Stereochemical study of the allylation of aldehydes with allyl halides in cosolvent/water(salt)/Zn and in cosolvent/water(salt)/Zn/haloorganotin media

Article abstract of DOI:10.1021/jo951562h

The stereochemical course of the allylation and propargylation of several aldehydes with crotyl and propargyl halides using zinc powder as the condensing agent in cosolvent/water(salt) media have been extensively studied. Radical ions of well-defined geom

Full text of DOI:10.1021/jo951562h

J . Org. Chem. 1996, 61, 2731-2737
2731
Ster eoch em ica l Stu d y of th e Allyla tion of Ald eh yd es w ith Allyl
Ha lid es in Cosolven t/Wa ter (Sa lt)/Zn a n d in Cosolven t/Wa ter (Sa lt)/
Zn /Ha loor ga n otin Med ia
Daniele Marton, Diego Stivanello, and Giuseppe Tagliavini*
Dipartimento di Chimica Inorganica, Metallorganica e Analitica, Universita´ di Padova, Via Marzolo, 1,
I 35131 Padova, Italy
Received August 25, 1995X
The stereochemical course of the allylation and propargylation of several aldehydes with crotyl
and propargyl halides using zinc powder as the condensing agent in cosolvent/water(salt) media
have been extensively studied. Radical ions of well-defined geometry such as [CH2dCHCH(CH3)Br]•-
are thought to be formed through an electron-transfer process, which seems to be accelerated by
the high interfacial energy developed by the water. These radicals are trapped by the aldehyde
molecules to form isomeric mixtures of threo/ erythro-â-methylhomoallylic alcohols or R-allenic/â-
acetylenic alcohols. The stereochemistry seems to be determined by the structure of the radical
anions adsorbed on the zinc particles and is independent of the geometry of the organic halides,
the cosolvents, or the salts. Reactions performed in the presence of Bu3SnCl show no discrimination
between the coupling reactions of the allyl bromide with aldehydes and with Bu3SnCl, both
homoallylic alcohols and allyltributylstannanes being formed. In contrast, in the presence of Bu2-
SnCl2 only â-methylhomoallylic alcohols are formed, this being a function of the differential
allylstannation of aldehydes and crotyldibutyltin halides, which are intermediate products of the
zinc-mediated process. The zinc-mediated method gives satisfactory results with acetals if workup
is made under the catalytic mediation of Bu2SnCl2.
In tr od u ction
Of special interest in recent years has been the use of
waterseither alone or with cosolventsas a medium for
reactions with organic or organometallic reagents.1-3 In
the organometallic field, particular attention has been
paid to the use of allyl- and allyllike-tin halides, as well
as BuSnCl3, as reactants in water for C-C4-7 and C-O-
C8-10 bond-forming reactions under homogeneous or
heterogeneous conditions. Also, silicon and tin allyl
derivatives have been used for aminoethanodesilylation-
cyclization11,12 and aminomethanodestannylation13 pro-
cesses, with water.
The same procedure has been recently employed to
prepare mixed allylstannanes17,18 by coupling reactions
of allyl or allyllike bromides with R4-nSnXn substrates
(eq 2).
A very versatile procedure based on the mediation of
zinc powder14-16 (eq 1) has been adopted for the prepara-
tion of homoallylic alcohols.
The zinc-mediated method has also been applied to the
preparation of distannanes18 and mixed benzylstan-
nanes.19
X Abstract published in Advance ACS Abstracts, March 15, 1996.
(1) Multzer, J .; Altenbach, H. J .; Braun, M.; Krohn, K.; Reissig, H.
U. Organic Synthesis Highlights; VCH: Weinheim, Germany, 1991; p
71.
(2) Grieco, P. A. Aldrichim. Acta 1991 24, 59.
(3) Chao-J un Li, Chem. Rev. 1993, 93, 2023 and references therein.
(4) Boaretto, A.; Marton, D.; Tagliavini, G.; Gambaro, A. J . Orga-
nomet. Chem. 1985, 286, 9.
(5) Boaretto, A.; Marton, D.; Tagliavini, G. J . Organomet. Chem.
1985, 297, 149.
(6) Furlani, D.; Marton, D.; Tagliavini, G.; Zordan, M. J . Organomet.
Chem. 1988 341, 345.
(7) Marton, D.; Tagliavini, G.; Vanzan, N. J . Organomet. Chem.
1989, 376, 269.
(8) Marton, D.; Slaviero, P.; Tagliavini, G.; Vanzan, N.; Zordan, M.
In Chemistry and Technology of Silicon and Tin; Kumar Das, V. G.,
Weng, N. S., Gielen, M. Eds.; Oxford University Press: Oxford, 1992;
277.
A close correlation exists between the reactions rep-
resented in eqs 1 and 2. In both cases, intermediate
radical ions of the type [CH2dCHCH2Br]•- Zn•+ are
thought to be trapped by electrophilic species, aldehydes
in eq 1 and R4-nSnXn in eq 2. A further similarity is the
observed high regioselectivity; aldehydes and organotin
derivatives react at the more substituted carbon of the
allyl halides, giving rise to â-methylhomoallylic alcohols16
and (R-methylallyl)tin derivatives20 respectively, when
crotyl halides are used. With regard to the stereochem-
ical outcomes of these reactions, those of eq 2 have been
(9) Tagliavini, G. J . Organomet. Chem. 1992, 437, 15.
(10) Boaretto, A.; Marton, D.; Tagliavini, G. J . Organomet. Chem.
1985, 286, 9.
(11) Larsen, S. D.; Grieco, P. A.; Fobare, W. F. J . Am. Chem. Soc.
