Mendeleev Communications Electronic Version, Issue 1, 2002 1 Dichlorosilane chlorination in the presence of propylene as an inhibitor at low pressures and 293 K Nikolai M. Rubtsov,* Georgii I. Tsvetkov and Victor I. Chernysh Institute for Structural Macrokinetics and Materials Science, Russian Academy of Sciences, 142432 Chernogolovka, Moscow Region, Russian Federation. Fax: +7 095 962 8025; e-mail: ab3590@mail.sitek.ru 10.1070/MC2002v012n01ABEH001537 The strong inhibiting effect of C3H6 on the self-ignition of dichlorosilane–chlorine mixtures points to silylenes as chain carriers of the reaction.A critical concentration of chemically inert SF6 (> 45%) drastically improves the inhibiting action of C3H6 indicating the occurrence of energy branching. The chlorination of inorganic hydrides [monosilane,1 dichlorosilane (DCS),2 phosphine,3 germane4] and hydrocarbons,5 the reaction of monosilane with nitrogen trichloride6 and the thermal decomposition of NCl3 7 show special features of branched-chain processes (BCP).These are the existence of self-ignition isothermal limits and their dependence on the state of contacting surfaces, as well as the sensitivity of the kinetics to small additives of promoters and inhibitors.8,9 The energy branching (EB, the participation of excited intermediates in chain branching) plays an important role in chain chlorination because the material branching, e.g., ·H + O2 ® ·OH + ·O·, is evidently ruled out.EB in the reactions of electronically excited particles was established for NCl3 decomposition.7 EB in the reactions of vibrationally excited particles was detected in the fluorination of H2 and fluorinated hydrocarbons.11 In fact, fluorination reactions differ from those of chlorination in that they are more exothermic.Moreover, HF has a higher first vibrational level than that of HCl. In fluorination, EB occurs if the energy released in a certain step is enough for unimolecular decomposition of the product of this step, e.g., for the CH2F2 + F2 reaction the following branching process is suggested: CF2H + F2 ® CHF3 * + ·F; CHF3 * ® :CF2 + HF.11 Related reactions were proposed for 1,1- difluoroethane5 and SiH4 1 chlorination.In these mechanisms, carbenes and silylenes, respectively, participate in the EB process. Ab initio calculations were used to construct a mechanism for the pyrolysis of chlorinated silanes.12 The insertion reactions of chlorinated silylenes into chlorinated silanes, which yield chlorinated disilanes, were examined.It was assumed that DCS decomposition could be accelerated by a chlorosilylene-catalysed cycle including EB [its energetics is shown in Figure 1 (ref. 12)]: It is known that olefins react very rapidly with silylenes.13 The rate constants vary from 10–10 (for SiH2) to 10–11 cm3 s–1 (for SiHCl and SiCl2).Since SiHCl (A1B1–X1A1) was detected in DCS chlorination,2 the inhibiting action of olefins (e.g., propylene) will indicate that the reactions of silylenes as chain carriers are of considerable importance in the BCP. If inhibition takes place, the uniform self-ignition over the reactor volume can be studied because the self-ignition of pure DCS and Cl2 at 293 K occurs immediately after mixing (induction period ti is < 300 ms).2 This work is aimed at the establishment of the important role of silylenes as chain carriers, as well as