1618 J. Org. Chem.

1992,57,1618-1621
the correaponding epoxide.% cis-t-Cyclohexyl-l,2epoxyoctane. NMR (CDCl,): 6 0.03 (s,6 H), 0.86 (8, 12 H), 1.2-1.4 (m, 16 H),
'H NMR (CDCl,): 6 0.81 (t, J = 6.9 Hz, 3 H), 0.9-1.5 (m, 12 HI, 1.4-1.8 (m, 12 H), 2.9-3.0 (m, 2 H), 3.62 (t, J = 5.9 Hz, 2 H).
1.3-1.6 (m, 4 H), 1.4-1.8 (m, 4 H), 1.8-2.0 (m, 1 H), 2.55 (dd, J Registry No. 1,765-03-7; 2-dz, 127863-12-1;3a, 87969-78-6;
= 4.0, 8.0 Hz, 1H), 2.8-2.9 (m, 1 H). 3b, 138384244; 4,125641-95-4;5,12113452-9; 6,6975-99-1;6-d2,
(2)-2,2-Dimethy1-3-undecene. Bp: 88-90 "C (bath tem- 125642-23-1;phenylacetylene, 536-74-3; cyclodecyne, 3022-41-1;
perature, 13 Torr). IR (neat): 2954,2922,1729,1466,1363,1273 1-phenyl-1-octyne,16967-02-5; 1-cyclohexyl-1-octyne,125641-943;
~m"'. 'H NMR (CDClJ: 6 0.88 (t, J = 6.7 Hz, 3 H), 1.08 (8, 9
(Z)-&dodecene, 7206-29-3; (2)-cyclododecene, 1129-89-1;(a-1-
H), 1.1-1.5 (m, 10 H), 2.0-2.3 (m, 2 H), 5.15 (dt, J = 12.0,7.0 Hz, phenyl-1-&ne, 42036-72-6;(2)-1-cyclohexyl-1-&ne, 62444546;
1H), 5.31 (d, J = 12.0 Hz, 1H). '3C NMR (CDCl,): 6 14.1,22.7, cis-l-cyclohexyl-1,2-epoxyoctane, 106262-67-3;cis-2,2-dimethyl-
28.4,29.3,29.4,30.4,31.2,31.9,33.1,129.1, 139.6. MS: m/z (re1 3,4-epoxyundecane, 138813-13-5;trans-2,2-dimethyl-3,4-epoxy-
intensity) 182 (M+, 3), 139 (3), 97 (20), 83 (loo), 69 (69),55 (48). undecane, 138813-146; (Z)-2,2-dimethyl-3-undecene,125642-19-5;
Anal. Calcd for C13HB: C, 85.63, H, 14.37. Found C, 85.61; H, (2)-1-(trimethylaily1)-1-dodecene,70875-30-8; (E)-1-cyclohexyl-
14.54. The E / Z ratio was determined by 'H NMR analysis of 1-octene,87393-89-3;(E)-2,2-dimethyl-3-udecene, 14033-68-2;
the corresponding epoxide.% cis-2,2-Dimethyl-3,4-epoxy- Q-l-(trimethylsilyl)-l-dodecene,7087531-9; (2)-1,6undecadiene,
undecane. 'H NMR (CDCl,): 6 0.87 (t, J = 6.8 Hz, 3 H), 0.98 91914-03-3; cis-6,7-epoxy-l-undecane,138813-17-9; (2)-1,7-oc-
(8, 9 H), 1.2-1.5 (m, 10 H), 1.6-1.9 (m, 2 H), 2.63 (d, J = 4.2 Hz,
tadecadiene,127863-11-0;cis-7,~epoxy-l-octadecene,l38813-15-7;
1H), 2.8-2.9 (m, 1H). (E)-2,2-Dimethyl-3-undwene. 'H N M R (2)-5-octadecen-l-ol,41207-39-0; cis-5,6-epoxy-l-(tert-butyldi-
(CDCl,): 6 0.92 (t,J = 6.7 Hz,3 H), 0.99 (e, 9 H), 1.2-1.5 (m, 10 methylsiloxy)octadecane, 138813-16-8; l-undecen-6-yne,
H), 1.9-2.1 (m, 2 H), 5.29 (dt, J = 15.7,6.3 Hz, 1H), 5.43 (d, J 127863-09-6; l-octadecen-7-yne,127863-10-9;5-hexadecyn-1-01,
= 15.7 Hz, 1H). t~s-2f-Dimethyl-3,4-epoxyundecane. 'H 72443-47-1; (Z)-5-hexadecen-l-o1,106463-48-3; niobium chloride,
NMR (CDCI,): 6 0.81 (t,J 6.8 Hz, 3 H), 0.85 ( ~ , H), 9 1.1-1.5 10026-12-7;zinc, 7440-66-6; tantalum chloride, 7721-01-9; 1-do-
(m, 10 H), 1.5-1.6 (m, 2 H), 2.45 (d, J = 2.3 Hz, 1H), 2.7-2.8 (m, decene, 112-41-4; styrene, 100-42-5.