1986, 108, 3512.
(17) Carofiglio, T.; Marton, D.; Tagliavini, G. Organometallics 1992,
11, 2961.
(18) von Gyldenfeldt, F; Marton, D.; Tagliavini, G. Organometallics
1994, 13, 906.
(12) Grieco, P. A.; Fobare, W. F. Tetrahedron Lett. 1986, 27, 5067.
(13) Grieco, P. A.; Bahsas, A. J . Org. Chem. 1987, 52, 1378.
(14) Petrier, C.; Luche, J . L. J . Org. Chem. 1985, 50, 910.
(15) Petrier, C.; Einhorn, J .; Luche, J . L. Tetrahedron Lett. 1985,
26, 1449.
(19) Marton, D.; Russo, U.; Stivanello, D.; Tagliavini, G.; Ganis, P.;
Valle, G. C. Organometallics 1996, 15, 1645.
(20) Bu3SnCl reacts with an isomeric mixture of (C4H7)Br (C4H7
stands for R-methylallyl (15%) and trans- (76%) and cis-crotyl (9%), in
cyclohexane/H2O (NH4Cl)/Zn medium to form selectively Bu3SnCH-
(CH3)CHdCH2 (see ref 18).
(16) Einhorn, J .; Luche, J . L. J . Organomet. Chem. 1987, 322, 177.
0022-3263/96/1961-2731$12.00/0 © 1996 American Chemical Society

2732 J . Org. Chem., Vol. 61, No. 8, 1996
Marton et al.
Ta ble 2. Ster eoch em ica l Cou r se of th e Rea ction s
Ta ble 1. Ster eoch em ica l Cou r se of th e Rea ction s
betw een P r op ion a ld eh yd e a n d Allyl Ha lid es in
Cycloh exa n e/H2O(Sa lt)/Zn Med iu m a
betw een Isobu tyr a ld eh yd e a n d Allyl Ha lid es In
Cosolven t/H2O(NH4Cl)/Zn Med iu m a
(CH3)2CHCHO + 9a or 9b f
2
(threo,erythro)-(CH3)2CHCH(OH)CH(CH3)CHdCH2
+
(trans,cis)-(CH3)2CHCH(OH)CH2CHdCHCH3
isolated
allyl product
isomeric composition, %
entry
no.
cosolvent halide yield, % threo erythro (trans + cis)
1
2
3
4
5
6
7
b
b
9a
9b
9a
9b
9a
9b
9a
9a
9b
9b
72
62
76
66
73
72
77
75
67
63
74
77
76
78
74
76
75
75
76
77
25
22
23
21
25
23
24
24
23
21
1
1
1
1
1
1
1
1
1
1
THF
THF
acetonitrile
acetonitrile
cyclohexane
isolated
product
isomeric composition, %
entry
no.
allyl
8c cyclohexane
salt
halide yield, % threo erythro (trans + cis)
9
cyclohexane
1
2
3
4
5
NH4Cl
NH4Cl
NaCl
9a
9b
9a
9a
9a
40
39
67
81
76
53
52
53
51
51
47
48
44
46
45
10d cyclohexane
a
3
3
4
Reactions conditions: 0.15 mol of 2 and 0.30 mol of 9a or 9b;
NaCl
25 mL of cosolvent; 50 mL of water (NH4Cl satd); 0.30 mol of Zn
NaClb
powder. Runs performed without cosolvent. c The system has
b
d
been cooled with an external ice bath. The mixture 9b/2 has been
added to the three-phase system cyclohexane/H2O(NH4Cl)/Zn.
a
Reactions conditions: 0.15 mol of 1 and 0.30 mol of 9a or 9b
in entries 1-3; 0.30 mol of 1 and 0.15 mol of 9a or 9b in entries
4 and 5; 25 mL of cyclohexane; 50 mL of water (saturated with
salt); 0.30 mol of Zn powder (325 mesh). NaCl saturated aqueous
solution with 1 mL of HClO4 70%, pH ) 0.63.
Ta ble 3. Ster eoch em ica l Cou r se of th e Rea ction s
betw een Ald eh yd es 3-5 a n d Allyl Ha lid es in
Cycloh exa n e/H2O(NH4Cl)/Zn Med iu m a
b
more extensively examined than those of eq 1. Therefore,
we have studied under different experimental conditions
the coupling of the aldehydes RCHO [R ) C2H5 (1),
(CH3)2CH (2), (CH3)3C (3), cyclo-C6H11 (4), C6H5 (5),
CH2dCH (6), CH3CHdCH (7), and CH3(CH2)2CHdCH
(8)] with CH2dCHCH(CH3)Cl (9a ) or (C4H7)Br (9b)21 and
reactions of aldehyde 2 and propargyl bromide (10). Also,
the effect of various cosolvents (THF, cyclohexane, ac-
etonitrile) or salts in the aqueous phase (NH4Cl, NHdC-
(NH2)‚HCl, NaCl, LiClO4, ZnCl2) has been studied.
Mixtures of (CH3)2CHCHO and Bu2SnCl2 have been
allowed to react with (C4H7)Br and zinc in cyclohexane/
H2O (salt) media. Thus, we have verified how the
stereochemical course of the coupling reaction leading to
â-methylhomoallylic alcohols via eq 1 is influenced by the
competitive allylstannation of the aldehyde6,7,22 on ac-
count of the crotyldibutyltin chloride formed via eq 2.
Following this line, we also have found that the acetal
CH3CH(OCH3)2 (11) reacts with 9b in the presence of
catalytic amounts of Bu2SnCl2.
