Observation of stereotopic group recognition in chiral borate complexes in solution

Article abstract of DOI:10.1002/(sici)1099-0690(199908)1999:8<1775::aid-ejoc1775>3.0.co;2-1

The borate ester of 1a and 2e exhibits complete quaternization of the boron center. An equilibrium is observed between the two diastereomeric complexes 3 and 4 formed by coordination to the boron center of either amino group of 2e. At -78°C, this equilibrium is slow on the NMR timescale and a diastereomeric ratio of about 3:1 is observed. At room temperature, a rapid equilibration is observed resulting in two NMR resonances (both ¹H and ¹³C) for the dimethylamino groups of 2e in 3/4. Comparison with the NMR characteristics of the analogous borate complexes of 1h and 1j and of 2c and 2d shows that these resonances each correspond to both methyl groups of a single dimethylamino group. The chemical shift difference occurs solely from the difference between the time fractions that the dimethylamino groups are coordinated to the boron center, and forms, therefore, a direct measure of the ratio 3/4. This ratio amounts to 63:37 at room temperature, which was confirmed by a protonation titration experiment. The elucidation of these NMR characteristics of B(1a)(2e) allowed the more rapid evaluation of the enantiotopic group recognition in many other borate complexes, which range from 57:43 to 67:33 at room temperature. The recognition of the (after complexation) diastereotopic dimethylamino groups of 2e in the borate complex B(1a)(2e) was utilized in a consecutive methylation reaction, which yielded enantiomerically enriched ammonium salt 7. The enantiomeric excess proved to be identical to the diastereomeric excess observed in the preceding recognition event. Strong evidence is presented that complex 3 is indeed the major diastereomer present in a solution of 3/4, as expected from molecular models.

Full text of DOI:10.1002/(sici)1099-0690(199908)1999:8<1775::aid-ejoc1775>3.0.co;2-1

FULL PAPER
Observation of Stereotopic Group Recognition in Chiral Borate Complexes
in Solution
[a]
[b]
Jurriaan Huskens* and Manfred T. Reetz*
Keywords: Borate esters / Molecular recognition / NMR spectroscopy / Enantioselectivity
The borate ester of 1a and 2e exhibits complete
quaternization of the boron center. An equilibrium is
observed between the two diastereomeric complexes 3 and
therefore, a direct measure of the ratio 3/4. This ratio
amounts to 63:37 at room temperature, which was confirmed
by a protonation titration experiment. The elucidation of
these NMR characteristics of B(1a)(2e) allowed the more
rapid evaluation of the enantiotopic group recognition in
many other borate complexes, which range from 57:43 to
4
formed by coordination to the boron center of either amino
group of 2e. At –78°C, this equilibrium is slow on the NMR
timescale and a diastereomeric ratio of about 3:1 is observed.
At room temperature, a rapid equilibration is observed 67:33 at room temperature. The recognition of the (after
1
13
resulting in two NMR resonances (both H and C) for the
complexation) diastereotopic dimethylamino groups of 2e in
the borate complex B(1a)(2e) was utilized in a consecutive
dimethylamino groups of 2e in 3/4. Comparison with the
NMR characteristics of the analogous borate complexes of 1h methylation reaction, which yielded enantiomerically
and 1j and of 2c and 2d shows that these resonances each
correspond to both methyl groups of a single dimethylamino
group. The chemical shift difference occurs solely from the
difference between the time fractions that the dimethylamino
groups are coordinated to the boron center, and forms,
enriched ammonium salt 7. The enantiomeric excess proved
to be identical to the diastereomeric excess observed in the
preceding recognition event. Strong evidence is presented
that complex 3 is indeed the major diastereomer present in
a solution of 3/4, as expected from molecular models.
Introduction
timescale was observed, a ratio of 7:1 between the two dia-
stereomeric complexes was found using the cis-1,2-cyclo-
Selective complexation of one of two enantiotopic groups hexanediammonium guest, but the question which of the
present in a prochiral molecule by a chiral agent may be two enantio-conformers bound preferentially to the host re-
conceptually the simplest,[ though both synthetically and mained unanswered. At room temperature, one set of six
analytically the hardest problem in the field of molecular averaged signals was observed owing to the fast exchange
recognition. Once the chiral compound becomes associated between the two diastereomeric complexes and the uncom-
with the prochiral molecule, the initially enantiotopic plexed guest. From the chemical shift differences between
groups of the latter become diastereotopic. It has been pos- the resonances, a ratio of 5:1 was deduced for this system
tulated that such a stereoselective recognition phenomenon at room temperature. An analogous picture was observed
occurs in the desymmetrization of prochiral or meso-com- for the cis-1,2-cyclopentane diammonium guest molecule.
