Infrared and Raman spectra of methyl- and phenyl-arsenic halides

Spectroohimlca A&, Vol. 26A,pp. 1581to 1693.Perg8monPrea!~ 1970.Printedin NorthernIreland

Infrared and Raman spectra of methyl- and phenyl-arsenic halides D. M. Department

REVITT and D. B. SOWERBY* of Chemistry, University of Nottingham, Nottingham

(Received17 June 1969) Abstract-Infiwed and laser Raman spectra for MeAsCl,, Me&&l, Me&C&, PhAsCl, PhAaBr,, Ph.&Cl, Ph&aBr, Ph.&, PhAsCl,, Ph&sC1, and Ph&Cl, are presented and possible assignments are discussed. The observation of five bands in the skeletal stretchingregion for PkAsCl, is con&tent with e trigonal bipymmidal structure with an equatorial phenyl group. INTRODUCTION THERE is little information available on the vibrational spectra of organoarsenic halides. The methylhaloarsines have been examined in the Raman [l] but no infrared data have been obtained. This paper reports infrared spectra of the methylchloroarainea and attempta to obtain spectra,of methylarsenic chlorides. Spectra of the latter are of interest aa it would be possible to assign with fairly high certainty vibrations of chlorine atoms in axial and equatorial positions. A variety of phenylarsenic(III) and (V) compounds have also been examined continuing earlier work on phenyl derivatives of elements in Groups IV and V[2].

Afet~larsenb(II1)

and (V) clblwides

The infrared and Raman spectra.of MeAsCl,, Me&Cl and Me&Cl, are summarized in Table 1 together with suggested a.a&gnments. The Raman data for the chlorides are generally in agreement with earlier work [l] except that the symmetric and asymmetric A&l, stretohes are aseigned to bands at 410 and 386 cm-1 respectively. These now compare favourably with those for analogous vibrations in arsenic trichloride [3] and the methylchlorogermr [4]. In dimethylchloroarsine, neither the infrared nor the Raman resolves the two expected’ arsenic-carbon stretching modes. Attempts have been made to obtain spectra of MeAsCl, and Me&Cl, to help in assigning spectra of the corresponding phenyl compounds. Here, a&gnments of the arsenic-chlorine vibrations are expected to be less ambiguous. Both compounds have been reported [S] but at room temperature the product which remained after chlorinating methyldichloroaraine at -78’ was the original arsenio(II1) compound. Dimethylchloroarsine, on the other hand, gave the expected arsenic(V) derivative as an unstable, moisture sensitive white solid. Dimethylarsenic trichloride is expected to be a trigonal bipyremidal molecule with methyl groups in equatorial positions and thus to belong to the C,, point group. [l] G. P. Vu DER I(ELEN and M. A. HER-, BUZZ.L&c.C?&n. Belgee85, 350 (1966); E. G. CAYS and G. P. VAN DER KELEN, S~ectvoclaim. Acta %,2103 (1966). [2] K. M. MACKAY, D. B. SO-BY and W. C. You~a, Spdvchim. Ada MA, 611 (1968). [3] J. E. D. DAVIES and D. A. LONQ, J. Ohem. Sot. A 1757 (1968). [4] D. F. VAN DE VONDEL and G. P. VAN DER KELEN, Bull. Sot. Cham. Be&ea 74,453 (1965). [6] A. BAEYER,Am. 107, 263 (1858). lb81

D. M. REVITT and D. B. SOWE~SY

1582

Table 1. Infrared and Raman spectra of methylarsenic M&&l, Ramsn

(liquid) Infrared 1417eh 1397B

1266VW

689w, p 413VEX. p 3898. p

12488 1224VW 332B 631B 410m 334“S 302w

226VW 199m, p 1628, dp(?)