RCHO + 9a or 9b f
(threo,erythro)-RCH(OH)CH(CH3)CHdCH2
+
(trans,cis)-RCH(OH)CH2CHdCHCH3
isolated
product
isomeric composition, %
entry
no.
allyl
RCHO halide yield, % threo erythro (trans + cis)
1
2
3
4
5
6
7
3
3
3
4
4
5
5
9a
9a
9b
9a
9b
9a
9b
55
65
76
80
77
64
55
76
76
75
73
76
42
49
20
20
22
26
24
58
51
4
4
3
1
a
Reactions conditions: 0.075 mol of aldehyde and 0.15 mol of
9a in entry 1; 0.15 mol of aldehyde and 0.30 mol of 9a or 9b in
entries 2-7; 25 mL of cyclohexane; 50 mL of water (NH4Cl satd);
0.30 mol of Zn powder.
Ta ble 4. Ster eoch em ica l Cou r se of th e Rea ction s
betw een Ald eh yd es 6-8 a n d Allyl Ha lid es in
Cycloh exa n e/H2O(Sa lt)/Zn Med iu m a ,b
isomeric composition, %
isolated
Resu lts a n d Discu ssion
entry
no.
allyl product
RCHO halide yield, % threo erythro
(trans +
cis)
salt
Tables 1-4 show the data resulting from the coupling
reactions between aldehydes 1-8 and allyl halides 9a
and 9b. Neither yields nor stereochemical outcomes are
affected by changing cosolvents and salts. In fact,
reactions can be performed in the absence of cosolvents
(see Table 2). Better yields are obtained when 9a is
employed. Yields are found in the range of 60-80% when
aldehydes 1-5 are employed (see Tables 1-3). They
decrease to 20-40% when R,â-unsaturated aldehydes
6-8 are employed (see Table 4). This is mainly due to
1
2
3
4
5
6
NH4Cl
NH4Cl
NH4Cl
NaCl
NH4Cl
NH4Cl
6
7
7
7
8
8
9b
9a
9b
9b
9a
9b
13
20
24
21
36
42
40
45
50
49
43
45
60
55
50
51
57
55
a
Reactions conditions: 0.15 mol of aldehyde; 0.30 mol of 9a or
9b; 25 mL of cyclohexane; 50 mL of water (saturated with salt);
0.30 mol of Zn powder. For the alcoholic isomeric mixtures see
b
Table 3.
side reactions that occur at the expense of the aldehyde.
In the case of crotonaldehyde, isomeric mixtures of
(meso,dl)-2,6-octadiene-4,5-diol were recovered. In the
case of benzaldehyde (see Table 3, entries 6 and 7), the
low yields of the homoallylic alcohol are also due to the
reduction of the carbonyl compound to benzyl alcohol.
(21) For the isomeric composition of (C4H7)Br see ref 20.
(22) Gambaro, A.; Marton, D.; Peruzzo, V.; Tagliavini, G. J . Orga-
nomet. Chem. 1982, 226, 149.
(23) The Sn/Al method refers to the allylation of aldehydes in an
ether/water medium by means of allyl bromide under the mediation
of tin and aluminum powder; see: Nokami, J .; Otera, J .; Sudo, T.;
Okawara, R. Organometallics 1983, 2, 191.

Allylation of Aldehydes with Allyl Halides
J . Org. Chem., Vol. 61, No. 8, 1996 2733
Ta ble 5. Com p a r ison of th e Ster eoch em ica l Resu lts in th e Syn th esis of â-Meth ylh om oa llylic Acoh ols in Aqu eou s Med ia
by Mea n s of Th r ee Differ en t Meth od s
isomeric composition, threo:erythro (yield, %)
alcohol
RCH(OH)CH(CH3)CHdCH2
allystannation
method
R
Zn methoda
Sn/Al methodd
C2H5
53:47 (39-81)
76:24 (62-77)
76:24 (55-76)
42:58 (55-64)
60:40 (87)b
78:22 (85)c
46:54 (67)c
61:39 (85)c
36:61 (100) Et2O/H2O
35:65 (92) Et2O/H2O
(CH3)2CH
(CH3)3C
C6H5
40:60 (87) Et2O/H2O
22:78 (81) THF/H2O
a
b
d
Present work. See ref 10. c See ref 6. See ref 23.
The same stereochemical results are seen with re-
agents 9a and 9b, showing that the stereochemistry is
independent of the geometry of the starting allyl halide.
The addition reactions are stereorandom for aldehydes
1, 5, and 6-8, the threo:erythro ratio being about 50:50.
With aldehydes 2, 3 and 4, which bear larger substituents
such as (CH3)2CH, (CH3)3C, and cyclo-C6H11, respectively,
moderate stereoselectivity in favor of the threo-isomer is
observed (threo:erythro ) 75:25).
To date, aqueous media have been employed in the
syntheses of â-methylhomoallylic alcohols, either by
direct addition of organometallic reagents generated
separately6,7 or by allylation of carbonyl compounds via
mediation of various metals and salts.3 A comparison of
the present results with those arising from both allyl-
stannation6 and Sn/Al methods is given in Table 5. It is
noteworthy that a moderate threo-selectivity is seen with
both the allystannation and zinc methods. In contrast,
erythro-selectivity is favored by the Sn/Al method.