1]
[2]
pounds using enzymes, chiral chemical reagents or cata- This system constitutes the first example ever for the direct
[
3]
lysts. Although examples of such reactions are amply observation of enantiotopic group recognition. However,
found, and their importance for the preparation of asym- visualization of this phenomenon by X-ray crystallography
metric compounds is generally accepted, no spectroscopic was not achieved. Furthermore, due to the low stability
evidence for the preceding enantiotopic group recognition constant of the complex, it was not possible to utilize the
events was ever presented.
The first direct evidence for enantiotopic group recog- meso-compound into a chiral product. It was therefore not
host/guest complex in a synthetic reaction to convert the
nition in solution was described by our group.[
study, one of two enantiotopic groups of a meso-diam- lective transformation.
In a short communication,[ we recently reported the di-
monium guest molecule was found to be complexed selec-
1][4]
In this possible to correlate the recognition event with a stereose-
5]
tively by a chiral crown ether host, as shown by NMR spec- rect observation of enantiotopic group recognition induced
troscopy. At Ϫ78 °C for which slow exchange on the NMR by a chiral borate complex. This appeared to be a unique
system, as one of the diastereomeric complexes crystallized
[
a]
out selectively in a quantitative fashion. Owing to the
strong guest-complexing nature of the compounds, the rec-
ognition in solution was employed to produce an enantiom-
erically enriched product. This system constituted the first
example for an enantioselective desymmetrization of a pro-
chiral substrate shown to proceed via an intermediate en-
antiotopic group recognition event.
University of Twente, Supramolecular Chemistry and Techno-
logy
P. O. Box 217, 7500 AE Enschede, The Netherlands
Fax: (internat.) ϩ31-53/ 4894645
E-mail: j.huskens@ct.utwente.nl
[
b]
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1, D-45470 Mülheim/Ruhr, Germany
Fax: (internat.) ϩ49-208/ 3062985
E-mail: reetz@mpi-muelheim.mpg.de
Eur. J. Org. Chem. 1999, 1775Ϫ1786

WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ193X/99/0808Ϫ1775 $ 17.50ϩ.50/0
1775

J. Huskens, M. T. Reetz
FULL PAPER
As mentioned above, not only finding systems exhibiting
enantiotopic (or diastereotopic) group recognition is diffi-
cult, also analyzing and quantifying the degree of recog-
nition remains a challenge. In the present paper we report
the unique (NMR) spectroscopic properties of the borate
complexes mentioned above which allowed us to quantify
the ratio of the diastereomeric complexes in solution, even
under fast exchange conditions. Attention is paid to the dy-
namic behavior of the complexes and the comparison with
other model compounds revealing the boundary conditions
for proper quantitative analysis. Furthermore, evidence is
presented for the exact configuration of the major isomer
in solution, confirming that this is the one selectively and
quantitatively crystallizing out under equilibrium con-
ditions.
Results and Discussion
The borate esters discussed in this study were prepared
Scheme 1. Preparation of borate esters of diols and amino alcohols
by mixing equimolar amounts of B(OiPr) , a chiral diol (to
3
induce the stereotopic selection), and a (di)amino alcohol
(Scheme 1; diols used are depicted in Scheme 2). Coevapor-
ating the evolving 2-propanol with toluene was necessary to
drive the reaction to completion. It was envisaged that an
equilibrium as shown in Scheme 1 would occur between
tri- and tetracoordinated boron. An easy way to study this
1
1
phenomenon appeared to be B-NMR spectroscopy since
the chemical shift of borate complexes is almost solely de-
pendent on the atoms directly coordinated to the boron
center and to the coordination geometry of the boron atom,
[
6]
especially the size of the chelate ring formed by the diol,
but not on residual groups further away. Typically, the 11B
shift (in CDCl ) changed from δ ϭ 23 to 11 (while the res-
3
onances commonly had a width of a few ppm) upon com-
plete coordination of an amine when a 1,2-diol was used,
whereas for 1,3-diols these limiting shifts were δ ϭ 18 and
9
1
, respectively. In general, borate complexes esterified with
,2-diols gave complete formation of tetracoordinated bo-
ron when an additional amino alcohol was complexed. In
contrast, borate-1,3-diol complexes gave only partial contri- Scheme 2. Diols employed in this study
butions of amino group coordination. The degree of partici-
pation depends in these cases on the basicity of the amino quaternization of the boron center, the NH proton reso-
2
1
1
group. For example, an B-NMR shift of δ ϭ 15.3 was nance occurred at δ ϭ 7.5, while for uncomplexed amino-
observed for B(1j)(2e), leading to the conclusion that about ethanol (2a) δ ϭ 1.5 was observed. Similarly, the NH2 1H-
7
0% of the complex is present in the tricoordinated state chemical shift for B(1e)(2b) (i.e. using a diamine) was δ ϭ
and 30% in the tetracoordinated form (divided between 4.5, showing that half of the NH groups are coordinated
2
1
1
both amino groups), while the B-NMR shift of B(1j)(2b) and the other half not, simply because only one group can
δ ϭ 17.2) indicates less than 10% participation of the latter be bound in the borate complex. It allows, however, no con-
(
coordination geometry. Similarly, B(1k)(2e) showed mainly clusion about possible stereotopic group recognition since
tricoordinated boron. Probably, the five-membered chelate the protons of the two NH groups may interchange be-
2
ring for boron-coordinated 1,2-diols provides a ring strain tween the NH groups.
