Mukaiyama Aldol Reactions Catalyzed by Zirconocene Bis(triflate) Complexes: Stereochemistry and Mechanisms for C-C Bond Formation

Article abstract of DOI:10.1021/jo971803g

The aldol condensations of α- and β-(benzyloxy) aldehydes with enol silanes, catalyzed by Cp₂Zr(OTf)₂THF or Cp₂Zr(OTf)₂, in a variety of solvents were studied. The simple diastereoselectivity of these reactions is modest and comparable to that observed using simple aldehydes of similar steric requirements. Studies have revealed that TMSOTf is directly produced on reaction with the zirconocene catalyst with the enol silane in nitroalkane solvent or is formed during catalysis in dichloromethane solution. Although TMSOTf is known to catalyze cross-aldol reactions under these conditions, the rate of this process is not always competitive with that observed using the zirconocene catalysts. In particular, sterically unhindered or aromatic aldehydes react via a mechanism that appears to be mainly Zr-catalyzed, based on both the difference in rate between Zr- and Si-mediated reactions as well as differences in enol silane/silyl triflate reactivity in crossover-type experiments. With sterically hindered aldehydes in dichloromethane or nitromethane, catalysis is mediated by Si. The Zr-catalyzed process occurs via formation of a Zr-aldolate complex from aldehyde and enol silane, with liberation of TMSOTf, followed by rate-limiting O-silylation of the metal aldolate by TMSOTf, as revealed by both model studies and in situ monitoring during catalysis.

Full text of DOI:10.1021/jo971803g

J . Org. Chem. 1998, 63, 1885-1892
1885
Mu k a iya m a Ald ol Rea ction s Ca ta lyzed by Zir con ocen e Bis(tr ifla te)
Com p lexes: Ster eoch em istr y a n d Mech a n ism s for C-C Bon d
F or m a tion
Shuqiong Lin, Georgiy V. Bondar, Christopher J . Levy, and Scott Collins*
Department of Chemistry, University of Waterloo Waterloo, Ontario, Canada N2L 3G1
Received September 29, 1997
The aldol condensations of R- and â-(benzyloxy) aldehydes with enol silanes, catalyzed by Cp2Zr-
(OTf)2‚THF or Cp2Zr(OTf)2, in a variety of solvents were studied. The simple diastereoselectivity
of these reactions is modest and comparable to that observed using simple aldehydes of similar
steric requirements. Studies have revealed that TMSOTf is directly produced on reaction with
the zirconocene catalyst with the enol silane in nitroalkane solvent or is formed during catalysis
in dichloromethane solution. Although TMSOTf is known to catalyze cross-aldol reactions under
these conditions, the rate of this process is not always competitive with that observed using the
zirconocene catalysts. In particular, sterically unhindered or aromatic aldehydes react via a
mechanism that appears to be mainly Zr-catalyzed, based on both the difference in rate between
Zr- and Si-mediated reactions as well as differences in enol silane/silyl triflate reactivity in crossover-
type experiments. With sterically hindered aldehydes in dichloromethane or nitromethane, catalysis
is mediated by Si. The Zr-catalyzed process occurs via formation of a Zr-aldolate complex from
aldehyde and enol silane, with liberation of TMSOTf, followed by rate-limiting O-silylation of the
metal aldolate by TMSOTf, as revealed by both model studies and in situ monitoring during
catalysis.
In tr od u ction
Several years ago, we described that the Mukaiyama
cross-aldol condensation reaction could be catalyzed by
cationic alkoxide complexes of Zr.1 Somewhat earlier,
Bosnich and co-workers reported that titanocene and
zirconocene bis(triflate) complexes could also be employed
as catalysts for this transformation (eq 1).2 In the former
the use of either R- or â-alkoxy aldehydes as substrates.
High levels of chelation-controlled diastereoselectivity are
frequently observed in reactions of enol silanes with such
aldehydes (and related compounds) when the latter are
activated by conventional Lewis acids.5
In this paper, we describe the aldol reactions of a
number of aldehydes with enol silanes, mediated by
zirconocene bis(triflate) complexes, and studies concern-
ing the mechanism of this C-C bond forming reaction.
This work complements that independently reported by
Bosnich and co-workers using titanocene bis(triflate)
complexes for simple aldehyde substrates6 and indicates
that the zirconocene analogues can behave quite differ-
ently in these reactions.
case, the simple diastereoselectivity of this reaction was
modest but could be rationalized by acyclic transition
states involving the enol silane and the aldehyde, acti-
vated by complexation to the metal center.
More recently, we have demonstrated that the Diels-
Alder cycloaddition reaction between oxazolidinone-based
dienophiles and cyclopentadiene can be efficiently cata-
lyzed using metallocene bis(triflate)complexes (eq 2).3 In
the case of chiral catalysts, significant asymmetric induc-
tion was observed3 and could be related to the formation
of complexes in which the dienophile engages the metal
center in a bidentate fashion under the reaction condi-
tions.4
Resu lts a n d Discu ssion
It had occurred to us that this latter feature might also
be exploited in the Mukaiyama aldol reaction through
Ald ol Rea ction s Em p loyin g r- a n d â-Alk oxy Al-
d eh yd es. The â- and R-(benzyloxy) aldehyde substrates
(1) Hong, Y.; Norris, D. J .; Collins, S. J . Org. Chem. 1993, 58, 3591.
(2) Hollis, T. K.; Robinson, N. P.; Bosnich, B. Tetrahedron Lett. 1992,
33, 6423.
(3) J aquith, J . B.; Guan, J .; Wang, S.; Collins, S. Organometallics
1995, 14, 1079 and references therein.
(5) For a review of synthetic applications see: Reetz, M. T. Angew.
Chem., Int. Ed. Engl. 1984, 23, 556. For relevant mechanistic work
see: (a) Reetz, M. T.; Raguse, B.; Seitz, T. Tetrahedron 1993, 49, 8561
and references therein. (b) Keck, G. E.; Castellino, S. J . J . Am. Chem.
Soc. 1986, 108, 3847.
(4) J aquith, J . B.; Levy, C. J .; Bondar, G. V.; Collins, S. Organome-
tallics, in press.
(6) Hollis, T. K.; Bosnich, B. J . Am. Chem. Soc. 1995, 117, 4570 and
references therein.
