Opening of tartrate acetals using dialkylboron bromide: Evidence for stereoselectivity downstream from ring fission

Article abstract of DOI:10.1021/ja012530g

Johnson-type acetals derived from dimethyl tartrate give, after opening with Me₂BBr and cuprate displacement, secondary alcohols with high diastereoselectivity (>30:1). The mechanism proposed for the induction of diastereoselectivity is downstream from the ring fission. It implies a direct participation of the Lewis acid as a source of nucleophile and the stereospecific transformation of the resulting bromo acetal through an invertive and temperature-dependent process. The acetals are prepared by reaction of the desired aldehyde with dimethyl tartrate. Removal of the auxiliary is accomplished through Sml₂ reduction or by an addition - elimination protocol using methoxide.

Full text of DOI:10.1021/ja012530g

Published on Web 12/14/2002
Opening of Tartrate Acetals Using Dialkylboron Bromide:
Evidence for Stereoselectivity Downstream from Ring Fission
Yvan Guindon,*,§ William W. Ogilvie,*,†, Jose´e Bordeleau,† Wei Li Cui,†
Kathy Durkin,‡ Vida Gorys,† He´le`ne Juteau,† Rene´ Lemieux,† Dennis Liotta,‡
Bruno Simoneau,† and Christiane Yoakim†
Contribution from the Boehringer Ingelheim (Canada) Ltd., Bio-Me´ga Research DiVision,
2100 rue Cunard, LaVal Que´bec, Canada H7S 2G5, Department of Chemistry, Emory
UniVersity, 1515 Pierce DriVe, Atlanta, Georgia 30322, Institut de recherches cliniques de
Montre´al (IRCM), Bio-organic Chemistry Laboratory, 110 aVenue des Pins Ouest,
Montre´al, Que´bec, Canada H2W 1R7, Department of Chemistry and Department of
Pharmacology, UniVersite´ de Montre´al, C.P. 6128, succursale Centre-Ville,
Montre´al, Que´bec, Canada H3C 3J7, and Department of Chemistry, UniVersity of Ottawa,
10 Marie Curie Street, Ottawa, Que´bec, Canada K1N 6N5
Received November 13, 2001 ; E-mail: wogilvie@science.uottawa.ca; guindoy@ircm.qc.ca
Abstract: Johnson-type acetals derived from dimethyl tartrate give, after opening with Me2BBr and cuprate
displacement, secondary alcohols with high diastereoselectivity (>30:1). The mechanism proposed for the
induction of diastereoselectivity is downstream from the ring fission. It implies a direct participation of the
Lewis acid as a source of nucleophile and the stereospecific transformation of the resulting bromo acetal
through an invertive and temperature-dependent process. The acetals are prepared by reaction of the
desired aldehyde with dimethyl tartrate. Removal of the auxiliary is accomplished through SmI2 reduction
or by an addition-elimination protocol using methoxide.
Scheme 1
The induction of stereogenic centers on acyclic molecules is
a topic of great research interest. The concept of using acetals
as chiral templates to achieve such an objective originated from
the seminal work of W. S. Johnson1 and was later extended
and enriched by the contributions of many other scientists.2-6
The opening of acetals is now a commonly used approach in
the synthesis of new molecules. A variety of reaction conditions
involving different types of acetals, Lewis acids, and nucleo-
philes have been considered. To date, three mechanisms have
* Corresponding authors. Y.G.: Telephone: (514) 987-5786. Fax: (514)
987-5789. W.O.: Telephone: (613) 562-5800. Fax: (613) 562-5170.
† Boehringer Ingelheim (Canada) Ltd.
‡ Emory University.
§ Institut e recherches cliniques de Montre´al (IRCM), Universite´ de
Montre´al, and University of Ottawa.
been envisaged for this process. The first mechanism proposed
was an SN2 process proceeding through a direct displacement
of the acetal/Lewis acid complex. In this mechanism, the Lewis
acid complexed selectively to the oxygen attached to the carbon
bearing the axial substituent (2, Scheme 1).1,2 Heathcock,
Bartlett, and Yamamoto,3 and then later Denmark4 and Davies,5
provided experimental evidence that an SN1 oxocarbenium ion
mechanism may be implicated in most instances. It was
suggested that a cyclic tight ion pair such as 3 might be present
in the transition state. In structure 3, the presence of tight ion
pairs produces facial selectivity during the approach of the
nucleophile. The third mechanism proposed involved a solvated
oxocarbenium ion such as 4 in a transition state leading to the
major product 5a, although the sense of diastereoselectivity in
this case was more difficult to explain.4
University of Ottawa.
(1) (a) Johnson, W. S.; Harbert, C. A.; Stipanovic, R. D. J. Am. Chem. Soc.
1968, 90, 5279-5280. (b) Johnson, W. S.; Harber, B. E.; Ratcliffe, B. E.;
Stipanovic, R. D. J. Am. Chem. Soc. 1976, 98, 6188-6193. (c) Bartlett, P.
A.; Johnson, W. S.; Elliot, J. D. J. Am. Chem. Soc. 1983, 105, 2088-
2089. (d) Johnson, W. S.; Kevson, A. B.; Elliott, J. D. Tetrahedron Lett.
1988, 3757-3760.
(2) See, for example: (a) Denmark, S. E.; Wilson, T. M.; Almstead, N. G. J.
Am. Chem. Soc. 1989, 111, 9258-9260. (b) Alexakis, A.; Mangeney, P.
Tetrahedron: Asymmetry 1990, 1, 477-511.
(3) Mori, I.; Ishihara, K.; Flippin, L. A.; Nozaki, K.; Yamamoto, H.; Bartlett
P. A.; Heathcock, C. H. J. Org. Chem. 1990, 55, 6107-6115.
(4) (a) Denmark, S. E.; Almstead, N. G. J. Am. Chem. Soc. 1991, 113, 8089-
8110. (b) Denmark, S. E.; Almstead, N. G. J. Org. Chem. 1991, 56, 6458-
6467. (c) Denmark, S. E.; Almstead, N. G. J. Org. Chem. 1991, 56, 6485-
6487.