2
energy release upon formation of tetracoordinated boron,
In both complexes discussed here, two resonances were
thus providing a driving force for amine coordination.
observed for the CHOϪB carbon atoms of the tartrate unit.
Apart from 11B NMR, also H NMR of the NH protons The de- and recoordination of amines, however, leading to
1
2
of primary amino alcohols was indicative for coordination the equilibrium shown in Scheme 1, is fast on the NMR
of NH2 groups to the boron center. For example for timescale. Therefore, it was concluded that the dynamics of
1
1
B(1e)(2a), for which B NMR (δ ϭ 11.3) showed complete these complexes is as depicted in Figure 1: the de- and reco-
1776
Eur. J. Org. Chem. 1999, 1775Ϫ1786

Observation of Stereotopic Group Recognition in Chiral Borate Complexes in Solution
FULL PAPER
ordination of a single amino group is fast on the NMR The 11B-NMR spectrum displays a signal at δ ϭ 10.6, in
timescale, but the rotation of the whole amino alcohol moi- the range expected for tetracoordinated boron, and leads to
ety in the tricoordinated state is slow most likely due to the the conclusion that in each complex one of the amino
low population degree of this coordination mode. For the groups is coordinated to boron while the other one remains
1
13
monoamine complex B(1e)(2a), this implies that the fast de- noncoordinated (cf. 3 and 4). The H- and C-NMR spec-
and recoordination of the amino group takes place at the tra are fully in accordance with this conclusion.
same side of the chelate ring formed by the chiral diol,
whereas the rearrangement of the amine to the other side
can occur only after the (slow) rotation around the
BϪOϪC bond in the tricoordinated state. As a result, the
two CHOϪB carbons of the diol become diastereotopic.
Analogously for B(1e)(2b), decoordination of an amino
group and its recoordination or the coordination of the
other amino group at the other side of the chelate ring (up-
per part of Figure 1) is rapid. This results, however, in the
fact that the noncoordinated aminomethyl group is con-
tinuously at the same side of the chelate ring rendering the
two CHOϪB carbons of the diol diastereotopic.
Scheme 3. Equilibration of 3 and 4 via the tricoordinated state
X-Ray structural analysis[5] showed the presence of a sin-
gle diastereomer in the solid state, namely 3 in which the
pro-(S) amino group is coordinated to the boron center. It
appeared that this crystallization occurred in quantitative
yield. A powder diffraction experiment[ of the whole batch
5]
of crystals showed that 3 was the only isomer crystallizing
out from solution. This leads to the conclusion that the
crystallization of 3 occurs with concomitant re-equili-
Figure 1. Dynamics of de- and recoordination equilibria relative to
the NMR timescale (postulated)
For B(1e)(2e) and B(1a)(2e), however, only one CHOϪB bration of 3 and 4 in the remaining solution. A process in
carbon signal is observed at room temperature. This cannot which a single compound crystallizes out quantitatively
be attributed to a higher population degree of the tricoordi- from an equilibrating mixture, such as the here observed
nated state (i.e. a lower stability of the BϪN coordination crystallization of 3 from 3/4, is called an “asymmetric trans-
bond), since the methylated amino alcohols are more nucle- formation of the second kind”.[ In cases described so far
ophilic, leading to higher degrees of tetracoordinated boron equilibration is brought about by heating or by a catalyst,
7]
1
1
complexes as confirmed by the B-NMR shifts of B(1j)(2b) usually an acid or base. In the example described here, how-
δ ϭ 17.2) and B(1j)(2e) (δ ϭ 15.3). This dynamics suggests, ever, equilibration in solution occurs spontaneously. As a
therefore, an overall more rapid de- and recoordination and consequence, equilibration of 3/4 is resumed as soon as the
rotation kinetics for the methylated derivatives. solid is redissolved, in contrast to the systems described in
(
Due to the rapid exchange processes depicted in Figure the literature for which no re-equilibration occurs after re-
, it was hard to quantify the stereotopic group recognition dissolution.