energy factors, in the chain branching in BCP of DCS chlorination in the presence of propylene as an inhibitor and chemically inert sulfur hexafluoride (SF6) as a deactivator.The method of ignition delay was used.14,15 The basis for this method is an expansion of self-ignition area in the time of consumption of an inhibitor. The dependence of self-ignition limits on the fraction of an inhibitor is shown in Figure 1.16 As is seen, a combustible mixture enters the self-ignition area when the inhibitor concentration attains a certain critical value zcrit.The time tC it takes for the mixture to enter the self-ignition area and then to ignite is combined from the time t required to reach the boundary of the self-ignition area (ignition delay) and induction period ti, that is, tC = t + ti. Subsequent calculations hold for t >> ti, or tC @ t. In case of the participation of excited intermediates in chain branching, the addition of an effective deactivator will provide an additional pathway of chain termination and affect t.Let us demonstrate that the dependence of t on the concentrations of initial reactants in the presence of both an inhibitor and a deactivator differs markedly from that in the presence of a single inhibitor by an example of the generalized kinetic model of BCP with EB.The mechanism of DCS chlorination is unknown; therefore, the qualitative model of BCP must include chain origination, propagation, EB and termination via both an inhibitor (In) and a deactivator (M). Here, Y1 and Y2 are the initial reactants, for instance fuel and an oxidizer; Y0 and Y3 are active centres, Y4 * is the excited species that causes branching via unimolecular decomposition.Since the lower limit of the DCS + Cl2 reaction is low,2 the heterogeneous chain termination is not taken into account.14 The amounts of initial reactants consumed and intermediates formed during ignition delay may be considered as negligibly small.14,15 Therefore, dYi/dt (i = 0,3,4) may be put equal to zero and [Y1]0 = [B] and [Y2]0 = [A].Then, by entering new variables y = Y0/[B], x = Y3/[B], u = Y4/[B], A = Y2/[B], z = [In]/[B], p = [M]/[B] into the corresponding equations for the kinetic mechanism (II), for the stationary concentrations of Y0, Y3 and SiHCl + SiH2Cl2 HCl2SiSiH2Cl*; HCl2SiSiH2Cl* HCl2SiSiCl + H2; HCl2SiSiCl Cl2SiSiHCl SiCl2 + SiHCl; HCl2SiSiH2Cl* deactivation. (I) Y1 2Y0 Y1 + Y3 Y0 + products Y0 + Y2 Y4 * + products Y3 + In chain termination Y4 * 2Y3 Y4 * + M deactivation Y0 + In chain termination (k0) (k1) (k2) (k3); (k4) (k5); (k6). chain origination; chain propagation; chain propagation; linear chain branching (EB); (II) PI PII 0 zcrit Fraction of inhibitor / z Total pressure / P Figure 1 The dependence of self-ignition limits of a combustible mixture on the concentration of an inhibitor (g).16Mendeleev Communications Electronic Version, Issue 1, 2002 2 Y4 we obtain: Equation (3) can be solved with respect to u.Substituting this value in equation (2) and inserting h = k4/(k4 + k5p[B]) we obtain: It is no doubt that parameter h changes from 1 to 0 with an increase in the dimensionless amount of deactivator p.Equation (4) suggests that if in certain experiment p amounts up to the partial pressure enough for h to take the value 0.5 chain branching is completely terminated. Therefore, at h > 0.5, ignition delays