S0022-3263(97)01803-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/26/1998

1886 J . Org. Chem., Vol. 63, No. 6, 1998
Ta ble 1. Ald ol Rea ction s Med ia ted by Zir con ocen e Bis(tr ifla te) Com p lexesa
Lin et al.
entry
triflate (mol %)
aldehyde (R)
solvent
syn:antib
T (°C)
t (h)
Yc (%)
1
2
3
4
5
6
7
8
9
10
10d
10
10
10d
10
1
1a BnO(CH2)2-
2-NO2Pr
2-NO2Pr
2-NO2Pr
2-NO2Pr
2-NO2Pr
2-NO2Pr
2-NO2Pr
CH2Cl2
1:1.48
1:1.35
1:1.05
1.29:1
1.57:1
2.51:1
2.00:1
2.26:1
2.04:1
2.23:1
2.24:1
2.07:1
-78
-78
-78
-78
-78
-78
-78
-78
-78
-78
-78
-78
8
5
1.5
10
5
5
10
5
90
90
92
85
88
89
92
85
90
97
94
95
1a
1b CH3CH2-
1c BnOCMe2-
1c
1d BnOCH2CMe2-
1d
1d
1d
1e Me3C-
1e
1e
10
1
CH2Cl2
10
1.5
8
10
11
12
10
1
2-NO2Pr
2-NO2Pr
2-NO2Pr
10d
1.5
a
Typical reaction scale: 0.5 mmol of aldehyde, 0.75 mmol of enolsilane, 2 mL of solvent. The aldehyde was added to the catalyst
b
solution followed by the addition of enol silane, and they were both added in one portion. Diastereoselectivity data is based on the
analysis of 1H and 13C NMR spectra in conjunction with GC analyses (see Experimental Section). c Yields were determined by GC with
d
reference to an n-decane as an internal standard. Cp2Zr(OTf)2 was the catalyst.
1a , 1d , and 1c, shown in eq 3, were synthesized using
either literature procedures or as described in the
Experimental Section. They were selected as substrates
“selectively”, whereas with the more encumbered susb-
trates 1c-e the syn isomer predominated. There was a
slight dependency of simple diastereoselectivity on cata-
lyst loading; lower loadings tended to lead to lower
selectivities. Thus, there appears to be very little dif-
ference in the level of diastereoselectivity using aldehydes
capable of chelation to the metal center compared with
their simpler counterparts. In fact, the stereochemical
consequences of these reactions are qualitatively similar
to those observed earlier using cationic alkoxide catalysts
and simple aldehydes.1
The catalytic activity of the bis(triflate) complexes in
reactions involving simple aldehyde substrates (i.e., 1b,e)
is similar to that of cationic alkoxide complexes in CH2-
Cl2 under equivalent conditions. However, in 2-nitro-
propane solution, these reactions are more rapid, al-
though an excess of enol silane was required for complete
consumption of the aldehyde substrates. We had earlier
reported that cationic alkoxide complexes catalyze the
isomerization of O-silylketene acetals to C-silyl esters
(which do not participate in aldol reactions) in CH2Cl2
solution at higher temperatures.1 Analogous experi-
ments using Cp2Zr(OTf)2 and enol silane 2 revealed that
this complex is also an effective catalyst for this process.
Thus, the requirement for the use of an excess of enol
silane in some of these reactions reflects competitive
conversion of this material to the unreactive C-silyl
isomer 4 (eq 4).
because the simple diastereoselectivity in their reactions
with enol silanes could be directly compared to that of
sterically similar, aldehydes 1b and 1e, respectively,
under equivalent conditions. Chelation-controlled addi-
tions to the former substrates should result in enhanced
2,3-syn selectivity that should, in any event, differ
significantly from that observed using simple aldehydes.
The results of a series of aldol reactions of aldehydes
1a -e with enol silane 2, conducted in (CH3)2CHNO2 or
CH2Cl2 solution at -78 °C in the presence of 1-10 mol
% of either Cp2Zr(OTf)2‚THF or Cp2Zr(OTf)2, are sum-
marized in Table 1. The relative stereochemistry of the
aldol products 5a -e was established (after deprotection
of the intermediate â-siloxy esters 3) from the 13C NMR
chemical shifts of the C-2 methyl group as originally
described by Heathcock and co-workers.7 It should be
pointed out that, in many cases, the intermediate â-siloxy
esters 3 also exhibit the same NMR behavior as the aldol
products themselves; the C-2 methyl group of the syn
diastereomer resonates at higher field than in the cor-
responding anti isomer.
To demonstrate that these reactions were under kinetic
control, a number of experiments were carried out using
substrate 1e. The syn/anti ratio of the product 3e did
not change with time during a typical reaction, as
monitored by GC, and 3e, depleted in the syn isomer
(anti/syn ∼1:1.2), did not undergo equilibration under the
conditions of the reaction, i.e., in the presence of enol
silane 2 and Cp2Zr(OTf)2‚THF. Some experiments were
also performed to determine whether the stereoselectivity
could be perturbed by the mode of addition of sub-
In all cases the simple diastereoselectivity of the
reaction is modest. Sterically unhindered aldehydes 1a
and 1b reacted to provide the 2,3-anti diastereomer
(7) Heathcock, C. H. In Asymmetric Synthesis; Morrison, J . D., Ed.;
Academic Press: New York, 1984; Vol. 3, pp 115-118 and references
therein.

Mukaiyama Aldol Reactions
J . Org. Chem., Vol. 63, No. 6, 1998 1887
strates: slow addition of 2 over 1 h to a mixture of 1d
and Cp2Zr(OTf)2‚THF in 2-NO2Pr failed to significantly
affect the anti/syn ratio (1:1.85 vs 1:1.95), whereas
inverse slow addition of 1d did (anti/syn 1:1.35). Neither
of these results is consistent with that expected for
chelation-controlled additions to, e.g., aldehyde 1d .
A more worrisome interpretation of the present results,
especially in view of the results reported by Bosnich
involving analogous reactions using Cp2Ti(OTf)2,6 is that
these reactions are not even mediated by the zirconocene
catalyst, but by TMSOTf formed in situ. Indeed, a control
experiment between aldehyde 1a and 2 in the presence
of TMSOTf (10 mol %) exhibited the same stereochemical
outcome as observed in the presence of Cp2Zr(OTf)2‚THF
(eq 5). Having said that, under these conditions, the
These experiments indicate that catalysis of the aldol
reaction by TMSOTf is a possibility, particularly in
nitroalkane solvents where significant amounts are
present. Mechanisms that have been proposed for aldol
condensations mediated by metal triflate catalysts are
shown in Scheme 1.6,10
In the first of these, the metal activates the aldehyde
to nucleophilic attack from enol silane and intramolecular
transfer of the silyl group to the â-oxygen of the aldolate
complex is rapid (path A). In a variation of this mech-
anism, the silyl group is rapidly transferred to another
substrate (e.g., TfO- ion) and it is reaction of the latter
compound with the aldolate that gives rise to product
formation (path B). Finally, TMSOTf, formed in situ by
path B, may activate the aldehyde via silylation in what
is basically a metal-initiated but Si-catalyzed pathway
(path C), particularly when reaction of the metal aldolate
with TMSOTf, via path B, is slow. Perversely, even the
aldolate complex 6 may serve as source of electrophilic
Si and could serve to activate aldehyde to nucleophilic
attack via silyl transfer.