1
occurring for the primary diamino alcohols. We switched to
At Ϫ78 °C, the exchange between the two tetracoordi-
tertiary amines such as 2e in order to eliminate the possible nated, diastereomeric complexes is slow on the NMR time-
interchanging of the amine protons. The borate ester scale, the ratio of diastereomers being roughly 3:1 (Figure
B(1a)(2e) proved to be a fruitful model system, largely be- 2, top; integration averaged 2.85:1). It was not possible to
cause the methyl resonances of the dimethylamino groups make unambiguous configurational assignments from these
became additional NMR probes. As discussed above, this spectra alone. We assume, however, on the basis of molecu-
complex was expected to provide the tricoordinated inter- lar models that the major diastereomer is 3, formed by the
mediate, which should spontaneously form the two possible coordination of the pro-(S) amino group. These models
(
diastereomeric) complexes 3 and 4 in a thermodynamically show that for 3 the noncoordinated dimethylaminomethyl
controlled process involving the complexation of the pro- group occupies one of the two quadrants, formed by the
S) and pro-(R) amino groups, respectively (Scheme 3). In- planes through the spiro-configured five-membered rings,
deed, compounds 3 and 4 were obtained as the sole prod- not occupied by a phenyl group of the chiral diol (Figure 3).
(
1
1
1
13
ucts as shown by B-, H-, and C-NMR spectroscopy. This would mean that the isomer crystallizing out in the so-
Eur. J. Org. Chem. 1999, 1775Ϫ1786
1777

J. Huskens, M. T. Reetz
FULL PAPER
called asymmetric transformation of the second kind is also quaternization of the boron center (as does 1a) and the oc-
the major one occurring in solution. This contrasts with currence of two signals for the former must therefore be due
most studies in this field in which it is usually observed that to the position of the N-methyl groups relative to the 2-
the minor isomer present in solution is the one crystallizing methyl group.
out preferentially (the so-called VanЈt HoffϪDimroth
[
7]
rule). More evidence supporting the assumption for the
exact configuration of 3 is presented below.
1
Figure 4. H-NMR shifts (200 MHz) of the dimethylamino groups
(singlets) of B(1a)(2e) relative to uncomplexed 2e
Figure 2. 13C NMR spectra of the propane resonances of B(1a)(2e)
(
3/4) at Ϫ78 °C (top) and of B(1a)(2e); MeOTf (5/6) at room tempe-
rature (bottom)
Figure 3. Schematical top (left) and side (right) views of 3, with
the planes through the two spiro-configured five-membered rings
indicated by dashes
At room temperature, a rapid exchange process occurs.
Two equally intense singlets for the dimethylamino groups
Table 1. Number of 13C resonances of the dimethylamino groups
1
13
were observed by both H- (Figure 4) and C-NMR spec-
troscopy. From these spectra alone it could not be con-
cluded if such a singlet corresponds to (two) methyl groups
from different or identical dimethylamino groups. There-
fore, several model systems were prepared and the number
of several borate model compounds, B(1)(2)
[
a]
3 2
No splitting of N(CH ) resonance(s) observed due to the occur-
rence of diastereomers (amino alcohol 2d was applied as the race-
mate). Ϫ [ Relatively broad (see text).
b]
With 2-methyl-2,4-pentanediol (1j) as the (chiral) diol, a
1
3
of C signals for the dimethylamino groups was deter- single resonance for the dimethylamino group was observed
mined (Table 1). For B(1a)(2d), two signals were observed. in all cases. This can be explained by the fact that only
This shows that, in this case, N-inversion is slow on the incomplete quaternization of the boron center occurs (as
NMR timescale due to (complete) coordination of the di- found commonly for 1,3-diols, see above) so that rapid N-
methylamino group to the boron center. The occurrence of inversion is possible whilst the amine is noncoordinated.
two signals might be due to the chirality of the hydro- For B(1h)(2e), only one resonance is observed also. In this
benzoin portion or to the stereocenter present in the amino case, no discrimination between the dimethylamino groups
alcohol. The latter appears to be true since B(1a)(2c) shows can occur since the diol is achiral, and therefore each group
only one (slightly broadened) signal. This proves that the is coordinated for 50% of the time. When a group is not
chirality in the latter system has little influence, and the two coordinated, rapid N-inversion occurs.