In the case of allylstannation, the activation of the
carbonyl group through a coordination step at the tin
site24 leads to a pericyclic transition state which explains
the moderate threo-selectivity when bulky groups are
present.22 In the zinc mediated method,14-18 intermediate
The hypothesis of a transition state with a cyclic
structure might explain the moderate threo-selectivity,
which is similar to that encountered in the allylstanna-
tion case, where the γ-carbon of the allyl group is
involved. On the contrary, the erythro-selectivity of the
Sn/Al method, which requires an oxidative addition of
the allyl halide to Sn(0) activated by Al(0), might be
explained by invoking an open transition state as previ-
ously reported.28
Reactions between isobutyraldehyde and propargyl
bromide performed in cyclohexane and water containing
various salts are characterized by very low yields (see
Table 6). Isomeric mixtures of â-acetylenic and R-allenic
alcohols with ratios â:R ) 90:10 are recovered. Therefore,
also in this case, the major component maintains the
acetylenic geometry of the precursor bromide, showing
that the aldehyde interacts with the more substituted
carbon atom (R-carbon) of the propargyl group.
A different stereochemical course is followed when a
substituted propargyl bromide, such as 1-bromo-2-octyne,
is used. The R-allenic alcohol is the major component of
the recovered isomeric mixture, the ratio being â:R ) 30:
70. In this case, the initial geometry of the precursor
organic bromide is not maintained, showing that a
rearrangement of the propargyl unit has occurred during
the process.
These findings need to be discussed in light of the
results arising from the use of either CH2dCHCH(CH3)-
Cl (9a ) or (C4H7)Br (9b), where the stereochemistry is
the same. As a reasonable explanation we hypothe-
size the formation of radical anions of the sole type
[CH2dCHCH(CH3)X]•- on the zinc surface. This means
that the isomeric trans- and cis-crotyl bromide also gives
radical ions of the type [CH2dCHCH(CH3)X]•- Zn•+
,
formed by a one-electron transfer mechanism, are thought
to interact with the carbonyl compound. The previously
proposed transition state for this Barbier-type reaction,
with geometric characteristics similar to those of the
corresponding structure in the SN2 displacement,25 does
not explain the observed threo-selectivity. In our opinion,
the stereochemical results, very similar to those obtained
by the allylstannation method (see Table 5), might be
better explained through the mechanism proposed as
follows.
In the transition state, the more substituted carbon
atom (R-carbon) of the allyl moiety interacts with an
aldehyde molecule adsorbed on the zinc metal surface.
The rigid arrangement of the transition state, having a
cyclic structure, favors the threo-form due to minor
interactions between the methyl group linked to the allyl
R-carbon atom and the bulky R group of the aldehyde.
Adsorbed carbonyl compounds can interact with the zinc
metal to give radical anions and then, by desorption,
radical species.25,26 As a matter of fact, the encountered
side reactions at the expense of the aldehydes 5-8, such
as dimerization and reduction, must be attributed to the
formation of radicals [RCHO]•.27
(27) Comparison of the values affinities shows that unsaturated
carbonyl compounds are more able to form carbonyl radical anions than
saturated carbonyl compounds. See ref 25.
(28) Hoffmann, R. W. Angew. Chem., Int. Ed. Engl. 1982, 21, 555
and references therein. (b) Uneyama, K.; Kamaki. N.; Moriya, A.; Torii,
S. J . Org. Chem. 1985, 50, 5396. (c) Wada, N.; Ohki, H.; Akiba, K.-y.
Bull. Chem. Soc. J pn. 1990, 63, 1738.
(24) Tagliavini, G.; Peruzzo, V.; Plazzogna, G.; Marton, D. Inorg.
Chim. Acta 1977, 24, L47.
(25) Moiano, A.; Pericas, M. A.; Riera, A.; Luche, J . L. Tetrahedron
Lett. 1990, 31, 7619.
(26) de Souza-Barboza, J . C.; Luche, J . L.; Petrier, C. Tetrahedron
Lett. 1987, 28, 2013.

2734 J . Org. Chem., Vol. 61, No. 8, 1996
Marton et al.
Sch em e 1. Ster eoch em ica l Step s of th e Cou p lin g Rea ction s of Allyl a n d P r op a r gyl Ha lid es w ith
Ald eh yd es or Or ga n otin s by Mea n s of th e Zin c-Med ia ted Meth od
Ta ble 6. Ster eoch em ica l Cou r se of th e Rea ction s
betw een Isobu tyr a ld eh yd e a n d P r op a r gyl Br om id e in
Cycloh exa n e/H2O(Sa lt)/Zn Med iu m a
reagent, either RCHO or Bu3SnCl (step d), occurs at the
more substituted carbon atom of the radical anion.
Isomerization, either of (R-methylallyl)- or propargyltin
compounds, takes place only after the coupling processes
(equilibria e).18 Further arguments support the above
scheme. Due to the π-electron system, the radical anions
have sufficiently long life30 to be trapped on the surface
by RCHO or Bu3SnCl. Trapping of these radical anions
must be relatively rapid, otherwise radicals, [CH2dCH-
CHCH3]•, would be formed. These radicals, adsorbed at
the metal surface or diffused freely in solution31 after
desorption, give rise to side reactions such as couplings
or reductions. Actually, when only (R, trans, cis)-(C4H7)-
Br (R:trans:cis ) 20:76:4) is allowed to react in a cyclo-
hexane/H2O(NH4Cl)/Zn system, a six isomeric mixture of
dimeric allyl species with a ratio of R-groups:(trans + cis)
groups ) 38:62 is recovered together with other hydro-
carbons.
Therefore, the zinc-mediated method can be success-
fully utilized to prepare either homoallylic alcohols or
allyltin derivatives such as allyltriorganotins, allyldior-
ganotin halides, and/or diallyldiorganotins. We have
examined reactions between allyl bromide and equimolar
mixtures RCHO/Bu3SnCl. Results from the systems
CH2dCHCH2Br/C2H5CHO/Bu3SnCl and CH2dCHCH2Br/
C6H5CHO/Bu3SnCl show that there is no discrimination
between the two coupling reactions 1 and 2, because both
homoallylic alcohols and Bu3SnCH2CHdCH2 are isolated.
