Copper-Mediated and -Catalyzed o-DPPB-Directed Allylic Substitution

Article abstract of DOI:10.1002/1615-4169(200107)343:5<429::AID-ADSC429>3.0.CO;2-K

Complete control of chemo-, regio- and stereoselectivity in the course of copper-catalyzed and -mediated allylic substitution could be obtained with me ortho-diphenylphosphanyl (o-DPPB) function as a reagent-directing leaving group. Complete chirality transfer by way of a syn-addition process has been achieved for cyclic and acyclic systems. Readily available Grignard reagents may be employed as nucleophiles and the directing o-DPPB group can be recovered quantitatively. The reaction requires neither cooling nor an excess of organometallic reagent.

Full text of DOI:10.1002/1615-4169(200107)343:5<429::AID-ADSC429>3.0.CO;2-K

COMMUNICATION
Copper-Mediated and -Catalyzed o-DPPB-Directed
Allylic Substitution
Bernhard Breit,* Peter Demel
Organisch-Chemisches Institut, Ruprecht-Karls-UniversitaÈt Heidelberg, Im Neuenheimer Feld 270,
69120 Heidelberg, Germany
Fax: (+49) 6221±54±4205, e-mail: bernhard.breit@urz.uni-heidelberg.de
Received April 24, 2001; Accepted May 26, 2001
This paper is dedicated to Professor David A. Evans on the occasion of his 60th birthday
Abstract: Complete control of chemo-, regio- and
stereoselectivity in the course of copper-catalyzed
and -mediated allylic substitution could be ob-
tained with the ortho-diphenylphosphanyl (o-
DPPB) function as a reagent-directing leaving
group. Complete chirality transfer by way of a syn-
addition process has been achieved for cyclic and
acyclic systems. Readily available Grignard re-
agents may be employed as nucleophiles and the
directing o-DPPB group can be recovered quantita-
tively. The reaction requires neither cooling nor an
Scheme 1.
excess of organometallic reagent.
Keywords: allylic substitution; asymmetric synth-
esis; organocopper reagents; synthetic methods
try upon reaction of acyclic derivatives is often unsa-
tisfactory. As a consequence, chirality transfer in
acyclic allylic systems may be incomplete.[2] Fre-
quently, an excess of organocopper reagent has to be
used which is undesirable, in particular, if valuable
organic residues are to be transferred. Additionally,
the directing leaving group is generally irreversibly
lost in the overall process.
We recently identified the ortho-diphenylphospha-
nylbenzoyl (o-DPPB)-group as a multifunctional re-
agent-directing group.[4] Several late transition
metal-catalyzed or -mediated processes such as hy-
droformylation,[5] rhodium-catalyzed domino pro-
cesses,[6] palladium-catalyzed atropselective biaryl
coupling[7] as well as conjugate addition of organo-
copper reagents[8] have been described. We now re-
port on a new function of this group, namely its use
as a reagent-directing leaving group for regio- and
stereoselective allylic substitution with organocopper
reagents.
Synthetic methods which allow the stereoselective
construction of a desired carbon skeleton in a predict-
able and reliable fashion are of great value to organic
synthesis. In this respect allylic substitution with or-
ganocopper reagents is a synthetically appealing op-
eration since hard nucleophiles such as alkyl- and
aryl-type substituents may be introduced into an ex-
isting carbon skeleton. However, efficient control of
regioselectivity, i.e., a (SN2) versus c (SN2') attack is
often a severe problem (see Scheme 1).[1] A solution
to this problem might be a directing leaving group
which may control the trajectory of the incoming cop-
per nucleophile to occur as an exclusive c-attack. Ad-
ditionally, a directing leaving group may reverse the
stereochemistry of allylic substitution to occur as a
syn-addition in opposition to the standard anti-attack
relative to the leaving group. Several leaving groups
have been evaluated for this purpose, among which
carbamates[2] and benzothiazoles[3] have proven to
be useful. However, both systems suffer from several
drawbacks. For instance, control over alkene geome-
Our study commenced with the allylic-o-DPPB es-
ter (±)-1. This substrate was chosen since it is structu-
rally not biased towards preference of either a- or c-
attack to give alkenes 3 and 4, respectively.
In a first experiment, o-DPPB ester 1 was treated
with two equivalents of dimethylcuprate at ±20 °C
Adv. Synth. Catal. 2001, 343, No. 5
Ó WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2001
1615-4150/01/34305-429±432 $ 17.50-.50/0
429