methyl resonances, although in principle diastereotopic, re-
Returning to B(1a)(2e), it is concluded that no perfect
main isochronous. This is confirmed by B(1h)(2d) and recognition (i.e. 100% coordination) of one of the dimeth-
B(1h)(2c): the catechol portion is achiral but gives complete ylamino groups takes place since both resonances are
1778
Eur. J. Org. Chem. 1999, 1775Ϫ1786

Observation of Stereotopic Group Recognition in Chiral Borate Complexes in Solution
FULL PAPER
shifted relative to the free amino alcohol (Figure 4). There- nated on average for a greater time. This was confirmed by
fore, both groups are noncoordinated for substantial frac- the titration experiment where this resonance was less affec-
tions of the time, during which rapid N-inversion can take ted by protonation. The ratio of the slopes of the lines
place, while the sum of the time fractions that the groups shown in Figure 5 equals the ratio between the dia-
are coordinated equals unity. Thus it is concluded that each stereomers, more correctly the ratio between the protonated
of the two signals observed for this system can be ascribed diastereomers. It was assumed that the protonation con-
to the two methyl groups of a single dimethylamino group. stants and the bound shifts were equal because the pro-
However, it was reasoned above that the chirality of 1a does tonation sites are far away from the chiral center. The ratio
not cause a substantial chemical shift difference. Also, the of the slopes gave 62:38, which is equal to the ratio given
emergence of a stereocenter on the CHOϪB carbon of the above. A protonation experiment using p-toluenesulfonic
2
e moiety cannot explain the observed shift difference be- acid as the proton source gave an identical ratio.
cause the methyl groups on a single dimethylamino group
are chemically equivalent, leading to a single methyl reso-
nance B(1j)(2e). Therefore, the observed chemical shift dif-
ference is attributed solely to the difference between the
time fractions that the two dimethylamino groups are coor-
dinated to the boron center. Thus the chemical-shift differ-
ence is a direct measure for the stereotopic group recog-
nition!
1
An analysis of the H-NMR shifts corroborates this pic-
1
ture. The sum of the changes of the H-chemical shifts (δ ϭ
0
.30 and 0.18, respectively; see Figure 4) of the dimethylam-
ino groups in the diastereomeric mixture of B(1a)(2e) (3
and 4) relative to noncomplexed 2e turned out to be equal
to the analogous chemical shift change (δ ϭ 0.48) of the
singlet of B(1a)(2c) relative to 2c. This supports the reason-
1
3
Figure 5. C-NMR shift changes of the dimethylamino groups of
B(1a)(2e) upon addition of triflic acid; more coordinated N(CH
؉) and less coordinated N(CH (∆) to boron
3 2
)
(
3 2
)
1
ing given above and, as a consequence, the observed H-
The dynamics of the B(1a)(2e) system shows the follow-
chemical shift changes of 3/4 can be quantitatively corre- ing picture. Exchange of the alkoxy group of the 2e portion
lated with the relative amounts of coordination of each of is slow on the NMR timescale at any temperature. The ex-
the two dimethylamino groups to the boron center. This change of the dimethylamino groups between the coordi-
direct measure of the stereotopic group recognition in solu- nated and noncoordinated states, i.e. the 3/4 equilibrium, is
tion under fast exchange conditions at room temperature rapid on the NMR timescale at room temperature, as also
amounts to 62:38. From the equilibria of 3 and 4 at Ϫ78 the rotation around the alkoxy bond in the boron-tricoordi-
and ϩ25 °C, K values (ϭ [3]/[4]) of 2.85 (by integration; nated intermediate as discussed above. In a variable tem-
slow exchange) and 1.63 (by shift analysis; fast exchange) perature NMR experiment, it appeared that the exchange
were thus obtained, respectively. Assuming that the heat ca- of the dimethylamino groups becomes slow only below
pacity is negligible over this temperature range, these data Ϫ60 °C (slow exchange was observed at Ϫ78 °C and very
lead to ∆H° ϭ Ϫ2.6 kJ/mol and ∆S° ϭ Ϫ4.8 J/molK. The broad but coalesced resonances at Ϫ50 °C) which leads to
Ϫ1
equilibrium is, therefore, enthalpy driven over this tempera- an exchange rate of about 10Ϫ100 s at this temperature.
ture range and these data suggest that a stronger BϪN From this result can be deduced that at Ϫ78 °C the ex-
bond is formed for 3 compared to 4 but that 3 is less flex- change is still chemically rapid, i.e. in the order of seconds.
ible.
This was confirmed by an experiment in which a sample of
In order to verify this ratio, a protonation titration was crystalline 3 was dissolved in partly frozen CD Cl
2
2
1
3
1
performed and monitored by C NMR. Protonation oc- (Ϫ95 °C) and measured ( H NMR) within two minutes
curs at a noncoordinated dimethylamino group and, there- after bringing the sample into the magnet at Ϫ78 °C. Here-
fore, the group which is least coordinated is more prone to after, a 13C spectrum was taken (30 min) and another H
protonation and concomitant chemical-shift change than spectrum (after warming to room temperature and recool-
1
1
3
1
the one coordinated to a larger degree. The C resonances ing to Ϫ78 °C). Both H spectra were identical and equal
1
of the dimethylamino groups were chosen because the H to the spectrum of a sample prepared at room temperature.
resonances showed smaller shifts and considerable line- The latter was true for the 13C spectrum also. Therefore,
broadening upon protonation. The titration experiment complete equilibration of 3 and 4 is reached within
using trifluoromethanesulfonic acid (triflic acid, TfOH) as 1Ϫ2 min after dissolving 3 at Ϫ78 °C.