Path A can be distinguished from either B or C through
an appropriate crossover experiment. In the present
case, the reaction of benzaldehyde with enol silanes 7a
and 7b was investigated.11 The product mixture was
analyzed by GC/MS, and the aldol products shown in eq
7 were formed in the amounts indicated after 1 h at -78
°C in 2-nitropropane. Thus, crossover is clearly exten-
sive, and this is consistent with reaction occurring via
either path B or C.
reaction of 2 with 1a was complete after 1 h using the
latter catalyst, whereas the conversion was only 40-50%
after a similar period of time using TMSOTf at identical
loadings.
In view of these uncertainties concerning the mecha-
nism of this reaction, we elected to conduct a number of
studies along these lines.
Mech a n istic Stu d ies
If TMSOTf is partially responsible for catalysis using
Cp2Zr(OTf)2 and enol silane 2, its in situ formation from
these materials could be a necessary prerequisite. With
this in mind, the reaction of these two compounds was
studied in CD2Cl2 and CD3NO2 solution under stoichio-
metric conditions. In CD2Cl2 solvent, the isomerization
reaction (eq 4) was not observed, and only trace quanti-
ties of trimethylsilyl triflate are present under these
conditions. In CD3NO2 solution, enol silane 2, in the
presence of equimolar Cp2Zr(OTf)2, is converted to C-silyl
ester 4, a process that is complete after 2 h at -29 °C.
Significant quantities (∼10 mol % with respect to Cp2-
Zr(OTf)2 initially present) of TMSOTf (δ 0.52 ppm) are
produced under these conditions, and unlike the situation
reported by Bosnich6 this material is not exclusively
formed as a result of adventitious hydrolysis of the metal
triflate complex. In particular, the expected hydrolysis
coproducts, methyl propionate and hexamethyldisiloxane,
were not formed in detectable quantities.
The results obtained in nitromethane solution suggest
that there is an alternative mechanism for TMSOTf
formation, possibly involving the (reversible) metathetical
reaction shown in eq 6. A broadened singlet at δ 6.5 ppm
was observed during these experiments and grew in
intensity at the same rate as that of TMSOTf; this may
correspond to the Zr(enolate) complex invoked below.8
Consistent with this interpretation, the addition of excess
TMSOTf resulted in the disappearance of this signal after
the isomerization reaction was complete.9
Analogous experiments in CH2Cl2 at -78 °C using
either 10 mol % Cp2Zr(OTf)2 or TMSOTf were quenched
after 4 h at -78 °C. There were dramatic differences in
(10) (a) Kobayashi, S.; Uchiro, H.; Shiina, I.; Mukaiyama, T.
Tetrahedron 1993, 49, 1761. (b) Carreira, E. M.; Singer, R. A.
Tetrahedron Lett. 1994, 35, 4323. (c) Denmark, S. E.; Chen, C. T. Ibid.
1994, 35, 4327.
(11) Qualitatively similar results were observed using 2 and PhCHO.
However, quantitation was not possible due to the formation of syn
and anti diastereomers, which were not adequately separated by GC.
(8) A number of other signals were present that might be assigned
to this intermediate, but their low intensity precluded unambiguous
identification.
(9) It is quite possible that TMSOTf is also a catalyst for this
isomerization process in nitroalkane solvents as the rate of isomer-
ization qualitatively increased as the amount of TMSOTf increased.

1888 J . Org. Chem., Vol. 63, No. 6, 1998
Lin et al.
Sch em e 1
is unlikely that path C is kinetically important. Given
the crossover observed, the present results suggest that
the aldol reaction involving PhCHdO and enol silanes
is Zr-catalyzed and proceeds via path B.
This result, although reasonably compelling, was less
satisfying since it was not possible to accurately measure
rates of the Zr-mediated reaction in this case. To that
t
end, we elected to study reactions involving BuCHdO
and enol silane 2 in the hope that these would prove
1
sufficiently slow for more quantitative work by H NMR
spectroscopy.
This was indeed the case; reaction of enol silane 2 with
1
tBuCHdO could be conveniently monitored by H NMR
in CD2Cl2 solution at -42 °C. As shown in Figure 1,
clean second-order kinetics (first order in both aldehyde
and enol silane) were observed in CD2Cl2 solution for
either the TMSOTf-catalyzed reaction or the “Zr-cata-
lyzed” process.
In the latter case, significant quantities of TMSOTf
were formed during reaction, the amounts of which were
effectively constant during the time period the reaction
was monitored (ca. 2-3 half-lives) and were essentially
equal to the amount of Zr catalyst initially present (i.e.,
ca. 10 mol % with respect to RCHdO). In addition, two
line-broadened peaks were observed for the Cp protons
during reaction, the appearance of which was dependent
on temperature but independent of conversion. This
behavior can be ascribed to formation of a Zr-based
intermediate, the most likely candidate being a Zr-
aldolate (Scheme 1, path B, vide infra). This species is
either formed in situ from substrates and catalyst, and
then no longer participates in the reaction (as in Bos-
nich’s studies involving Ti “catalysts”6), or is capable of
turnover by the mechanism outlined in path B (i.e., by
reaction with TMSOTf) in analogy to results obtained
using PhCHdO.
product distribution. In particular, in the case of TM-
SOTf, although both products 8a and 8b were present
(and in approximately the same amount), very little 8c
and 8d were present, indicating the nBuMe2SiOTf is a
less effective catalyst under these conditions. The ob-
servation of all four products being (more rapidly) formed
in the Zr reaction under the same conditions implies that
there is a pathway for catalysis involving Zr and that it
proceeds via path B. However, this experiment simply
demonstrates that a Zr-catalyzed pathway is more rapid
than an analogous process mediated by nBuMe2SiOTf; it
provides no insight into the importance of Zr-mediated
processes compared with those catalyzed by TMSOTf.
To gauge the relative importance of the TMSOTf vs
Zr-mediated pathways, it is necessary to quantitatively
determine the rates of reaction. Although this was
qualitatively alluded to earlier (eq 5), we felt it important
As shown in eq 9, the calculated values of the rate
constant for catalysis in CD2Cl2 are essentially identical
for the Zr- and Si-mediated reactions; it seems clear, in
this case, that path C dominates. In CD3NO2 solvent at
1
to study this more carefully by H NMR spectroscopy. It
was not possible to study the Zr-mediated reaction in
CD3NO2 solvent as the rates of reaction of PhCHO with
enol silane 2, even in dilute solution, were too fast for
quantitative work (i.e., complete on mixing at -29 °C).
