199
Journal of Organometallic 0 Elsevier
Sequoia
S.A.,
Chemistry. 84 (1975) 199-229 L,ausanne - Printed in The Netherlands
THE REACTION OF PERHALOGENOKETONES DERIVATIVES OF SILICON AND TIN
WITH ALLYUC
E.W. ABEL and R.J. ROWLEY Department
of Chemistry.
(Received July l’ith,
Unwers~ly
of Eseter.
Exeter
EX? -fQD (Great
Britain)
1974)
Summary For the purpose of this investigation a wide range of all:lic derivatives cf both silicon and tin were required, and we report tile preparation of these in greatly improved yield by a low temperature modification of the hydrolysis stage of the Grignard procedure. The non-catalysed reaction of (3-alkenyltlns with a number of perhalogenoacetones has been examined, and in all cases alkenyloxy derivatives have been isolated as esclusive products. The interaction of P-alkenylsilanes with perhalogenoacetones foilowvs a different course in the absence of catalysts, yielding alcohols of the general type R,SiCH=CH(R’ )C(Rh3,):OH. In the presence of aluminium chloride, however, significant yields of /I-alkenylosysilanes are formed. An allylic rearrangement accompanies the fi-alkenyltin insertion reaction, but in the case of the fl-alkenylsilane formation, no such rearrangement takes place. From these observations possible mechanisms for the reactions are proposed.
Introduction The reaction of perhalogenoketones and more particularly 1,1,1,3,3,3hesafluoropropan-2-one (HFA) with organometallic reagents has in recent years been a source of considerable interest * and an extensive range of products have now been isolated in which ketone insertion has occurred into a metal-heteroatom bond. This work has been particularly fruitful in the fie!ds of organosilicon and organotin chemistry where Si--H 12-51, Si-_O 161, Si-_N 17 1, Si-_P 181, Si-S [9], Si-As [lo], Sn-H [5], Sn-0 Ill, 121 and Sn--N [7] represent a few of the bonds with which perfluoroketones have been found to react. Ob* For a review
on Lhe lnsenion
mbstrates see ref. 1.
reacl~ons
of organo-metals
and -meMolds
involving
unsrrturated
200 TABLE
I
YIELDS
FROhT
GRIGNARDSYNTHESIS
OF
Compound
ALLYLIC
DERIVATIVES
Tbs
(%I
work
OF SILICON
AND
Previously yield (5)
Me3SlCH,CH=CH~
89
51 [Z]
Me $CH$I(Me)=CH~
61
JO [Zf@l
96
50 (361
89
61.1261
.%(CH,CH=CH2)4
89
35 1271
FhJSiCH:CH=CH:
96
90 [28l
Me$%CH:CH=CH,
85
48 I211
hle?S~(C.H~CH=CH~)~ MeSI(CH:CH=CH~)
3
SII(CH~CH=CH:)~
75
53 [171
PbjSnCH,CH=CH:
3-l 68 53
88 1191 a Ll
hle3SnCHzC(Ne)=CH? MezSn(CH?CH=CH2)? a No yreld reported
TIN report4
preb3ously.
servations of aldehyde insertion into a Si-C bond [ 131, stem from the esperiments of Birkofer et aJ., who showed that reaction of trimethyIsilylmethy1 esf or nitriles with aryl aidehydes in the presence of base afforded trimethylsilyl alJhoxides. Subsequently, esters [ 141, and other aldehydes [ 14-191 have been shown to insert into activated Si-C and Sn-C bonds. Representative of these are the insertions of unsaturated molecules such as a.Uyltrialkyltins [14, 151.
‘ii
-+ R-C-CH&H=
-i- R’ ,SnCH2CH=CH
CH,
ASnRls In contrast to these observations, there are to our knowledge no well authenticated examples of ketone insertion into either Sn-C or Si-C bonds, Koenig and Neumann [15] observed that acetophenone, p-bromoacetophenon and isobutenyl methyl ketone reacted incompletely with allylh-iethyltin at ZOO”, although no products were identified; benzophenone was said to be unreactive. In the present paper we describe reactions of aUylic derivatives of silicon and tin with perhalogenoketones to give products resulting either from metal-carbon or carbon-hydrogen insertion_ Although ailylsilanes 1201 and allyltins 1211 have been known for many years, little attempt has been made to optimise their method of preparation vi the Grignard method, and reported yields are often indifferent (Table 1). n CH2=CHCH2MgX
-i- R,_“MX,
--t R,.,M(CH&H=CH&
+ n MgXz
(n = 1-4, M = Si or Sn) it occurred to us that the reason for these indifferent yields could most reasonably be ascribed to the instability of allylsilanes [f22] and allyltins [23] in the presence of acids and bases. RjSiCHzCH=
CH, -
H3@
R&O&R,
+ CH,CH=CH,
201
This could become of major importance during the hydrolysis stage at the end of the reaction. In keeping with this proposal, we found that when the reaction products of trimethylchlorosi.Iane and aIlytmagnesium bromide were hydrolyzed at -2O”, an 89% yield of allyttrimethylsilane could be isolated. When hydrolysis was carried out at ambient temperatures the yields of allyltrimethylsilane from this preparation were drastically reduced to 30-4075, and the by-product was hexamethyldisiloxane. as expected from an allylic fission reaction in the polar medium. Using this preparative modification, we have been able to greatly improve upon previously reported yields for a large number of allylic derivatives of silicon and tin (Table 1). The only esceptions to this are the tnphenylsityland triphenyltin-derivatives, where reported yields were already high due to their much greater stability to hydrolytic fission [24]. Results
and discussion
Allyltrimethyttin reacts vigorously with hesafluoroacetone --2O”, to give 4,4-bis( trifluorwnethyt)-4-trimethylstannosy-l-butene, uct resulting from a tin-allytic insertion reaction. Me,SnCH2CH=CHz
+ (CF3)1C0
(HFA)
even at the prod-
-, Me,SnOC(CF3)$H,CH=CHz
The perhalogenoketones, l-chloro-1,1,3,3,3-pentafluoro-2-propanone (MCPFA), 1,3-dichtoro-1,1,3,3-tetrafluoro-2-propanone (DCTFA) and 1,1,3-trichloro-1,3,3trifluoro-2-propanone (TCTFA) also react in a completely analogous fashion. The spectroscopic properties of these compounds (Tables 2 and 3) are entirely consistent with the proposed structures. The adducts are hydrotysed to give the expected alcohol and trimethyltin hydroside.