Homoallylic alcohols arise only from the coupling reaction
of eq 1, because allyltributyltin formed through reaction
2 is unable to react with aldehydes.32 To the contrary,
easy allylstannation of carbonyl compounds can be
achieved using allyltin halides neat22,24 or in the presence
of water.4,6 Thus, the system (C4H7)Br/(CH3)2CHCHO/
Bu2SnCl2 has been investigated. Results listed in Table
(CH3)2CHCHO + CHtCCH2Br f
(CH3)2CHCH(OH)CHdCdCH2 (R) +
(CH3)2CHCH(OH)CH2CtCH (â)
isomeric composition, %
isolated
entry
no.
product
yield, %
propargyl
isomer (â)
allenyl
isomer (R)
salt
1
2
3
4
NH4Cl
NH4Cl
ZnCl2
21
33
21
22
91
89
91
84
9
11
9
b
c
16
a
Reactions conditions: 0.15 mol of aldehyde and 0.30 mol of
propargyl bromide in entry 1; 0.30 mol of aldehyde and 0.15 mol
of propargyl bromide in entries 2-4; 25 mL of cyclohexane; 50
b
mL of water (saturated with salt); 0.30 mol of Zn powder. ZnCl2,
5 M. c 25 mL of NH4Cl, 5 M, and 25 mL of NaOH, 5 M, pH ) 9.
rise to the above radical anions.29 It is worthy of note
that these radical anions with a well-defined geometry
are also operating in the reaction between the isomeric
mixture (C4H7)Br and Bu3SnCl, the sole recovered isomer
being Bu3SnCH(CH3)CHdCH2 when cyclohexane is em-
ployed as cosolvent.18
Scheme 1 summarizes all the results obtained so far
by means of the zinc-mediated method for couplings of
allyl halides either with aldehydes or triorganotin ha-
lides.17,18 It accounts for the stereochemistry of the
reactions involved. Steps a-c deal with the formation
of radical anions via an electron transfer from the zinc
metal to the organic halide. A rearrangement process
must occur in step b because addition of the trapping
(29) Rearrangements of radicals adsorbed on a metal surface have
been thought to occur during Grignard reagent formation, that is
during the electron-transfer step. Presently, there is no support about
this kind of rearrangement of radical anions. See: Hill, E. A. J .
Organomet. Chem. 1975, 91, 123 and references therein. (b) Walborsky,
H. M.; Aronoff, M. S. J . Organomet. Chem. 1973, 51, 31.
(30) Walborsky, H. M. Acc. Chem. Res. 1990, 23, 286. (b) Walling,
C. Acc. Chem. Res. 1991, 24, 255.
(31) Garst, J . F. Acc. Chem. Res. 1991, 24, 95.
(32) Nishigaichi, Y.; Takuwa, A.; Naruta, Y.; Maruyama, K. Tetra-
hedron 1993, 49, 7395 and references therein.

Allylation of Aldehydes with Allyl Halides
J . Org. Chem., Vol. 61, No. 8, 1996 2735
Ta ble 7. Ster eoch em ica l Cou r se of th e Rea ction s
betw een Isobu tyr a ld eh yd e (2) a n d Allyl Ha lid e 9b in
Cycloh exa n e/H2O(Sa lt)/Zn /Bu 2Sn Cl2 Med iu m a ,b
Ta ble 8. Ster eoch em ica l Cou r se of th e Rea ction s
betw een Aceta l 11 a n d Allyl Ha lid e 9b in Cycloh exa n e/
H2O (LiClO4, 5 M)/Zn /Bu 2Sn Cl2 Med iu m a
isolated
product
isomeric composition, %
C2H5CH(OCH3)2 + 9b f
entry
no.
11
salt
NH4Cl
NHd
yield, % threo erythro trans cis
(threo,erythro)-C2H5CH(OH)CH(CH3)CHdCH2
+
1
2
81
79
65
67
24
23
1
1
10
9
(trans,cis)-C2H5CH(OH)CH2CHdCHCH3
C(NH3)2‚HCl
LiClO4
LiClO4
3
4
78
88
66
57
23
25
1
3
10
15
isolated
entry molar ratio product
isomeric composition, %
no.
acetal:Sn
yield, % threo erythro trans
cis
a
Reactions conditions: 0.15 mol of 2; 0.30 mol of 9b; 25 mL of
cyclohexane; 50 mL of water (saturated with salt); 0.30 mol of Zn
powder; 0.075 mol of Bu2SnCl2 in entries 1-3; 0.15 mol of
1
2
3
4
5
6
32
16
8
4
2
55
58
65
52
63
65
46
44
43
43
41
39
52
53
52
51
49
45
2
3
5
6
8
b
Bu2SnCl2 in entry 4. For the alcoholic isomeric mixtures, see
Table 2.
2
3
Sch em e 2. Sch em a tic Rep r esen ta tion of th e
Over a ll P r ocess for th e System (C4H7)Br /
(CH3)2CHCHO/Bu 2Sn Cl2/Zn in Cycloh exa n e/H2O
(NH4Cl)
1
13
a
Reactions conditions: 0.15 mol of 11 in entry 1-5; 0.075 mol
of 11 in entry 6; 0.30 mol of 9b, 25 mL of cyclohexane, 50 mL of
salt aqueous solution, and 0.30 mol of Zn powder; 4.7, 9.4, 18.8,
37.5, 75.0, and 75.0 mmol of Bu2SnCl2 in entries 1-6, respectively.
the same ratio (cf. Table 5). However, we think that step
a is prevalent because its rate is certainly greater than
the overall rates of steps b, d, and e.