Me&Cl Reman 1416w 1390VW 1261w, p

(liquid) Infrered 1421m 1406 m 1400 > 1264m

681x8,p

901m 931m 684m

3628, p

364s

243mw, dp(?) 212m, dp 199m, dp

Assignment

Me&Cl, Ramen

870 VW

6. CH, P

VAE-c va A&l, vaaAs’& Y As-cl s CA& BanCA&l 68 CA&l 6 A&l,

(solid) Infwred 1268w 108Ow 916s

&a CH, &a CH,

C=a P CHa

chlorides

642mw 681ms 413s 270vs 219me 192166“VW 126w

786 776jrn 722w 670VW

Aaeignment 6. C=, P C=a P CR,

678m 411Ins

VAE-C VAs-cIeq

279w

v, A&J ax

acAaci

6Ascl,sx

Three arsenic-chlorine stretches are expected, two of these are associated with the axial atoms (A, + B,) and one with the equatorial chlorine (A,). By analogy with related molecules such as MePCl, [6] and Me,PF, [7], the latter should occur at the highest energy and is assigned to the band at 410 cm-l. Two bsnds at 310 and 270 cm-l can then be associsted with the axial stretches. That at lower energy is very strong in the Raman and is assigned to the symmetric (A,) mode while the 310 cm-l band, which is observed only in the infrared, is the asymmetric component. Seven skeletal deformations, distributed as 2A, + A, + 2B, + 2B,, are predicted for Me&Cl,. The AsC, deformation (A,) is expected to occur in the Ramcm near 240 cm-l, at a similar position to that found for Me&Cl but this region is obscured by strong Raman bands at 270 (due to an axial AsCl, stretch) and 219 cm-l. This latter band and the two weak ones at 190 and 155 cm-l are assigned to three of the four expected CAsCl bending modes. Finally, deformation of the two axial chlorine atoms is assigned to the band at 125 cm-l. Phenylarsenic(II1) halides Spectroscopic data and suggested assignments are summarized in Table 2. In compounds of this type there is no single absorption that can be assigned to phenylarsenic stretching. The spectra can be understood, however, if coupling occurs between C-As vibrations and mess-sensitive stretching vibrations of the phenyl group. The latter are Whiffen’s q, r and t vibrations [8] and with arsenic strongest coupling occurs with t [2]. A similar conclusion has been reached for other arsenic compounds [9] and has been shown to be valid for a vsriety of phenylsilanes [lo, 111. [0] [7] [8] [9] [lo] [ll]

I. R. BEATTIE, K. LIVINQST~N and T. GILSON, J. Chem. Sot. A 1 (1968). A. J. DOWNS and R. SC~ZLER, Spectrocbim. Acta 28A, 681 (1967). D. H. WIB~EN, J. Chem. Sot. 1350 (1956). J. H. S. GREEN, W. KYNASTON and G. A. RODLEY, Spectrochim. Acta %A, A. L. SMITH, Speotrochim. Acta BSA, 1075 (1967). A. L. SMITH, Spectrochim. Acta MA, 695 (1968).

853 (1968).

254 mw, p 227 w, p

190 VW

191 w

39ova 368 va 307 V8

466 VW

393 vs, p 369 8, p(m) 312 w, p

460 a

677 mw, p 619 mw, dp

1067 sh 1023mw 999 s 986 vw 966 VW

1482 ms 1436 va 1381 mw 1332 m 1304 m 1275 w 1262 w 1182 mw 1160 w 1093 w 1076 vs

1575 mw

914 w 841 w 799 w 738 VB 700 eh 688 WI1 672 mw 616 vw

743 w

921 VW

1016 ms, p 1002 vs, p

1077 s, p

1186 w, p 1163 mw, dp

1286 mw

1482 vw 1439 w 1381 VW 1335 w

1580 s, dp

PhAeCl, (liquid) Infrared Raman

248 m, p 218~

320 w 302 w

312 B 301 s

372 8, b

372 VB, pr

479 vw 405 vw

692 YS 668 w 615 VW 544 vw 473 s 460 I

911 w 842 w 793 vw 735 vs

1578 mw 1516 w 1480 m 1434 s 1382 VW 1333 w 1304 w 1272 vw 1261> 1181 mw 1155 w 1090 eh 1077 sh s 1072 1068 sh1 1022 m 998 ms 986 vw 967 vw

(liquid) Infrared

695 vw 672 8, p 618 s, dp

741 w

946 VW 915 w 849 w

1027 8. p 1002va, p

1082 E, p

1188m. p 1162 s, dp

1276 w

1484w 1437 w 1396 VW 1335 w

1582vs, dp

PbAsCl Rfbmen

246 mw, p 221 va, p 193 VW

291 a, p

312 sh 299sh s, br 288

474 VIY 459 vs

474 vw 460 vw 395 VW

315 ah

674 mw, p 619 w, dp

225 vs, p 206 mw, p(t)