We thus elected to look at reactions in CD2Cl2 solution
at lower temperatures, although admittedly, formation
of TMSOTf is less problematic in this solvent.
As shown in eq 8, the aldol reaction between PhCHO
and enol silane 2, catalyzed by Cp2Zr(OTf)2 (1 mol %),
was extremely rapid even at -80 °C in CD2Cl2; the
reaction was complete in the time required for the first
acquisition following mixing (<5 min). In contrast, the
- 30 °C, reaction was much more rapid (at identical
loadings) and it was not possible to distinguish between
the Si- and Zr-mediated reactions as they were es-
sentially complete on mixing at this temperature.
A
lower estimate for the rate constant for both processes
in CD3NO2 at -30 °C is ca. 0.6 M-1 s-1 based on >95%
conversion within 5 min at [RCHdO]o ) [2]o ) 0.11 M.
The viability of path B (vs path C) as a mechanism
for catalysis can, in part, be gauged by examining the
reaction of putative, metal aldolate product with
TMSOTf. If the former compound does not react with
TMSOTf to provide aldol products, it can be concluded
that path B is not important and that the aldol reaction
is Si-catalyzed (path C). In the present case, the likely
intermediate involved is the metal aldolate 10 shown in
eq 10b.
TMSOTf-mediated process is at least four times slower,
reaching 95% conversion after ∼20 min under the same
conditions. Thus, in dry dichloromethane solution, where
only trace quantities of TMSOTf are initially present, it

Mukaiyama Aldol Reactions
J . Org. Chem., Vol. 63, No. 6, 1998 1889
F igu r e 1. Aldol reactions between tBuCHO ([tBuCHO]o
)
0.161 and 0.170 M) and enol silane 2 ([2]o ) 0.158 and 0.126
M) catalyzed by Cp2Zr(OTf)2 (0.026 M, circles) or TMSOTf
(0.040 M, squares), respectively. The slopes are equal to kobs
) k[Cp2Zr(OTf)2] and kobs([tBuCHO]o - [2]o) with kobs
k[TMSOTf], respectively.
)
Preparation of the metal aldolate 10 was achieved
through reaction of Cp2ZrMe2 with 1 equiv of the anti
aldol derived from pivaldehyde (prepared via the lithium
enolate followed by chromatographic separation of the
anti isomer) in toluene (eq 10a), followed by reaction of
aldolate 9 with 1 equiv of TfOH in diethyl ether at -78
°C (eq 10b). Complexes 9 and 10 were characterized by
1H NMR spectroscopy but were not obtained as analyti-
cally pure compounds.
F igu r e 2. 1H NMR spectra of zirconium aldolate 10 (300 MHz,
CD2Cl2) at (a) 298, (b) 243, (c) 223, and (d) 203 K. The signal
marked with an X is an unknown zirconocene impurity.
at the same temperature (vide supra). Thus, although
turnover of 10 by reaction with TMSOTf is feasible,
catalysis by TMSOTf (path C) dominates under these
conditions.
In CD2Cl2 solution at -40 °C, aldolate 10 did react with
TMSOTf, but at a much slower rate than in CD3NO2. This
again, in conjunction with the kinetic results reported
earlier, strongly implicates that the reaction of tBuCHdO
with enol silane 2 is Si-catalyzed. This may be a
reflection that the turnover limiting step in the Zr-
catalyzed process involves silylation of a quite sterically
hindered, Zr-aldolate species and that this process is
simply not competitive with the analogous silyl-transfer
process. With sterically less hindered aldehydes (e.g.,
PhCHdO or 1a ), this does not appear to be the case.
Interestingly, complex 10 is fluxional in solution on the
NMR time scale. At room temperature, two line-
broadened signals, due to the diastereotopic Cp protons,
are present in toluene-d8 while only a single, line-
broadened peak is seen in CD2Cl2 solution. We tenta-
tively attribute this behavior to reversible coordination
of the CdO group to the metal center in this complex
(eq 10b). At lower temperatures (in toluene-d8 or CD2-
Cl2), only two sharp signals for the Cp protons were
observed, which suggests that one of these isomers is
highly favored at equilibrium (see Figure 2). The ap-
pearance of this spectrum in CD2Cl2 is very similar to
that observed during catalysis and provides evidence for
the intermediate formed during catalysis as being com-
plex 10.
Con clu sion s
The Mukaiyama aldol condensation between enol si-
lanes and aldehydes, mediated by zirconocene bis(triflate)
complexes, can proceed by both Zr- and Si-catalyzed
pathways, in contrast to similar reactions involving
analogous titanocene complexes, which are predomi-
nantly, if not exclusively, Si-catalyzed.6 With aromatic
aldehydes (and possibly, unhindered, aliphatic aldehydes,
eq 5), the Zr-mediated process appears to dominate,
whereas with sterically hindered, aliphatic aldehydes,
both processes are competitive, but their relative impor-
tance is dependent on solvent polarity. In our opinion,
the Zr-mediated reactions are favored, compared to Ti-
catalyzed pathways, by the more open coordination
When complex 10 (∼0.05 M in CD3NO2 solution) was
treated with a stoichiometric amount of TMSOTf at -30
°C, it was slowly transformed to anti-3e and Cp2Zr(OTf)2.
An estimate of the second-order rate constant for this
process was obtained and is 7.5 ( 1.5 × 10-3 M-1 s-1
.
This is roughly 2 orders of magnitude slower than the
lower limit estimated for the effective, second-order rate
t
constant for reaction of BuCHdO with [2] in CD3NO2

1890 J . Org. Chem., Vol. 63, No. 6, 1998
Lin et al.
2-(Ben zyloxy)-2-m eth ylp r op a n a l (1c). This compound
was prepared from 2-hydroxy-2-methylpropanoic acid by the
following sequence: benzylation (2 equiv of NaH, excess BnBr,
DMF, 25 °C) to provide a mixture of the benzyl ether and
dibenzyl ether/ester, reduction (LAH, THF, reflux) and oxida-
tion (PCC, CH2Cl2). The product was obtained as a colorless
oil on distillation (90 °C, 1 mmHg): IR (thin film) 3064, 3031,
2980, 2933, 2869, 2804, 1735, 1456, 1386, 1170, 1058, 740, 697
cm-1; 1H NMR (250 MHz, CDCl3) δ 9.64 (s, 1H), 7.35 (m, 5H),
4.46 (s, 2H), 1.35 (s, 6H); 13C NMR (62.5 MHz, CDCl3) δ 204.6,
138.7, 128.9, 128.2, 128.0, 80.9, 67.0, 21.4. Anal. Calcd for
environment provided by the larger Zr atom in the
aldolate intermediate. The rate of reaction involving a
Zr-aldolate intermediate with TMSOTf should be sensi-
tive to steric effects of the aldehyde substituent, and this
is qualitatively supported by the relative importance of
the two different pathways (i.e., Si- vs Zr-catalyzed) as a
function of substrate structure.