(5) Bull, S. D.; Correia, L. M. A. R. B.; Davies, S. G. J. Chem. Soc., Perkin
Trans. 1 1998, 2231-2233.
(6) (a) Sammakia, T.; Smith, R. S. J. Org. Chem. 1992, 57, 2997-3000. (b)
Sammakia, T.; Smith, R. S. J. Am. Chem. Soc. 1992, 114, 10998-10999.
(c) Sammakia, T.; Smith, R. S. J. Am. Chem. Soc. 1994, 116, 7915-7916.
9
428
J. AM. CHEM. SOC. 2003, 125, 428-436
10.1021/ja012530g CCC: $25.00 © 2003 American Chemical Society

Dialkylboron Bromide To Open Tartrate Acetals
A R T I C L E S
Scheme 2
Table 1. Reaction of Tartrate Acetal 11 with Me2BBr and
Me2Cu(CN)Li2 at Various Temperatures
entry
T (°C)
1
T (°C)
2
ratioa (12a :12b)
yieldb (%)
1
2
3
4
5
6
-78
25
25
25
25
-78
-78
-40
-30
-20
0
1:1
8:1
14:1
22:1
15:1
6:1
34
52
77
76
64
38
Results offered by Sammakia6 indicated that selective com-
plexation did not take place between the Lewis acid and the
oxygen atoms of the acetal.7 Using a labeled compound,
Sammakia clearly showed that there was a similar distribution
of 9 and 10 in the products, a result inconsistent with selective
complexation (Scheme 2). Since either of the carbon-oxygen
bonds may be cleaved in the opening of an acetal, it could be
concluded that equilibration occurs between the oxocarbenium
intermediates formed during the reaction.
25
a Determined by capillary GC. b Isolated yield.
Table 2. Reaction of Various Tartrate Acetals with Various
Cuprates and Dialkylboron Bromides
Our studies have shown that secondary alcohols can be
obtained with high selectivity when tartrate acetals are opened
using dimethylboron bromide (Me2BBr) and a mixed organo
cuprate, Me2Cu(CN)Li2 (eq 1).8
entry
acetal
R
1
R
2
products A:B
ratioa (A: B)
yieldb
1
2
3
4
5
6
7
8
11
13
15
17
11
11
11
13
nC9H19
cC6H11
(CH3)2CH CH3
nC6H13
nC9H19
nC9H19
nC9H19
cC6H11
CH3
CH3
12a:12b
14a:14b
16a:16b
18a:18b
19a:19b
20a:20b
21a:21b
22a:22b
34:1
17:1
12:1
18:1
>20:1
14:1
26:1
6:1
80
83
82
77
67
45
62
68
CH3
Bu
Ph
However, good selectivity was realized only when a particular
set of experimental conditions was respected, suggesting that
an unusual mechanism was operative in the reaction. The present
study on the opening of tartrate acetals will show that the origin
of diastereoselectivity is downstream from the opening of the
acetal. This involves nucleophilic displacement of the bromide
originating from the Me2BBr to form intermediate bromo ethers
that equilibrate under thermodynamic control before reacting
with organo cuprates in a stereospecific and temperature-
dependent manner. We believe that competing SN1 and SN2
mechanisms are involved in the cuprate displacement. These
findings will be elaborated in the following pages.
CH2CHc
PhSd
a Determined by GC. b Combined isolated yield. c Cuprate added at -50
°C. d PhSH replaced R2Cu(CN)Li2 as a nucleophile.
Scheme 3: Removal of Tartrate Auxiliarya
a Method A: (i) 1.3 equiv of CH3SO2Cl, 2.0 equiv of Et3N, 0 °C, 30
min; (ii) 10.0 equiv of DBU, CH2Cl2, 0 °C, 30 min: (iii) 5.0 equiv of
NaOMe, 65 °C, 30 min. Method B: 5.0 equiv of SmI2, 10.0 equiv of MeOH,
6.0 equiv of HMPA, THF, 25 °C, 2 h.
Results and Discussion.
Addition of Me2BBr to acetal 11 at -78 °C followed by the
introduction of Me2Cu(CN)Li28,9 gave a 1:1 mixture of diaster-
eomers (Table 1, entry 1). Modest selectivity for product 12a
was achieved only when, after the addition of Me2BBr (at -78
°C), the reaction mixture was warmed to room temperature for
1 h and then recooled to -78 °C for the addition of cuprate
(entry 2). Increasing the temperature at which the cuprate was
added resulted in an increase in selectivity, the highest ratios
being noted at -30 °C (entries 3 and 4). Above -30 °C,
diastereoselection decreased (entries 5 and 6).
had much impact on the diastereoselectivity obtained. A survey
of various reagents was conducted to find the optimal conditions
for carbon-carbon bond formation. Almost no selectivity was
observed with Me2CuLi, Me3CuLi2,10 Me3Cu2Li,11 or Me5Cu3-
Li2.10 Attempts to use Grignard-derived cuprates met with failure
as did the use of MeCuCNLi. Only Me2Cu(CN)Li29 offered an
excellent ratio for carbon-carbon bond formation. The use of
thiophenol as a nucleophile (Table 2, entry 8) gave hemithio-
acetal products with a modest 6:1 diastereoselectivity.12
Auxiliary removal was accomplished using one of two
possible methods (Scheme 3). The first approach involved the
Neither the nature of the R1 group on the acetals (Table 2,
entries 1-4) nor the size of the nucleophile (entries 1, 5-7)
(7) Sammakia observed high selectivity when Lewis acid was added slowly
and when milder Lewis acids were used.
(10) (a) Still, W. C.; MacDonald, T. L. Tetrahedron Lett. 1976, 31, 2659. (b)
Ashby, E. C.; Noding, S. A. J. Org. Chem. 1979, 44, 4371. (c) Ashby, E.
C.; Lin, J. J. J. Org. Chem. 1977, 42, 2805.
(11) Ashby, E. C.; Lin, J. J.; Watkins, J. J. J. Org. Chem. 1977, 42, 1099.
(12) Crystallographic analysis of the minor isomer (22b) indicated that the thiol
addition had occurred with the same facial selectivity as was the case for
the cuprate addition (see Supporting Information).