1
1
the proton source is shown in Figure 5. B NMR con-
The B(1a)(2e) system constituted the first example for
firmed that the boron center remained tetracoordinated as which stereotopic group recognition was observed and
long as less than one equivalent of acid was used. Relative quantitatively assessed in both solution and the solid state.
to 2e, the 13C resonances of the dimethylamino groups in Selective functionalization of the noncoordinated amino
nonprotonated B(1a)(2e) were shifted upfield, leading to group in the complexes 3/4 and consecutive release of the
the conclusion that the more shifted one is the one coordi- product from the boron moiety was attempted. To this end
Eur. J. Org. Chem. 1999, 1775Ϫ1786
1779

J. Huskens, M. T. Reetz
FULL PAPER
a solution of 3/4 was treated with methyl trifluoromethane-
The methylation at Ϫ78 °C took several minutes to 1 h
sulfonate (MeOTf) and subsequently with water (Scheme to reach completion, and it is therefore slower than the
). Upon addition of one equivalent of MeOTf, complete equilibration of 3/4. This was confirmed by several experi-
chemoselectivity in favor of the monomethylated products ments in which crystalline 3 was subjected to MeOTf before
4
1
1
5
/6 was observed. The B spectrum showed still complete dissolution. None of these experiments gave other selectivit-
quaternization of the boron center. Only an excess of Me- ies than those mentioned above. Therefore, methylation oc-
OTf led to the formation of the (undesired) dimethylated curs only after dissolving 3, and the equilibration to 3/4 is
product. Therefore, it can be concluded that the noncoordi- (at all temperatures) more rapid than the methylation. Be-
nated dimethylamino group is methylated more rapidly, and cause the methylation is slower than the equilibration and
that coordination of the remaining dimethylamino moiety because the selectivities observed after methylation are
to the boron center results in protection against further identical to the degrees of stereotopic group recognition,
methylation. This was confirmed by the methylation of this strongly suggests that the diastereomers 3/4 have iden-
B(1j)(2e) for which the addition of one equivalent of Me- tical reactivity towards methylation.
OTf led to a mixture of mono- and dimethylated products.
Whether the coordinated or the noncoordinated dimeth-
This is attributed to the fact that this complex is present in ylamino group is being methylated has been suggested by
a tetracoordinated form for only 30% (see above) so that two pieces of evidence already. First is the observed protec-
protection after monomethylation does not occur.
tion effect of coordination to the boron center of the re-
maining dimethylamino group which proves the (much)
lower methylation rate for boron-coordinated amino groups
versus noncoordinated ones. Second are the identical meth-
ylation rates found for 3/4 which suggests that methylation
takes place away from the chiral center. The third piece of
evidence is found in Figure 2: the relative positions of the
13
C resonances of the alkoxy and the coordinated BϪNCH2
carbon atoms of the amino alcohol moieties are identical
for the major and minor isomers of 3/4 and 5/6. From these
spectra is concluded that the noncoordinated group (either
the dimethylamino group of 3/4 or the trimethylammonium
group of 5/6) assumes the same position with respect to the
chelate ring of the boron-coordinated amino alcohol por-
tion for both major isomers (and for both minor isomers
also).
As discussed above, molecular models suggest the lower
steric hindrance of 3 compared to 4. Another way to answer
the question of the major isomer 3 or 4 in solution is to
try to elucidate the absolute stereochemistry of the major
enantiomer of 7. This proved to be a nontrivial problem,
Scheme 4. Methylation of 3/4 and subsequent hydrolysis
Diastereomeric ratios 5/6 of 74:26 were observed by 13C however. All attempts to eliminate trimethylamine from this
NMR when the methylation reaction was performed at compound to prepare (dimethylamino)methyloxirane
Ϫ78 °C (see Figure 2, bottom), and 62:38 at 25 °C. These failed. The other direction, i.e. to make the compound from
values were in agreement with the observed ee values of optically pure epichlorohydrin and dimethylamine, also
monomethylated 2e (7) which were measured to be 48% at failed because the ring-opening reaction is not (completely)
Ϫ78 °C and 26% at 25 °C. The ee determinations were per- selective; probably reclosure under elimination of HCl oc-
1
formed by H NMR using (S)-1-(9-anthryl)-2,2,2-trifluoro- curs followed by reopening and addition of a second equiv-
ethanol as the (diamagnetic) chiral shift reagent.[ The sel- alent of dimethylamine. This impossibility to stop the reac-
ectivity values found here are identical to the observed dia- tion at the intermediate N,N-dimethyl-3-amino-1-chloro-2-
stereomeric ratios 3/4 at the respective temperatures. Thus, propanol renders this approach useless. Attempts to crys-
the ee of 7 results from the thermodynamic process of en- tallize 7 as a diastereomeric salt from the 5/6 mixture were
antiotopic group recognition of the dimethylamino groups also fruitless. No other chiral ligands were attempted to ob-
8]
of 2e in the diastereomeric complexes 3/4.