1 H).
(2)-1-(Trimethylsily1)-1-dodecene. Bp: 76-78 OC (bath Supplementary Material Available: Spectroscopicdata (IR,
temperature, 1Torr). IR (neat): 2922, 2852, 1607, 1466, 1248, lH NMR, 13C NMR, MS) for internal alkynes and two experi-
837, 761,688 cm-'. 'H NMR (CDCl,): 6 0.08 (8, 9 H), 0.93 (t, J mental procedures (5pages). Ordering information is given on
= 6.8 Hz, 3 H), 1.1-1.5 (m, 16 H), 2.0-2.2 (m, 2 H), 5.45 (d, J = any current masthead page.
14.0 Hz,1H), 6.29 (dt, J = 14.0,7.3 Hz,1H). '% N M R (CDCld:
6 0.58, 14.1, 22.7, 29.4, 29.8, 32.0, 33.6, 128.7, 149.3. M S m/z
(re1 intensity) 240 (M+,0.8),225 (50),114 (27), 73 (loo), 59 (61).
Anal. Calcd for Cl6H3$i: C, 74.91; H, 13.41. Found C, 74.48; High Regioselectivity in the Alternative,
H, 13.57. The E / Z ratio was determined by 'H NMR analysis. Reductive Cleavages of Terminal Epoxides with
(E)-1-(Trimethylsily1)-1-dodecene.'H NMR (CDClS): 6 0.02 A1uminum Reagents
(8, 9 H), 0.85 (t,J = 6.7 Hz, 3 H), 1.1-1.5 (m, 16 H), 2.0-2.2 (m,
2 H), 5.59 (d, J = 18.0 Hz, 1 H), 6.02 (dt, J = 18.0,6.1 Hz, 1H). John J. Eiech,* Zhi-Rong Liu, and Mona Singh
(2)-1,6-Unde~adiene.~'H NMR (CDCla): 6 0.90 (t, J = 7.0 Department of Chemistry, State University of New York at
Hz, 3 H), 1.2-1.4 (m, 4 H), 1.46 (dd, J = 6.9,7.3 Hz,2 H), 1.9-2.2 Binghamton, P.O.Box 6000,
(m, 6 H), 4.95 (d, J = 10.3 Hz, 1H), 5.01 (d, J = 17.0 Hz, 1H), Binghamton, New York 13902-6000
5.3-5.5 (m, 2 H), 5.82 (ddt, J = 10.3, 17.0,6.8 Hz, 1H). The E / Z
ratio was determined by 'H NMR analysis of the corresponding Received August 29, 1991
epoxide.% cis-6,7-Epo.y-l-undecene.'H NMR (CDClS): 6 0.93
(t,J = 6.8 Hz, 3 H), 1.2-1.8 (m, 10 H), 2.0-2.3 (m, 2 H), 2.9-3.0 Introduction
(m, 2 H), 4.98 (d, J = 17.1 Hz, 1H), 5.01 (d, J = 10.3 Hz, 1H), One of the potentially most versatile methods for
5.82 (ddt, J = 10.3, 17.1, 6.6 Hz, 1 H). functionalizing a carbon skeleton is the Prilezhaev ep-
(2)-1,7-Octadecadiene. Bp: 103-105 OC (bath temperature, oxidation of a carbon-carbon double bond (1)2and the ring
0.20 Torr). IR (neat): 2922, 2852, 1735, 1642, 1460, 1271, 991 opening of the oxirane 2 by a polar reagent, A-B (eq l ) ?