(8) Guindon, Y.; Simoneau, B.; Yoakim, C.; Gorys, V.; Lemieux, R.; Ogilvie,
W. Tetrahedron Lett. 1991, 5453-5456.
(9) (a) Lipshutz, B. H. Synthesis 1987, 325. (b) Lipshutz, B. H.; Wilhelm, R.
S.; Kozlowski, J. A. Tetrahedron 1984, 40, 5005. (c) Lipshutz, B. H.;
Moretti, R.; Crow, R. Organic Synthesis 1990, 69, 80.
9
J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003 429

A R T I C L E S
Guindon et al.
Table 3. Ratio of Bromo Ethers Observed at Various
Temperatures
entry
acetal
R
T (°C)
ratioa
1
2
3
4
5
6
7
13
13
11
23
15
13
13
cC6H11
cC6H11
nC9H19
CH3
(CH3)2CH
cC6H11
cC6H11
-78
1.2:1
8:1
5.3:1
6.5:1
7:1
1.5:1
8:1
20
20
20
20
-78 to -30
20 to -30
a Determined by NMR.
formation of an enol ether by mesylation and elimination
followed by exposure to NaOMe in MeOH at reflux to give
the secondary alcohols in good yield. The second and more
convenient approach consisted of exposing the tartrate auxiliary
to SmI2 in the presence of HMPA using methanol or ethylene
glycol as the hydrogen source.13 In all cases, the configuration
of the major alcohol isomer was R when L-tartrate was used,
indicating that the reaction proceeded with consistent facial
selectivity.14
Figure 1. Plot showing change in ratio of bromo ethers derived from 15
and Me2BBr as a function of time at various temperatures. Curves generated
at -35 °C (0), -25 °C (O), -20 °C (]), and -15 °C (∆) are shown.
Table 4. Effect of Tartrate Functional Groups on the Ratio of
Products Obtained
Mechanistic considerations. In proposing a mechanistic
rational for the various levels of selectivity obtained, we had to
account for three significant observations: (1) good selectivity
was achieved only when the reaction mixture was warmed
briefly following the addition of Me2BBr and before cuprate
addition, (2) overall ratios were optimal when the cuprate was
added at -30 °C, (3) the best nucleophiles tested were the
Lipshutz-type cuprates9,15 and thiolates. The fact that acetals
are normally cleaved by Me2BBr at -78 °C16 and that good
selectivity could be achieved only when the reaction was
warmed at an early stage, implied that the diastereoselectivity
observed was dependent upon events occurring after ring fission.
To explore this conclusion further, we considered the role of
acyclic bromo acetal derivatives in the reaction as well as their
selective displacement by cuprates.
Acyclic Bromo Acetal Derivatives and Thermodynamic
Control. NMR experiments indicated that the acetals reacted
rapidly with Me2BBr to give the corresponding bromo ethers
(borate esters) in an equal ratio at -78 °C (Table 3, entry 1).
Upon warming to room temperature, the bromo ethers equili-
brated rapidly to give ratios of between 5:1 and 8:1 (entries
2-5). Warming the solution from -78 to -30 °C preserved
the ratio obtained at -78 °C (entry 6). Cooling the solution
from room temperature to -30 °C preserved the bromo ether
distribution obtained at room temperature (entry 7).
entry
acetal
R
products
ratioa
yieldb
1
2
3
4
5
11
24
26
28
30
CO2Me
12a:12b
25a:25b
27a:27b
29a:29b
31a:31b
34:1c
25:1c
27:1c
2:1
80
74
24
17
42
i
CO2 Pr
CONMe2
CH2OMe
Me
1:1
a Ratio determined by NMR. b Total combined yield. c Ratio determined
by GC.
ethers, following equilibration, was critical for obtaining a highly
diastereoselective ratio of final products. Second, the ratio of
final products was greater than the ratio of bromo acetals noted
(cf. Tables 2 and 3).
The rate at which the acyclic bromo acetal derivatives
equilibrated was found to be highly temperature-dependent.
Figure 1 illustrates, for the opening of acetal 15 using Me2-
BBr, the change in ratio of bromo ethers as a function of time
at various temperatures.17 At -35 °C, no equilibration was
observed. As the temperature was increased to -20 °C and then
to -15 °C, a slow equilibration was observed. At 0 °C, an 8:1
ratio was achieved after only 20 min, whereas at 20 °C,
equilibrium was reached almost instantaneously.
From these results, two important observations could be made.
First, achieving a diastereomeric excess of one of the bromo
The different structural features of the tartrate moiety were
then considered. As seen in Table 4, the presence of esters or
simple amides gave diastereoselective reactions (entries 1-3),
while replacing the esters with methoxymethylenes or methyl
groups led to a complete loss of diastereoselection (entries 4
and 5).
The cyclic and acyclic acetals used in the next part of our
study were chosen for their capacity to produce comparable
bromo acetal derivatives. As seen in Table 5, cyclic acetal 23
(13) Kusuda, K.; Inanaga, J.; Yamaguchi, M. Tetrahedron Lett. 1989, 30, 2945.
For a review, see: Soderquist, J. A. Aldrichimica Acta 1991, 24, 15.
(14) The absolute configuration of the final products was assigned by correlation
to literature values of specific rotation or by the preparation of Mosher
esters and comparison of the resulting NMR spectra with those of authentic
samples prepared from commercially available chiral alcohols (Dale, J. A.;
Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34, 2543-2549). The
corresponding D-tartrate acetals produced secondary alcohols with S
configurations.
(15) For complete results, see Supporting Information.
(16) (a) Guindon, Y.; Yoakim, C.; Morton, H. E. J. Org. Chem. 1984, 49, 3912-
3920. (b) Guindon, Y.; Anderson, P. C.; Yoakim, C.; Girard, Y.;
Berthiaume, S.; Morton, H. E. Pure Appl. Chem. 1988, 60, 1705-1714.
(17) Me2BBr was added to acetal 15 in CD2Cl2 at -78 °C, and the samples
were monitored by NMR at various probe temperatures.