asc.wiley-vch.de
Table 1: Results of allylic substitution of esters (±)-1 and (±)-2 with organocopper reagents.
entry
equiv. [MR]
X =
T [°C]
t [h]
yield (%)[a]
ratio of 3[b]:4:5
1
2
3
4
5
6
7
8
9
2 Me2CuLi
2 Me2CuLi
2 Me2CuLi
2 Me(CN)CuLi
1 CuBr´SMe2/1.1 MeMgI
1 CuBr´SMe2/1.1 MeMgI
2 MeMgI
1 CuBr´SMe2/1.1 BuMgCl
1 CuBr´SMe2/1.1 i-PrMgCl
1 CuBr´SMe2/1.1 PhMgBr
P
P
CH
P
P
CH
P
P
P
P
±20
±80
±20
±20
rt
rt
rt
rt
rt
3
6
3
> 95
> 95
> 95
> 95
> 95
80[c]
> 95
98[d]
84[d]
94[b]
75:25:-
74:26:-
78:22:±
95:5:±
>99:<1:-
42[e]:15:21
±:±:>99
>99:<1:±
97:3:±
65
2 min
6
12
2
2
2
10
rt
88:12:±
[a] GC yield.
[b] E/Z > 99:1.
[c] Conversion.
[d] Isolated yield after aqueous work-up and flash chromatography.
[e] The SN2' product 3 was obtained as a E/Z mixture (64:36).
(Table 1, entry 1). A quantitative conversion was ob- excellent levels of regioselectivity and with complete
served to give a 3:1 mixture of SN2' product 3 and SN2 control of olefin geometry (E:Z > 99:1 in all cases, en-
product 4. Decreasing the reaction temperature did tries 8±10). The reaction is compatible with the pre-
not effect the regioselectivity (entry 2).
sence of an ester group and can even enforce decon-
To test whether this result is due to a directing ef- jugation as shown for the reaction with styrene
fect of the o-DPPB group the CH-derivative 2 was em- derivative (±)-6 to give E-alkene (±)-7 in excellent
ployed (entry 3). Thus, both benzoate functions yield and selectivity (Scheme 2). Notably, in all cases
should have the same steric demand whereas the the ortho-diphenylphosphinobenzoic acid could be
phosphine-free derivative 2 should have lost its ability recovered in the work-up process almost quantita-
to coordinate the organocopper reagent. However, in tively.
both cases almost the same regioselectivity and reac-
tivity were observed (entries 1 and 3). Hence, at least
for the dimethylcuprate reagent a directing effect of
the o-DPPB group may be excluded. We next
switched to a lower order cyanocuprate which is
known to provide an intrinsically higher SN2' selectiv-
ity.[1] Accordingly, the c-substitution product 3 was
obtained in a 97:3 ratio (entry 4). However, reaction
rate was rather low. The best result was obtained
Scheme 2.
when we pretreated the o-DPPB-ester 1 with one
To probe the stereochemistry of the o-DPPB-direct-
equivalent of copper bromide ´ dimethyl sulfide.[9]
ed allylic substitution we chose the cis- and trans-cy-
Subsequent addition of just one equivalent of methyl
clohexenes (±)-8 and (±)-10 as model systems. In both
Grignard reagent in ether provided even at room
cases a completely syn-selective nucleophile transfer
temperature within less than two minutes the SN2'
was observed to give cis- and trans-menthene [(±)-9,
product exclusively with complete control of alkene
(±)-11], respectively (Scheme 3 and Scheme 4).
geometry (entry 5). In a control experiment the CH-
benzoate 2 was used. A significant decrease of che-
mo- and regioselectivity as well as E/Z-selectivity
was found (entry 6). Furthermore the reaction is sig-
nificantly slower in the absence of the phosphane
function. Six hours were required to achieve a con-
version of 80% whereas the reaction for o-DPPB deri-
vative 1 was completed in less than two minutes un-
der the same reaction conditions.[10] Hence, both
Scheme 3.
observations, the differences in chemo-, regio- and
diastereoselectivity as well as the significant rate ac-
celeration underline the role of the o-DPPB group to
function as a directing leaving group.
To explore chirality transfer in acyclic systems the
enantiomerically pure allylic o-DPPB ester derivative
(±)-12 was selected.
In addition to a methyl group, n-butyl, iso-propyl,
and a phenyl substituent could be transferred with
The substitution reaction proceeded smoothly un-
der standard conditions at room temperature to fur-
430
Adv. Synth. Catal. 2001, 343, 429±432