Me3SnOC( Rhal)$H2CH=CH2
H2
CH2=CHCH2C( Rhd )@H
Me,SnOH
Dimethyldiallyltin in a fashion to give both mono- and di-insertion products in good yield, dependant upon relative reagent ratios. An attempted reaction of tetrallyltin with HFA afforded only non-distil!able polymeric material. The formation of polymeric materials has also been noted [ 301 in a similar reaction of tetrallyttin with sulphur dioside. It is interesting to note that this insertion reaction can also be estended to cyclic alkenes; thus indenyltrimethyltin and cyctopentadienyltrimethyltin react with HFA and MCPFA to give the expected Sn-C insertion products_
(CF,),CO
-
CF,
ae
3ava
2-Butenyltrimethyttin upon reaction with HFA the expected products horn a simple Sn-C insertion rearrangement of the allytic fragment.
or MCPFA
9:
did not give
reaction, but a complete (continued
on JJ. 221)
202
x 0
Y
8
2
(CHJ)~SI(CHZCH=CH~)~ CF2CICOCPC12 (l/l)
CFTCI
CF>
____.___._
---p--
CF2CI
irlti-1.18~0.03
107 10.36
lO!)fO.O3
\
jxCI:
SKCH:CH=CH$~ + CF3COCF3 (112)
CH~CH-rCH / KH&i-CH=CHCHzC’-OH
15010.01rl
7310.03
U.P. (“Clnm) I hl.lr.1 (“Cl
_____-l_-_--_-*----
127-12W0.01
(l/4.4)
.---_-
_
CH=CHCH+IH
--
CF~CI / ICH~)ZSI(CH=CHCH?C-OI~ \ CFCI~) 2
I
Product
Sl(GH:!CH=CH~),~ + CF,COCFJ (112.6)
CF3COCF3
).I *
+
SlWH2 CH=C&
t
t
we--
(CH3)2S1(CH2CH=cHI): CF$XOCFCI~ I t1/2.lt
CH,Si(CH2CH=CH2)3 CF&OCF:CI (l/l)
Renctonts (rcogenl ratlo)
TABLE 2 (conrlnurd)
41
42
4b
30
40
30
.----_______-_I--
Yivld (%I
lOO”l4Bh
100” 14811
10b”/48h
too” I24h
100G/46h
100” /2&r
(cm-’ 1
1R
35u5. MOtj
3695.349G
3GSO. 3800
3GGO. 3600
3G60.3600
36&0.3bOO
II
Yclecrctl
colldltiolls~
-._-_. RCWtiOl~
1634,lGlO
1634.1607
1604
163L, 1607
X06
1631, 1 tiO!I
laxa
--
alimrptions
+
(CHJ)~SICH?CI~=CHCH~ CFJCOCF~CI (l/1.1)
FHI,CFJ
CF2Cl
CF?Cl
CF3
cr3
W
:,F?,,
+
(CgH~)3SICH2CH=CH~ CF2ClCOCF:Cl (l/2)
(C&)$ICH=CHCII2C-OH
/ U&H5)jSlCH=CHCH$-OH \
+
CF.I / -CII:CH=CH, \ CF.\
(Cf,HI;)3SICH2CH=CH2 CFJCOCFZCI (111.3)
(cH~)JsIoc
CHI CF> I / KH>)+ICH=CHCIl-011 \ CF2CI
(CHJ) ISICH=CHCHC-OH \ CF1
-_-_____
P_---_--_pe
/ (C~,HS)JSICH-CHCH:C-OH \
+
-------_
Product
(C(, HI;)$XH2 CH=CH2 + CFlCOCF3 (l/1.1)
(CH3)$XH:!CH=CH;? CFJCOCF~ (111.1)
VXJ)$KH~CH=CH~ + CFJCOCF~ (l/l.l/O.Ol~ICl~~
+
(C)IJ)JSICII:CH=CHCH:, CFJCOCB~ (111.1)
Rcnc1nnt.n (rcogonl riltlo)
TABLE 2 (continued)
162-641
(GO-711
f55.671
G8/14
38/l 6
37/0.00
3811
B.p. (‘C/mm) IhlmlfC)
4n
63
GG
41
61
66
60
Ylcld
(‘%I
--
lU0” /24h
80p/24h
GO’ l2811
-20°/241,
?6”/2411
1OG0/2GOh
1Q0~110011
Roncllon condillon5a
---
3676.3620
3670.3620
3696.3620
1GOG
1GOG
1607
lG42
1604
1602
lJ(C-=a
1R nbsorwlons
3G76.3BOO
3580.3510
Non)
Sclcyd (cm )
+
CFJCOCF~CI (l/1.1)
(CH$$inCH2C=CHz +
YHI
(CHj)3SnCH:C=CHz CFJCOCF~ (l/3.1)
FH3
(CHJ)3SnCH2CH=CHz ccI~cocc13(1/2)
+
(CH3)3SnCH2CH=CH + CF~CICOCF’CI~ (111.6)
(CH#inCH~CH=CH2 CF2CICOCB:Cl (1/1.2G)
(CH3)3SnCH2CH=CH2 CF3COCF2CI (l/1.1)
(CHJ)3SnCH2CH=CH2 CFJ COCFj (111.1)
(CHg3SICgHS + CF3COCF2CI (111.1)
(CH3)3SIC5H5 + CFJCOCF~ (111.1)
W-3
CF:CI
CF.l
4511
-.-__ -~ ..-.-.