9
430 J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003

Dialkylboron Bromide To Open Tartrate Acetals
A R T I C L E S
Table 5. Bromo Ether Ratios and Final Product Ratios of Acyclic
Acetals Derived from Tartrates
entry
acetal (A)
ratio (B)
products (C)
ratio (C)
1
2
3
4
5
6
7
8
23
33
35
37
39
41
43
45
6.5:1
9:1
a
32a:32b
34a:34b
36a:36b
38a:38b
40a:40b
42a:42b
44a:44b
46a:46b
15:1
7:1
1.2:1
1.1:1
7:1
1.6:1
1.6:1
1.6:1
a
9:1
4:1
4:1
4:1
a Ratio could not be determined.
Figure 2. Low-energy conformations optimized for bromo ethers 47 and
48.
The configuration of the thermodynamically preferred bromo
acetal was subsequently considered. Our initial experience with
such molecules had indicated that they were extremely tem-
perature- and air-sensitive. Theoretical calculations had to be
used to determine the thermodynamic stability of benzylated
adducts 47 and 48. A conformational search was carried out
using Monte Carlo methods in MACROMODEL. The low-energy
conformations thus obtained were optimized using the AM1
Hamiltonian in MOPAC, and the resulting optimized structures
were used to estimate the Boltzmann energy (Figure 2). Isomer
47, bearing an S configuration at the bromo ether carbon, was
found to have a ∆H°f of -26.5 kJ/mol, 10.3 kJ/mol lower than
the ∆H°f for 48, indicating that the S bromo ether was the
preferred configuration.
Closer inspection of these calculations provided insight into
the reason for the S configuration preference. The tartrate portion
of the molecules appeared to be extremely rigid, bearing
conformations as shown in Figure 2. The tartrate torsional angles
had relatively high barriers of rotation as indicated by MOPAC.
There were two potential reasons for this: (1) the two esters
tended to adopt an anti orientation to minimize electrostatic
repulsions, and (2) the two electron-withdrawing groups tended
to adopt a gauche relationship to take advantage of the gauche
effect.19 These two factors explained our previous results from
the structural studies on cyclic and acyclic acetals (Tables 4
and 5). In entries 3 and 4 (Table 5), the removal of one ester
resulted in no diastereoselection, indicating that both esters were
required for the minimization of the dipole-dipole interaction.
Unlike the methyl-substituted acetal, the fluorine-substituted
acetal gave good diastereoselection (cf. entries 5 and 6, Table
5), results consistent with the need for electronegative groups
at the Z position. Finally, altering the tartrate configuration
(entries 7 and 8) eliminated the possibility of a stabilizing gauche
effect and dipole opposition occurring simultaneously. In
summary, a favorable combination of interactions between the
X, Y, and Z components locked the tartrate into a rigid
conformation. Any change to the harmony of the interactions
would disrupt the rigidity of such derivatives (Table 5)
translating into lower selectivity.
and acyclic acetal 33 gave similar ratios of bromo ether
derivatives (B) (entries 1 and 2). Both substrates gave selective
reactions, but acetal 33 did not show the magnification of ratio
(vide supra) normally observed for cyclic tartrate acetals (entries
1 and 2, cf. ratios B and C).18 This result suggested that the
borate ester moiety formed during the opening of cyclic acetal
23 with Me2BBr played a role in the magnification of the ratio
seen for the cuprate addition.
A more detailed investigation was carried out using a variety
of acyclic precursors to determine the relative contribution of
each of the chemical features of the tartrate. With acyclic
compounds, we were able to selectively alter X, Y, and Z (Table
5) and thus study the relative contribution of these functionalities
to diastereoselection. Each ester was systematically replaced
with a methyl group (Table 5, entries 3 and 4). The results
implied that both of the ester functions were required to obtain
high stereoselection, as the deletion of either resulted in a 1:1
mixture of products. The presence of an electronegative sub-
stituent at the Z position also proved necessary (entries 2 and
5), since replacing the benzyl ether with a fluorine maintained
selectivity while the use of a methyl group at this position gave
a nonselective reaction (entry 6). The relative configuration of
the tartrate acetal was also found to be important. Cis tartrate
acetals such as 43 gave no diastereoselection (entry 7), as was
the case with meso cyclic acetal 45 (entry 8).
The energy of 48, being significantly higher than that of 47,
arose from van der Waals and torsional components between
(18) Hydrogenation of a 7:1 final product mixture of 34a:34b with Pd(OH)2
gave a 7:1 mixture of 32a:32b, indicating that the same facial bias was
maintained with the acyclic acetal system.
(19) Labelle, M.; Morton, H. E.; Guindon, Y.; Springer, J. P. J. Am. Chem.
Soc. 1988, 110, 4533.
9
J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003 431

A R T I C L E S
Guindon et al.
Scheme 4 a
Scheme 5
a Reagents and conditions: (a) 1.2 equiv of Me2BBr, CH2Cl2, -78 °C
and then allowed to warm to 25 °C, 15 min, 40%; (b) 1.5 equiv of Et3SiCl,
2.0 equiv of imidazole, DMF, 25 °C, 2 h, 95%; (c) Br-C6H4-COCl, py; (d)
HF‚py, py, THF, 1 h, 100%.
the bromide and the nearest carbonyl group as indicated by MMX
calculations. The distribution of bromo ethers 47 and 48 was
therefore a result of the conformational bias of the tartrate and
of the steric effects involving the bromide. These results were
later confirmed by single-crystal X-ray diffraction analysis of
the major isomer of the bromo ethers derived from the reaction
of 13 with Ph2BBr.20 This study indicated that the major isomer
took on the predicted S configuration at the bromo ether carbon.
In addition, the solid-state conformation, being very similar to
the low-energy conformation of 47, further validated the
calculations that had initially been used to arrive at 47.
That the reaction required both esters suggested that anchi-
meric participation may have been involved during the equili-
bration of the bromo ether derivatives. To investigate this, we
prepared tartrate 49 containing two tert-butyl esters that could
potentially trap any intermediate oxocarbenium species as
lactones. When bis-tert-butyl ester 49 was treated with Me2-
BBr at -78 °C and then allowed to warm to room temperature,
a 40% yield of lactones 50ab was obtained as an inseparable
mixture of isomers in a 2:1 ratio (Scheme 4). These compounds
were converted to the corresponding triethylsilyl ethers 51a and
51b, which were readily separable by flash chromatography.
