Dramatic rate enhancement with preservation of stereospecificity in the first metal-catalyzed additions of allylboronates

Article abstract of DOI:10.1021/ja027453j

This communication successfully challenges the perception that the additions of allylbonates to aldehydes cannot be catalyzed effectively by added Lewis acids. Using a novel class of isomerically pure, tetrasubstituted 2-alkoxycarbonyl allylboronates (1),

Full text of DOI:10.1021/ja027453j

Published on Web 09/06/2002
Dramatic Rate Enhancement with Preservation of Stereospecificity in the First
Metal-Catalyzed Additions of Allylboronates
Jason W. J. Kennedy and Dennis G. Hall*
Department of Chemistry, UniVersity of Alberta, Edmonton, AB T6G 2G2, Canada
Received June 25, 2002
Allylboronates constitute a very useful class of synthetic
1
intermediates for accessing functionalized homoallylic alcohols.
2
They belong to the type I class of allylation reagents (Figure 1,
ML
n
2
) B(OR) ), whose additions onto aldehydes are believed to
occur via closed, six-membered cyclic chairlike transition states
characterized by internal activation of the aldehyde by the boron
center.3 This mechanism contrasts with the type II reagents
exemplified by the popular allylsilane and allylstannane analogues
Figure 1. Accepted mechanisms for type I and II allylation reagents.
(M′L
n
3 3
) SiR , SnR ), which react with aldehydes generally under
the activation of an external Lewis acid catalyst and proceed by
way of open transition structures. Because of their compact and
organized transition structure, allylborations with γ-substituted
reagents tend to demonstrate a level of stereoselectivity superior
to that of type II reagents and in a highly predictable fashion.
There are no reports on the use of Lewis acids to catalyze the
4
addition of allylboronates to aldehydes. Because the carbonyl is
activated by the allylboronate itself, there is no apparent reason to
believe that these reactions could be accelerated by added Lewis
acids. Moreover, by potentially inducing a changeover from a type
I mechanism toward open transition structures, the use of Lewis
acids could be detrimental to the reaction’s stereoselectivity. Herein,
we challenge the preVailing perception that allylboronates are not
suited to Lewis acid catalysis. Using our recently described
5
isomerically pure tetrasubstituted allylboronates 1 (eq 1), we now
report that some metals allow these reagents to add onto aldehydes
to yield γ-lactone products 2 at temperatures almost 100 °C lower
than the corresponding uncatalyzed reactions. Moreover, the
stereospecificity observed in the uncatalyzed allylborations is
preserved, suggesting that type I behavior is maintained in this
unprecedented catalytic reaction manifold.
Figure 2. Comparative reaction speed for different Lewis acids. Conversion
of 1a into lactone 2a (%) is plotted against time.
We initiated our investigation by screening a large number of
representative Lewis acids of moderate strength. Most metals were
employed as their triflates for optimal solubility in the apolar
noncoordinating solvents typically used in allylboration reactions.
These experiments were carried out in NMR tubes using equimolar
amounts of pinacol allylboronate 1a and benzaldehyde with 0.5
2 2
equiv of the Lewis acid as a dilute solution in CD Cl (0.07 M
concentration) at 25 °C (eq 2).
Our design approach takes advantage of the 2-alkoxycarbonyl
While a number of Lewis acids induced a modest acceleration
over the background uncatalyzed reaction, two of them, Cu(OTf)
and Sc(OTf) , provided a dramatic rate enhancement (Figure 2).
Several assays involving Lewis acids of limited solubility in CD
group of allylboronates 1 and is based on the premise substantiated
by the work of Brown and co-workers that boron electrophilicity
6
2
3
is determinant to the reactivity of allylboronates. We envisioned
that formation of a metal ion chelate (for example, 3 in Figure 3)
between the carboxyester and one of the boronate’s oxygens could
increase the acidity of boron and, in turn, strengthen its interaction
with the aldehyde at the transition state level. In fact, recent ab
initio MO calculations conclude that the strength of B-O coordina-
2
-
Cl
2
were repeated in 1:1 CD
2
Cl
2
/THF-d
8
to rule out the influence
, which
8
of solubility. The outcome was similar, except for Yb(OTf)
3
was later found to be efficient in THF and toluene. Further
evaluation of the three best metals (Yb(III), Cu(II), Sc(III)) was
carried out in different solvents, and the use of Cu(OTf) or Sc-
2
tion is the main factor lowering the activation energy in the addition
of allylboronates to aldehydes.7
(OTf) in either dichloromethane or toluene were the combinations
3
found to provide the fastest rate of formation of R-exomethylene
*
To whom correspondence should be addressed. E-mail: dennis.hall@ualberta.ca.
γ-lactone 2a. Some of these experimental conditions were validated
1
1586 J. AM. CHEM. SOC. 2002, 124, 11586-11587
9
10.1021/ja027453j CCC: $22.00 © 2002 American Chemical Society