are not liable to exist. Let us consider that the initial mixture always contains 8% In with respect to A. Then, [A] = z/0.08. Solving equation (1) and equation (4) for x and y, we derive If the denominator of these latter equations is equal to 0, x and y increase infinitely, i.e., self-ignition occurs.Therefore, zcrit is determined by the equality We divide the equation d[In]/dt = –k6Y0[In] – k3Y3[In], which determines the rate of inhibitor consumption, by [B]2: Let us substitute x and y into equation (5) and integrate equation (5) numerically with respect to z between the limits z and zcrit and with respect to t between 0 and t for obtaining the dependence of t on z and h.The results of numerical calculation are shown in Figure 2(a). The chosen values of the rate constants are close to those of silylenes13 being considered as chain carriers in the process. It can be seen that the linear dependence of t on [In]/[B] is expected, the shape of this dependence is determined significantly by the value of k0.It can be seen in Figure 2(a) (curve 1) that at h < 0.5 there are no ignition delays in the presence of the deactivator in the wide range of the ratios [In]/[B], whereas for h = 1 (no deactivation), i.e., in the absence of EB [Figure 2(a), curve 2] this effect is missing. To simplify the treatment, the possible occurrence of material branching along with EB was ignored.However, there are also no ignition delays at h < 0.25 if material branching, e.g., Y0 + Y3 ® 3Y1 (km ~ k2), is included, i.e., the critical character of the dependence of t on h also occurs. The result was tested by the numerical solution of a nonstationary system of kinetic equations for mechanism (II) using the forth-order Runge–Kutta method.It was shown that there are no ignition delays at h < 0.5: the system does not tend to self-ignite at all. Therefore, the qualitative calculations predict a drastic change in the dependence of t on the amount of a deactivator in the presence of both an inhibitor and the deactivator. The experiments were carried out under static conditions at 293 K and total pressures of 0.27–2 kPa.Two cylindrical quartz reactors (15 cm long, 3 cm in diameter, reactor I and 15 cm long, 12 cm in diameter, reactor II) had inlets for gas evacuation and optical windows. Chlorine was initially allowed to bleed into a reactor up to a necessary pressure (0.07–0.4 kPa). Then, the mixture containing DCS + C3H6 + SF6 or Kr (if necessary) was admitted up to a total pressure of 0.4–2 kPa; t was taken as an interval from this moment up to that of a severe decrease in Cl2 concentration.Mixtures containing SF6 were investigated in parallel with mixtures containing the same amount of Kr instead of SF6. The mixtures of DCS + 3–8% C3H6 + 20–80% SF6 or 20–80% Kr, as well as analogous mixtures without C3H6, were prepared before experiments.The concentration of Cl2 was measured by the absorption at 300 nm using a VM-25 monochromator (Germany), a FEY-39 photomultiplier and a Hg–He lamp. The emission and absorption spectra of self-ignition were recorded with an OSA-500 optical spectra analyser (Germany) with a resolution of 0.2 nm per channel. The required number of scans (1 scan = 500 channels per 32 ms) was stored in a computer.The change in total pressure P during self-ignition in reactor II was measured by a mechanotrone. It