The examined reactions deal with systems where
competitive coupling reactions occur. In the system CH3-
CH(OCH3)2/(C4H7)Br/Bu2SnCl2 no competition occurs
because only the organotin chloride undergoes the cou-
pling reaction.35 Results from this system are given in
Table 8. Homoallylic alcohols are recovered in 50-60%
yield as mixtures containing mostly the threo and erythro
isomers together with variable amounts of cis-isomer.
The content of cis-isomer gradually increases from 2 to
13% on account of the increase in Bu2SnCl2. In fact,
when the organotin dichloride considerably exceeds both
the reactants (C4H7)Br (9b) and acetal,36 the isomeric
composition of the recovered mixture is threo:erythro:
trans:cis ) 31:46:1:22. Rates of production of homoallylic
alcohols are independent of the Bu2SnCl2 amount. There-
fore, activation of the allylation process via an acid-base
interaction of the acetal with Bu2SnCl2 can be excluded.37
Thus, allyltin halides produced via steps b and d, as
shown in Scheme 2, seem to be responsible for the
allylation of the acetal, as has been previously found in
aqueous medium.6
7 show that â-methylhomoallylic alcohols are isolated in
high yield (77-88%) without any recovery of allyltin
species. The following points are relevant: (a) the
isolated alcoholic mixtures contain the four isomers threo,
erythro, trans, and cis with ratios varying from 66:23:1:
10 to 57:25:3:15, respectively; (b) the cis-isomer percent-
age increases with the increase of the amount of Bu2-
SnCl2 (see Table 7, compare entry 4 with entries 1-3);
and (c) these mixtures have different isomeric composi-
tion with respect to those arising from runs listed in
Table 2. Scheme 2 explains these results.
Step a deals with the coupling reaction (see eq 1)
forming the (threo,erythro)-â-methylhomoallylic alcohols.
Step b competes with step a to form Bu2(Cl)SnCH(CH3)-
CHdCH2. This compound can interact through step c
with the aldehyde to give only alcohol in the cis-form.33
Alternatively, it undergoes redistribution with unreacted
Con clu sion
The zinc-mediated method recalls the synthetic pro-
cedures based on the heterogeneous reactions occurring
at a solid-liquid surface such as the Grignard and
Barbier reactions. However, unlike these methods, it
makes use of a metal together with a two-phase cosol-
vent-water (salt) system. The water phase is very
important, since the coupling reactions are highly acti-
vated by its presence.1,2,18 Under such conditions, the
zinc particles, located mainly at the boundary surface of
the emulsified system, are forced to interact with the allyl
bromide, present with the aldehyde in the organic phase,
34
Bu2SnCl2 to give trans- and cis-Bu2(Cl)SnCH2CHdC-
HCH3 (step d) which react with aldehyde via step e to
give a mixture of threo- and erythro-â-methylhomoallylic
alcohols. It is noteworthy that the cis-alcohol arises only
via steps b and c, while the threo- and erythro-alcoholic
mixture can be formed through step a, or alternatively
via steps b, d, and e. We cannot differentiate the extent
of the contributions of the two processes leading to the
threo and erythro alcoholic mixture because either step
a or steps b, d, and e lead to alcoholic mixtures having
(35) Under these conditions acetaldehyde- and benzaldehyde di-
methyl acetals as well acrolein aldehyde diethyl acetals do not give
reactions with the zinc-mediated method.
(36) Using the same amount of entry 43, the mixture acetal/(C4H7)-
Br is added in 25 min to the system cyclohexane/H2O (LiClO4)/Zn/Bu2-
SnCl2.
(33) Gambaro, A.; Ganis, P.; Marton, D.; Peruzzo, V.; Tagliavini, G.
J . Organomet. Chem. 1982, 231, 307.
(34) Gambaro, A.; Marton, D.; Tagliavini, G. J . Organomet. Chem.
1981, 210, 57.
(37) 1H- and 13C-NMR measurements show that acid-base equilib-
ria between Bu2SnCl2 and acetal are not operating.

2736 J . Org. Chem., Vol. 61, No. 8, 1996
Marton et al.
equipped with a condenser and dropping funnel was added
the appropriate aldehyde to a cosolvent/water (salt saturated)
mixture. Then, zinc powder was added (quantities are given
in Table 1-4). Under magnetic stirring, CH2dCHCH(CH3)-
Cl (9a ) or (C4H7)Br (9b) (generally in 1:1 stoichiometric ratio
with respect to zinc) was added at a rate sufficient to maintain
a gentle reflux due the exothermicity of the reactions. The
addition lasts about 30-40 min, a time to note the disappear-
ance of the most part of the zinc powder or its total disap-
pearance, as it has been noted in many runs. Then, the
progress of the reactions was monitored by means of GC: the
complete standing of the product peaks at constant values
marked the end of the reactions. Completion of the reactions
was verified in a time of about 30-60 min. During this time,
formation of a white solid was observed. The solid together
with the zinc residue was filtered off, and the cosolvent layer
was separated from the aqueous one. Extraction over the
aqueous layer was made with two portions (20 mL each) of
cyclohexane or ethyl ether. The organic fractions were col-
lected together and dried over Na2SO4, and then the organic
solution separated from the salt was submitted to fractional
distillation. In all cases, the isomeric composition of the
isolated alcoholic mixtures was the same as that checked
before workup. Yields, listed in Tables 1-4, have been
calculated from the amounts of the pure separated mixture of
alcohols, and the isomeric compositions have been determined
by GC.
