Photochemistry of silicon compounds. 5. The 147-nm photolysis of

The 147-nm photolysis of dimethylsilane. A. G. Alexander, and O. P. Strausz. J. Phys. Chem. , 1976, 80 (23), pp 2531–2538. DOI: 10.1021/j100564a001...
0 downloads 0 Views 829KB Size
T H E

J O U R N A L

OF

PHYSICAL CHEMISTRY Registered in U. S. Patent Office 0 Copyright, 1976, by the American Chemical Society

VOLUME 80, NUMBER 23 NOVEMBER 4, 1976

Photochemistry of Silicon Compounds. 5. The 147-nm Photolysis of Dimethylsilane A. G. Alexander' and 0. P. Strausz* Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada

(Received March 18, 1976)

Publication costs assisted by the National Research Council of Canada

The 147-nm gas phase photolysis of dimethylsilane yields ten retrievable products in addition to a solid polymeric material. The nature of the competing, simultaneous primary steps, their quantum yields, and secondary reactions of the primary radicals were deduced from the observed products and the effect of pressure, exposure time, added scavengers, and deuterium labeling on product yields. At least 11primary steps are required to explain the experimental observations. The major mode of decomposition is molecular elimination and, although the 147-nm photolysis line lies in the Si-H absorption region, Si-C cleavage predominates. Three silylene diradicals, SiH2, SiHCH3, and Si(CH&, are implicated in the reaction. They all readily insert into the Si-H bonds of the substrate to give vibrationally excited methylated silanes but neither one reacts with nitric oxide or ethylene. The presence of primary and secondary HID kinetic isotope effects in the primary decomposition was established by a detailed analysis of the kinetic data.

In earlier s t u d i e ~on ~ , the ~ 147- and 123.6-nm photoIysis of monomethylsilane it has been shown that the decomposition proceeds by eight simultaneous primary steps. The nature of the primary steps and their quantum yields along with the ensuing secondary reactions of the primary radicals were deduced from product measurements and detailed kinetic studies of the effect of pressure, exposure time, wavelength of photolysis, added radical scavengers such as nitric oxide and ethylene, and deuterium labeling on the reactions. The primary steps are characterized by the predominance of molecular modes of elimination over the free radical modes. Polyfragmentations are less important in the 147-nm photolysis than in the 123.6-nm photolysis. The intermediacy of the two silylene diradicals SiH2 and CH3SiH was demonstrated in the reactions and furthermore it was shown that both silylenes readily insert into the Si-H bonds of the substrate molecule to yield disilanes but they are unreactive with nitric oxide or ethylene. Thus, the use of these scavengers made it possible to remove mono free valent silicon radicals from the system without affecting the reactions of silylenes. That nitric oxide and olefins are effective scavengers of mono free valent silicon radicals has been demonstrated earlier.4 The present study was undertaken as part of a continuing effort to extend our knowledge of the photochemistry of silicon hydrides and the chemistry of silicon hydride free radicals.

Experimental Section

Standard high vacuum techniques were employed. All apparatus was stopcock grease-free and equipped with mercury float valves, Hoke C415K and TY440 valves. A cylindrical Pyrex reaction cell 28 mm in diameter and 185 mm long with a cold finger and a LiF end window was used. The latter was sealed to the reaction cell with black wax and was polished with cerium oxide after each run to remove polymer that formed on the surface. The xenon lamp was an air cooled electrodeless discharge operated by a 2450 MHz microwave generator. The lamp, emitting radiation3 at 147.0 nm (and at 129.5 nm where the intensity is ca. 2% of that at 147.0 nm), was filled with approximately 0.2-0.3 Torr of Xe and 1.0-2.0Torr of neon as a carrier gas. All gases used were obtained from Air Reduction Co. Barium getters (Kemet Co.) were used to eliminate residual gases. The window of the lamp was placed in contact with the window of the cell and surrounded with a column o f flowing nitrogen gas in order to remove oxygen which absorbs strongly below 200 nm. The light intensity of the Xe lamp, based on COz actinometry [taking 4(CO) = l.05-9], was determined to be 6.0 X 1015quantals. (CH3)2SiHZ (Peninsular) and (CH&SiD2 (Merck) were purified by low temperature distillation at -126 "C and the 2531