cm-'. 'H NMR (CDCl,): 6 0.88 (t, J = 6.9 Hz, 3 H), 1.2-1.5 (m,
16 H), 1.3-1.5 (m, 4 H), 1.9-2.2 (m, 6 H), 4.95 (d, J = 10.3 Hz,
1 H), 5.00 (d, J = 16.9 Hz,1 H), 5.3-5.5 (m, 2 H), 5.82 (ddt, J
10.3, 16.9,6.6 Hz, 1H). 13CNMR (CDCl,): 6 14.1, 22.7, 27.0,
27.2,28.6,29.2,29.4,29.6,29.7,29.8,31.9,33.7,114.2, 129.6,130.1, R R'
139.0. MS: mlz (re1 intensity) 250 (M+, 4), 123 (9), 96 (75), 82 3 4
(loo), 67 (76), 55 (79). Anal. Calcd for ClaS(: C, 86.32; H, 13.68. With an unsymmetrical alkene, even if the ring opening
Found: C, 86.18; H, 13.61. The E / Z ratio was determined by
'H NMR analysis of the corresponding epoxide?' cis-7,8-Ep- were to occur in a stereospecific, SN2manner (e.g., 3 and
oxy-1-octadecene. 'H NMR (CDCIS): 6 0.88 (t, J = 6.8 Hz, 3 4), there remains the further complication of a nonre-
H), 1.2-1.5 (m, 14 H), 1.3-1.6 (m, 10 H), 2.Cb2.1 (m, 2 H), 2.9-3.0 gioaelective ring opening leading to various proportions of
(m, 2 H), 4.95 (d, J = 10.2 Hz, 1 H), 5.01 (d, J = 17.1 Hz, 1 H), regioisomers 3 and 4. A significant improvement in the
5.81 (ddt, J = 10.2, 17.1, 6.6 Hz, 1 H). utility and versatility of this method for organic synthesis
(2)-6-Octadecen-l-ol. Bp: 146-147 OC (bath temperature, would be the discovery of A-B reagenta and experimental
0.30 Torr). IR (neat): 3318,2922, 2850, 1652, 1466, 1067 cm-'. conditions that cleave oxiranes in a highly regioselective
'HNMR (CDCl& 6 0.88 (t,J = 6.8 Hz, 3 H), 1.1-1.5 (m, 21 H), manner and thereby lead either to product 3 or to product
1.4-1.5 (m, 2 H), 1.5-1.7 (m, 2 H), 1.9-2.2 (m, 4 H), 3.6-3.7 (m, 4 in high yield.
2 H), 5.3-5.5 (m, 2 H). 13CNMR (CDCl,): 6 14.2,22.7, 25.9,26.9, For this study, we have chosen oxiranes derived from
27.3,29.4,29.6,29.7,31.9, 32.4,62.9, 129.3, 130.4. MS: m/z (re1
intensity) 250 (M+ - HzO, 6), 123 (10) 96 (58),82 (loo), 41 (66). terminal olefins as our teat substrates because functionality
Anal. Calcd for Cl8HSO C, 80.53; H, 13.52. Found C, 80.56; at or near the end of a carbon chain is often the most
H, 13.63. The E / Z ratio was determined by 'H NMR analysis valuable site for carbon-carbon chain elongation. As a
of the epoxideHof the tert-butyldimethyldyl ether of the alcohol. potentially valuable cleavage process, we have selected
cis-5,6-Epoxy-1-( tert-butyldimethylsi1oxy)octadecane. 'H reduction with sources of aluminum hydride.
(1) Part 48 of the series Organometallic Compounds of Group III.
(24) Epoxidation was conducted with mCPBA in CH2C12at 0 OC. The Part 47: Eiech, J. J.; Caldwell, K. R.; Werner, S.; Kriiger, C. Organo-
following procedure was modified. Paquette, L. A.; Barrett, J. H. Organic metallics 1991, 10, 3417.
Synthesis; Wiley: New York, 1973; Collect. Vol. V, p 467. (2) House, H. 0. Modern Synthetic Reactions; W. A. Benjamin: New
(25) Barluenga, J.; Yus, M.; Concellon, J. M.; Bernad, P.; Alvarez, F. York, 1965; pp 109-123.