by the high cohesive energy density developed by the
water phase.38 Radical ions are formed by a one-electron
transfer from zinc metal to (C4H7)Br (see Scheme 1). The
transfer, accelerated by the high interfacial energy
developed by water, occurs under kinetic control. Among
the several transition states, the more compact will be
favored, with the more negative ∆Vq.39 Namely, the
geometry of a branched allyl group will have a volume
smaller than that of cis- and trans-linear groups. Thus,
the branched radical ion adsorbed on the zinc surface will
be trapped by the RCHO species to form regioselectively
threo- and erythro-â-methylhomoallylic alcohols. The
regioselection encountered in the case of a (C4H7)Br
mixture is not a unique example, since other substituted
allyl bromides such as prenyl always react at the more
substituted carbon.14,15 On the other hand, regioselection
is observed in the case of propargyl substrates, mainly
with propargyl bromide. Also, in the case of 1-bromo-2-
octyne, the isolated alcoholic mixture is richer in the
branched R-allenic alcohols (70%) than the linear â-acet-
ylenic alcohol (30%). Therefore, rearrangements of the
organic reactants arise only during the electron-transfer
step to give radical anions of well-defined geometry which
are responsible for the observed regioselectivity.
3-Meth yl-1-h exen -4-ol (Ta ble 1). In all cases, the isolated
3-methyl-1-hexen-4-ol consists of isomeric mixtures having the
same ratio, threo:erythro ) 52:48. These isomeric mixtures
have been distilled in the range 110-114 °C/150 mmHg. 13C-
NMR chemical shifts are in agreement with those already
published.22 Yields are greater when NaCl is used (67-81%)
with respect to NH4Cl (39-40%).
3,5-Dim eth yl-1-h exen -4-ol (Ta ble 2). The isomeric ratio
threo:erythro ) 75:25 is maintained in all runs independently
on the employed cosolvent. An isomeric mixture with a ratio
threo:erythro:(cis+trans) ) 74:24:1 boils in the range 150-156
°C/760 mmHg. 13C-NMR chemical shifts are in agreement
with those already published.22 Yields are in the range 62-
77%.
With regard to the applicability of this method, in spite
of some limitations, it is successful in many cases of
addition of allyl and propargyl halides to carbonyl
compounds. The method has also been applied to cou-
pling reactions of organotin derivatives with allyl, pro-
pargyl, and benzyl bromide. On the basis of its rapidity
and convenience together with the possibility of using
aqueous commercial solutions, it offers important advan-
tages over other procedures.
Exp er im en ta l Section
All manipulations were carried out at room temperature,
under air atmosphere. Solvents, salts, and zinc powder (325
mesh), commercially available, were used as received. All
chemicals, commercially available, were distilled before use.
After distillation of technical crotyl bromide, from Aldrich, the
resulting product (C4H7)Br was an isomeric mixture of R-me-
thylallyl (15%), trans-crotyl (76%) and cis-crotyl (9%) bromide.
Bu2SnCl2 was purified by recrystallization before use.
Characterization of all prepared compounds was made by
means of IR spectroscopy and 13C-NMR. IR spectra were
registered as liquid film using KBr optics. 13C-NMR spectra
of samples were recorded as pure liquid using Me4Si as
internal standard. Analysis of the isolated mixtures was based
on the integrated intensities of the appropriate 13C-NMR
signals, as previously described.22,33,40 Quantitative determi-
nations from 13C-NMR spectra were made by using sufficient
long pulse intervals (at least in the range 20-30 s) in order to
avoid saturation of the nuclear spins, together with the gated
decoupling method in order to suppress the nuclear Overhause
effect (NOE).41 Additional GC analyses were performed using
a gas-chromatograph equipped with a flame ionization detector
3,5,5-Tr im eth yl-1-h exen -4-ol (Ta ble 3, En tr ies 1-3).
Isomeric mixtures of this alcohol have been recovered with a
ratio threo:erythro:(cis + trans) ) 76:20:4. 13C-NMR chemical
shifts are in agreement with the previous data.22 Yields are
in the range 55-76%.
3- Meth yl-4-cycloh exyl-1-bu ten -4-ol (Ta ble 3, En tr ies
4 a n d 5). In both cases, the alcoholic mixtures (bp 115 °C/20
mmHg) are recovered in a threo:erythro ratio of around 75:25.
13C-NMR: threo-isomer δ 17.7, 26.4, 26.6, 26.8, 40.7, 41.2, 79.2,
114.8, 140.6; erythro-isomer δ 14.9, 26.6, 28.5, 29.9, 40.9, 40.9,
79.0, 113.7, 142.6. Yields are in the range 77-80%.
3-Meth yl-4-p h en yl-1-bu ten -4-ol (Ta ble 3, En tr ies 6 a n d
7). The isolated mixtures show a threo:erythro ratio of 42:58
and 49:51 for entries 6 and 7, respectively. These mixtures
boil in the range 93-104 °C/10 mmHg. 13C-NMR chemical
shifts are in agreement with those previously reported.22
Yields are between 55 and 64%.
3-Meth yl-1,5-h exa d ien -4-ol (Ta ble 4, En tr y 1). The
isomeric composition is threo:erythro ) 40:60. 13C-NMR
chemical shifts are in agreement with the previous data.42 The
yield in this case is very low, 15%.