2532

A. G. Alexander and 0. P. Strausz

purity checked by gas chromatography. The extent of deuteration of (CH3)zSiDZ on the silicon atom was determined to be at least 99% (D/(D H)) by 100-MHz NMR. Nitric oxide (Matheson) was distilled at -186 "C (Liquid Ar) and degassed at 196 "C. In the series of quantitative measurements, the noncondensable fraction at -196 OC was removed by a Toepler pump and analyzed by a thermal conductivity gas chromatography on a 10-ft molecular sieve 13X column using argon as carrier. The condensable fraction was analyzed on three columns: (1) 10 ft X 0.25 in. stainless steel medium activity silica gel, (2) 16 f t X 0.25 in. stainless steel 10%silicon oil DC 200 on Chromosorb W.A.W. 60-80 mesh, (3) 16 ft X 0.25 in. aluminum 10% T.C.P. on Chromosorb W, 30-60 mesh, using a HewlettPackard 5750 dual flame gas chromatograph. Response factors relative to substrate were determined for each compound. Identification of products was done using MS2, MS9, and MS12 mass spectrometers and isotopic analysis of hydrogen and methane using MS2 and C.E.C. 21-614 mass spectrometers.

+

Results The photolysis yielded ten products. In decreasing order of importance they were: hydrogen; i,l,2-trimethyldisilane (Tri-MDS); methane; ethane; 1,1,2,2,-tetramethyldisilane (Tet-MDS); trimethylsilane (Tri-MS); 1,l-dimethyldisilane (DMDS); monomethylsilane (MMS); methylethylsilane (MES); and ethylene. Two other products detected in trace amounts, which could not be identified, have been neglected in the kinetic treatment. A solid polymeric material formed during the photolysis was deposited on the cell walls and caused a decrease in the transparency of the LiF window. The effect of substrate pressure on product yields was studied in 16-min photolyses. The relative yields are shown graphically in Figures 1 and 2. The observed pressure dependence of CzH6, C2H4, CH4, Hz, MES, Tri-MS, DMDS, Tri-MDS, and Tet-MDS can be rationalized in terms of pressure stabilization of hot silane insertion products:

)*

silane (insertion products

M --f

3

A4

PRESSURE [TORR)

Figure 1. Relative yields of H2, Tri-MDS,Tri-MS, and C2H4 as a function of pressure from the xenon lamp photolysis of dimethylsilane: (0)H2 (Ihs), (0)Tri-MDS (Ihs), (A)Tri-MS (Ihs), (V) C2H4 (rhs).

6

n

n

ICHJ,

n

SiHSiH,

0

0

50

100

150

200

300

250

350

400

PRESSURE (TORR)

Figure 2. Relative yields of DMDS, MES, Tet-MDS, CH4, CPH6,and MMS as a function of pressure from the xenon lamp photolysis of dimethTet-MDS, ( 0 )CH4, ( 0 )C2H& (A) ylsilane: ( V ) DMDS, ( M ) MES, (0) MMS.

I

6

H,H,,CH,,CH4,CP,, silane fragments

1

silanes

The silane fragments would presumably contribute to the formation of the polymer. MMS is unaffected by substrate pressure and it will be shown later that most of it is formed directly in a primary process. To avoid complications due to pressure-dependent fragmentations, all subsequent experiments were conducted a t a pressure of 100 Torr or higher. In order to correct for attenuation of the incident light intensity by polymer deposition in the determination of the quantum yields, a time study of product yields was carried out. Carbon dioxide actinometry was done before and after each run. Light intensities were corrected to a constant initial photon input and a plot was made of the post run absorbed intensity corrected vs. time. This graph, Figure 3, was integrated to yield the total photon input a t any given time. A graphical extrapolation of the data to zero exposure time gave the following quantum yield values: +(Hz)= 0.90, +(Tri-MDS) = 0.35,4(CH$ = 0.20, I$(C~HC~) = 0.15, $(Tet-MDS) = 0.135, +(Tri-MS) = 0.11, 4(DMDS) = 0.075, +(MMS) = 0.035, 4(MES) = 0.018, and +(CzH4) = 0.015. The time study also enabled us to distinguish between products of primary and secondary origin. The results given in Table I show that within The Journal of Physical Chemistry, Vol. 80, No. 23, 1976

TIME (MINUTES)

Flgure 3.

time.