J. Chem. Res. (S)1984,122. (3) Parker, R. E.; Isaac, N. S. Chem. Rev. 1959,59, 737.

0022-3263/92/1957-1618$03.00/00 1992 American Chemical Society

Notes J. Org. Chem., Vol. 57, No. 5, 1992 1619

Table I. Relative Cleavage of Terminal Epoxides with Neutral Organoaluminum Reagents"

RCH-cHz I. i-BuZAIH
C R-CH-CH, + R-CHz-CHz-OH
2. HzO
'0'
6
I
OH 7
5

epoxide (R) entry reagent solventb temp ("C) (time (h)) yield (%) ratio 6:7
n-octyl 1 i-Bu,AlH n-C7H16 25 (8) 95 1oo:o
i-Bu,Al-0-a-Bu
2 i-Bu,AlH n-C7H16 50 (20) 85 99:l
N-methylpyrrolidine'
3 i-Bu,AlH n-C7H16 100 (20) 91 96:4
4 i-Bu3Al n-C7H16 100 (24) 55 55:45d
5 i-Bu3A1 Et20 35 (30) 60 23:77
6 i-Bu& THF 60 (6) 93 0100
phenyl 7 i-Bu,Al n-C7H16 100 (2) 35 01W
8 i-Bu3Al THF 25 (10) 85 7:93
9 i-Bu,AlH n-C7H16 25 (20) 95 1486
i-Bu,Al-0-s-Bu
10 i-Bu,Al n-C6H14 25 (12) 90 1882
EgNC
11 i-Bu2AlH n-C7H16 80 (10) 96 3268
12 i-BuzAIH nC7Hl6 25 (12) 85 65:35
N-methylpyrrolidine'
13 LiAlH, THF 25 (1) 95 97:3'
triphenylailyl 14 i-Bu3Al THF 25 (18) 75f 65:35
15 i-Bu3Al n-C6H14 25 (18) 9d 6040
16 i-Bu,AlH n-C6H14 25 (1) 90 0100
17 i-Bu3Al n-C6H14 25 (18) 08
Et,"
a The reagents were stirred in the stated medium at the indicated temperature and for the given time. The cooled solution was cautiously
treated dropwise with water bas euohtion!) and then with 1 N aqueous HCl. The product was extracted into diethyl ether, the ether
extracts were dried over anhydrous MgS04, and the volatiles were evaporated. The reaction product mixture was analyzed or separated by
gas chromatography or flash column chromatography. bThe reagents, either commercially available (i-BudH and i-Bu,Al (Texas Alkyls)
or generated in situ i-Bu2Al-0-s-Bu from i-BuzAIHand s-BuOH), were employed in 10% excess. 'The tertiary amine employed was equi-
molar with the aluminum reagent. dThe action of i-Bu&l on epoxide 5a or Sb, when conducted in hydrocarbon solution, led to Considerable
amounts of RCH(i-Bu)CH20H (cf. ref 13) but the use of donor solvents suppressed such isobutylated producta and gave excellent regiose-
lectivity for the primary alcohol 7 (Cf. runs 6 and 8). eThishighly regioselective conversion to 1-phenylethanol by LiAlH, in THF, repro-
duced in this work, was already observed by Yoon and Brown ( J . Am. Chem. SOC. 1968,90,2927). 'The balance of the completely consumed
epoxide was found aa triphenylvinylailane (25% in entry 14,10% in entry 15). This olefin probably results from a metalation of the epoxide
carbon a to the silicon and a carbenoid a elimination." 8The epoxide was consumed but no alcohol was isolated. The principal producta
were triDhenvlvinvlsilane(cf. footnote f).Ph8iOH.
" and Ph8iOCH=CH2. The latter two products undoubtedly stem from a base-promoted
isomeri&ion of the epoxide (cf. ref 15).
The reductive cleavage of terminal epoxides has been Scheme I
examined previously in studies that have variously em- 0
ployed LiAlH4 in ether$5 LiA1H4-A1Cl3,4t6AlH3-AlC13,5 R - CH- CH, 4 '.2. MH
H20
*".