Treatment of each product with HF‚py led to the re-installment
of isomers 50a and 50b. The relative configuration of the two
adducts was determined by NOE measurements on the p-
bromobenzoate derivatives 52a and 52b. These experiments
established that there was a cis stereochemical relationship of
dioxolane substituents for major product 50a. A temperature
profile was established to determine the rate of interconversion
of the bromo ethers (vide supra). Exposure of 49 to Me2BBr at
-78 or -30 °C did not produce more than a trace of lactones,
indicating that equilibration occurred only at higher temperatures
and through the anchimeric participation of the esters.
the bromide and the proximal ester, a consequence of the
conformational locking of the tartrate moiety (Figure 2).
A few preliminary conclusions could be drawn at this point.
First, the ring fission of the tartrate acetal was nonselective at
-78 °C, suggesting that preferential complexation with one of
the oxygen atoms of the acetal did not occur and that this step
proceeded through an SN1 mechanism. Second, enriching the S
bromo ether proved essential for achieving a high diastereose-
lectivity of final products. Third, the level of diastereoselectivity
for the final products, when cuprates were used as nucleophiles,
was much higher than that of the bromo ethers. The increase in
selectivity for the final products suggested that the cuprate
displacement had an effect on the stereochemical outcome of
the reaction.
Selective Displacement by Cuprates. Elucidation of the
carbon-carbon bond-forming step involving the cuprate reagent
began with an examination of the relationship between the ratio
of the bromo ethers and the overall diastereoselectivity. A variety
of bromo ether mixtures, of varying ratios, had to be accessed
for this part of the study. This was accomplished by mixing
different volumes of stock solutions containing different ratios
of bromo acetals. The first solution, containing a 1:1 mixture
of bromo ethers, was obtained by opening the acetal at -78
°C. The second stock solution, containing an 8:1 mixture of
bromo ethers, was obtained by warming the previous 1:1 mixture
to room temperature and then recooling the solution to -78
°C. Various amounts of each stock solution were mixed at -78
°Cssince the rate of equilibration was negligible at this
temperaturesto arrive at the variety of bromo ether ratios needed
to carry out a rigorous analysis. The resulting solutions were
then warmed to -30 °C for the addition of the cuprate reagent.
Figure 3 illustrates the results obtained. The first graph (Figure
3a) shows the mol % fraction of 16a in relation to the mol %
of 59a, the major S isomer. This experiment indicated that a
linear relationship existed between the ratio of the bromo ethers
and the diastereoselectivity obtained for the carbon-carbon
bond-forming reaction, the increase of 16a being proportional
to the increase of 59a. Interestingly, an increase in yield
proportional to the mole fraction of 59a was also found, as
shown in Figure 3b.
Scheme 5 illustrates our rationale for the bromo acetal
equilibrium. The opening of tartrate acetal 53 at -78 °C
provided an equal amount of the two bromo esters 54 and 55.
The mixture of isomers was warmed to room temperature, and
equilibrium was achieved via the intermediacy of either one of
the two esters to form the five-membered ring oxoniums 56
and 58. Bromo ether 55 was thermodynamically preferred over
54 as the latter was destabilized by steric interactions between
These results, along with the fact that equilibration of the
bromo ether derivatives was required for inducing diastereose-
(20) Drouin, M.; Michel, A. G.; Guindon, Y.; Ogilvie, W. Acta Crystallogr.,
Sect. C 1993, C49, 75.
9
432 J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003

Dialkylboron Bromide To Open Tartrate Acetals
A R T I C L E S
Scheme 6
Et2BOMe was used as a model for the tartrate-derived borate
ester. This compound displayed a broad resonance in THF at
54.4 ppm21,22 and a small resonance at 31.9 ppm (attributable
to small amounts of EtB(OMe)(OEt)).22 When 2.0 equiv of Me2-
Cu(CN)Li2 were added, the signals at 54.4 and 31.9 ppm were
no longer observed, but three new resonances at 7.7, 0.1, and
-17.9 ppm appeared. These new resonances were assigned to
Li[MeEt2BOMe], Li[Et2B(OMe)2], and Li[Me2Et2B], respec-
tively.23 The latter two species had originated from a dispro-
portionation of an initially formed -ate complex, this being the
Li[MeEt2BOMe].24 To verify this result, the experiment was
repeated using another alkylating agent. Et2BOMe and 2
equivalents of MeLi gave similar results with signals at 7.9,
0.3, and -17.6 ppm, confirming that the formation of an -ate
complex had indeed taken place.
This model study was then extended to the borate ester
derived from tartrate 15, for which a resonance at 54.8 ppm
was observed. 11B NMR showed that this resonance was
replaced by resonances at 8.5 and -23 ppm when a slight excess
of Me2Cu(CN)Li2 was added. The fact that a resonance near 0
ppm did not appear on the spectra suggested that dispropor-
tionation did not occur for the tartrate-derived borates.25 These
observations suggested that the first equivalent of Me2Cu(CN)-
Li2 was consumed as a source of MeLi for the formation of a
boron -ate complex from the borate ester function.
Figure 3.
Table 6. Effect of Amount of Cuprate on Reaction Yield for the
Conversion of 15 to 16
entry
equivalents Me2Cu(CN)Li2
yield (%)
1
2
3
4
5
6
0.5
1.0
1.5
2.0
2.5
3.0
0
0
0
44
72
76
We had to question whether a significant amount of the final
products achieved for 15 had been the result of intramolecular
delivery of an alkyl group from the -ate complex to the bromo
acetal. To elucidate the alkyl delivery, a crossover experiment
was performed using MeLi and a butyl cuprate reagent (Scheme
6). The required bromo ethers were prepared using standard
conditions. The bromo ether mixture was cooled to -30 °C
and 1 equiv of MeLi was added followed immediately by the
addition of 2.0 equiv of Bu2Cu(CN)Li2. After workup, the only
product observed was 60ab, which resulted from the introduc-
tion of a butyl group. Similarly, only methylated adduct 16ab
was isolated when BuLi was added followed by Me2Cu(CN)-
lectivity, were inconsistent with an SN1 mechanism being
operative in the cuprate displacement. An SN1 mechanism would
involve a common oxocarbenium intermediate for the two
acyclic bromo ether acetals, and thus the final ratio of products
would be independent of the intermediate ratio of bromo ethers.