C O MMU N I C A T I O N S
Table 1. Lewis Acid Catalyzed Formation of γ-Lactones 2 (Eq 2)a
boronate
(R ,R )
aldehyde
(R3)
yield
productd (%)e
1
2
L.A.b
solventc
entry
1
2
3
4
5
6
7
8
9
1a (Et, Me)
C6H5
Sc(OTf)3
CH2Cl2
2a
2a
2a
2a
2b
2c
2d
2d
2e
2f
72
93
67
91
55
53
64
66
62
53
54
1a
1a
1a
C6H5
C6H5
C6H5
Sc(OTf)3 toluene
Cu(OTf)2 toluene
Yb(OTf)3 toluene
Sc(OTf)3 toluene
Sc(OTf)3 toluene
Figure 3. Mechanistic studies and postulated transition structure.
1b (Me, Bu) 3-I-C6H4
1b
1a
1a
1b
1b
1a
C6F5
principle, coordination of the metal to one of the alkoxy groups
increases boron’s acidity, which in turn compensates through
increasing its interaction with the aldehyde and concomitantly
lowering the reaction’s activation barrier.6 Further support for
electrophilic boronate activation and for ruling out open transition
structures comes from the corresponding allylsilanes, which are
PhCH2CH2 Sc(OTf)3 toluene
PhCH2CH2 Yb(OTf)3 toluene
Bu
CH2-i-Pr
C6H11
Sc(OTf)3 toluene
Sc(OTf)3 toluene
Cu(OTf)2 toluene f
,7
1
1
0
1
2g
a
Reaction scale: approximately 0.4 mmol (>100 mg) of 1, and 1.0-
unreactive with Sc(OTf)
3
and react slowly even with a strong
to give mixtures of diastereomeric
8
b
1
0
(
.5 equiv of aldehyde, room temperature, 6-24 h. 10 mol %. c Typically
8 d
multivalent Lewis acid like TiCl
4
.5-1.0 M concentration in allylboronate.
The dr was usually over 19:1
1
8,11
determined by H NMR). In some cases, allylboronate of lesser isomeric
lactones. Another potential mode of activation, acting alone or
in concert with the above, may arise from twisting the ester out of
conjugation with the alkene, thereby increasing the nucleophilicity
of the allylboronate’s γ-carbon.
e
purity was employed, but the final lactones were of identical dr. Unop-
timized yields of pure products isolated after flash chromatography.
Performed at 60 °C for 16 h.
f
and further optimized on a practical scale with 10% catalyst loading
at 0.5-1.0 M concentration (Table 1, entries 1-4). Lactone 2a
was isolated in high yield, thereby confirming the capacity of the
Lewis acids to turnover in these reactions. The stereoselectivity
reflected the isomeric purity of allylboronates 1 and was usually
equal or superior to a 19:1 ratio favoring the indicated diastereomer,
In summary, this work on 2-alkoxycarbonyl allylboronates reports
the first examples of Lewis acid-catalyzed additions of allylbor-
onates to aldehydes. The huge rate enhancement over the uncata-
lyzed reaction provides a highly improved practical approach to
access aldol-like adducts with a stereogenic quaternary carbon
center. This novel catalytic reaction manifold opens exciting
possibilities for the development of substoichiometric methods of
absolute stereocontrol. Moreover, further to the remarkable rate
enhancement, the fact that the stereospecificity observed in the
uncatalyzed process is preserved raises intriguing mechanistic
questions. Ongoing investigations on the nature of this unprec-
edented mode of allylboronate activation may unveil similar
opportunities for catalyzing other reactions of organoboronates.
5
the same as in the uncatalyzed process. Unbranched aliphatic
aldehydes, notorious for self-condensing in the presence of Lewis
acids, provided good yields of products (entries 7-9). Gratifyingly,
branched aldehydes that were ineffective without metal activation
are now suitable substrates (for example, cyclohexanecarboxalde-
hyde, entry 11). It is noteworthy that reactions catalyzed by Sc-
3
(OTf) are tolerant of moisture; rates and yields were not appreciably
affected by the addition of up to 1 equiv of water. Most reaction
examples of Table 1 were complete within 12 h at room temper-
ature, a feature viewed as a significant practical improvement over