_.__
TH?
(CH3)>SnOC-CH~C=CII~ \ CF-I Cl
/ CF?
Y’ F”-’ (CH~)~SnOC -CH:C=CIIz \ WI
ccl>
5310.9
--
68
67
46 1J
/ ccl)
(CHJ)3SnOCq-CHzCH-CH:
85
81
12
35
56
86 1’
/
GO/O.006
4611
ti9113
6310.04
5G10.3
(ClI])~SnOC\C~,~~ICB=CH?
cm2
CF~CI
CPlCI / (CHJ)JS~OC\CH$H=CHZ
(CHj)jSnOC
CF3 / -CH~CH=CH;? \ CFzCI
CF.I
(CH3I$nOC
CFJ
/ (CHJ),SI(C~IIJK-OH \
(CH3)$%CsHsqlC--OH \ CF,
/
26“/12h
2’;’ /I 211
GO=/24h
26” /24h
26” /24h
80“ 1311
2G012411
100’/48h
so” 14811
3680.3600
3596,349G
1646
lG4b
1641
1641
1641
1641
1633
208
f .2
m P
B
---
+ (l/1,:1)
(CH3)JSnCl,Hy + CF~COCFJ (1 n.1)
(CH3)$hCgH5 + cF3cocF2c1(1/l.1)
CF3COCF2CI
Wh)$SnC5H5
Reeclsnls (rongent ratio)
TABLE 2 (continued)
--
--
\ CF,CI
H
/ OSn (CHJj3
F3CHc
Producl
h
83/0.01
G110.006
G110.006
1M.p.l (‘C)
R.P. (‘C/mm)
86
33
34
__--
Ylold (W)
2ti0131h
25” 12511
25” /24h
Reaction condlllons” 140H)
(cm”) ..__.-_-_ _---
u(C=C)
Srlecwd IR nbsarptlone
N l-a 0
(l/1.1)
*
__c----
/\ CF3 OSn(CH3)3 _----
9210.005
79
2tFl24h
_-_I__-.
_.-“l._s.-
.“-l--
..I_,.“-l... .._.
a Renctlon condllions are tnbulnlcd es trmpcramrelreucrion llmc. f~These products wete unstohle Lo d~stilln~lon, und figures xepreicnl yields after the xcmovul of volntUo mnterinls under vncuum.
CF~COCF:CI
(CH3f35nCgH7
._“_..^“.” ..”..-.. II-^.
212
--
/_CHb=CHcCHz
CH;S;;CH;CH’=clI;)~
(CH’;,2
C\
d/
____
CF2CI
OH’
CFJ
-_______-______
9.70. s: (b, c) 3.83. rni 7.00. d. J 7Hz; (0 LCO. P
9.Gl. 5; (b. c) 3.66. m: 7.12. d, J 6112. (c) 5.55. s
(I)
_---_.--
---
.-.-
8.3f.i. d. J 811~;(h) 5.11. m
(U) 9.79.9: (b. c. F) 3 90. m: (d) 6.92, d. J 7Hz; (c) 6.67. :n:
_ ..-.-
I.---_
-
CFJ CF,
___.._
--_
73.9. ~‘5 12147. Gl 6. q. ‘J 12Hz
----
74.4. dl. ‘5 12Hr 61.7. dq. “J 12Hz CFI CF:
9.GG. 9; :b. c. 6) 3 90. In; 7.1G. d. .I t;.!iMz: (c) 6.70. 5; 8 33.11,.I 7.6ik. (11)5.1 5, rn
(u)
(d) (f)
74.0. IL. ‘J 12Hz 61.7. t.q. ‘J 12Hr
cFJ CF:
9.51. s: (b. c) 3.70. m: 7.12. d. J 6.5tf1: (c) 6.77. s
(a) (d)
76.3.s
63.3.5
CF)
CF-
(11) 9.79.9: (b. c. 6) 3.!10. m: (rl) 7.19. d. .I G.514~; (c) 6.83. s. (I) 8.33. d. J 7.6H:; (Ii) 6.08. m
(0) 9.73. s; (b. c. 6) 3.82. m: (11) 7.35. d. .J 6 5Hz; (e) G.67. s: (I) 8.23. d, .I 7.5Hz; (11)5.12. m
(id (d)
(n) 9.81. s; (b. c. 6) 3.83. m: (dl 7.02. d. J 7Hz; (e) ti.66, s: (f) 8.37. d. J 7.6Hz: (II) 5.15. m
(n) (d)
fconllrrllcdl
13
E;
.. x h
P
-a
215
CF.3
(CH$,SnOC
&H;CHC=CH; \ CF2Cl
CF2CI
=CIiz
-CIlbCHC-CHd \ z -: CF3
CFJ
‘CF? Cl
CF:,
/ bc (Cll~)3SnOC-CH2CH \ CF:CI
(C1$)lSnOC
/
/ (CH’;)~Si(C5Mjb)c,~lic
CFq
OHC
CI?J
d
2 ‘CF:CI
(C I3H’? 5 3SICHb=CHcCHd~/~~C
a (CH3),SI(C5H3) bC/ \
F2CI
,CF:CI
\c
CFJ d/ (C,,H”s)~S,CHb=CHCCHIC-OHC
9.96. s: (b) fLA1. m; 5.01. m; (d) 7.2G. m
(T vulues)”
-----
2 GG,In. (Il. r) 3.G3, m; 7.39. d. J 6Hz; cc’) 7.G3. 3
9.Gl. s. “J GGHz: (b) 7.37. d. J 71I7; 4.11. m; (d) 4.90. m
9.G.l. 3. “J GGHz: (b) 7.44. d. J 7H:: 4.11. m. (d) 4.91. m
9.98. s: 9.81, s: (b) 3.34. m: 6.63. m; (cl ti.71, s
(10 9.GO. Y. “.I GGH?; (b) 7.33. d, J 71-11: k) 4.18, m; (d) 4.OG. m
(cl
(0)
(0) k)
(il)
(II) 9.97.~: D.79. Y: (b) 3.30, m: G.tiG. m: (c) LGG. s
(II) 2-M. m: (b. c) 3.65. m: (d) 7.41 d. J 7117; (e) 7.36. s
(dl
(it)
(10 2 GO. m: (b. 49 3.60. In; (d) 7.44. d. J GHr; (I?) 7.73. s
(c)
(II)
(CI~)JSICHbCHc-CHd !