was shown that self-ignition occurs immediately after the mixing of pure DCS and Cl2 (see ref. 2). However, the addition of > 4% C3H6 caused ignition delays more than 5 s pointing to the fact that C3H6 acts as an inhibitor, i.e., t >> ti. By this means that the reactants may be considered mixed well; therefore, the uniform self-ignition over the reactor volume occurs in the presence of C3H6.Moreover, the strong inhibiting action of C3H6 suggests that linear chain branching gives rise to short ti rather than chain initiation. The only known very rapid step silylene + C3H6 ® chain termination (10–10–10–11 cm3 s–1) is responsible for this inhibiting action.An example of the simultaneous recording of chemiluminescence (470 nm) and absorption (300 nm) is shown in Figure 3. As can be seen from the absorption curve, two intermediates form in self-ignition. The long-lived species has a structureless spectrum at 270–480 nm; it is observed only at [DCS]/[Cl2] £ 1. There is no evidence for the assignment of this spectrum in the literature. The spectra of the short-lived intermediate are shown in Figure 3.The emission bands are due to SiHCl (A1B1–X1A1)17 and the absorption bands, due to SiCl2.18 The silylenes detected are chain carriers of the BCP because C3H6, which reacts rapidly with silylenes, has a marked inhibiting effect on this BCP. It was shown that t increases linearly with an increase in [C3H6]/[DCS] at constant [Cl2].It was also shown that t does not depend on the total pressure, but it depends on the ratio ([DCS] + [C3H6])/[Cl2]. This dependence is almost linear. Since in each set of experiments [DCS]/[C3H6] = const, t depends only on the ratio between an inhibitor and Cl2 [Figure 2(b)]. The data 2k0/(k1[B]) + x – k2y[A]/k1 – k6yz/k1 = 0 –x + 2k4u/(k1[B]) – k3xz/k1 = 0 k2y[A]/k1 – k4u/(k1[B]) – k5up/k1 = 0 (1) (2) (3) –x + 2k2hy[A]/k1 – k3xz/k1 = 0 (4) x = –4k0(2k1 + k3z)/{z[B](50hk1k2 – 25k1k2 – 25zk2k3 – 2k6k1 – 2zk6k3)} y = –2k0(50hk2 + 25k2 + 2k6)/{[B](50hk1k2 – 25k1k2 – 25zk2k3 – – 2k6k1 – 2zk6k3)} zcrit = k1(50hk2 + 25k2 + 2k6)/[k3(25k2 + 2k6)].dz/dt = –[B](k6yz + k3xz). (5) 140 120 100 80 60 40 20 0 0.4 0.3 0.2 0.1 0 1 0.8 0.6 0.4 0.2 1 2 t/s z h 1 2 3 4 5 6 - Kr (%) 7 200 150 100 50 0 0.5 0.4 0.3 0.2 0.1 0 0 20 40 60 80 SF6 (%) [C3H6]/[Cl2] t/s (a) (b) Figure 2 (a) Calculated dependence of ignition delays t on the concentration of inhibitor z and deactivator h.B = 5×1016 cm–3, k0 = 2×10–3 s–1, k1 = = 10–12 cm3 s–1, k2 = 10–14 cm3 s–1, k3 = 10–12 cm3 s–1, k6 = 10–11 cm3 s–1. (1) generalized model (II), EB occurs; (2) generalized model (II), EB does not occur, h = 1 (no deactivation); (b) experimental dependences of ignition delays t on the ratio between an inhibitor and Cl2 and concentration of SF6 (closed circles) and Kr (open circles) Reactor I.(1) 0.20 kPa Cl2 + mixture of DCS + 8% C3H6; (2) 0.28 kPa Cl2 + mixture of DCS + 8% C3H6; (3) 0.12 kPa Cl2 + mixture of DCS + 8% C3H6; (4) 0.20 kPa Cl2 + mixture of DCS + 6.4% C3H6 + 20% SF6; (5) 0.20 kPa Cl2 + mixture of DCS + 5.4% C3H6 + 32% SF6; (6) 0.20 kPa Cl2 + mixture of DCS + 8% C3H6 + 20%, 50%, 60% Kr; 0.20 kPa Cl2 + mixture of DCS +80% Kr; (7) 0.20 kPa Cl2 + mixture of DCS + 4.5% C3H6 + 45% SF6.Mendeleev Communications Electronic Version, Issue 1, 2002 3 shown in Figure 2(b) were obtained from the time dependence of [Cl2] in