/ -\
c - c ".," H 1. MH
___)
2. HzO
R - CHz- CHz- OH
LiBH4-MeOH,' NaBH3CN,B NaA1H2(OR)2,9NaBH4,10 I
OH "d H' 7
LiBH4-Ti(OR)4:1 n-Bu3SnH-NaI,12and i-Bu3Al.13 Some 6 3

a : n-C8H,,
high regioselectivities were observed in these investigations
b : C6Hs
but no type of reagent was capable of yielding both the c : (C,HS)$i
primary or the secondary alcohol in high regioselectivity
by a simple modification of experimental conditions. reagents would also form etherate complexes when em-
Moreover, most such reductionswere conducted in ethereal ployed in etheral media," they are freely soluble in hy-
solution whose Lewis basic character might influence the drocarbons without complexation>&l6Thus,we have been
reactivity of the hydride source.
To explore the influence of the one substituent on the
regiochemistry of terminal epoxide ring cleavages, we have (14) The varying amounts of triphenyl(viny1)silane formed in these
reactions (entries 13 and 14,lO-25%) probably stem from metalization
examined the reduction of alkyl-, aryl-, and silyl-substi- of the epoxide carbon a to silicon and an a-elimination process: Eiech,
tuted epoxides, namely 1,2-epoxydecane (Sa),styrene oxide J. J.; Galle, J. E. J. Organomet. Chem. 1988,341, 293.
(Sb), and (epoxyethy1)triphenylsilane (Sc), by such neutral
hydride sources as i - B u M and i-Bu3Al. Although these 0
/ \
m3Si - CH- CH] -%
RH
!.L I
/ \
Ph$i -C -'3%
&
0
- 0-AfR
1
ph3si--c--cH2

H
,

(4) (a) Eliel, E. L.; Delmonte, D. W. J. Am. Chem. SOC.1958,80,1744. p 0

(b) Eliel, E. L.; Rerick, M. N. J. Am. Chem. SOC.1960, 82, 1362.
(5) Yoon, N. M.; Brown, H. C. J. Am. Chem. SOC.1968, 90,2927. Ph,L-CH-CH,
(6) Andrejevic,V.;Bjelakovic,M.; Mihailovic,M. M.; Mihailovic,M.
L. Helu. Chim. Acta 1986,68, 2030. (15) Rearrangements of a,&epoxyalkylsilaneshave been previously
(7) Soai, K.; Ookawa, A. J. Org. Chem. 1986,51,4000. reported Eisch, J. J.; Galle, J. E. J. Org. Chem. 1976, 41, 2615.
(8)Hutchins, R. 0.;Taffer, I. M.; Burgoyne, W. J. Org. Chem. 1981, (16) Eiech, J. J.; Chiu, C. S. J. Orgonomet. Chem. 1988, 358, C1.
46, 5214. Aluminum halides can cleave epoxides, often with the initiation of car-
(9) Jones, T. K.; Peet, J. H. J. Chem. Ind. (London) 1971, 35, 995. benium ion rearrangementsleading to carbonyl or skeletal isomers. For
(10) Ookawa, A.; Soai, K. Heterocycles 1988,27, 213. this reason, we had wished to avoid using combinations of LiAlH, or
(11)Dai, L.-X.; Lou, B.-L.; Zhana,Y.4.;Guo, G.4. Tetrahedron Lett. i-Bu,AlH with AlC1, to achieve regioselective reductions of epoxides.
1986,27,4343. However, this procedure hae tmen employed with considerablesuccess by
(12) Bonini, C.; DiFabio, R. Tetrahedron Lett. 1988,27,819. Eliel and co-workers;cf. Eliel, E. L.; Rerick, M. N. J. Am. Chem. SOC.
(13) Namy, J.-L.; Abenhaim, D. J. Organomet. Chem. 1972,43,95. 1960,82,1362.
 1620 J,Org. Chem., Val. 57, No. 5, 1992 Notes
able to vary systematically both the solvent and the Lewis Thus, by judicious use of the appropriate aluminum
base (R20or R3N) employed in such epoxide reductions. reagent all three types of epoxides 5a-5c can be trans-
As a result, we have now observed that the regiochemistry formed into either the corresponding primary alcohol or
of these reductive cleavages can be advantageously steered secondary alcohol in high yield and with great regiose-
by a rational choice of experimental conditions. lectivity.