(DB 225 polar column, 15 m, 0.25 mm i.d., Ti ) 200 °C, Td
)
tr a n s-3-Met h yl-1,5-h ep t a d ien -4-ol (Ta b le 4, E n t r ies
2-4). The isomeric ratios threo:erythro of the isolated products
are about 50:50. 13C-NMR chemical shifts are in agreement
with the previous data.42 Yields are in the range 20-24%. In
these cases it has been ascertained that the low yields are
mainly due to the dimerization of the crotonaldehyde leading
to 1:1 isomeric mixtures of (meso,dl)-2,6-octadiene-4,5-diol. A
pure sample (4 g) of this compound with an isomeric ratio
meso:dl ) 42:58 (bp 105-107 °C/9 mmHg, (lit.43 113-114 °C/9
mmHg)) has been isolated from the collected samples. 13C-
220 °C, appropriate temperature programms of the column,
Tc, were chosen for the analysis of the different alcohols,
nitrogen as carrier gas at 10 psi).
Rea ction s of Ald eh yd es 1-8 a n d Allyl Ha lid es 9a a n d
9b (Ta bles 1-4). A general procedure has been used in all
these cases. In a round-bottom two-necked flask (250 mL)
(38) Water is thought to have a cohesive energy density of 27 kbars.
See, Lubineau, A; Meyer, E. Tetrahedron 1988, 44, 6065.
(39) Dack, M. R. J . Chem. Soc. Rev. 1975, 4, 211.
(40) Gambaro, A.; Boaretto, A.; Marton, D.; Tagliavini, G. J .
Organomet. Chem. 1983, 254 293.
(41) Freeman, R.; Hill, M. D. W.; Kaptein, R. J . Magn. Reson. 1972,
7, 327.
(42) Marton, D.; Vanzan, N. Ann. Chim (Rome) 1984, 79, 479.
(43) William, G. Y.; Levanas, L.; J aisaitis, Z. J . Am. Chem. Soc.
1936, 58, 2274.

Allylation of Aldehydes with Allyl Halides
J . Org. Chem., Vol. 61, No. 8, 1996 2737
NMR chemical shift values are in agreement with those
already reported.44
aldehyde 2 and allyl bromide 9b were performed in the
presence of Bu2SnCl2 (quantities are given in Table 7) following
the procedure above adopted for the reactions between alde-
hyde 2 and 5 and CH2dCHCH2Br in THF/H2O (NH4Cl)/Zn/
Bu3SnCl. Isomeric mixtures of homoallylic alcohols are iso-
lated in very high yields (78-88%) without any recovery of
allyltin species. Both diastereomeric pairs were present:
(threo,erythro)-3,5-dimethyl-1-hexen-4-ol as major product to-
gether with a small amount of (trans,cis)-3-methyl-5-hexen-
3-ol. Indeed, the alcoholic mixtures have an isomeric ratio
threo:erythro:trans:cis ) 66:23:1:10 for entries 1-3 and 57:25:
3:15 for entry 4. The increase in cis-isomer in entry 4 is related
to the increase of the amount of Bu2SnCl2.
tr a n s-3-Meth yl-1,5-n on a d ien -4-ol (Ta ble 4, En tr y 5 a n d
6). In both cases, the isomeric alcoholic mixtures have a threo:
erythro ratio roughly around 44:56. 13C-NMR chemical shifts
are in agreement with those previously reported.42 Yields are
in the range 36-42%.
Rea ction s of Isobu tyr a ld eh yd e (2) a n d P r op a r gyl
Br om id e (11) (Ta ble 6). The above-described procedure has
also been adopted in these cases. The appropriate amount of
propargyl bromide was added to the cyclohexane/H2O (salt)/
Zn system using an 80% weight solution in toluene. Therefore,
the cosolvent was represented by a mixture of cyclohexane and
toluene.
Rea ction s of Aceta ld eh yd e Dim eth yl Aceta l (11) a n d
(C4H7)Br (9b) in th e P r esen ce of a Ca ta lytic Am ou n t of
Bu 2Sn Cl2 (Ta ble 8). Following the above procedure, runs
have been performed on varying amounts of Bu2SnCl2 from
4.7-75 mmol. Yields of homoallylic alcohols are found in the
range 52-65%. Under these conditions, the isomeric composi-
tions of the isolated homoallylic alcohols depend on the amount
of Bu2SnCl2. Indeed, the percentage of the cis-CH3CH(OH)-
CH2CHdCHCH3 isomer increases from 2 to 13% on varying
the molar ratio acetal:Sn from 32 to 1, respectively.
2-Meth yl-5-h exyn -3-ol. The isomeric alcoholic mixtures
propargyl isomer:allenyl isomer ) 90:10 boil in the range 134-
136 °C at 760 mmHg. 13C-NMR chemical shifts were in
agreement with those previously reported.10 Yields are found
in the range 21-33%.
Rea ction s of Ald eh yd es 2 a n d 5 w ith CH2dCHCH2Br
in th e P r esen ce of Bu 3Sn Cl. Workup was the same, as
above described. Addition of the components was made in the
following order: THF (25 mL), H2O, NH4Cl saturated (50 mL),
aldehyde (0.075 mol), Bu3SnCl (24.7 g, 0.075 mol), Zn powder
(6.5 g, 0.1 mol), then allyl bromide (12.1 g, 0.1 mol). In both
runs, nearly equimolar mixtures of homoallylic alcohols and
Bu3SnCH2CHdCH2 were obtained.
Ack n ow led gm en t. We are indebted to the Murst
(Rome) and Consiglio Nazionale delle Ricerche (Rome),
Progetto Finalizzato Chimica Fine II, for financial
support.
Rea ction s of Isobu tyr a ld eh yd e (2) w ith (C4H7)Br (9b)
in th e P r esen ce of Bu 2Sn Cl2 (Ta ble 7). Reactions between
(44) Danechpajouh, H. C. R. Acad. Sci. Paris, Ser. C 1978, 286, 595.
J O951562H