the Cl2 + DCS + C3H6 + SF6 (or Kr) system.Typical data are shown in Figure 4. In the presence of 8% C3H6 and 50% SF6, t > 30 min whereas t in the presence of 8% C3H6 and 50% Kr is no longer than 110 s. As is expected, the deactivating action of Kr, if any, is markedly weaker than that of SF6. Actually, as can be seen in Figure 2(b), ignition delays are observed at 80% Kr (open circles), i.e., in the context of our qualitative consideration, hKr is close to 1.However [Figure 2(b), closed circles], the critical amount (~45%) of SF6 exists such as the dependence of t on [SF6] undergoes a drastic change. A comparison of Figures 2(a) and 2(b) shows that the result can be qualitatively rationalised on the basis of the generalized model of BCP with EB; in this case, in the context of kinetic mechanism (II), hSF6 ¡í 0.5.The self-ignition in reactor II at [Cl2] > 0.03 kPa and P > > 0.20 kPa is always followed by a sharp sound. The time dependence of the total pressure during self-ignition is shown in Figure 3. As can be seen, an almost three-fold increase in P during the self-ignition is observed. It means that the BCP at low surface to volume ratios occurs in a chain-thermal explosion regime,19 and the warming-up makes up .T = 293..P/P ¡í ¡í 1000¡Æ. Therefore, the set of reactions (I) of a chlorosilylenecatalysed cycle may occur.In a series of experiments, H2 was added into a reactor before Cl2 (Figure 4). It was shown that t in the presence of H2 slightly increases (Figure 4, curve 2).Note that t does not depend on the total pressure in our conditions. SiHCl and SiCl2 do not react with H2,13 the reaction Cl + H2 ¢ç H + Cl2 is chain propagation, and it cannot cause an increase in t. However, the rapid chain termination SiH2 + H2 ¢ç SiH4 (10.13 cm3 s.1, ref. 13) can give rise to an increase in t. It may serve as indirect evidence for SiH2 formation in DCS chlorination.Then, a heat emission can take place in the fast reactions (rate constant of the reaction SiH2 + Cl2 is 1.4¡¿10.10 cm3 s.1, ref. 13): Note that the energy released in (III) is enough to obtain excited SiHCl (A1B1). SiH2 can result from the steps12 SiCl2 + SiH2Cl2 ¢ç ¢ç Cl3SiSiH2Cl* ¢ç SiH2 + SiCl4. The set of reactions (I), (III) and (IV) represents possible steps of DCS chlorination including silylenes as chain carriers and formation of SiHCl (A1B1) and EB.This presumption is still to be refined by the establishment of the composition of reaction products, especially, detection of H2. We are grateful to Professor V. V. Azatyan for many useful discussions. This work was supported by the Russian Foundation for Basic Research (grant no. 00-03-32979a). References 1 E. N. Chesnokov and V. N. Panfilov, Dokl. Akad. Nauk SSSR, 1981, 261, 925 [Dokl. Chem. (Engl. Transl.), 1981, 261, 865]. 2 N. M. Rubtsov, S. M. Temchin, V. I. Chernysh and G. I. Tsvetkov, Kinet. Katal., 1997, 38, 495 [Kinet. Catal. (Engl. Transl.), 1997, 38, 441]. 3 V. V. Azatyan, S. G. Gagarin, V. I. Zakharyin, V. A. Kalkanov and Yu.A. Kolbanovskii, Khim. Fiz., 1983, 2, 201 (in Russian). 4 R. G. Aivazyan and V. V. Azatyan, Khim. Fiz., 1997, 16, 621 [Chem. Phys. Reports (Engl. Transl.), 1997, 16, 599]. 5 I. R. Begishev, O. L. Gromovenko and V. A. Poluektov, Zh. Fiz. Khim., 1994, 68, 1099 (Russ. J. Phys. Chem., 1994, 68, 992). 6 V. V. Azatyan and E. A. Markevich, Kinet. Katal., 1988, 27, 1291 [Kinet. Catal.