Results Discussion
By the reaction of either i-Bu,AlH or i-Bu3Alwith ter- Examination of the organoaluminum hydride reducing
minal epoxides 5a-5c, in the presence or the absence of agents employed and the attendant experimental condi-
Lewis bases such as ethers or amines, a highly regiose- tions (Table I) permits certain generalizations concerning
lective cleavage into either the primary alcohol 7 or the the resultant dominant regioselectivity: (1)with the al-
secondary alcohol 6 (Scheme I) was observed (Table I). kyl-substituted epoxides, the use of i-Bu,AlH, without
With the alkyl- and aryl-substituted epoxides 5a and 5b, (entry 2) or with (entries 1 and 3) strong donors, favors
it was found that the use of i-BuzAIH,specifically in the the formation of the secondary alcohol, while the use of
presence of a Lewis base (R3N,R2A10R', R,O), favors the i-Bu3Al favors the formation of the primary alcohol only
formation of the secondary alcohol 6. Indeed, a comparison with a strong donor (THF, entry 5); (2) with the phenyl-
of entries 1 and 4 shows that the regioselectivity can be substituted epoxide, the use of i-Bu3Al, with or without
made exclusive. On the other hand, the action of i-Bu,Al, strong donors, always favors the formation of the primary
especially in the presence of a donor such as THF, favors alcohol, while the use of an aluminum hydride source fa-
the formation of the primary alcohol 7. Again, such re- vors the secondary alcohol only with the strongest donors
gioselectivity can become nearly exclusive, as reflected in (R3N with i-Bu2A1Hin entry 12 and H- with LiAlH4 in
entries 6 and 8. The use of a donor solvent with i-Bu3Al entry 13); and (3) with the silyl-substituted epoxide, i-
also serves to sharply reduce the amount of isobutylation Bu2AlHfavors the primary alcohol and i-Bu3Alfavors the
that accompanies reductive ring cleavage in hydrocarbon secondary alcohol.
media.13 These empirical observations can be interpreted to mean
Although essentially exclusive regioselectivity leading that i-Bu2A1H,which is actually present as a trimer in
to either 6 or 7 was attained with the alkyl-substituted hydrocarbon media" but as a Lewis complex of the mo-
epoxide 5a (entries 1and 6), the regioselectivity in reducing nomer in donor media, acts as a nucleophilic hydride
the aryl-substituted epoxide 5b was not completely source whose site of attack is governed by steric factors
steerable. The primary alcohol could indeed be formed (9). With i-Bu3Al, complexation with the epoxide prob-
exclusively with i-Bu3Al (entry 7), but the secondary al-
cohol could only be obtained in a 65% regioselectivity from
the action of i-Bu2A1H complexed with N-methyl-
pyrrolidine (entry 12). However, the solution to this
problem of obtaining 7 with high regioselectivity was al-
ready available: Yoon and Brown had shown that LiAlH4
reduces 5c to 6 in a 98% regi~selectivity.~ Their finding ably precedes isobutene elimination and the generation of
is reproduced in this study (entry 13). the AI-H bond. The transition state for ring opening may
The regioselectivities observed with the silyl-substituted be attained only with considerable stretching of a C-0
epoxide 5c ran counter to those found with the alkyl- and bond and the development of carbenium-ion character
aryl-substituted epoxides 5a and 5b. The action of i- (lo), hence the favoring of primary alcohol formation.
BuzAIH in the absence of any donor led exclusively to the That the behavior of the silyl-substituted epoxide 5c
primary alcohol (entry 16), while the action of i-Bu3Al in toward i-Bu,AlH and i-Bu,Al is the reverse of that shown
THF led to the formation of the secondary alcohol in a by alkyl and aryl-substituted epoxides 5a and 5b is not
65 % regioselectivity (entry 14). This regioselective for- surprising. Electronic effects operative in the ring-opening
mation of the secondary alcohol could not be improved; reactions of a,&epoxyalkylsilanes can rationally account
when a stronger Lewis was employed (Et3N,entry 17), the for this behavior. Nucleophilic attack of LiAlH4or L W . ,
epoxide apparently was consumed by a combination of on 5c is known to favor attachment of the hydride ion to
metalation14and base-promoted isomeri~ation'~ processes the carbon CY to silicon and the exclusive formation of the
that led to the observed products vinyltriphenylsilane, primary alcohol (eq 4):18
triphenylsilane, and (viny1oxy)triphenylsilane. 0
The challenge of converting (epoxyethyl)triphenylsilane Ph,Si
/
- CH- \
CH,
1. LiAID4
Ph$i - CH- CH2OH (4)
2. H20
(5c) into the secondary alcohol with high selectivity had 5C
I
D 6a
to be met, therefore, in another manner. Noting our recent
finding that treatment of 5c with MeAlC1, and then H20 This regiochemistry is proposed to result via a transition
produces exclusively 8 in >90% yield (eq 2),16 we were state stabilized by partial pentacoordination of the hydride
n nri
ion with silicon.16 In the present study we could accord-
Ph,Si - CH-/-\ CHI 1. McAIC12 I
P ~ ~ -SCH-
I CH2- CI
ingly propose that the analogous reduction of 5c with i-
(2)