I I2 o-c(cFJ):
I H Ihsonuncrs
---
Compound
TAULE 3 (continued) ----_.--_
7G.3. s
70.0.5
GB.8. s
73.2. L. “J l”llz 61.7. (1. “J 12H~
76.!1. 9
79.6. Ill
---. (ppm upflcld from CFCIJ)
CF2
CF:
57.9.5
60.0. q. ‘5 12Hz
CF.3 72.4. L, ‘J 12Hz
CPJ
CFJ
CF:
c I,‘2
CF’>
Cc>
CI:J
“‘1: Rcronanccr
-
15 w 0,
CF3
-CHbCHC=CHd \ 2 CFJ
/
(C,I$),SnOC-
(C,H$JSnOC
’ \
‘CF? Cl
CF3
C@-CHd=CHf
CFJ ,CH;
CF2 Cl / -CHbCHc=CHd \ = CF2 Cl
(CgHt)3SnOC- <%&d=CH;
(Q,Ht)$nOC
Compound
TABLE 3 (continued)
?
2
2.W m:(b) 7.42. d. J 7Hr: 8.06. m: (d) 5.04. m
2.67. m;(b) 7.2G. m; 8.81.d.J7Hz:~d)~l.lI(.m. 5.16. m
2.68. III; (b) 7.27. d. J 7142 4.12. m:(d) 6.01. m
2.ti6. m;(b) 7.02. m; 8.71, d. J 7Hz: (d) 4.16, m: G.13. m
01) 6.88. s (lx): (b) 3.41. m. (c) 9.19. s. OJ 56Hz
(n) (c) (e)
(II) 2.67 1rn.3(b) 7.1”. . m’. (c) 8.78. d. J ~HL; (d) 4.08. m; k) 5.16. m
(u) (c) (c)
(0) (c)
(II) 2.61. m; (b) 7.3G. d, J 7Hr: (c) 4.16. m: (d) 5.12. m
(n) (r)
’ II Resonances (T vnlues)”
CFJ
CFx CP2
CFj
CF2
CFJ
7G.l.s
68.5. m 56.0. m
71.3.9.
JJ 8Hz
71.5, g. ‘3 &Hz
GG.8.s
7G.O. :h
“‘I.. Resnnonres (ppm upflcld from CFCl3)
218
N
-_ 2
219
6.13. s(bt); (b) 3.61. m;
couplmg constnnl In tho appropwtr
(01 C.04, s(br); (b) 3.4h, m; Cc) 2.87, m: (dl2.17, n; (c, 9.W s. 4I 52%
(cf 2.V2. m: Cd) 2.32, m: ‘$1 9.6% s, aJ 8211/.
(a)
’ tl Rcsonuncc (7 s~lues)~
a frotono nte indicrtted by subscriprs: “J Is Lhe unrerolved 119~1’7Sn-C~~
Compound
TABLE; 3 kontlnued)
72.8. 10.2, 69.0, 56.8,
msertian produc(.
CF:
cF>
t, ‘I5 12Hz t, ‘J 12Hi m m
221
Me$nCH,-CH=CHMe
(CF:,),CO
+ Me$Q0-C(CF3)~CH(Me)CH=CI-IL,
Similar with 2-butenyltriphenyltin result in skeletal rearrangement. of this have been previously in reaction of [ 14, and sulphur [ 321 P-alkenyltins, and have been in terms of cyclic transition state. This mechanism seem to be equally the perhalogenoketone insertions, where reaction would be promoted by the electron deficient and weakly polar-
ised carbonyl group of the substrate. The observation that other labile groups attached to tin, such as benzyl, fail to react with HFA in this way is additional evidence favouring this mechanism, since these other substituents are unable to rearrange in an allylic fashion. In complete contrast, when allyltrimethylsilar~e is allowed to react with HFA in the absence of catalyst three products may be isolated in varying yield dependent upon conditions. When the two reactants are heated at 140” for 48 h, the alcohol I is the exclusive volatile reaction product (85% yield). The heating of I with a three-fold excess of HFA at 120” for 40 days failed to induce any further reaction. Allyltrimethylsilane and allyltriphenylsilane react at 100” in a similar fashion with the perhalogenoketones HFA, MCPFA and DCTFA to give the analogous alcohols as products (Tables 2 and 3). Similarly dia.lIyldimethylsilane, triallylmethylsi!ane and tetraallylsilane react with HFA and MCPFA with suitable variation of reagent ratios to produce good yields of analogous alcohols which result from partial and complete ketone insertion (Tables 2 and 3). [Ne,SiCH=CHCHC(CF&OH (I) Me,biCHCH=
CHz + (CFJ)CO
MeSiCH2
-CH-CCH? I I 0 --c(CFd, w
When the reaction between allyltrimethylsilane and HFA is allowed to proceed at 50”) a second product, the oxetane II is obtained (15% yield) in addition to I; and the yield of osetane may be raised to 40% when the reaction is carried out at -20”.