An excess of cuprate also proved to be necessary for
achieving high yield. This was illustrated by an experiment in
which the number of equivalents of cuprate was varied. As
shown in Table 6, significant quantities of product could only
be obtained when at least 2 equiv of cuprate was used. This
observation, together with the previous observation of ratio
magnification only in the presence of borate esters (Table 5,
entries 1 and 2), suggested the possibility that an intermediate
had formed by interaction of the cuprate with the borate ester
function.
(21) BF3‚OEt2 (1.0 M in CD2Cl2) was used as an external reference.
(22) Reported for Et2BOMe (neat): 53.6. Reported for EtB(OMe)2 (neat): 31.5.
(a) No¨th, H.; Vahrenkamp, H. Chem. Ber. 1966, 99, 1049-1067. (b)
Dahlhoff, W. V.; Ko¨ster, R. Liebigs Ann. Chem. 1975, 1625-1636.
(23) Reported for Li[BMe4] (Et2O): -20.2 to -21.1. Reported for Li[BEt4]
(Et2O): -17.5. Reported for K[Me2BOMe] (MeOH): -1.0. No¨th, H.;
Wrackmeyer, B. Nuclear Magnetic Resonance Spectroscopy of Boron
Compounds; Springer-Verlag: New York, 1978.
Further analysis of the cuprate displacement was conducted
to determine the reactive species responsible for the increase
in diastereoselectivity in this step of the process. 11B NMR
investigations shed light on the interaction between the dialkyl-
borates and the higher-order cuprates. In the control experiment,
(24) (a) Brown, H. C.; Cole, T. E. Organometallics 1983, 2, 1316-1319. (b)
Brown, H. C.; Cole, T. E.; Srebnik, M. Organometallics 1985, 4, 1788-
1792.
(25) The resonance at -23 ppm could be attributed to the slight excess of Me2-
BBr used to open the acetal.
9
J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003 433

A R T I C L E S
Guindon et al.
Table 7. Selective Labeling of Bromo Ether Diastereoisomer
Scheme 7
s
entry
TH (°C)
TD (°C)
initial ratio (D:H)
final ratio (D:H)
1
2
3
4
5
6
-78
-78
-78
25
25
25
25
25
25
-78
-78
-78
1:1
3:1
1:5
1:1
3:1
1:5
1.3:1
4:1
1:4.3
1:1.7
1.8:1
1:8.6
product ratio could then be rationalized by the reactivity of the
-ate complexes.
Li2. This experiment showed that the formation of the -ate
complex was irreversible under the reaction conditions and that
an intermolecular alkyl delivery did in fact occur.
As depicted in Table 7, treatment of deuterated cyclic acetal
65 with Me2BBr gave, after equilibration at room temperature
(TD ) 25), an 8:1 ratio favoring the S isomer. This stock solution
was mixed at -78 °C, a temperature where no equilibration
had previously been detected (vide supra), with different
volumes of a stock solution containing nondeuterated -ate
complexes in a 1:1 ratio (obtained after opening of 15 at TH )
-78 °C). The resulting solutions were allowed to react with
cuprates at -30 °C. Increases in the amount of labeled relative
to unlabeled products were noted, suggesting that the nonlabeled
-ate complexes had decomposed (entries 1 and 2, respectively).
Other permutations were realized to confirm the observation.
As seen in entry 3, decomposition of nonlabeled -ate complexes
was also observed when a 1:5 ratio of bromo ethers (D:H) was
used for the cuprate displacement. From these results, and
assuming that the deuterated and nondeuterated isomers would
react in the same way given their similar structures, we suggest
that a decomposition of the nonlabeled R -ate complex was
responsible for the increase in ratio of the label D over the label
H.
Three additional experiments were performed to take into
account any potential isotopic effects. A stock solution of S:R
nonlabeled bromo ethers in an 8:1 ratio (TH ) 25) was mixed
at -78 °C with different volumes of a stock solution (obtained
after the opening of 65 at TD ) -78 °C) containing deuterated
bromo ethers in a 1:1 ratio. The resulting solutions were allowed
to react with cuprates at -30 °C. As seen in entries 4-6, an
increase in ratio of H over D was consistently observed,
suggesting that the deuterated -ate complexes had decomposed.
The arguments given above for the previous series of experi-
ments could also apply here. Once again, the R isomer, this
time deuterated, seemed to react with lower conversion. The
percentage of decomposition was higher for this series, indicat-
ing that an isotopic effect perhaps took place. Overall, the results
obtained support the fact that selective conversion at -30 °C
was less efficient with the R -ate complex than with the S isomer.
Figure 3 could now be analyzed in greater detail. As noted
before, a linear relationship existed between the ratio of bromo
The mechanistic hypothesis for the cuprate displacement is
shown in Scheme 7. Two -ate complexes, 61 and 62, were
generated from bromo ethers 54 and 55, respectively, following
the addition of cuprate at low temperatures. Given that the
cuprate additions performed at -78 °C led to a major product
with an inverse configuration from that of the major bromo
ether, and that the bromo ether ratio and final product ratios
were identical, it could be suggested that an SN2-like process
(an SN1 process involving an intimate ion pair) takes place for
both -ate complexes at -78 °C. However, magnification of the
final product ratio at -30 °C suggested that the two -ate
complexes, 61 and 62, reacted differently with the cuprate at
this temperature, meaning that two different pathways were
involved (61 f products; 62 f products). The fact that both
the increase of the ratio of final products (Figure 3a) and the
increase in yield (Figure 3b) were proportional to the amount
of major bromo ether 55 suggested that the cuprate displacement
of S isomer 62 was highly selective and thus SN2-like. To
explain the increase in diastereoselectivity (i.e. magnification
of final product ratio), we proposed that the displacement of
the R isomer 61 was less efficient in terms of both selectivity
and total conversion to final products. It follows that, at -30
°C, 61 may react through an SN1 process involving an
oxocarbenium intermediate. Further support of this may be
drawn from our previous calculations, which showed that 61
was higher in energy than 62, as a result of steric constraints,
suggesting that the ionization of 61 may have been favored
relative to the ionization of 62.