Intensity of the xenon lamp as a function of post photolysis

experimental errors all products are invariant with respect to exposure time, consequently they are of primary origin. It has been shown in the mercury photosensitization of silanes4 and in the direct photolysis of rn~nomethylsilane~~~ that nitric oxide is an efficient scavenger of monovalent silyl radicals. Silylenes, on the other hand, are not affected by nitric oxide and relatively small concentrations of nitric oxide would

Photochemistry of Silicon Compounds

2533

TABLE I: Relative Yields of Products as a Function of Time from the Xenon Lamp Photolysis of Dimethylsilane

Relative vield. % Products

1

2

3

5

8

10

12

16

20

HP CH4 C2H6

46.0 10.5 7.1

47.6

43.2 10.9

45.9

7.2

C2H4

0.7

0.7

7.4 0.6

43.7 13.7 7.7

44.4 14.1 6.9 0.3

1.5

5.8 0.8 3.9 17.6 7.4

5.4 0.7 3.3 15.2 6.4

45.9 14.5 7.0 0.3 3.4 5.4 0.7

47.5 15.8 6.3 0.4

1.8

47.9 14.9 6.8 0.5 1.4 4.9 0.6 3.0

3.5

2.8

2.6

14.2

14.7 6.1

13.9 5.7

13.9 6.1

~

Time,min

~~

12.2

6.9 0.8 1.4 4.6 0.7 3.6 15.0 6.7

0.9 5.6 0.7 3.6 17.1 6.8

MMS Tri-MS MES DMDS Tri-MDS Tet-MDS

f2.9

0.5

2.5 5.5 0.7 3.3 15.4 6.5

5.2

2.9

5.6

0.8

1.1

4.9 0.7

TABLE 11: Micromoles of Products as a Function of Added Nitric Oxide from the Xenon Lamp Photolysis of Dimethylsilane

Nitric oxide, %

-~ ~

a

Products

0

0.97

4.19

H2 Tri-MDS CH4 CzH6 Tet-MDS Tri-MS DMDS MMS MES CZH4 Tri-MDSO Tet-MDSO N2

0.96 0.31 0.27 0.16 0.13

1.37

1.73

a

a

a

a

0.27

0.36

0.62

0.069

0.59 0.058 0.057 0.016 0.079

0.032

0.052

0.019 0.012

0.013

0.13 0.024

0.11

0.00 0.00 0.00

0.022 0.10

6.94

a

0.009 0.070 0.022

0.002 0.047

0.10

0.086 0.019 0.018 0.026 0.13 0.23

0.092

0.14

a

a

0.45

0.90

15.5

a

0.004 0.087

a

a

0.12 0.073 0.032 0.051 0.26

0.76

a

a

Not determined.

TABLE 111: Isotopic Composition of Hydrogen and Methane from the Xenon Lamp Photolysis of Mixtures of Dimethylsilane and Dimethylsilane-dz

Mole fraction Substrate

Hydrogen

Methane

XHe

XH2

XHD

XD,

XCHa

XCH~D

0.000

1.000

0.119

0.102 0.250 0.498 0.754 0.900

0.898 0.750 0.502 0.146

0.203

0.403 0.419 0.369

0.100

0.921

0.478 0.379 0.241 0.126 0.043 0.016

0.265 0.414 0.539 0.690 0.860 0.937

0.735 0.586 0.461 0.310 0.140 0.063

XL

0.389 0.586 0.813

0.286

0.144 0.063

effectively suppress all products arising from monovalent silyl radicals. In order to differentiate products originating from monoradical and diradical precursors, experiments were carried mt in the presence of a few Torr of nitric oxide. The results are tabulated in Table I1 and, as can be seen, 51% of the Tet-MDS, 83% of the Tri-MS, and 100%of the C2H6 are scavengeable. The unscavengeable Tet-MDS, Tri-MS, TriMDS, DMDS, and MES must arise from insertion of various diradicals into the substrate. The increase in the Ha, CH4, C2H4, and N2 yields with the increasing concentration of nitric oxide is indicative of the chain nature of the NO reaction.The