222
The reaction of 2-butenyltrimethylsilane with HFA afforded the alcohol no skeletal rearrangement had occurred. This observation is in sharp contrast to that with 2-butenyltrimethyltin: Ii1 in which
C ha1
Chd
Me,SiCH,CH=
CHCH,
+ F.c,”
= 0 - Me,SiCH=
, C hal = CF,
CHCH( CH,)C-
/
\:,”
(III)
3
or CF,CI
The incursion of steric strain in the Lransition state is evidenced by the slower rate of reactlon of Zoutenyltnmechylsilane relative to the unsubstituted precursor, and the fact that 2-butenyltriphenylsilane fails to react with HFA at 140’. In our view, these observations are most reasonably accomodated within a mechanism involving a six-centre process in which there is a significant polar contribution to the transition state. CFs CF3
-
-I
(CH3)3Si
(C H313 SI
4s an explanation for double bond migration in olefin/HFA reactions, a concerted mechanism has previously been proposed [ 33-361 involving a siu-membered cyclic transition state and this is illustrated below for propylene. CF3
&-&CF3 CH I
II
p
CH2=CHCH2C,-OH
/
,=F,
Adelman [37] has criticlsed this mechanism and proposed instead a fourmembered cyclic dipolar intermediate IV, on the basis of product studies with a sensitive dlfunctional trapping agent, ally1 glycidyl ether. (CF,),CO
+ CH,=CHCH,R
-
CH2T&CH:R KC\/ 7
: &
j
CH,-CH=CHR W,
I
C-OH
F,C’ (IV) An appreciable solvent effect was observed in the reaction of DCTFA with 2,4,4-trImethyl-1-pentene, and large differences in rates of adduct formation were observed for olefins containing electron withdrawing and donating substituents. F,C
323 It is also pointed out that the high rate of reaction of olefins with HFA has not been observed Previously in non-catalysed 1,5-hydrogen shift reactions believed to proceed in a concerted fashion [ 38-401. The observation that there was no rearrangement in the reaction of o-pinene with HFA (rearrangement in fl-pinene reactions has previously been taken as good evidence for carbonium ion intermediates), was related to geometrical limitations to orbital interpenetration rather than the absence of carbonium ion intermediates. However, evidence [ 33, 361 that reactions of HFA with terminal olefins lead esclusively to 1,1-bts(trifluoromethyl)-3-alken-1-01s seems difficult to rationalise in terms of carbonium ion formation. Likewise, the formation of (3-silyl carbonium ions in the reaction of allylic silanes with perhalogenoketones would seem unreasonable because of the known instability of these Intermediates [22, 411. In the mechanism we propose, a significant increase in rate would be espetted, relative to concerted 1,5-hydrogen shift reactions, but free carbonium ions need never be produced and rearrangement would not be expected. Whilst oxetsnes prepared previously 1421 by photoinitiated additron of fluoroketones to fluoroolefins were suggested to arise via diradrcal species, the osetane IT more probably forms via a four-centre process which again has polar characteristics. Oxetanes derived from vinyl ethers have been shown to isomerise thermally to give substttuted a!kenes.
F3C
F3C,
F,C
OH
CH2-CH-OR
However, this is discounted as a significant route to the formation of I since the osetane II fails to isomerise to I at measurable rates below 180”. The product from reaction of cyclopentadienyltrimethylsilane and HFA shows a strong IR absorption at 3500 cm-’ again suggesting alcohol formatton. The proton NiVlR spectrum is interpreted in terms of a mixture consisting predominantly of V and VI although other isomers may be present in smaller quantities.
I c?
H
\
i-
(CF3J2C0
-
SIMe,
3
(PI
(Eln, IV, VI produczd in ratlo
1011)
Cyclopentadienyltrimethylsilane and MCPFA react in a similar manner give the espected alcohol which again exists as a mixture of isomers. When aliyltrimethylsilane is allowed to react with HFA in the presence 1% aluminium chloride, I (13%) is formed together with two other products VII (51%) and VILI (16%).
to of
r Me,SiCH,CH=CH,
+ (CF,),CO
15T-,A’c’3
I I
i
/c
,F,
) Me,SiCH=
CHCH,C-
(I, 13%)
\CZH
3
,Ch j Me,SiOCeCHCH=CH, (VII,
512;’
hle,SiCH?CH=CHC-OH (VIII,
16%)
/’
CF,
‘\CF,
It has been suggested [ 401 that the aluminium chloride-catalysed reaction of olefins with HFA proceeds through a Friedel-Crafts type of reaction involvin a free carbonium ion (or equivalent associated species). Reaction then proceeds via an initial electrophilic attack of the HFA/AICI, comples at the unsaturated centre. This intermediate can then rearrange by proton elimination or by tri-
3
methylsilyl migration to &e the observed products. This product distribution is also consistent with the absence of carbonium ion intermediates in the non-catalysed reaction. The non-catalysed and aiuminium chloride catalysed reaction of allyltrimethyltin and HFX afford 4,4-bis(trifuoromethyI)-4-trimethyl-stannoxy1-butene in each case. However. in the latter reaction, considerable quantities of trimethyltin chloride (30%) are also formed, suggestive of polar transition state intermediates leading to trialkyltin fission.