the previous uncatalyzed process. Indeed, to reach completion
without a catalyst, the allylboration between 1a and benzaldehyde
necessitated 14 days at room temperature or 16-24 h at 110 °C!5
Acknowledgment. This work was funded by the Natural
Sciences and Engineering Research Council (NSERC) of Canada
and the University of Alberta. J.W.J.K. thanks NSERC and the
Alberta Heritage Foundation for Medical Research (AHFMR) for
graduate scholarships. The authors thank Glen Bigam, Gerdy Aarts,
and Lai Kong for the NMR analyses.
Here, comparative kinetic experiments with the same substrates
showed that while Sc(OTf)
room temperature, the background uncatalyzed reaction reaches only
-4% completion. This issue of rate enhancement has important
3
-catalyzed runs are over within 6 h at
Supporting Information Available: Additional NMR screening
(1
11
1
11
H and B) of the best catalysts, spectra and details of H and
B
3
NMR studies of the 1a-Sc(OTf)
3
complex, full experimental details
repercussions toward the eventual use of chiral catalysts.
Preliminary investigations were carried out to gain insight on
the possible mode of activation in these metal-catalyzed allylbo-
for all entries of Table 1 and control compound 4, characterization
data and spectra for all new lactone products 2 (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
4 3
rations. Control reactions with Bu NOTf, and with Sc(OTf) in the
presence of a base (DIPEA), ruled out the possibility that either
References
8
triflate ion or adventitious TfOH are the activating species.
(
1) For reviews, see: (a) Matteson, D. S. Stereodirected Synthesis with
Organoboranes; Springer-Verlag: Berlin, Heidelberg, 1995; Chapter 7.
(b) Roush, W. R. StereoselectiVe Synthesis, Houben-Weyl, 4th ed.;
Thieme: Stuttgart, 1995; Chapter 1.3.3.3.3 in Vol. E21b.
9
Allylboronate 4 (Figure 3) lacking the 2-alkoxycarbonyl group was
reacted with benzaldehyde, and although the rate of product
3
formation was enhanced in the presence of Sc(OTf) , it was found
(
(
2) Denmark, S. E.; Weber, E. J. HelV. Chim. Acta 1983, 66, 1655-1660.
3) Li, Y.; Houk, K. N. J. Am. Chem. Soc. 1989, 111, 1236-1240.
to be small as compared to the case of allylboronates 1.8 These
results clearly highlight the crucial role of the ester group and the
metal ion at providing optimal rate acceleration. NMR studies on
,10
(4) Exceptionally, the related allyltrifluoroborate salts add to aldehydes with
high diastereoselectivity using strong Lewis acid catalysts: Batey, R. A.;
Thadani, A. N.; Smil, D. V.; Lough, A. J. Synthesis 2000, 990-998.
3
mixtures of 1a and Sc(OTf) provided evidence for the formation
of a defined 1:1 complex, and proton chemical shift data lend
(
(
5) Kennedy, J. W. J.; Hall, D. G. J. Am. Chem. Soc. 2002, 124, 898-899.
6) Brown, H. C.; Racherla, U. S.; Pellechia, P. J. J. Org. Chem. 1990, 55,
1868-1874.
support to a chelate of type 3 implicating coordination of scandium
8
(7) Omoto, K.; Fujimoto, H. J. Org. Chem. 1998, 63, 8331-8336.
both to the carboxyester and to the boronate groups. These
(8) See Supporting Information for details.
observations and the fact that the catalyzed reactions preserve the
stereospecificity of the uncatalyzed type I process led us to propose
hybrid transition structure model A (Figure 3). This model involves
a seven-membered metal-activated complex assembled within the
usual chairlike transition structure of type I allylmetal reagents. In
(9) Hoffmann, R. W.; Schlapbach, A. Tetrahedron 1992, 48, 1959-1968.
(
3
10) Relative decrease of half-life reaction times between Sc(OTf) -catalyzed
and background uncatalyzed reaction: 3× for 4, >35× for 1a.
(
11) Zhu, N.; Hall, D. G., unpublished results.
JA027453J
J. AM. CHEM. SOC.
9
VOL. 124, NO. 39, 2002 11587