(Engl. Transl.), 1988, 27, 1191]. 7 V. V. Azatyan, R. R. Borodulin, E. A. Markevich, N. M. Rubtsov and N. N. Semenov, Izv. Akad. Nauk SSSR, Ser. Khim., 1976, 1459 (Bull. Acad. Sci. USSR, Div. Chem. Sci., 1976, 25, 1396). 8 N. N. Semenov, O nekotorykh problemakh khimicheskoi kinetiki i reaktsionnoi sposobnosti (On Some Problems of Chemical Kinetics and Reactivity), Akad. Nauk SSSR, Moscow, 1958, p. 685 (in Russian). 9 E. T. Denisov and V. V. Azatyan, Ingibirovanie tsepnykh reaktsii (Inhibition of Chain Reactions), Ross. Akad. Nauk, Chernogolovka, 1997, p. 265 (in Russian). 10 V. V. Azatyan, R. R. Borodulin and N. M. Rubtsov, Fiz. Gorenia Vzryva, 1980, 5, 34 (in Russian). 11 V. I. Vedeneev and A. E. Shilov, in Fizicheskaya khimiya, Sovremennye problemy (Physical Chemistry, Contemporary Problems), Khimiya, Moscow, 1985, p. 7 (in Russian). 12 M. T. Swihart and R. W. Carr, J. Phys. Chem., A, 1998, 102, 1542. 13 I. Safarik, V. Sandhu, E. M. Lown, O. P. Strauss and T. M. Bell, Research Chem. Intermed., 1990, 14, 108. 14 A. V. Parijskaya and V. I. Vedeneev, Kinet. Katal., 1973, 14, 1116 [Kinet. Catal. (Engl. Transl.), 1973, 14, 981]. 15 N. M. Rubtsov, V.V. Azatyan and R. R. Borodulin, Izv. Akad. Nauk SSSR, Ser. Khim., 1980, 1234 (Bull. Acad. Sci. USSR, Div. Chem. Sci., 1980, 29, 1165). 16 V. V. Azatyan, Kinet. Katal., 1975, 16, 567 [Kinet. Catal. (Engl. Transl.), 1975, 16, 487]. 17 C. P. Conner and E. W. Stewart, J. Am. Chem. Soc., 1977, 99, 2544. 18 F.-T. Chau, D. C. Wang, E. P. F. Lee, J. M. Dyke and D. K. W. Mok, J. Phys. Chem., A, 1999, 103, 4925. 19 V. V. Azatyan, Kinet. Katal., 1999, 40, 818 [Kinet. Catal. (Engl. Transl.), 1999, 40, 791]. 20 L. V. Gurvich, G. V. Karachentsev and V. N. Kondratiev, Energii razryva khimicheskikh svyazei. Potentsialy ionizatsii (Chemical Bonds Energies. Ionization Potentials), Nauka, Moscow, 1976 (in Russian). 1 2 3 SiHCl SiCl2 P/Torr 2 6 10 10 30 50 t/ms t/s Chemiluminescence Absorbance 6 4 2 4 8 (arb.units) (arb. units) 310 315 320 325 Absorbance SiCl2 A.X 13 12 11 10 9 8 7 6 5 9 8 7 6 5 4 l/nm Intensity 450 460 470 480 490 500 l/nm 013-010 012-010 011-010 010-010 030-000 020-000 010-000 000-000 000-010 SiHCl (A1B1.X1A1) Figure 3 The time dependences of (1) chemiluminescence at 400 nm, (2) absorption at 300 nm and (3) total pressure in the self-ignition of 0.13 kPa Cl2 + 0.13 kPa (DCS + 8% C3H6) and experimental spectra of SiHCl (A1B1.X1A1) and SiCl2 [10 scans, 10 accumulations, 0.20 kPa Cl2 + 0.20 kPa (DCS + 8% C3H6)]. Reactor II. Curve 3 demonstrates the inherent vibrations of a mechanothrone (300 Hz) under the impact of an expanding gas. 10 1 20 n 20 n SiH2 + Cl2 SiHCl + HCl + 363 kJ mol.1 (ref. 20) SiH2 + Cl2 SiH2Cl + Cl + 221 kJ mol.1 (ref. 20) (III) (IV) 10 5 1 2 3 4 5 0 20 40 60 80 100 18000 10 5 10 5 t/s bleeding-up self-ignition pumping out self-ignition DCS-additives bleeding-up DCS-additives self-ignition bleeding-up DCS-additives Absorption at 300 nm (%) Figure 4 The experimental time dependences of Cl2 concentration: (1) 0.20 kPa Cl2 + 0.56 kPa (DCS + 8% C3H6); (2) 0.13 kPa H2 + 0.20 kPa Cl2 + 0.56 kPa (DCS + 8% C3H6); (3) 0.20 kPa Cl2 + 0.20 kPa (DCS + 8% C3H6 + 50% Kr); (4) 0.20 kPa Cl2 + 0.20 kPa (DCS + 8% C3H6 + 50% SF6); (5) 0.20 kPa Cl2 + 0.18 kPa (DCS + 80% SF6). Received: 5th December 2001; Com. 01/1863