Speclroscopy The reactions of perhalogenoketones with /3-alkenyi derivatives of tin lead in all cases to products resulting from an insertion reaction into the allyl-tin bond. The structures of the products are established unambiguously by means of their ‘H NMR and IR spectra. Typicaily, NMR spectra show terminal allylic patterns and methylene shifts consistent with direct. attachment to -C(C,,,),rather than to oxygen. The IR spectra show a strong C=C stretch in the 1639-1646 cm-’ region consistent with a terminal alkenyl-linkage well removed from the tin atom. The product from reaction of crotyltrimethyltin and HFA has a ‘H NMR spectrum which shows a typical three proton, terminal allytic pattern, fully in accordance with a complete rearrangement of the aI.Iylic fragment. The downfield shift of the methine proton r 7.31 also supports the proposed structure. The non-catalysed reaction of aUyltrlmethyisiIane with HFA affords the
225
alcohol I as the major product. The proton NMR spectrum of I shows the pattern of an isopropenyl linkage centred at r 3.80 and the large downfield methylene shift to r 7.17 is consistent with the close proslmity of the hydrosyl and trifluoromethyl substituents. The absorption at 7 6.92 disappears upon eschange with deuterium oslde and is assigned to the hydroxyl proton. The IR absorptions at 3595, 3515 (OH stretch) and 1602 cm- ’ (vlnylic C=C stretch) also support this structure. The silyl et.her VIII predominates in the renctron product mixture when the interactlon of allyltrimethylsilane with HF.4 is catalysecl by aluminium chloride. The ‘H NMR spectrclm of VIli bears strong resemblances to the allyltin insertion product spectra and the absence of an IR OH absorption and the wit,h shift of the C= C stretching frequency to 1641 cm- ’ are also in agreement this structure. By examining the ‘H NMR spectrum of the product from interaction of crotyltrlmethylsilane and HFA it is immediately apparent that no rearrangement has taken place; of particular note In this spectrum, is the two-proton olefinic absorption centred at ; 4.00 and the large downfield shift of the methine proton to r 6.96, indicating direct attachment to the -Cl,CF,)!OH substituent. The product from reaction of cyclopentadienyltrimethylsilane and HFA shows a strong iR absorption at 3500 cm- ’ , again suggesting alcohol formation. The proton NMR spectrum is Interpreted in trlrms of a misture consisting primarily of V and VI although other isomers may be present in smaller proportlons. The isomer V shows a sharp singlet at r 9.97, a singlet nt 7 6.66 which exchanges with deuterium oxide, and broad absorptions at r 6.50 and 3.15. Absorptions at 7 9.79, 6.66 and 3.40 are assigned to V. Previous studies on cyclopentadienyltrimethylsilane (341 have shown that at room temperature the 1 isomer predominates converting by prototropic
with small quantitk of the 2 and 5 isomers interand metallotroplc shifts. The downfield shift of the
trimethylsilyl protons to ; 9.79 is characteristic of this group changing its point of attachment from an sp” to an sp’ carbon atom [ 451. The insertion products of hlCPFA have previously shown [16, 471 interesting features in their ” F NMR spectra due to the asymmetry developed in the adduct and they have been analysed in terms of ABX; systems. The ’ 'F NhIR spectrum of IX shows this typical pattern; CC is the centre of asymmetry in the molecule and this renders the two fluorine atoms on C” non-equivalent. Coupling within the AB system (““F-J gives two doublets Lvhich are further spl!t into four interpenetrating quartets by coupling wit11 .X3 (COF,). The CbF, signal itself appears as a triplet by virtual coupling with the C”F? group. This feature is not generally observed in the other nlkene Insertion products with RICPFA and In these cases it is likely that AB is too small for splitting to be observed. ,,C”F&I hle,Si(C,H,)C’,-OH (IX)
‘CbF,
Ekperimentd All melting
points
and boiling
points
are uncorrected.
IR spectra
were
ob-
226
tamed with a Perkin-Elmer 257 gratrng spectrometer and NMR spectra were determined using a Perkin-Elmer RlO spectrometer operating at 6OMHz for protons and 56.44 MHz for “) F nuclei, and with a Jeol JNM-MH-100 1OOMHz spectrometer, either as neat liquids or solutions in carbon tetrachloride. The analytical data for a representative cross-section of insertion products are given in Table 4. All reactions and subsequent manipulations, as a matter of course, were conducted under an atmosphere of dry nrtrogen and solvents were dried prior to use. The following were prepared using previously reported methods; 2-butenyltrimethylsilane [ 481, cyclopentadienyltrimethylsi)ane [ 331, (trrmethy!stannyl)cyclopentadiene [ 491 and benzyltriphenyltin [ 501. TABLE-l ELE>IENTAL
.\NALYSES
OF REPRESENTATIVE
Compound
COhlPOLINDS Anaibsls
found
(calcd.)
(‘G)
C
H
38.7 (38.6)
(5.0)
36.5 (36.4)
(4.7)
3-l -l (31.5)
(4.4)
314 (32.3)
t-t.?)
43 7 (43 -!)