The validity of these hypotheses seemed apparent. Indeed, if
both diastereoisomers reacted in the same way in a stereose-
lective process, how could the final ratios be higher than the
starting ratios? To verify that the two -ate complexes, 61 and
62, reacted differently, we used solutions containing different
ratios of labeled (D) and nonlabeled (H) bromo ethers for our
study of the cuprate displacement. The analysis of the final D:H
9
434 J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003

Dialkylboron Bromide To Open Tartrate Acetals
A R T I C L E S
Table 8. Mathematical Reconstructions of Figure 3
process with a range in selectivity from 35:65 to 94:6,
surprisingly in favor of 16a, and a maximum yield of 51% (16a
+ 16b).27
Results of a similar experiment and simulation, in which the
cuprate was added at -78 °C, are presented in entries 3 and 4.
At this temperature, both intermediates were displaced selec-
tively, although in low yield, with an inversion of configuration.
Taken together, the data obtained from these calculations offered
support that an SN2-like mechanism was operative for 67 and
68 at -78 °C. This result supported the hypothesis that when
the cuprate displacement took place at -30 °C, the behavior of
the minor (R) isomer was responsible for the increase in
diastereoselectivity.
Conclusions
minimum
selectivity
(a:b)
maximum
selectivity
(a:b)
minimun
yield
(a + b)
maximum
yield
(a + b)
The opening of acetals derived from tartaric esters using Me2-
BBr as the Lewis acid and Me2Cu(CN)Li2 as the alkylating agent
is a valuable method for the synthesis of chiral secondary
alcohols. Good selectivity was realized only when a particular
set of experimental conditions was respected, following the
opening of the acetal, suggesting that the mechanism of this
reaction was different from those published previously for
Johnson-type acetal substitutions. The diastereoselectivity origi-
nated from equilibration of bromo ether derivatives formed
during the opening of the tartrate acetals with Me2BBr. With
L-tartrate as the chiral auxiliary, the most favorable isomer takes
on the S configuration at the bromo ether carbon, a result
supported by calculations and confirmed by X-ray diffraction
analysis. The preference observed for the S isomer was dictated
by the minimization of the ester dipole and by the gauche effect
between the two oxygen atoms, both interactions being respon-
sible for the conformational locking of the tartrate moiety.
Initial introduction of Me2Cu(CN)Li2 resulted in the formation
of -ate complexes through a trans-metalation process that was
apparently faster than the displacement of the bromine. When
cuprate displacement was realized at -78 °C, the S and R bromo
ethers (the -ate complexes) reacted through SN2-like mecha-
nisms. At -30 °C, the two isomers behaved differently with
Me2Cu(CN)Li2. The major S bromo ether reacted in a highly
selective manner and with high yield through an invertive SN2-
like mechanism. By contrast, the minor R isomer, for which
oxocarbenium formation seemed to be favored in conjunction
with an increase in temperature, reacted inefficiently through
an SN1 mechanism, resulting in oVerall increased selectivity
during the cuprate displacement. This conclusion is supported
both by the experimental data that was obtained and by the
mathematical reconstruction that was subsequently performed.
These results illustrate the importance of considering the role
of a potentially competing nucleophile as well as the source
and fate of minor products, aspects often overlooked in
investigations of selectivity and mechanistic elucidations.
bromo
ether
entry
temp
1
2
3
4
-30
-30
-78
-78
67
68
67
68
93:7
98:2
94:6
100: 0
0:100
77
34
25
30
83
51
33
45
35:65
87:13
20:80
ethers 59a:59b and the ratio and yield of the final products.
We had hypothesized that the two -ate complexes reacted
through two separate chemical reactions: (1) the high-yielding
and stereospecific displacement of the S bromo ether 59a into
16a, and (2) the low-yielding and unselective displacement of
the R isomer 59b into both 16a and 16b. We felt that our
hypothesis could be further supported through a mathematical
reconstruction of the experimental findings in Figure 3.
The scheme presented in Table 8 gives an empirical descrip-
tion of all the reactions that were performed using the
intermediates derived from 15. Both bromo ethers 67 and 68,
upon reaction with Me2Cu(CN)Li2, give final products 16a and
16b. A computer program was devised to process all of the
possible values that could be achieved for 67 and 68 in terms
of yield and selectivity. The resulting data was plotted on a
curve using standard linear regression. The program then verified
each curve and returned sets of parameters that reproduced the
experimentally derived results.26 Different sets of results that
were consistent with our experimental observations were
generated by the program, and several trends became apparent.
The bromo ether 67, in simulations at -30 °C, appeared to
be a very good substrate for the cuprate displacement, giving
overall combined yields of 16a and 16b in the range of 77-
83%. Furthermore, this isomer showed good selectivity for 16a
(Table 8, entry 1). The mathematical reconstruction that
presented the closest match to the original experimental results
indicated 67 was converted to 16a by inversion of configuration
(an SN2-like process) with a minimum selectivity of 93:7 (16a:
16b) and a minimum yield of 77% (16a + 16b).
Experimental Section28
General Procedure for the Preparation of Tartrate Acetals. To
a mixture of aldehyde (10.0 mmol) and trimethylorthoformate (10.0
mmol) at 0 °C was added p-toluenesulfonic acid (0.5 mmol) in one
Conversely, intermediate 68 at -30 °C appeared to be a poor
candidate for the cuprate displacement, giving low overall yields
of 16a and 16b and moderate selectivity for 16a (entry 2). The
mathematical reconstructions for 68 are summarized in entry 2
and show that this isomer was displaced in an unselective SN1
(27) This type of enhancement, where selectivity is generated on two or more
levels, has been more rigorously described. Buschmann, H.; Scharf, H.-
D.; Hoffmann, N.; Esser, P. Angew. Chem., Int. Ed. Engl. 1991, 30, 477-
515. See also: Ward, D. E.; Liu, Y.; Rhee, C. K. Synlett 1993, 561-563.
(28) General experimental comments, physical properties, and analytical data
are given in Supporting Information.