290 sh 276 vs, p

312 m, p

461 vw

743 VW

691 vs 670 mw 615 VW

696 VW 669 mw, p 618 w, dp

741 vw

916 VW

1025 mw, p 1001 vs, p

1074mw, p

1185w, p 1162 w

1272 vw

1333 vw

1480 vw 1438 VW

1580 mw, dp

PbAsBr, R8men

734 vs

912 mw 844 mw

1070 vs 1066 1077 sh1 1022 8 999 VB 986 vw 969 w

1183 m 1159 mw 1094 w

1580 m 1634 VW 1481 vs 1435 VB 1381 mw 1333 m 1305 m 1273 w

(liquid) Infnwed

913 VW 845 VW

1029 m, p 1003 vs, p 988 sh

1079 mw, p

1185w, p 1159 w, dp

1479 w, p 1434 w, dp 1383 VW 1332 vw 1301 VW 1272 VW

1583 ms, dp

Pb,AsBr Raman

>

1

>

289 s 276 s 1

> 308 ah

1111

458 s

736 VB 7ooall 686 B > 670 m 615 vw

1181 m 1168 mw 1093 mw 1073 vs 1 1065 ah 1021 m 998 a 986 vw 966 w 938 VW 913 w 841 w

1574 m 1528 vw 1477 ma 1434 a 1378 w 1330 ms 1302 m 1272 w

(liquid) I&kWZd

Table 2. Infrared and Raman spectra of phenylarsenic(III) halides from 1600 cm-l

2, +

w(T)

x

eo

u+

6CAeCl

t+ vAs-Br

t

v80 &cl,

v A&l Ye A&l,

Y

h

P i

d b

9

C

a

c

w+i

0

w+i

n

II)

iandk

Aeeignment

1684

D. Pd. REVITT end D. B. SOWERBY

Similar interactions between mass-sensitive bending modes and skeletal deformations are in general to be expected and SMITB[ll] has reported that phenyl vibrations y, u and 2 may all be involved. In addition to the compounds in Table 2, the Raman spectrum of triphenylarsine has been re-examined. The strong bands reported by GREEN and his co-workers [O] have been confirmed and, by using concentrated solutions in carbon disulphide, weaker bands, assigned as shown in parentheses, were observed at 1482 (m), 1379 (W + jj), 1332 (o), 1187 (a), 911 (i), 855 (g), 734 cf), and 700 cm-l (w). Vibration y at 475 cm-l was obtained from a solid state spectrum as this region is obscured by carbon disulphide. The solution spectrum showed that bands at 1187 (a), 1082 (a), 1025 (b), 1001 (p), 314 (t) and 238 (u) were polarized. Non mass-sensitive vibrations The assignment of these vibrations is not discussed in detail as it follows the pattern observed previously. In general, the assignments are confirmed by the polarization data reported here. There is a complication however in assigning vibration k as it occurs close to 1. l?or an isolated phenyl group with C,, symmetry, k has A, character and I belongs to B,. In the hdobenzenes both occur in the 1580 cm-l region but are separated only for the fluorine compound [8]. Two bands are found for the phenylchlorosilanes [lo] and following the accepted pattern k has been assigned to that at the higher frequency with the higher intensity in both the Reman and infrared. In the arsenic(II1) compounds, the 1580 cm-l Raman band coincident with a strong infrared band, is depolarized and the major contribution must therefore be from vibration 1. The A, mode, k, is then much weaker and may not be separated from 1. There is some ambiguity about vibrations d (B,) because in arsenic compounds, the mass sensitive A, mode 41occurs in the same region (~1080 cm-l). The Reman shows a single strongly polarized line in this region which must be Qbut the infrared is complex. In polyphenyl compounds some of the complexity may arise from the carbon-arsenic stretching component of this mode but this is not a possibility for monophenyl species. Vibration d most probably gives rise to the 1068 cm-l absorption often seen in these compounds as a low frequency shoulder on the stronger p absorption. Mass sensitive phenyl and skeletal modes

If the phenyl group is considered as a point mass, the vibrational modes for triphenylarsine and the mono- and di-halides are as shown in Table 3. The spectra of the methyl analogues have been assigned on this basis, but with the phenyl derivatives, the situation is more complicated because of the interaction between skeletal and phenyl modes and, when more than one phenyl group is present, because of the effect due to the symmetry at the arsenic atom. ChJorinecompounds Absorptions due to arsenic-chlorine stretching vibrations occur in expected positions as strong, broad infrared bands coincident with intense Raman bands. The positions and separation (22 cm-l) of the symmetric and asymmetric components for phenyldichloroarsine are close to those found for the methyl analogue. It seems