5C
2 . H,O
8 Bu2AlHwould be determined by a transition state like ll.
CH
i-Bu2Ai' \2
. g M e Z

encouraged to seek a means of reductively dehalogenating

8 to produce 7c (eq 3). Indeed, we have found that n-
Bu3SnH in toluene smoothly reduces 8 to 7c in an isolated Conversely, the tendency of i-Bu3Al in THF to favor for-
yield of 81%. Therefore,eqs 2 and 3 depict a high-yielding,
two-step process for the exclusive regioselective conversion (17) Eisch, J. J.; Rhee, S. G. J . Organomet. Chem. 1972, 42, C73.
of 5c into its secondary alcohol. (18) Eisch, J. J.; Traiinor, J. T. J. Org. Chem. 1963, 28, 2870.
 J. Org. Chem. 1992,57,1621-1622 1621

mation of the secondary alcohol 6c would be consistent The Gas-Phase Basicity of Hydroxamic Acid
with a transition state, as in 10, again resembling a Derivatives
structure leading to the more stable carbenium ion 12a.
In this situation, however, the more stable carbenium ion Michele Decouzon, Otto Exner,' Jean-FranGois Gal,* and
site is the primary carbon, because the positive charge can Pierre-Charles Maria
be stabilized by carbon-silicon u-bond hyperconjugation Laboratoire de Chimie Physique Organique, Groupe
( 1 2 ~ 9 FT-ICR, Universite de Nice - Sophia Antipolis,
Parc Valrose, F-06108 Nice Ceder 2, France
Experimental Section
Received September 19,1991
General Procedures. All steps in the preparation, transfer,
and main reactions of the organometallic reagents studied here Introduction
were conducted under an atmosphere of anhydrous and oxy-
gen-free argon. All solvents and apparatus were likewise freed One can understand at first sight, already from the name
of traces of dissolved or adsorbed moisture and oxygen and then "hydroxamic acids" that the basic properties of these
maintained under argon. Methods and techniques for working compounds (1) have been less investigated.' The crys-
under anhydrous and anaerobic conditions and for conducting tallized hydrochloride of benzohydroxamic acid (2) was
standard chromatographicand spectrometric analyses have been prepared: and pK values in sulfuric acid have been re-
described previously.20 ported for several substituted protonated benzohydroxamic
Starting Materials and Products. Diisobutylaluminum acids.3 As far as the site of protonation is concerned,
hydride and triisobutylaluminum were obtained as neat reagents several kinetic studies3-+have assumed that it is the car-
from Texas Alkyls Inc., Deer Park, TX. Diisobutylaluminum
sec-butoxide was generated in situ by the cautious treatment bonyl oxygen (see Scheme I, formula la or l b with possible
of i-Bu2A1Hin heptane with 1 equiv of sec-butyl alcohol. hydrogen bonds) rather than the nitrogen atom (IC), but
1,2-Epoxydecaneand styrene oxide were CommerciaUy available without any proof, simply from an analogy with amides.