It is clear that the sensitivity of the mechanism of this
reaction to changes in aldehyde structure (i.e., path B or
paths B and C, with either being dominant under
appropriate conditions) as well as experimental condi-
tions such as solvent polarity does not bode well for
significant asymmetric induction using chiral zirconocene
catalysts of this type.12 Also, the level of simple diaste-
reoselectivity in these reactions is disappointingly low,
even for substrates capable of chelation to the metal
center. It is possible that the aldehyde substrates
examined do not readily form chelates with these cata-
lysts under these conditions13 and/or the rate of reaction
of the unchelated form of the aldehyde complex with enol
silane is competitive and/or faster. Future work, involv-
ing triflate-free catalyst systems, will be directed toward
these problems.14
C
11H14O2: C, 74.13; H, 7.92. Found: C, 73.96; H, 8.11.
3-(Ben zyloxy)-2,2-d im eth ylp r op a n a l (1e). This com-
pound was prepared from 2,2-dimethyl-1,3-propanediol as
described in the literature:18 1H NMR (250 MHz, CDCl3) δ
9.55 (s, 1H), 7.30 (m, 5H), 4.49 (s, 2H), 3.44 (s, 2H), 1.08 (s,
6H).
Mu k a iya m a Ald ol Rea ction s. Typ ica l P r oced u r e. A
solution of Cp2Zr(OTf)2‚THF (0.1 mmol) in 2.0 mL of dry
solvent was cooled to -78 °C under nitrogen. The aldehyde
(1.0 mmol) and enol silane (1.5 mmol) were then added via
syringe. The reaction was monitored by either TLC or GC for
consumption of the aldehyde. After the reaction was complete,
the solution was diluted with hexane:ether 9:1 and filtered
through a short pad of silica gel to remove catalyst, washing
with additional hexane:ether 9:1. The filtrate was concen-
trated in vacuo to dryness to provide the crude aldol products
(as the TMS ethers), which could be further purified by flash
chromatrography on silica eluting with hexane:ether 9:1. The
simple diastereoselectivity was determined either by 1H NMR
spectroscopy or GC prior to purification.
To determine stereochemistry, the crude mixture was
directly converted to the free aldols (THF/AcOH/H2O 8:1:1, 25
°C for 3a ,b; TBAF, THF/H2O 9:1 for 3c-e 25 °C), which could
be purified by flash chromatrography on wet silica gel eluting
with hexane/ethyl acetate 8:1 and whose spectroscopic and
analytical data are summarized below:
Exp er im en ta l Section
All solvents and chemicals were reagent grade and purified
as required. Tetrahydrofuran, diethyl ether, hexanes, and
toluene were dried by distillation from sodium-benzophenone
ketyl. Methylene chloride was dried by distillation from CaH2.
Nitromethane and 2-nitropropane were dried by distillation
from CaCl2 and were stored over activated MS 4 Å prior to
use. All synthetic reactions were conducted under an atmo-
sphere of dry nitrogen in dry glassware unless otherwise noted.
Cp2Zr(OTf)2‚THF and Cp2Zr(OTf)2 were prepared via literature
procedures.15 Enol silanes 2, 7a , and 7b were prepared by
silylation (R3SiCl, THF, -78 to 0 °C) of the lithium enolates
derived from the corresponding esters (LDA, THF, -78 °C).16
Routine 1H and 13C NMR spectra were recorded in CDCl3
or C6D6 solution at 200 and 50 MHz or 250 and 62.5 MHz,
respectively. Low-temperature 1H NMR spectra were recorded
in CD2Cl2 or CD3NO2 solution at 200 MHz using a properly
calibrated thermocouple. IR spectra were recorded on an FTIR
spectrometer. Mass spectra were obtained using an instru-
ment at the University of Guelph. Gas chromatography was
performed on a 0.25 mm × 50 m, SE-30 capillary column. GC-
MS analyses were obtained on a 0.32 mm × 25 m HP-5
column. Elemental analyses were determined by M.H.W.
Laboratories of Phoenix, AZ.
Meth yl 5-(ben zyloxy)-3-h yd r oxy-2-m eth ylp en ta n oa te
(2,3-syn - a n d -a n ti-5a ): IR (thin film) 3464, 2948, 2867, 1728,
1455, 1363, 1201, 1097, 741, 699 cm-1
C
.
Anal. Calcd for
14H20O4: C, 66.65; H, 7.99. Found: C, 66.62; H, 7.85.
2,3-syn -5a : 1H NMR (250 MHz, CDCl3) δ 7.32 (m, 5H), 4.52
(s, 2H), 4.07 (m, 1H), 3.69 (s, 3H), 3.67 (m, overlapping, 2H),
3.19 (s, 1H, syn-OH), 2.57 (pseudo quint, superimposed, J ≈
7.0 Hz, 1H), 1.83 (m, 2H), 1.19 (d, J ) 7.0 Hz, 3H); 13C NMR
(62.5 MHz, CDCl3) δ 175.7, 137.9, 128.3, 127.5, 127.4, 72.8,
72.0, 68.4, 51.5, 44.8, 33.7, 11.4 (syn-2-CH3). 2,3-a n ti-5a : 1H
NMR (250 MHz, CDCl3) δ 7.32 (m, 5H), 4.52 (s, 2H), 3.94 (m,
1H), 3.72 (m, overlapping, 2H), 3.70 (s, 3H), 3.28 (s, 1H, anti-
OH), 2.57 (pseudo quint, overlapping, J ≈ 7.1 Hz, 1H), 1.83
(m, 2H), 1.21 (d, J ) 6.6 Hz, 3H); 13C NMR (62.5 MHz, CDCl3)
δ 175.7, 137.9, 128.3, 127.5, 127.4, 73.1, 70.8, 68.2, 51.5, 45.4,
33.7, 13.5 (anti-2-CH3).