Infrared and Raman spectra of methyl- and phenyl-arsenichalidea Table 3. Vibretionel modes for Ph&a, PhgsX

p$As 4

and PhAsX,

Ph&X E

A’

1685

PhAsX, A”

A’

A”

clear that the higher energy band should be assigned to the symmetric stretching vibration because of its low depolarization ratio. But the band at 369 cm-1 also has a depolarization ratio less than O-75 and thus also appears to be polarized. This may not be real and could perhaps result from difficulties in resolving the two peaks. If the band is in fact polarized the simplest rationalization would be that the molecule is following the selection rules for C, symmetry. Although it is quite likely that the overall symmetry of PbAsCl, is C, in systems similar to this the vibr&,ional spectra can usually be described in terms of the local symmetry at the central atom. This is unlikely to be a complete explanation as the results for MeAsCI, and PhAsCl, are very similar. In both VAN DEB KELEN’S work [l] and the present reinvestigation of MeAsCl, both of the bands attributable to amen&chlorine stretching appeclr to be polarized, with the higher energy component having the lower depolarization ratio. For the methyl compound, it is less likely that the C, selection rules should be followed. At the present no further explanation can be offered. With diphenylchloroaraine, the intense polarized Ramsn band at 372 cm-r is very close in energy to the arsenic+chlorine stretch in the dimethyl anslogue and ms,y indicate that there is little modiGcation of this mode in the phenyl compound. The A&l, deformation (A, in C,) can be assigned to the moderately intense band at 160 cm-r but mixing with CA&l deformations is possible. The higher energy mass-sensitive modes, q (A,) and r (A,), do not involve large C-M stretching components for heavier central atoms. In the monophenyl compound, q is a single, intense, polarized Raman band at 1077 cm-l with a strong infrared coincidence. Similarly, one polarized band at slightly higher energy occurs for the diphenyl compound, but the infrared, shows two bands separ&.ed by ~5 cm-l which are the expected asymmetric and symmetric components of q. Vibration r gives a strong, polarized Raman band but is weak in the infrared ; there is no evidence of splitting in the diphenyl compound. Vibration y, an out-of-plane deformation, is strong in the inf&red but weak in the Raman. The monophenyl compound shows a single absorption at 460 cm-l, while in the diphenyl asymmetric and symmetric components occur separated by 13 cm-l. For triphenylarsine, y also occurs as a doublet in the infrared but the sepa;ration is only 4 cm-l. SMITH[lo] has shown that there are similarities in the vibrational spectra of phenyl compounds and their bromine analogues and that the latter aid the assigning of phenyl spectra. In the present case, analogies with the spectra of chlorobromoarsines are sought. The individual compounds hrtve not been isolated, but Raman

1686

D. M. REVITT end D. B. SOWERBY

spectra of equilibrated mixtures of AsCl, and AsBr, are available [12]. A mixture rich in the trichloride showed extra lines at 129, 142, 178 and 282 cm-l assigned to AsC1,Br; while lines ascribed to AsClBr, at 103,154 and 380 cm-l occurred when the tribromide was in excess. As arsenic-bromine stretching vibrations occur in the region of 280 cm-l, those associated with phenyl-arsenic stretches are expected in a similar position. Vibration t of the phenyl group also occurs close to this energy, thus a composite stretching mode is expected. In polyphenyl compounds, this would be resolved into the components demanded by the symmetry at the arsenic atom. In phenyldichloroarsine the intense infrared absorption at 307 cm-l and the depolarized Reman line at 312 cm-l are assigned to t, while for the diphenyl compound both the Raman and infrared show the required two components. The 10 cm-l separation is consistent with that for the two arsenic-bromine stretching modes in both methyldibromoarsine [l] and arsenic tribromide [3], but should be contrasted with the large separations of the components of t found for lighter central atoms such as silicon [ 10, 111. The region below 300 cm-l contains absorptions due to vibrations compounded from skeletal PhAsCl deformations and phenyl group bending modes, in particular vibration U. A similar suggestion has recently been made by DURIG and his coworkers [ 131 for the spectra of phenylhalogermanes. The two chloroarsines have very similar Raman spectra between 200 and 300 cm-l and of the two bands found the one at higher energy is the more intense and is polarized. In the monophenyl compound both bands cannot arise solely from vibration u and must therefore be a consequence of strong interaction with PhAsCl deformations. The polarized band at 254 cm-l is then the more highly symmetric vibration derived from the A’ (in C,) deformation coupled with U, while the lower intensity 227 cm-l band could arise similarly from the A” deformation. The Raman band at 190 cm-l in the spectra of the chlorides and triphenylarsine can be assigned to vibration Z. Bromine compounds