COMMUNICATION
by chiral GC, Betadexä 110) [a]2D0: ±54°, c 1.9 in CHCl3] in 25
mL of freshly dried diethyl ether (0.05 M) was added
CuBr´SMe2 (257 mg, 1.25 mmol, 1.0 equiv., 99%) in one por-
tion and the resulting yellow suspension was stirred for
5 minutes at room temperature. MeMgI (1.3 mL, 1.38 mmol,
1.1 equiv., 1.06 M solution in diethyl ether) was added drop-
wise within 5 minutes and the bright yellow suspension was
stirred for further 2 hours at room temperature. The reac-
tion was quenched by successive addition of a saturated aqu-
eous NH4Cl solution (8 mL) and an aqueous ammonia solu-
tion (12.5%, 10 mL) followed by the addition of pentane/
diethyl ether (1:1, 10 mL). The mixture was stirred at rt for
10 min during which a yellow precipitate formed. The mix-
ture was filtered and the residue washed with additional
pentane/diethyl ether (5:1, 20 mL). The organic phase of
the filtrate was separated and the aqueous phase was ex-
tracted with three further portions of diethyl ether (10 ml).
The combined organic phases were washed with brine,
dried (MgSO4), and the solvent removed under vacuum.
Flash chromatography (SiO2, eluent pentane) furnished
(S)-2-phenyl-3-hexene (13); yield: 184 mg (92%); [a]2D0: +10°
(c 1.5, pentane); ee > 99% (GC, permethyl-b-cyclodextrin
column).
Scheme 4.
nish the E-alkene (+)-13 in 92% isolated yield (regio-
selectivity > 99:1, Scheme 5). The enantiomeric purity
of the product alkene (+)-13 was determined by chiral
GC to be > 99%.[11] Accordingly, the o-DPPB directed
allylic substitution occurs also for acyclic systems
with complete chirality transfer by way of a syn-addi-
tion process. The formation of the substitution pro-
duct (+)-13 is readily rationalized via reactive confor-
mation 14 which should be favored based on steric
and stereoelectronic arguments. Notably, the reac-
tion can be run with substoichiometric amounts of
copper salt (20 mol %) without significant loss of se-
lectivity (Scheme 5).
Selected physical data of 13: 1H NMR (CDCl3, 500.13 MHz):
d = 0.98 (t, J = 7.4 Hz, 3 H), 1.33 (d, J = 7.0 Hz, 3 H), 2.03 (m,
2 H), 3.42 (mc, 1 H), 5.49 (dt, J = 15.2, 6.0 Hz, 1 H), 5.59 (dt, J =
15.4, 6.7 Hz, 1 H), 7.16±7.22 (m, 3 H), 7.28±7.36 (m, 2 H);
13C NMR (CDCl3, 125.76 MHz): d = 13.9, 21.6, 25.5, 42.2,
125.9, 127.2 (2 C), 128.3 (2 C), 130.8, 133.9, 146.6; HR-MS
(EI): calcd. for C12H16: 160.1252, found 160.1238.
Recovery of the ortho-Diphenylphosphanyl-
benzoic Acid (o-DPPBA)
The yellow residue obtained in the filtration process above
was dissolved in dichloromethane (20 mL), washed with
water (10 mL), followed by addition of pH 4.75 buffer in
methanol/water (1:1, 20 mL) and KCN (600 mg, 9.2 mmol)
upon which the color of the organic phase changed from yel-
low to colorless. The phases were separated and the aqu-
eous phase washed with three more portions of dichloro-
methane (10 mL each). The combined organic phases were
washed with water (20 mL), dried (MgSO4), and the solvent
removed under vacuum to give o-DPPBA as a pale yellow so-
lid; yield: 303 mg (85%).
Scheme 5.
In summary the copper-mediated (and -catalyzed)
o-DPPB directed allylic substitution with Grignard nuc
-
leophiles occurs with complete control of regio- and
stereochemistry as well as alkene geometry for both
cyclic and acyclic allylic alcohol derivatives. Readily
available Grignard reagents may be employed as nuc-
leophiles and the directing o-DPPB group can be re-
covered quantitatively. The reaction requires neither
cooling nor an excess of organometallic reagent and
even catalytic amounts of copper may be sufficient.
Reaction Procedure for Employment of Catalytic
Amounts of Copper(I) Bromide ´ Dimethyl Sulfide
To a solution of o-DPPB-ester 12 (276 mg, 0.61 mmol, ee
>99%) in diethyl ether (12 mL) was added copper bromide ´
dimethyl sulfide (25.1 mg, 0.12 mmol) in one portion and the
resulting yellow suspension was stirred for 5 min at room
temperature. A 0.96 M solution of methylmagnesium iodide
(0.70 mL, 0.67 mmol) was added within 5 min and the bright
yellow suspension was stirred for further 3 h at room tem-
perature. The reaction was quenched by successive addition
of a saturated aqueous NH4Cl-solution (7 mL) and an aqu-
eous ammonia solution (12.5%, 4 mL) followed by the addi-
tion of diethyl ether (12 mL). The organic phase was sepa-
rated and the aqueous phase was extracted with three
further portions of diethyl ether (10 mL). The combined or-
Experimental Section
Representative Procedure:
Synthesis of (S)-2-Phenyl-3-hexene (13)
To a solution of o-DPPB-ester 12 (565 mg, 1.25 mmol, 1.0
equiv.) [ee > 99% (determined at the stage of the allyl alcohol
Adv. Synth. Catal. 2001, 343, 429±432
431