(26) We used the largest experiment error to construct a “window” within which
calculated curves were accepted.
9
J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003 435

A R T I C L E S
Guindon et al.
portion. After stirring for 1 h the mixture was diluted with benzene
(60 mL), dimethyl tartrate was added (11.0 mmol), and the mixture
refluxed overnight using a Soxlith packed with molecular sieves to
remove methanol. The solution was then cooled to room temperature
and diluted with ether. Washing with saturated NaHCO3 and brine was
followed by drying over MgSO4 and flash chromatography to afford
the desired acetals.
temperature for 15 min. Et(iPr)2N (4 equiv) or Et3N (4 equiv) was then
added dropwise followed immediately by a solution of PhSH in CH2-
Cl2 (1.0 M). It was imperative that the PhSH solution be injected as
rapidly as possible. After stirring for 1 h at -78 °C, the solution was
poured into a solution of NaHCO3 in water and stirred 15 min.
Extraction with EtOAc was followed by washing with brine and drying
over MgSO4. Flash chromatography afforded the pure products.
General Procedures for the Removal of Tartrate Auxiliary.
Method A. To a solution of substrate (0.5 mmol) and Et3N (3 equiv)
in CH2Cl2 (0.1 M) was added methanesulfonyl chloride (1.5 eq). After
the mixture stirred for 30 min, DBU was added (10 equiv), and the
solution was stirred at 0 °C until TLC indicated that the reaction was
complete (30 min). The solution was then diluted with ether, washed
with water, 10% HCl, saturated NaHCO3, and brine, dried (MgSO4),
and concentrated in vacuo. The residue was then taken up in methanol
(2 mL), NaOMe was added (1 mL of a 25% solution in MeOH), and
the resulting solution refluxed 30 min. Brine was added (20 mL) after
cooling to room temperature, and the solution was extracted with EtOAc
(5 ×). Drying (MgSO4) and concentrating were followed by flash
chromatography to afford the pure alcohols. Method B. To a mixture
of substrate (ca. 0.5 mmol), alcohol (MeOH or ethylene glycol, 10
General Procedure for the Preparation of Mixed Acetals. Method
A. To a cold (-78 °C) solution of acetaldehyde dimethyl acetal (1
equiv) in CH2Cl2 was added a solution of Me2BBr in CH2Cl2 (1.2 equiv,
ca. 1.5 M) and the resulting solution stirred at -78 °C for 10 min.
Et(iPr)2N (2.5 equiv) was then added followed by a solution of the
appropriate alcohol (1 equiv). After 1 h at -78 °C, the solution was
poured into saturated NaHCO3 and extracted with EtOAc. The organic
phase was washed with brine, dried (MgSO4), and concentrated in
vacuo. Flash chromatography gave the desired products. Method B.
A solution of methylvinyl ether was prepared by sparging methylvinyl
ether into a cold (0 °C) solution of triethanolamine (ca. 40 mg) in CH2-
Cl2 (20 mL). Typically 1-3 g was collected. A portion of this solution
(1.3 equiv) was added to a solution of tartrate alcohol in CH2Cl2 (0.1
M) at 0 °C. Camphorsulfonic acid (0.02 equiv) was then added and
stirring continued until TLC indicated that the reaction was complete
(30 min). The mixture was then diluted with ether and washed with
saturated NaHCO3 and brine and dried (MgSO4). Flash chromatography
afforded the desired acetal.
General Procedure for Me2BBr-Promoted Opening of Tartrate
Acetals by Higher-Order Cuprates. A solution of Me2BBr (ca. 1.5
M, 1.25 equiv) in CH2Cl2 was added dropwise to a cold (-78 °C)
solution of acetal in CH2Cl2 (0.1 M). The resulting mixture was stirred
at room temperature for 1 h before being cooled to -30 °C and stirred
at this temperature for 15 min. A solution of Me2Cu(CN)Li2 (3 equiv)
in THF (0.25 M) was then introduced via cannula (5 min addition)
and the resulting solution stirred at -30 °C for 1 h before being
quenched with MeOH (10 equiv). The mixture was then poured into
an aqueous solution of NH4Cl:NH4OH (4:1), diluted with EtOAc, and
stirred at room temperature for 15 min. Brine was added, the phases
were separated, and the aqueous layer was extracted with EtOAc. The
combined organic extracts were washed with brine, dried (MgSO4),
and concentrated in vacuo. The crude material was purified by flash
chromatography. Diastereomeric ratios were determined by capillary
gas chromatography.
equiv), and HMPA (6.0 equiv) was added 0.1 M SmI2 in THF.30
29
When TLC indicated that the reaction was complete (0.5-6 h), the
septum was removed, and the mixture was stirred to destroy excess
SmI2 (blue solution becomes yellow). The mixture was then diluted
with ether and washed with 10% HCl, 10% Na2S2O3, saturated
NaHCO3, and brine, and dried (MgSO4). Flash chromatography afforded
the pure alcohols.
Acknowledgment. We thank Dr. Brigitte Gue´rin and Ms.
LaVonne Dlouhy for their assistance in the preparation of this
document as well as NSERC (Canada) for the industrial
postdoctoral fellowship of B.S.
Supporting Information Available: Experimental procedures
and characterizations for all new compounds, X-ray crystal-
1
lographic structure for 22b as well as H or 13C NMR spectra
for those compounds without elemental analysis, and 1H NMR
spectra for the Mosher ester derived from secondary alcohols
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
General Procedure for Me2BBr-Promoted Opening of Tartrate
Acetals Using Thiophenol. A solution of Me2BBr (ca.1.5 M, 1.25
equiv) in CH2Cl2 was added dropwise to a cold (-78 °C) solution of
acetal in CH2Cl2 (0.1 M). The resulting solution was stirred at ambient
temperature for 1 h before being cooled to -78 °C and stirred at that
JA012530G
(29) Evans, D. A.; Hoveyda, A. H. J. Am. Chem. Soc. 1990, 112, 6447-6449.
(30) All reagents were degassed by sparging with dry nitrogen. Commercially
available (Aldrich) SmI2 gave satisfactory results only if the solution was
very fresh.
9
436 J. AM. CHEM. SOC. VOL. 125, NO. 2, 2003