It is not possible to assign the arsenic-bromine stretching vibrations with the same certainty as for the arsenic-chlorine modes. Although bands occur in the expected positions [l, 31, this is also the region where vibration t is expected. It seems clear that, owing to the similarity in the masses of arsenic, bromine and the phenyl group, all absorptions below 400 cm-l will result from vibrations of the complete molecule. There should however be similarities between the spectra of the phenylbromoarsines on one hand and those of arsenic tribromide and triphenylarsine on the other (see Fig. 1). For phenyldibromoarsine, there are three Reman bands with infrared coincidences in the 300 cm-l region. The highest energy band at 312 cm-l is in the same position as for the corresponding chloride and may therefore be mainly associated with vibration t. The two lower energy bands could then be mainly AsBr, stretching modes, but in this case it is the lower energy band that is strongly polarized in contrast to the situation with arsenic tribromide, methyldibromoarsine and phenyldichloroarsine. With diphenylbromoarsine there is a strong, broad band at -290 [12] M. L. DELWAULLE and G. SCHILLINO,Coopt. Rend. 244,70 (1967). [13] J. R. DURIG, C. W. SINK and J. B. TURNER,Spectrochim.Acta 25A, 629 (1969).

Infrared and Raman speutraof methyl- and phenyl-arseniahalides

1587

cm-1 in both the infrared and Raman with at least one shoulder on the high frequency side. This is very similar to the triphenylarsine spectrum. In the simplest analysis, three absorptions would be expected here, i.e. two components oft and an arseniobromine stretch. But, due to extensive coupling, the major Raman signal is expected to result from a composite symmetric skeletal stretch coupled with vibration t. The higher frequency shoulder may then be t coupled with the symmetric skeletal stretch.

Ph3As t

f

!

Fig. 1. Spectra of phenylarsenicbromides,triphenylarsineand arsenictribromide. Infi-ared bands are shown above and Raman bands below the line (crossed lines indicate polarized bands, solid lines, depolarized and broken linea those not measured).

The Raman band at 225 cm-l in the dibromide is -signed to vibration u, which in the corresponding dichloride occurred as a doublet at 227 and 254 cm-l. The single line probably implies that there is little coupling between u and the skeletal PhAsBr deformations which now occur at lower energies. For the monobromide, the Raman lines at 246 and 221 cm-l can be assigned to the asymmetric and symmetric components of u. Vibration 2 can be assigned to bands at 195 cm-l though this may now be coupled with skeletal deformations. A medium intensity band at 100 cm-l in the spectrum of the dibromide is associated with the symmetric AsBr, deformation; the analogous vibration for the dichloride occurs at 160 cm-l. 16

1688

D. M. REP

and D. B. SO-Y

Phenyl cr7senk(V) &3&k.s Absorption bands for three phenyl arsenic(V) chlorides are summarized in Table 4 and the spectra below 1600 cm-l are reproduced in Fig. 2. The assignment of absorptions above 460 cm-l follows closely that for the phenylhaloarsines and is not discussed further. The region below 460 cm-l is considered first for PI&Cl, M polarized Raman data Table 4. Infrared and Raman spectra of PhAsCl,, Ph&C1, PhAsCl,

1446 vw

1190 w 117ow

1024 m, p 1001 ma, p

(melt) Infrared 1667 mw 1568 ah 1 1478 * 1440 vs 1383 w 1326 m 1303 mw 1273 w 1182m 1161 mw 1090 mw 1074 mw 1054 m 1019 m 994 a

Ph@sCl,(solid)

InfrlWed

Ram&II

Infrared

1676 m

1666 mw 1667 sh 1 *

1680 m

1569 mw

1440 8

1442 VW

1480 8 1444 YS 1437 m *

1480 VW 1440 VW

* 1330 vw 1280 VW 1185W 1166 VW 1071 w 1022 m 1002 ms

1326 m 1303 mw 1283 w 1179 mw 1160 mw 1090 w 1073 vw 1066 mw 1014 mw 998 sh 991 ma1 983 sh