4.7 (4.6)
11 ” (41 1)
4.6 (4 4)
61.4 (61.8)
4.4 (-1.3)
59.8
1.3 (-?.?-)
4.6
,.CF:, hlqS,CH=CHCH&---OH \
CF,CI
4.;
4.3
_CF:CI \le$SiCH=CHCH2C/OH \
CFCI
2
-1.1
CF:,
/
hIe3SICjHAC-OH \
CFzCI
Ph$fGiCH=CHCH,c-OH
/
‘\
CF3
CFx CF.J
P~,SIC”=CHCH+-OH CF:CI
(59.i) 5i.8 (57.7) 29
hTe3SnOC-
/’ \
4.0 (-l.O)
4
3.7
(19.2)
(3.8)
27.9 (2i.9)
(3.6)
CFI CH$H=CH:! CF,CI
3.5
227
The synthesis of allylic precursors via the reaction of fl-alkenyl Grlgnard reagents with the appropriate chlorosilane or chlorostannane is illustrated by the reaction of allylmagnesium bromide with silicon tetrachloride. The yields of other materials using this modification are shown in Table 1. Silicorz tetrachloride with allylrnagnes~um bromide 30 g (0.176 mol) of silicon tetrachloride was added to a solution of 0.9 mol of allylmagnesium bromide in 600 ml of diethyl ether at a rate sufficient, to maintain gentle reflus. After addition the slurry was reflused for 2 h. The mLxture was then cooled to -20” and a 10% solution of ammonium chloride with efficient stirring over a period of (= 200 ml) was slowly added dropwise 2 h. After this time two clear layers had developed and the organic phase was separated. The aqueous phase was estracted with 3 X 100 ml of diethyl ether and the combined organic phases were dried (MgSO.,) and fractionated. The yield of tetraallylsilane was 7O.Sg (88%); h.p. 86.5”/10 mm [lit. [ 271 b.p. 87”/10 mm]. NMR: T 4.26 (m, 4H, C=CHC), 5.13 (m, SH, CH?=C) and S.52 (d, SH, J SHz, CCH,Si). The above procedirre was followed in the preparation of the following; aIlyltrimethylsi!ane, b.p. 84-85.5” [lit. [22] b.p. 84.9’1; (2-methylallyl)trimethylsilane, b.p. log”/747 mm [lit. [25] b.p. 110.5-112”]; dimethyldiallylsilane, b.p. 134-136” [lik [26] b-p. 136.8”/759 mm]; methyltriallylsilane, b.p. 67”/50 mm [lit. 1261 b.p. 68”/50 mm]; allyltriphenylsilane, m.p. 88-89” [lit. [ 281 m-p. 8%89”]; allyltrimethyltin, b.p. 50”/35 mm [lit. [ 511 b.p. 128-129”/ 767 mm]; tetraallyltin. b.p. 52”/0.2 mm [lit. [27] b.p. 52”/0.2 mm]; allyltri-
phenyltin, m.p. 72-74” [lit. [29] m.p. 73.5-74.5”]; (2-methylallyl)trimethyltin, b-p. 46’/18 mm and dimethyldiallyltin, b.p. 67”/16 mm. 2-ButenyItriphenylsilane A solution of 3.17 g (0.035 mol) of l-chloro-2-butene in 10 ml of tetrahydrofuran was added ciropwise to a solution of 0.032 mol of triphenylsilyllithium in 60 ml of tetrahydrofursn cooled in ice. After addition the misture was hydrolysed with 30 ml of water and the organic phase was separated, dried (MgSO,) and concentrated. Recrystallisation of the residue from light petroleum (t3.p. 60.80°) gave 6.4 g (60%) of product m.p. 69-71” NMR: 7 2.67 (m, 15H. Ph,Si), 4.65 (m, 2H,CH=CH), 7.78 (d, 2H, J 5 Hz, SiCH$) and 8.46 (d, 3H, J 4.5 HZ, CHjC). Z-Butenyltriphenyltin Using a similar procedure, 2-butenyltriphenyltin was prepared from lchloro-2-butene and triphenyltinlithium in 73% yield (as a misture of cis and tram isomers): m.p. 56-58” [tit. [48] m.p. 57-59”] NMR: 7 2.74 (m, 15H, Ph,Sn), 4.55 (m, 2H, CH=CH), 7.68 (d, 2H, J 6 Hz, J (“4”‘7 Sn-CH?) 66 Hz not resolved) and 8.44 (d, 3H, J 5 Hz, CH&). 2-Butenyltrimethyltin A solution of 0.09 mol of trimethyltinlithium was prepared from l$?,g (0.1 mol) of trimethyltin chloride and 1.74 g (0.25 mol) of lithium wire in SO ml of 1,2-dimethosyethane. The solution was filtered from excess lithium,
cooled m ice and 9 g (O-l mol) of I-chloro-2-butene was added with stirring. Work up in the usual fashion afforded 14.2 g (65%) of product (as a mixture of cis and trans isomers), b.p. 53-55”/6 mm [lit. [48] b.p. 148-154”] NMR T 4.72 (m, 2H,CH=CH), 8.44 (m, 5H, CH,C, CH$%) and 9.94 (s, 9H, 52.5 Hz, J(“‘Sn--CH3) 49 Hz, MeJSn). J( “9Sn-CHJ)
(TrimettzylstannyL)indene Dimethylaminotrimethyltin 1491 5.0 g (0.024 mol) was added dropwise to 11.6 g (O-l mot) of redistIlled indene. The misture was then fractionated to gwe 5.25 g (83%) of product, b.p. 84”/0.07 mm [lit. [52] b.p- 64”/0.015 mm] NMR 7 3.74 (m, 4H, C,HJ), 3.42 (m, 2H, CH=CH), 6.42 (m, IH, CHSn) and 10.15 (s, 9H, J( “9Sn-CH,)54 Hz, J( It7Sn-CHJ) 50 Hz). The reactions of perllalogenoketones with p-alkenylsiianes and stannanes were carried out using degnssed mixtures of the appropriate reactants, sealed in Carius tubes and heated to the required temperature. Interactions with solid reactants were carried out in benzene as 20% solutions. 1,1,1,3,3,3-hexafluoro2-propanone and 1-chloro-1 ,1,3,3,3-pencafluoro-Zpropanone were handled using conventional vacuum transfer techniques. The physical and spectroscopic properties of the products are reported in Tables 2 and 3.
2. .%Bis(trifluoromethyi)-4-(t-rime thyisilyimeth yl)oxetane(Il) .A mixture of 0.01 mol of allyltrimethylsiIane and 0.011
mol of 1,1,1,3,3,3herafnuoro-2-propanone were allowed to react at -20” for 24 h. Distillation afforded a misture of Ii and l,l-bis( trifluoromethyI)-5-trimethylsilyl-4-pentenel-01 (I). This mixture was chromatographed on alumina and elution with light peYoleum/benzene (20/I) gave the product in 30% yield (a crude yield of 40% was estimated from NMR analysis of the initial distillate).
References 1 2 3 4
h1.F. G.W. A.F. A.F.
5 6 7 8 9
W.R. R..% EW. E.W. EW.
10 11
12 13 l-l I5 16 17
18 19
20 21 22
bppert and 8. Parshall. lnorg. Janzen and C.J. Jsnzen and C.J.