Three questions remained to be answered for this system:
tain selective crystallization.
The only other way was viewed to be to apply the H-
1
(i) what are the relative rates for the equilibration process NMR methodology using the above described (S)-1-(9-
of 3/4 versus methylation; (ii) is the noncoordinated dimeth- anthryl)-2,2,2-trifluoroethanol shift reagent for identifying
[
8][9]
ylamino group really the one being methylated; (iii) which the absolute stereochemistry of 7. In earlier papers,
it
diastereomer (3 or 4) forms preferentially in solution. The was shown that classes of compounds (e.g. lactones)[ gave
last question can, together with the answer to the second identical coordination behavior to this reagent so that defi-
one, be reduced to determining the absolute stereochemistry nite conclusions about their stereochemistry could be made.
8]
of the major enantiomer of 7.
Since it was not clear beforehand in which geometry an
1780
Eur. J. Org. Chem. 1999, 1775Ϫ1786

Observation of Stereotopic Group Recognition in Chiral Borate Complexes in Solution
FULL PAPER
analogous chelate ring would be formed by the shift reagent thyl group so that 5 results from methylation of 3 and 6 of
and 7, the corresponding isoelectronic analog N,N-di- 4, this suggests that 3 is indeed the major isomer present in
methyl-2-hydroxy-4,4-dimethyl-1-aminopentane 8 was pre- the equilibrating diastereomeric mixture before meth-
pared (Scheme 5). The first step, starting from 4,4-dimethyl- ylation, in accordance with the reasoning given above de-
1
-pentene was the enantioselective epoxidation using the Ja- pending on molecular models.
[
10]
cobsen procedure.
Although the epoxidation of primary
aliphatic alkenes occurs slower and with less stereoselectiv-
ity, the (S)-epoxide was expected to be formed preferen-
[
11]
tially.
The goal of this study was not to reach a high ee
for this reaction anyway, because both enantiomers had to
1
be visible in the H-NMR experiment. Therefore, the reac-
tion was performed at room temperature to reach a
comfortable reaction rate. After epoxide ring-opening with
dimethylamine, the pure amino alcohol product 8 was ob-
tained in an overall, isolated yield of 23%. The ee as deter-
1
mined by H NMR proved to be about 15% (see below).
1
Figure 6. H-NMR methyl singlets of 7 (left two parts) and 8 (right
two parts) in the presence of (S)-1-(9-anthryl)-2,2,2-trifluoroetha-
nol
It appeared that variation of the chiral 1,2-diol did not
bring about substantial selectivity differences: B(1g)(2e)
(
62:38), B(1e)(2e) (63:37), and B(1f)(2e) (61:39) all gave
similar values as determined by protonation titrations. In-
structive is the lower value for B(1b)(2e) (57:43) because
here only one face of the chelate ring formed by the diol is
able to differentiate between the amino groups, leading to
5
0% less selectivity compared to B(1a)(2e). Using 1c, it was
attempted to get one phenyl group closer to the noncoordi-
nated amino group, but it resulted in incomplete quaterniz-
ation at the boron center (94%) resulting in a somewhat
Scheme 5. Preparation of the isoelectronic analog 8 of 7
1
Figure 6 shows parts of the H-NMR spectra of 7 (left) lower stereoselectivity for B(1c)(2e) (60:40). This steric hin-
and its isoelectronic analog 8 (right) in the presence of the drance was confirmed by B(1d)(2e), for which only 88%
chiral shift reagent. For 7 the best separation is observed for tetracoordinated boron was observed. As already men-
ϩ
the (CH ) N resonance, whereas for 8 the corresponding tioned above, 1,3-diols (1j and 1k) did not give complete
3
3
(
CH ) C signal shows only a poor resolution. It should be quaternization either so that these were not investigated
3 3
mentioned that the spectra for 7 were taken with 2.4 equiva- further.
lents of chiral alcohol, whereas for 8 more than 13 equiva-
Changes were attempted at the substrate side, also, to see
lents were required to obtain this resolution, but this effect whether an increase of the bulkiness of the noncoordinating
may be due to a smaller formation constant for the latter amino group would result in better recognition. 2g ap-
coordination complex. In favor of a similar binding mode peared to be too bulky: the 11B-NMR shift (δ ϭ 23.1) ob-
for 7 and 8 coordinating to the chiral shift reagent, as is served for B(1a)(2g) indicated the complete lack of coordi-
obligatory for analysis of the absolute stereochemistry, are nation of the amino groups. For B(1a)(2f), a slightly better
the identical shift separations (∆δ Ϫ ∆δ )/(∆δ ϩ ∆δ ) for selectivity compared to B(1a)(2e) was observed by pro-
R
S
R
S
the N(CH ) resonances and the fact that all shifts occur in tonation at room temperature (67:33), and clean monome-
3
2
the same direction (upfield) relative to the chemical shifts thylation was observed after methylation at Ϫ78 °C.