833 VW

827 w 790w 776 w

737 v8

749 ww

674 B

682 VW 671 mw 616 w

739 “S 720 ah > 678 me 666 sh 1 606 VW 468 ms 426 s

432 ma 430 mw 405 8, p

420 8 398 B

304 V8, p

376 sh 342 sh 329 ma 306 sh 1

1336 vw 1282 VW 1186 w 1162 w 1074 w 1026 m 1004 “S

917 sh 911 mw 1

826 VW

606 VW 464 s

Ph&Cll (solid)

Raman

982 sh 969 m 916 w

872 mw. p 610 w, dp

and Ph&sC!l, from 1600 cm-1 Assignment

1333 w 1306 mw 1280 VW 1182 mw 1160%~~ 1096 vw 1073 m 1066 w 1019 mw 1000 sh 994 m 1 986 VW 962 VW 916 9121 vw 841 8331 vw

746 ms 736 vs

667 m 616 mw

681 B 666 rsh> 610~~ 469 8

396 VW

268 m, dp

292 w 263 w

220 m. dp 196 mw, p 162 VW

364 s

‘I

YAsCl*ex

267 vs 249 sh 234 m 220 vw 203 ms 127 m

* Obscured by nujol.

t+

(see disonssion)

298 va

102 w 7KW

362 va

86w 7ow

274 s

266 B 236 w 219 w 179 VW 166 VW 128 8 106 sh

266 w u+

6CABCl

2+

dCAscl

8 ASCI,

Inbred

I

1689

and Reman ape&n of methyl- and phenyl-erseniohalidea

I

I

I

I

I

v,

I

500

1000

1500

cm-’

Fig. .28(i) I-

1

___________fl._____. . ..__.._..__._~------/c---___ I

1500

1

I

I

I

1000

I

500 Ah

c 0

cm-’

Fig. 2e(ii) Fig. 2(e). (i) Infrared and (ii) Raman spectra of PhAsCl,.

for the melt 8re 8v8ilable; in the other c&888,the compounds wem not sufficiently soluble for such meclsurements to be msde. If the compound h8s 8 trigonal bipyr8mid81 structure with 8n equ8torial phenyl group (0, symmetry), the stretching inodes would be represented by 3A, + B, + B,. By 8n8logy with the d8ta for BIeBAscI,rendthe phenylchloroarsines, 8ll five vibx&ions 81~8expected in the region between 260 8nd 460 cm-l. Three pol8rized Rrtman b8nds 8re expected which would result, in the 8bsence of interaction, from symmetic stretching of the equ8tori81 8nd axial chlorine 8toms 8nd stretching of the 8rseukphenyl bond (vibrcltion t).

D. NT. REVXTT and D. B. SOWERBY

1690

n

1500

I

I

I

I

I

500

1000 Y. cm-’

Fig. 2b(i)

Av,

cm-’

Fig. 2b(ii) Fig. 2(b).

(i) Infrared and (ii) Raman

spectra of Ph&sCl,

Two polarized Raman bands with infrared coincidences are observed at 406 and om-l. In addition there is a depolarized band at 268 cm-l and two strong bands at 420 and 329 cm-l occur in the infrared. It seems probable that the two bands near 400 cm-l arise from stretching of the equatorial chlorine atoms. Mixing with other modes is probably small as the equatorial stretches occur relatively far removed (100 cm-l) from other fundamentals. Further, the diphenyl compound shows a single band as expected in this region and no absorptions occur near 400 cm-l for the triphenyl analogue. It is much more probable for mixing to occur between vibrations 304

1691

Infrared and Raman spectra of methyl- and phenyl-arsenichalide3

1500

500

1000 v,

cm-’

Fig. 2c(i)

1500

500

1000

0

Au. cm-’ Fig. 2c(ii) Fig. 2(c). (i) Infrared and (ii) Raman spectra of Ph&CI,.