Prolix. .4d%an Orgsnometal. Cbem.. 5 (1967) Chem.. -l (1965) 52. Wdhs. Can J Chem . 13 (1965) 1963. Wti. lnorg. Cbem.. 10 (1967) 1900.
225.
Cullen and G.E. Stvan. Inorg. Cbcm . 1(1965) 1437. Braun. Inorg. Cbem.. 5 (1966) 1831 Abel and J.P. Crow. J. Chem. Sot. A. (1968) 1361 Xbel and I.H. Sabherwal. J. Chem. Sot A. (I 96s) 1105. Abel. D.J. WAkerand J.N. Wingfield. J Chem. Sot. A. (1968) 26.42. EW. .\bel tid S.hl. Illux.vorth. J. 0rg;momeral. Chrm., 17 (1969) 161. A.J. Bloodworth ;ind A.G. Dab%%. PTOC. Chem. SIX.. (1963) 325. X.G. Davies and W.R. Symes. J. Cbem. Sot. C. (1967) 1009. L. Bukofer. .4. Rltter and 1-i. Wleden. Chem. Bcr.. 95 (1962) 971. C. Sevens and hi. Pereyre. J. Or@nometd. Chem.. 26 (1971) C-l. K. Roerug and W.P. Neumann, Tetrahedron Lelt.. (1967) 495. S.U. Pooomuev ad I.F. Lulsenko. Zb. Obticb. am.. 34 (1964) 3450. J.G. Noltes. H M.J.C. Crecmers and C.J.M. van der Kerk, J. OrganomeW. F.H. Fmherion and SF. Thamcs. J. He:eroc>cl. Chem.. 9 (1972) 67. F.H. Pmkerton nnd S.F. Thames. J. Pamt Technol.. 43 (1971) 6i: Chem.
Cbem.. Absrr..
11 (1968) 76 (1972)
D.T. Hurd.J. Amer. Chem. Sot.. 67 (1945) 1813. A.D. Petrov. U.F. hlrronov and I.E. Dolpu. Izv. Akad. Nauk SSSR. (1956) 1146; Cbem. 51 (1957) 4938. L.H. Sommer. L.J. Tyler and F.C. Whltmore. J. Amer. Cbem. Sot.. 70 (19SS) 2872.
P21. 25629u.
Abstr..
229 33 2-f 35 16 27 28 29 30
H.G. Ku~vda and J.A. \‘erdone, Terrabedron LetL.. (1964) 119. F. Car+. R. Comu and B. Henner. J. Organomelal. Cbem . . 22 (1970) 589. A.D. Pelrov and G.I. Nikiahin. Izv. Akad. Nauk SSSR. Otd. Khlm. Nauk. (1956) 3-f3. AD. Petrov and V.F. Muonov. Dolil. Abad Nauk SSSR. 80 (1951) 761. Chem. AbsLr.. 46 (195% 11102b. S. O’Brien. M. Fishwxk. B hUdermat1. h1.G.H. WallbrIdge and G.A. Wrlghr. Inorg. Svn.. 13 (1973) A.W.P. Jarvle and R.J. Rowley, J. Chem. Sot. B.. (1971) 2439. i-L Guman and J. 5ch. J. Org. Chem.. 20 (1955) 763. U. Kunze. E. Lmdner and J. fioola. J. OrganomeLsf Cbem.. 57 (1973) 319.
31 32 33 31
hf. Lesuan and G. Gullerm. J. Orgxtometal. Chem.. 5-I (1973) 153. C.W. Fong and W. LilLchlng. J. Organometal. Chem , 1113 (1970) 107. N.P.Gambaryan. El. hl. Rokhbna and Yu.v. Zcrfman. IZV. Akad. Nauk SSSR. (1965) 1466. N.P. Gambawan. EM. Rokblm. Yu.V Ze~fman. C. Chine-Yun and I.L. Knunyanrs. Anger. Chem.. Int. Ed. En&. 5 (1966) 9-17.
35 36 37 38 39 -lO 41 42 13 44 35 46 17 18 49 50 51 53
C.G. Krespan and W.J. Mrddleron. Fluor. Chem. Rev.. 1 (1967) 145. V.A. PalLison. J. Org. Chem.. 34 (1969) 3651. R.L. Adelman. J. Org. Cbem.. 33 (1968) 1401. R.T. Arnold and P. Veeravagu. J. Amer. Chem. Sot.. 82 (1960) 5411. A.G. Snuth and B.L. Yales. J. Org. Chem., 30 (1365) 3067. J. Wolensky. B. Chollar and M.D. Bard. J. Amer. Chem. SIX.. 84 (1962) “1775. N.P. Gambaryan. L.A. Srmonyan and P.V Petrovskil. Izv AF;lld Nauk SSSR. (1967;) 918. J.F. Hams Jr. and D.D. Cofiman. J. Amer. Cbem. Sot . 84 (1963,) 1553. H.R. Da-s. U.S. PAL. 3.164.610: Chem. Abstr.. 63, (1965) 7737d. E.W. Abel and hl.O. Dunrrer. J. Organomeljl. Chem.. 33 (1971) 161 Yu. A. UsLvnyuk. A.V. Kisln and O.E. Okinsora. Zh. Obshch. Khlm.. 38 (1968) 391. E.W. Abel. hl. A. Cooper. R.J. Goodfe’low and A.J. Resl. Trans Feadar Sot.. -i-I (1969) 1697. EW. Abel. N G&s. D.J. Walker and J.N. Wmgiield. J. Chcm. Sot. 4.. (1971) 1391 D Sevfrrlh. T.F. Jula. H. Derboujos and hl. Perevre. J. Organomecal. Chem . 11 (1968) 63. K Jones and M.F. Lapperl. Proc. Chem. Sot.. (1964) 21. C. Tamborski. F.E. Ford, W.L. Lehn. G.J. Moore and E.J. Solo
7.