and (epoxyethy1)triphenylsilane was prepared from the ep- The actual structure of the cation was seldom examined
oxidation of triphenylvinylsilanewith m-chloroperbenzoic acid.2l in contrast to the rather extensive discussion of the
The reaction products 1-decanol,2-decanol, 1-phenyl-1-ethanol, structure of the and of the neutral mole~ule.'~~J~
and 2-phenyl-1-ethanolwere purchased, while the 1-(triphenyl- To our knowledge only the salt 2 was investigated by IR
sily1)-1-ethanoland 2-(triphenylsilyl)-l-ethanolwere synthesized in dioxan" and by XPS in crystal,12in either case in favor
by known procedures.22 of the structures la,b. In a recent theoretical s t u d 9 of
General Reduction Procedure. On a 1.0-10.0-mmol scale, formohydroxamic acid the structure 3a was found slightly
the epoxide was dissolved in 10-50 mL of the solvent and then
added slowly to 1.1equiv of the aluminum hydride source, which more stable than 3b but distinctly more than 3c. N-A1-
was contained in a like volume of solvent and, in certain cases, kyl-substituted hydroxamic acids seem to be more basic,
was previously complexed with an equivalent of a Lewis base (KN and two forms of the cation (lb and IC; la not considered)
or W O R ' ) . The reaction mixture was allowed to stir at the given were ~1aimed.l~In aminohydroxamic acids the amino
temperature for the stated reaction period (Table I). group is protonated first.14
If necessary, the reaction mixture was cooled to 25 "C and then In this paper we report the gas-phase basicities of ace-
slowly and cautiously treated with water (gas evolution!). tohydroxamic acid (4) and of its N-methyl (5) and 0-
Thereupon,dilute, aqueous HC1 was added to give two clear layers. methyl (6) derivatives (Table I). In addition to the general
The separated organic layer was dried over anhydrous MgSO, interest-placing these compounds into the basicity
and then the volatile8 were removed in vacuo. The residue was
analyzed by gas chromatography and separated by flash column
scale15-we also hoped to get some information concerning
chromatography. the structure of the cation, particularly from a comparison
Reductive Dehalogenation of 2-Chloro-l-(triphenyl- with isosteric amides 7-9.
silyl)-l-ethanol(8). A solution of 203 mg (0.61 "01) of 8,which
was prepared from 5c and MeA1C12,16and 120 mg (0.59 mmol) Results and Discussion
of tri-n-butyltin hydride in 15 mL of anhydroustoluene was heated The gas-phase basicities are collected in the table. The
at reflux for 8 h. The solvent was removed in vacuo and the first conclusion emerges that hydroxamic "acids" are also
residue subjected to flash chromatography on silica gel with 5% relatively strong bases. In the whole scale15 of the gas-
ethyl acetate in hexane. The isolated 1-(triphenylsily1)-1-ethanol phase acidity of organic acids (range of -100 kcal), hy-
(7c) was obtained pure in 81% yield and was identified by com-
parison of its mp and its IR, 'H NMR, and 13C NMR spectral droxamic acids are placed relatively near the strongest
properties with those of an authentic sample.'*
-
acids, viz. at 27% of the whole range. But even in the scale
of organic based5 ( 110 kcal) they fall in the vicinity of
strong bases (at 35% of the whole range). Compared with
Acknowledgment. The research was supported by amides hydroxamic acids are stronger acids (by 15 kcal)
Texas Alkyls Inc. of Deer Park, TX, whose fostering of and weaker bases (by 7 kcal). In all cases the reason is
research on synthetic applications for organoaluminum evidently the electron-attracting inductive effect of the
reagents has been most appreciated. oxygen atom.16 In the case of the acidity of hydroxamic
acids this effect is corroborated by a change of the me-
someric structure of the anion: the substitution by an
oxygen induces an electron attraction from N and renders
(19) Any electrophilic attack on 50, such as where E = MeAlC12,is the mesomerism within the O=C-N unit better balanced.
known to favor formation of the carbenium ion @ t o silicon
The two effects seem to strengthen each other in an ef-
/OE
ficient way." In water solution' the difference in acidity
wi- CH- BH, between hydroxamic acids and amides is leveled. In the
due to stabilization arising from carbon-silicon u-bond hyperconjugation:
case of basicity even the order is reversed. A comparison
Hanetein, W.; Berwin, H. J.; Traylor, T. G.J . Am. Chem. Soc. 1970,92, with carboxylic acids is less easy to interpret: hydroxamic
829.
(20) Eisch, J. J. Organometallic Synthesis; Academic Press: New
York, 1981; Vol. 2, p 194. Permanent address: Institute of Organic Chemistry and Bio-
(21) Eisch, J. J.; Galle, J. E. J. Org. Chem. 1976, 41, 2615. chemistry, Czechoslovak Academy of Sciences, CS-16610 Prague 6,
(22) Eisch, J. J.; Trainor, J. T. J . Org. Chem. 1963,28, 2870. Czechoslovakia.
0022-3263/92/1957-1621$03.00/00 1992 American Chemical Society