involving the axial chlorine atoms and the phenyl-arsenic bond. This kind of interaction has been shown by DURIGIand his co-workers [ 131 to occur for triphenylchloro- and bromo-germanes. The three stretching vibrations of the T-shaped PhAsCl, entity are represented by 2A, + B,. One of the A, modes is easily assigned to the polarized band at 304 cm-l while the B, vibration is assigned to the depolarized 268 cm-l band. There is no evidence for a second polarized Raman band, but a medium-strong intensity band occurs in the infrared at 329 cm-l. This is assigned to the second symmetric vibration and, if this is mainly vibration t, its absence in the Reman can be rationalized. The spectra of both PhAsCl, and Ph&Cl (see Table 2)

1692

D. M. REVITT end D. B. SOWERBY

show that while vibration t is weak in the Raman even in the monophenyl compound a strong infrared band results. The occurrence of five fundamentals is strong evidence in favour of the C,, structure assumed. If the phenyl group were in an axial position (C,) only four stretching modes would be expected and for an ionic formulation [PhAsCl,]+Clthree bands (two polarized) would result. The data in Table 4 show that as the number of phenyl groups in the molecule increases, the infrared spectra in particular become less complex. Data for tetraphenylarsonium compounds are not relevant here because of their ionic constitution. Pentaphenylarsenic, however, would represent the last member of the series but this compound has been examined in the infrared to 260 cm-i only [2]. The simplicity of the Ph&sCl, spectra is not solely a result of its higher symmetry (QJ ; three Raman and two infrared bands with one coincidence would in fact be expected in the stretching region. However, two bands only are found. One reasonable explanation of this is that with the inctreased number of equatorial phenyl groups more complete interaction takes place between the arsenic-phenyl t modes and the axial chlorine stretching modes. This results in the appearance of one symmetric vibration of the whole molecule (265 cm-l) and the corresponding asymmetric mode (362 cm-l). The spectra for Ph&sCl, represent an intermediate stage. The single arsenicequatorial chlorine stretch can be assigned to the infrared and Raman band at 430 cm-l but the vibrations of the remaining Ph&sCl, unit will be mixed. Four modes (2~4, + B, + B,) active in the infrared and Raman, are calculated for this C,, unit. The two A, modes are assigned in the absence of confirmation by polarization measurements to strong Raman bands at 270 and 298 cm-l. If there were no mixing, these would be described as symmetric stretches of the axial chlorine atoms and the Ph-As bonds respectively. Only one strong infrared band is observed which can be assigned to one of the asymmetric stretching modes. Confident assignments in the region below 260 cm-l are difficult. Interactions between skeletal deformations and the phenyl group bending modes u and x are expected to occur. In Table 4 only the general position of these composite vibrations is indicated. EXPERIMENTAL

Methyldichloroarsine was prepared by heating to 100” a mixture of arsenic t&chloride and tetramethyllead [la]. Dimethylchloroarsine resulted by reducing dimethylarsinic acid with hypophosphorous acid [El. Phenylarsonic acid reduced with sulphur dioxide in the presence of the appropriate acid gave the dihalides [16], while the monohalides were obtained from the redistribution of mixtures of triphenylarsine and the arsenic trihalide [17]. Arsenic(V) chlorides resulted from chlorination of the appropriate arsenic(II1) compound at -78” in dichloromethane. The di- and triphenyl compounds were recrystallized from toluene. [14] M. S. KIZARASCH,E. V. JENSONand S. WEINEOUSE,J. Org. Chms. 14,429 (1949). [IS] G. P. VAN DEEC I&LEN, Bu.22.Sot. C&n. Bebe 65, 343 (1966). [16] R. L. BARKER,E. BOOTH,A. F. MILLIDQE and F. N. WOODWARD,J. Soo. C?ms. Itad. 68, 289 (1949). [17] A. G. EVANS and E. WARHURST,Tram. Faraday Soo. 44,189 (1948).

Infrared and Raman spectra of methyl- and phenyl-arsenic halides

1693

Infrared spectra were obtained on either liquid tima or nujol mulls between silver chloride and polyethylene (R&ides Type 60; B.P. Chemicals Ltd.) windows using a Perk&Elmer 52 1 spectrometer. Raman spectra were measured using a Cary 81 spectrometer with He-Ne laser excitation. Polarization data were not measured for Ph&Cl,, Ph&Cl, and Me&&I,. Aokwwl.edgemmats-We thank Dr. K. M. MACKAY for helpful discussions,Mr. M. A. HEALY for help in obtaining the spectra, the Associated Octel Company for the gift of tetramethyllead and the S.R.C. for a Research Studentship (to D. M. R.).