trend in the yield of the new products 1,1,2-trimethyldisiloxane (Tri-MDSO) and 1,1,2,2-tetramethyldisiloxane (TetMDSO) also supports the operativeness of short chains as well as demonstrates the presence of the radicals CH3SiH2 and (CH&SiH. The photolysis of dimethylsilane can lead to various types of atomic and molecular loss of hydrogen and methane. In order to sort out these possibilities, experiments were carried out with isotopically labeled substrate (CH&SiD2. The results are presented in Table 111. For convenience we will use the symbol L (light) for dimethylsilane and the symbol He (heavy) The Journal of Physical Chemistry, Voi. 80, No. 23, 1976

A. G. Alexander and 0. P. Strausz

2534

for dimethylsilane-dz. The following kinetic,scheme for the primary loss of hydrogen in the photolysis of dimethylsilane should be considered:

+ hu H He + hv-D He + hv- Dz He + hv HD He

+

(4)

-+

He

+ hu

+

Hz

L+hv+H L

+ hu+

H2

where X D ~X, H ~and , XHDare the experimentally determined isotopic yields, X H and ~ XL are the mole fractions of the substrate, and 24 is the sum of experimental quantum yields. The extinction coefficients of the He and L compounds are identical12 and cancel out from the equations. A plot of X D ~ I Xvs. H~ XL and XHD/XHe vs. XLis given in Figure 4.It can be shown that for XL = 0

(5)

+ 2') (3' + 4' + 5') (1'

The hydrogen and deuterium atoms can carry various amounts of excess translational energy but are assumed to be thermalized in the kinetic treatment. The following abstraction and exchange reactions with both He and L are then possible (for clarity, the resulting silane fragments will not be written): H+L-+Hz H

+ He

-+

HD

(6)

(7)

D+L-HD

(8)

D+He+D2

(9)

H+He-+D

(10)

D+L-+H

(11)

+L H + He D +L D + He H

-

+

+

CH3

(12)

CH3

(13)

CH3

(14)

CH3

(15)

Since the yield of MMS was small, 4 = 0.035, the exchange reactions 12-15 are not important. It has been shown earlier that the exchange reactions 10 and 11are equally unimportant.1° Abstraction from the methyl groups of the substrate has not been included in the scheme because it has been shown that abstraction from the silicon end of the molecule is several orders of magnitude faster.l' Similarly, exchange with the methyl hydrogens was omitted because it should be even less probable than with the silicon hydrogens.1° Steady state treatment of the above reaction scheme gives the following expressions for the fractional yields of isotopic hydrogens:

and

(1x1

XHZ/XHe =

From these equations the following quantum yield values can be determined: 42 = +(D) = 0.295,43 = ~ ( D z=) 0.135, 4 5 = ~ ( H z=) 0.107, and 41 $4 = 4(H) $(HD) = 0.363. These figures are considered to be accurate to the second decimal and the additional significant figures are carried through the computations only to minimize round-off errors. Rearranging eq I gives

+

+

From the plot of left hand side of eq X against the ratio XL/XHe, Figure 5, a kinetic isotope effect of 3.6 was determined for 1281129 and it is assumed that the isotope effect for k 6 / k 7 would be the same. Owing to the primary isotope effect for hydrogen loss between He and L, 41 and 4 4 must be determined independently from ethylene scavenging studies. To this end mixtures of dimethylsilane-dz and ethylene were photolyzed and the isotopic compositions of hydrogen and methane were measured. The results are presented in Table IV. The kinetic scheme which is applied involves steps 1-5, the photolysis of C2H413 C2H4

+ hu

CzH4*

-

+

+

CZH4*

+ CzHz Hz + CzHz 2H

(16) (17) (18)

and the competing reactions of H and D atoms steps 7 and 9 as well as

+ CzH4 D + C2H4 H

and

+

[ 24

XHz 1 =-

XHe

+

46

XL +4(3!+4