Meth yl 3-Hyd r oxy-2-m eth ylp en ta n oa te (2,3-syn - a n d
-a n ti-5b).1 2,3-syn -5b: 1H NMR (250 MHz, CDCl3 with 1%
formic acid) δ 3.86 (m, 1H), 3.71 (s, 3H), 2.58 (pseudo quint, J
≈ 7.0 Hz, 1H), 1.50 (m, 2H), 1.30 (s, 1H), 1.21 (d, J ) 7.3 Hz,
3H), 0.96 (t, J ) 7.3 Hz, 3H). 2,3-a n ti-5b: 1H NMR (250 MHz,
CDCl3 with 1% formic acid) δ 3.71 (s, 3H), 3.64 (m, 1H), 2.58
(pseudo quint, J ≈ 7.0 Hz, 1H), 1.50 (m, 2H), 1.25 (br s, 1H),
1.18 (d, J ) 7.3 Hz, 3H), 0.98 (t, J ) 7.4 Hz, 3H).
3-(Ben zyloxy)p r op a n a l (1a ). This compound was pre-
pared from 1,3-propanediol as described in the literature.17 The
product was obtained a clear oil on Kugelrohr distillation (100
°C, 1 mmHg): 1H NMR (250 MHz, CDCl3) δ 9.78 (t, J ) 2.0
Hz, 1H), 7.30 (m, 5 H), 4.53 (s, 2H), 3.82 (t, J ) 6.8 Hz, 2H),
2.70 (dt, J ) 6.8, 2.0 Hz, 2H).
(12) The use of optically pure (S)-ethylenebis(H4Ind)2Zr(OTf)23 as a
catalyst with simple aldehyde substrates gave rise to aldol products
in <20% ee. Collins, S. Unpublished results.
Meth yl 4-(Ben zyloxy)-3-h yd r oxy-2,4-d im eth ylp en ta n -
oa te (2,3-syn - a n d -a n ti-5c): IR (thin film) 3459, 3064, 2977,
2948, 2880, 1726, 1455, 1370, 1197, 1061, 741, 710 cm-1. Anal.
Calcd for C15H22O4: C, 67.65; H, 8.33. Found: C, 67.84; H,
8.39.
(13) 13C NMR spectra of mixtures of Cp2Zr(OTf)2 and aldehyde 1a
in CD3NO2 at -30 °C revealed the presence of two sets of signals due
to free and bound 1a . However, it was not possible to unambiguously
determine the binding mode (i.e., mono- vs bidentate) from these
spectra.5b
2,3-syn -5c: 1H NMR (250 MHz, CDCl3) δ 7.31 (m, 5H), 4.45
(s, 2H), 3.86 (t, J ) 6.2 Hz, 1H), 3.57 (s, 3H), 2.76 (pseudo
quint, J ≈ 6.9 Hz, 1H), 2.55 (d, J ) 6.2 Hz, 1H, syn-OH), 1.34
(s, 6H), 1.27 (d, J ) 6.8 Hz, 3H); 13C NMR (62.5 MHz, CDCl3)
δ 176.1, 138.9, 128.1, 127.5, 127.2, 77.9, 76.6, 63.7, 51.4, 38.3,
22.5, 21.2, 13.7 (syn-2-CH3). 2,3-a n ti-5c: 1H NMR (250 MHz,
(14) Alternatively, the use of enol silanes that give rise to less
reactive silyl triflate intermediates should circumvent the Si-catalyzed
pathway as shown for nBuMe2SiOTf.
(15) (a) Siedle, A. R.; Newmark, R. A.; Glenson, W. B.; Lamanna,
W. M. Organometallics 1990, 9, 1290. (b) Thewalt, U.; Klein, H. P. Z.
Kristallogr. 1980, 153, 307.
(16) (a) Ainsworth, C.; Chen, F.; Kuo, Y. J . Organomet. Chem. 1972,
46, 59. (b) Ireland, R. E.; Mueller, R. H.; Willard, A. K. J . Am. Chem.
Soc. 1976, 98, 2868.
(17) Kozikowski, A. P.; Stein, P. D. J . Org. Chem. 1984, 49, 2301.
(18) Stummp, M. C.; Schmidt, R. R. Tetrahedron 1986, 42, 5941.

Mukaiyama Aldol Reactions
J . Org. Chem., Vol. 63, No. 6, 1998 1891
CDCl3) δ 7.31 (m, 5H), 4.43 (d, J AB ) 10.8 Hz, 1H), 4.37 (d,
J AB ) 10.8 Hz, 1H), 3.68 (dd, J ) 6.9, 3.4 Hz, 1H), 3.27 (s,
3H), 2.84 (pseudo quint, J ≈ 7 Hz, 1H), 1.58 (d, J ) 3.4 Hz,
1H, anti-OH), 1.31 (s, 6H), 1.27 (d, J ) 6.8 Hz, 3H); 13C NMR
(62.5 MHz, CDCl3) δ 176.7, 139.0, 128.0, 127.3, 127.1, 81.2,
77.1, 64.0, 51.1, 41.2, 21.8, 21.7, 16.8 (anti-2-CH3).
Met h yl 5-(Ben zyloxy)-3-h yd r oxy-2,4,4-t r im et h ylp en -
ta n oa te (2,3-syn - a n d -a n ti-5d ): IR (thin film) 3484, 2926,
2876, 1728, 1434, 1263, 1095, 1024, 804, 699 cm-1. Anal. Calcd
for C16H24O4: C, 68.55; H, 8.63. Found: C, 68.55; H, 8.49.
2,3-syn -5d : 1H NMR (250 MHz, CDCl3) δ 7.32 (m, 5H), 4.51
(s, 2H), 3.87 (d, J ) 5.1 Hz, 1H), 3.67 (s, 3H), 3.51 (d, J AB
)
8.6 Hz, 1H), 3.45 (d, J AB ) 8.6 Hz, 1H), 2.65 (pseudo quint, J
≈ 6.7 Hz, 1H), 1.44 (s, 1H), 1.25 (d, J ) 7.0 Hz, 3H), 0.94 (s,
6H); 13C NMR (62.5 MHz, CDCl3) δ 176.7, 137.7, 128.3, 127.4,
127.2, 79.7, 76.9, 73.5, 51.5, 41.5, 39.1, 22.7, 20.2, 13.3 (syn-
CH3). 2,3-a n ti-5d : 1H NMR (250 MHz, CDCl3) δ 7.32 (m, 5H),
4.48 (s, 2H), 3.89 (d, J ) 4.1 Hz, 1H), 3.62 (s, 3H), 3.34 (d, J AB
) 8.9 Hz, 1H), 3.25 (d, J AB ) 8.9 Hz, 1H), 2.77 (pseudo quint,
J ≈ 7.2 Hz, 1H), 1.60 (s, 1H), 1.33 (d, J ) 7.2 Hz, 3H), 0.90 (s,
6H); 13C NMR (62.5 MHz, CDCl3) δ 177.3, 138.3, 128.1, 127.6,
127.3, 79.6, 76.5, 73.1, 51.4, 39.6, 39.0, 22.1, 20.6, 17.7 (anti-
CH3).
