjunction with large basis sets, the errors encountered in the com1 puted values amount to, at most, 10 kJ mol­ with respect to the accurate experimental values. In summary, the most signi® cant chemical result of the present theoretical study is a prediction on the preferential formation of HCN in the H3 C+ NS reaction. Calculations show the complex nature of the [CH3 NS] potential energy surface involving multiple channels and multi-step routes leading to HCN with thionitrosyl methane, thionitrone and thioformaldoxime as low energy intermediates.

-

-

The authors are indebted to the Flemish Government and the KULeuven Quantum Chemistry group for supporting an `Inter-university Program for Education in Computational Chemistry in Vietnam’. M.T.N. thanks the FWO-Vlaanderen and GAO-program for continuing support.

1822

Research note References

[1] Wilson, C., and Hirst, D. M., 1996, Progr. Reaction Kinetics, 21, 69. [2] Wayne, R. P., 1991, Chemistry of Atmospheres, 2nd Edn (Oxford University Press) p. 26. [3] Peeters, J., van Look , H., and Ceursters, B., 1996, J. phys. Chem., 100, 15 124. [4] Brownsword , R. A., and Hancock , G., 1997, J. chem. Soc. Faraday Trans., 93, 1279. [5] R oggenbuck , J., and Temps, F., 1998, Chem. Phys. L ett., 285, 422. [6] Sengupta, D., Peeters, J., and Nguyen, M. T., 1998, Chem. Phys. L ett., 283, 91. [7] Sumathi, R., and Ngytyen, M. T., 1998, J. phys. Chem. A, 102, 3175. [8] Bunte, S. W., R ice, B. M., and Chabalowshi, C. F., 1997, J. phys. Chem. A, 101, 9430. [9] Hassanz adeh, P., and A ndrews, L., 1992, J. Amer. chem. Soc., 114, 83. [10] Lifshiftz , A., Tamburu , C., Frank, P., and Just, J., 1993, J. phys. Chem., 97, 4085. [11] Nguyen, M. T., Sengupta , D., and Vanquickenborne, L. G., 1996, J. phys. Chem., 100, 10 956. [12] Becke, A. D., 1993, J. chem. Phys., 98, 5648. [13] Lee, C., Y ang, W., and Parr , R. G., 1988, Phys. Rev. B, 37, 785. [14] Frisch, M. J., Pople, J. A., and Binkley, J. S., 1984, J. chem. Phys., 80, 3265. [15] Frisch, M. J., Trucks, G. W., Schlegel , H. B., Gill , P. M. W., Johnson, B. G., A l -Laham, M. A., Zakrewski, V. G., Ortiz , J. V., Foresman, J. B., Cioslowski, J., Stefanov, B. B., Nanayakkara , A., Challacombe, M., Peng , C. Y., A yala , P. Y., Chen,

[16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

W., Wong, M. W., A ndres, J. L., R eplogle, E. S., Gomperts, R., Martin, R. L., Fox , D. J., Binkley, J. S., Defrees, D. J., Stewart, Head -Gordon, M., Gonzalez , C., and Pople, J. A., 1995, Gaussian 94 (Pittsburgh, PA: Gaussian, Inc.). Lie, G. C., Peyerimhoff, S. D., and Bunker, R. J., 1985, J. chem. Phys., 82, 2672. A ltmann, J. A., Handy, N. C., and Ingamells, V. E., 1996, Intl. J. Quantum Chem., 57, 533. Somasundram, K., and Handy, N. C., 1996, J. phys. Chem., 100, 17 485. A ltmann, J. A., Handy, N. C., and Ingamells, V. E., 1997, Molec. Phys., 92, 339. Nguyen, L. T., Le, T. N., and Nguyen, M. T., 1998, J. chem. Soc. Faraday Trans, 94, 3541. Nguyen, M. T., Sengupta, D., and Ha , T. K., 1996, J. phys. Chem., 100, 6499. Sana , M., and Nguyen, M. T., 1992, Chem. Phys. L ett., 196, 390. Smith, B. J., Pople, J. A., Curtius, L. A., and R adom, L., 1992, Aust. J. Chem., 45, 285. Holmes, J. L., Lossing, F. P., and Mayer, P. M., 1992, Chem. Phys. L ett., 198, 211. Nguyen, M. T., R ademakers, J., and Martin, J. M. L., 1994, Chem. Phys. L ett., 221, 149. Lias, S. G., Bartmess, J. E., Liebmann, J. F., Holmes, J. L., Levine, R. D., and Malland, W. G., 1988, J. phys. Chem. Ref. Data, 17, Suppl. 1. Nguyen, M. T., Vanquickenborne, L. G., and Flammang , R., 1994, J. chem. Phys., 101, 4885. Mai, H. T., and Nguyen, M. T., unpublished. Nguyen, M. T., R aspoet, G., and Vanquickenborne, L. G., 1995, J. chem. Soc. Perkin Trans. ii, 1791.