asc.wiley-vch.de
ganic phases were washed with brine, dried (MgSO4), and
the solvent removed under vacuum. Column chromatogra-
phy (SiO2, eluent pentane 9:1) furnished (+)-13; yield:
83 mg (85%, ee >99%). Analytical GC (CP SIL 5CB, Chrom-
pack; 40 °C, 2 min, 10 °C/min, 200 °C, 12 min): Rt[(+)-13] =
11.36 min (91%), (Z)-alkenic isomer, Rt = 11.12 min (5%)
Chem. Soc., Perkin Trans. I, 1983, 2953±2956; (c) S.
Valverde, M. BernabeÂ, S. Garcia-Ochoa, A. M. GoÂmez,
J. Org. Chem. 1990, 55, 2294±2298.
[4] B. Breit, Chem. Eur. J. 2000, 6, 1519±1524.
[5] (a) B. Breit, Angew. Chem. 1996, 108, 3021±3023; An-
gew. Chem. Int. Ed. Engl. 1996, 35, 2835±2837; (b) B.
Breit, Liebigs Ann./Recueil 1997, 1841±1851; (c) B.
Breit, M. Dauber, K. Harms, Chem. Eur. J. 1999, 5,
2819±2827; (d) B. Breit, Chem. Commun. 1997, 591±
592; (e) B. Breit, Eur. J. Org. Chem. 1998, 1123±1134.
[6] (a) B. Breit, S. K. Zahn, Angew. Chem. 2001, 113,
1964±1967; Angew. Chem. Int. Ed. 2001, 40, 1913±1916;
(b) B. Breit, S. K. Zahn, Angew. Chem. 1999, 111,
1022±1024; Angew. Chem. Int. Ed. 1999, 38, 969±971;
(c) B. Breit, Tetrahedron Lett. 1998, 39, 5163±5166.
[7] B. H. Lipshutz, J. M. Keith, Angew. Chem. 1999, 111,
3743±3746; Angew. Chem. Int. Ed. 1999, 38, 3530±3533.
[8] (a) B. Breit, Angew. Chem. 1998, 110, 535±538; Angew.
Chem., Int. Ed. Engl. 1998, 37, 525±527; (b) B. Breit, P.
Demel, Tetrahedron 2000, 56, 2833±2846.
and regioisomer, (E)-1-phenyl-3-methyl-1-pentene Rt
=
12.29 min (4%). Chiral GC (Permethyl-b-cyclodextrin,
Chrompack; 65 °C): Rt[(+)-13] = 67.58 min.
Acknowledgements
This work was supported by the Fonds der Chemischen Indus-
trie, the Deutsche Forschungsgemeinschaft and the Alfried-
Krupp Award for young university teachers.
References and Notes
[9] 31P NMR spectroscopy showed formation of a cop-
[1] Y. Yamamoto in Houben-Weyl, Methods for Organic
Synthesis, 1995, Volume E21b, 2011±2040.
per(I) phosphine complex with o-DPPB ester
[
1
31P NMR (202.46 MHz, CDCl3): d = ±1 (s, br)].
[2] (a) C. Gallina, Tetrahedron Lett. 1982, 23, 3093±3096;
(b) H. L. Goering, S. S. Kantner, C. C. Tseng, J. Org.
Chem. 1983, 48, 715±721; (c) J. H. Smitrovich, K. A.
Woerpel, J. Am. Chem. Soc. 1998, 120, 12998±12999;
(d) J. H. Smitrovich, K. A. Woerpel, J. Org. Chem.
2000, 65, 1601±1614.
[3] (a) P. Barsanti, V. CaloÁ, L. Lopez, G. Marchese, F.
Naso, G. Pesce, J. Chem. Soc., Chem. Commun. 1978,
1085±1086; (b) V. CaloÁ, L. Lopez, W. F. Carlucci, J.
[10] A similar rate acceleration in the course of a Sim-
mons-Smith reaction caused by the presence of an al-
cohol directing group has been noted. See: S. Win-
stein, J. Sonnenberg, J. Am. Chem. Soc. 1961, 83,
3235±3244.
[11] (a) E. L. Eliel, J. P. Freeman, J. Am. Chem. Soc. 1952,
74, 923±928; (b) C. Spino, C. Beaulieu, J. LafrenieÁre, J.
Org. Chem. 2000, 65, 7091±7097.
432
Adv. Synth. Catal. 2001, 343, 429±432