The GC peaks were assigned on the basis of the (M+ - 15)
peak (the parent ion was absent in all cases) as well as
expected fragmentation patterns.
Similar experiments with TMSOTf and Cp2Zr(OTf)2 were
conducted in CH2Cl2 at the same temperature. After 4 h, the
reactions were quenched and analyzed in the same fashion
(GC-MS data is summarized in the Supporting Information).
P r ep a r a tion of Zir con ocen e 10. A solution of anti-5e
(prepared from the lithium enolate of methyl propionate and
pivaldehyde; anti/syn ∼15:1 after chromatography) in toluene
(1.0 mmol in 1.0 mL) was added via syringe to a solution of
Cp2ZrMe2 in toluene (1.0 mmol in 2.0 mL) at 0 °C. The syringe
was rinsed with 3 × 0.5 mL of toluene, which was added to
the mixture. The solution was warmed to room temperature
and was stirred at this temperature until methane evolution
ceased (ca. 10-15 min). The solvent was removed in vacuo to
provide crude 9 as a pale oil that was sufficiently pure (1H
NMR spectroscopy) for further use: 1H NMR (250 MHz, C6D6)
δ 5.87 (s, 5H), 5.79 (s, 5H), 4.04 (d, J ) 3.4 Hz, 1H), 3.41 (s,
3H), 2.54 (qd, J ) 7.0, 3.4 Hz, 1H), 1.11 (d, J ) 7.0 Hz, 3H),
0.82 (s, 9H), 0.33 (s, 3H).
Meth yl 3-Hyd r oxy-2,4,4-tr im eth ylp en ta n oa te (2,3-syn -
a n d -a n ti-5e).1 2,3-syn -5e: 1H NMR (250 MHz, CDCl3) δ 3.69
(s, 3H), 3.68 (d, J ) 3.5 Hz, 1H), 2.76 (qd, J ) 7.0, 3.5 Hz,
1H), 1.40 (br s, 1H), 1.25 (d, J ) 7.0 Hz, 3H), 0.94 (s, 3H).
2,3-a n ti-5e: 1H NMR (250 MHz, CDCl3) δ 3.69 (s, 3H), 3.28
(d, J ) 2.0 Hz, 1H), 2.74 (qd, J ) 7.0, 2.0 Hz, 1H), 2.08 (s,
1H), 1.35 (d, J ) 7.0 Hz, 3H), 0.89 (s, 3H).
Isom er iza tion of MeCHdC(OMe)(OTMS) to MeCH-
(TMS)CO2Me. CD3NO2 was distilled from CaCl2 and then
dried for several days over powdered, activated, 4 Å molecular
sieves. MeCHdC(OMe)(OTMS) (0.05 mmol, 8 mg, 9 µL) was
added to a -29 °C CD3NO2 (0.70 mL) solution of Cp2Zr(OTf)2
1
(0.048 mmol, 25 mg), and H NMR spectra were recorded. The
A solution of 9 in diethyl ether (ca. 1.0 mmol in 2.0 mL)
was cooled to -90 °C (ether, liquid N2). A solution of triflic
acid in diethyl ether (0.50 M) was cooled to the same temper-
ature. Two mL (1.0 mmol) of this solution was added to the
solution of 9 via precooled syringe with vigorous stirring.
Methane evolution was rapid, and the mixture was warmed
to 0 °C and kept at this temperature for 30 min. The solvent
was removed in vacuo at 0 °C to provide compound 10 as a
formation of TMSOTf (δ 0.52 ppm, ca. 10 mol % of Cp2Zr(OTf)2
present) was evident from the first spectrum obtained. Isomer-
ization was evident after 10 min at -29 °C from the appear-
ance of signals at δ 3.6 (s, OMe), 2.2 (q, MeCH(TMS)-), 1.2
(d, MeCH(TMS)-), and 0.12 (s, TMS) at the expense of those
due to 2 [δ 3.5 (overlapping s, OMe), 1.5 (overlapping d,
MeCHd) and 0.24 (s, OTMS)]. This process was ∼50%
complete after 50 min at -29 °C, while the amount of TMSOTf
had increased about 3-fold and thereafter remained constant.
In addition, a broadened peak at δ 6.5, present in the original
spectrum, increased in intensity at the same rate as TMSOTf,
reaching ca. 5-10 mol % of Cp2Zr(OTf)2 after 1 h. After 2 h
of reaction, the isomerization was virtually complete and
excess TMSOTf was deliberately added; this resulted in the
disappearance of the signal at δ 6.5 ppm (see Figure S7 in the
Supporting Information). At 0 °C in CH3NO2 solvent, the
conversion of 2 to 4, in the presence of Cp2Zr(OTf)2 (10 mol
%), was complete after 2 h.
1
white solid, substantially pure by H NMR spectroscopy: 1H
NMR (250 MHz, C6D6, 25 °C) δ 6.15 (br s, 5H), 6.04 (br s, 5H),
4.14 (d, J ) 3.0 Hz, 1H), 3.35 (s, 3H), 2.46 (qd, J ) 7.0, 3.0
1
Hz, 1H), 1.02 (d, J ) 7.0 Hz, 3H), 0.78 (s, 9H). The H NMR
spectra of this material in CD2Cl2 at various temperatures are
depicted in Figure 2.
Rea ction of Com p lex 10 w ith TMSOTf. A solution of
complex 10 (17 mg, 0.032 mmol) in dry CD3NO2 (0.60 mL) was
prepared. A 1H NMR spectrum of this solution was recorded
at -29 °C. To the cold solution was added TMSOTf (5.9 µL,
0.032 mmol) via syringe. After the tube was briefly shaken,
it was returned to the spectrometer probe; the first spectrum,
obtained within 7 min after addition, revealed ca. 60%
consumption of both 10 and TMSOTf, as revealed by integrals
of relevant signals with respect to residual CD2HNO2, and the
appearance of signals due to anti-3e and Cp2Zr(OTf)2. Spectra
obtained after 20 and 35 min at -29 °C were recorded, and
the results are summarized below:
A similar study was conducted in dry CD2Cl2 in the presence
of equimolar bis(triflate) catalyst. In this case, isomerization
was not observed after 1-2 h at -30 °C, and only traces [<1
mol % based on Cp2Zr(OTf)2] of TMSOTf were present.
Rea ction of Me2CdC(OMe)(OSiMe3) (7a ) a n d Me2CdC-
(OE t )(OSiMe2Bu ) (7b ) w it h P h CH O. Me2CdC(OMe)-
(OSiMe3) (0.5 mmol, 87 mg, 100 µL) and Me2CdC(OEt)(OSiMe2-
Bu) (0.5 mmol, 115 mg, 131 µL) were added to a flask
containing Cp2Zr(OTf)2 (0.05 mmol, 30 mg) in 2.0 mL of
2-nitropropane cooled to -78 °C. Benzaldehyde (1.2 mmol,
128 mg, 122 µL) was added in one portion to the cold flask,
and the solution was stirred for 1 h. Aqueous buffer (pH 7,
100 µL) was added rapidly via syringe, the suspension was
diluted with dry hexane (10 mL), and the suspension was
allowed to warm to room temperature. The suspension was
filtered through a Florisil plug, and the solvent was then
removed in vacuo. A dilute solution was made up in ether,
and GC-MS analysis revealed the presence of four products
in the following relative amounts:
t (s)
conversion (C)
C/1-C
420
1200
2100
0.586
0.638
0.675
1.415
1.762
2.078
A plot of C/1-C vs t was linear with a slope of 3.94 × 10-4
s-1 ) k[10]o and thus k ) 7.5 × 10-3 M-1 s-1
.
Rea ction of P h CHO a n d MeCHdC(OMe)(OTMS) Ca ta -
lyzed by Cp 2Zr (OTf)2 or TMSOTf. Solutions of PhCHO
(0.40 mmol, 41 mg, 39 µL), MeCdC(OMe)(OTMS) (0.40 mmol,
64 mg, 74 µL), Cp2Zr(OTf)2 (0.020 mmol, 10 mg), and TMSOTf

1892 J . Org. Chem., Vol. 63, No. 6, 1998
Lin et al.
Ta ble 2. Rea ction of En ol Sila n e 2 w ith tBu CHdO
(0.020 mmol, 5 mg, 4µL) were prepared in CD2Cl2 in 1.00 mL
volumetric flasks. Into each of two NMR tubes was placed
0.70 mL CD2Cl2 along with 50 µL of the PhCHO solution and
10 µL of catalyst solution-TMSOTf in one tube and Cp2Zr-
(OTf)2 in the other. After each tube was cooled to -80 °C in
a dry ice-filled dewar, the MeCdC(OMe)(OTMS) solution (50
µL) was added to the tube and it was immediately placed in
the -80 °C probe of the NMR spectrometer. 1H NMR spectra
were gathered immediately and at various intervals during
the reactions. With Cp2Zr(OTf)2 as catalyst the reaction had
gone to completion (>95% conversion) before the first spectrum
could be acquired (ca. 3 min).
Ca ta lyzed by TMSOTf a n d Cp 2Zr (OTf)2
[aldehyde] [enol silane] [TMSOTf] [Cp2Zr(OTf)2]
TMSOTf
t (s)
391
991
1591
2191
2791
0.149
0.113
0.103
0.094
0.088
0.101
0.056
0.045
0.031
0.026
0.041
0.040
0.040
0.040
0.039
n/a
n/a
n/a
n/a
n/a
Cp2Zr(OTf)2
0.158
0.093
0.081
0.063
0.050
0.047
0.041
0
736
0.160
0.091
0.083
0.061
0.052
0.048
0.038
n/a
0.030
0.029
0.028
0.028
0.030
0.030
0.029
0.026
0.025
0.025
0.025
0.025
0.026
In the case of TMSOTf, the reaction was slow enough to be
followed by NMR at -80 °C. The first spectrum showed the
conversion to be about 50%, but the reaction was not complete
until ∼20 min after addition. A small aldehyde resonance (δ
9.99 ppm) could be observed but its relative intensity did not
change with time (or increased temperature), indicating that
it was due to a slight excess of PhCHO in the reaction mixture.
The effective rate constants kobs were estimated, assuming
second-order kinetics (first order in aldehyde and enol silane),
using the time taken to reach g 95% conversion and are
summarized in eq 8.
1336
1936
2536
3136
3736
intervals during the reactions. Both reactions were essentially
complete in the time the first spectrum could be acquired.
A similar procedure was followed in CD2Cl2 solution except
the spectrometer and solutions were kept at -42 °C. The data
obtained are plotted in Figure 1, while concentration data as
a function of time are summarized in Table 2.
t
Reaction of Bu CHO an d MeCHdC(OMe)(OTMS) Cata-
lyzed by Cp 2Zr (OTf)2 or TMSOTf. Solutions of PhCHO
(0.20 mmol, 22 µL in 0.25 mL CD3NO2, 0.80 M) and mesitylene
(28 µL, 0.20 mmol, added as an internal standard in 0.125 mL
CD3NO2, 1.60 M), MeCdC(OMe)(OTMS) (0.20 mmol, 37 µL
in 0.25 mL CD3NO2, 0.80 M), Cp2Zr(OTf)2 (0.020 mmol, 12.2
mg in 0.125 mL CD3NO2, 0.16 M), and TMSOTf (0.040 mmol,
7.24 µL in 0.25 mL CD3NO2, 0.16 M) were prepared. Into each
of two NMR tubes was placed 0.38 mL of solvent along with
80 µL each of the PhCHO and enol silane solution and 40 µL
of the mesitylene solution to give a final concentration of
aldehyde (or enol silane) of ca. 0.1 M. After recording a t ) 0
spectrum at -30 °C, each tube was separately cooled to -30
°C in a dry ice/CH3NO2 filled Dewar, 4 µL of catalyst solution
(i.e., 1 mol %) was added to the tube, and the tube briefly
shaken and immediately placed in the -30 °C probe of the
NMR spectrometer. 1H NMR spectra were gathered im-
mediately (i.e., within ∼2-4 min of mixing) and at various
Ack n ow led gm en t. The authors would like to thank
the Natural Sciences and Engineering Research Council
(NSERC) for financial support of this work. G.V.B.
thanks NSERC for a Postgraduate Scholarship.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
various catalytic reactions, spectra of reaction of enol silane 2
with Cp2Zr(OTf)2, spectra of aldolate 10 and of its reaction with
TMSOTf and GC-MS data for reactions of enol silanes 7a and
7b with PhCHdO in the presence of TMSOTf or Cp2Zr(OTf)2
(16 pages). This material is contained in libraries on micro-
fiche, immediately follows this article in the microfilm version
of the journal, and can be ordered from the ACS; see any
current masthead page for ordering information.
J O971803G