of 7 and 8 without the chiral shift reagent present. Figure Overall, these borate model systems provided valuable
shows that for both 7 and 8 the major enantiomer gives new information regarding the possibilities for studying
6
the more upfield resonance. When a similar coordination stereotopic group recognition in solution. The B(1a)(2e)
mode is assumed, (R)-7 is the major enantiomer [(R)-7 is system showed three comfortable possibilities to study the
isoelectronic to (S)-8 for (R/S)-configurational assignment (rapid) stereotopic group recognition events occurring in
reasons] formed during the methylation and subsequent hy- solution for such systems: (i) NMR shifts relative to the
drolysis of B(1a)(2e) so that 5 must be the major component noncomplexed amines (as in Figure 4), (ii) NMR pro-
of the diastereomeric mixture 5/6. Combined with the fact tonation titrations using the shift movements of easily ob-
that methylation occurs at the noncoordinated aminome- served resonances upon addition of acid (as in Figure 5),
Eur. J. Org. Chem. 1999, 1775Ϫ1786
1781

J. Huskens, M. T. Reetz
FULL PAPER
and (iii) the methylation of tertiary to quaternary amines at antiomeric signals. The ee was determined by integration of these
resolved signals.
any desired temperature so that the resulting, and nonequil-
ibrating, product mixture (before or after hydrolysis) can be For the determination of the stereotopic group recognition occur-
investigated by NMR at room temperature (as in Figure ring in the borate complexes in solution under fast exchange con-
2
, bottom).
ditions (usually > Ϫ50 °C), titrations with acid (trifluoromethanes-
ulfonic acid, TfOH, or p-toluenesulfonic acid, TsOH) were em-
ployed. Generally 50Ϫ75 mg of a borate complex were dissolved in
CDCl
3
and the acid was added in portions of 0.20Ϫ0.25 equiva-
Conclusions
1
13
lents. Changes in H- and C-NMR chemical shifts were moni-
tored, while B spectra were recorded to verify the coordination
number (3 or 4) at the boron center.
1
1
This study describes in detail the first example of enanti-
otopic group recognition observed in both solution and the
solid state, implemented into an asymmetric reaction
scheme in a consecutive methylation step. Thus, a product
was obtained after hydrolysis of which the enantioselectivity
was shown to result directly from the foregoing stereotopic
group recognition event. Besides this phenomenon of en-
antiotopic group recognition, the spectroscopic properties
observed in solution have provided more general tools for
measuring the degree of stereotopic group recognition in
solution. Hopefully, this will allow more systematic and de-
tailed studies in this interesting part of the field of molecu-
lar recognition.
(
R,R)-(؉)-Hydrobenzoin (1a): 1a was prepared by enantioselective
dihydroxylation of trans-stilbene according to a procedure by
Sharpless.[ Ϫ H NMR (200 MHz, CDCl
12]
1
3
3
/CD CN ϭ 1:1): 7.05
3
4
(
m, 10 H, Ph), 4.45 (dd, 2 H, CH, JCHOH ϭ 2.0 Hz, JCHOH
ϭ
1
3
1
CD
7
.1 Hz), 3.45 (dd, 2 H, OH). C NMR (50.3 MHz, CDCl
CN ϭ 1:1): 140.0 (α-Ph), 127.2, 126.5 (o/m-Ph), 126.9 (p-Ph),
8.1 (CHOH).
3
/
3
(
±)-1,3-Diphenyl-1,3-propanediol (1k): The racemate of 1k was pre-
[13]
pared as reported by reduction of dibenzoylmethane to give 19%
of an off-white solid. Melting point: 130Ϫ132 °C (reported:
[13]
1
1
30°C).
3
Ϫ H NMR (200 MHz, CDCl ): 7.25Ϫ7.00 (m, 10 H,
3
Ph), 4.77 (dt, 2 H, CH), 4.64 (d, 2 H, OH, JCHOH ϭ 4.1 Hz), 1.89
3
, JCHCH2 ϭ 5.9 Hz). Ϫ 13C NMR (50.3 MHz, CDCl
(
t, 2 H, CH
2
3
):
1
44.8 (α-Ph), 127.7, 125.3 (o/m-Ph), 126.4 (p-Ph), 70.3 (CHOH),
Experimental Section
2 2
47.0 (CH ). MS (CI): 228 [M], 210 [M Ϫ H O].
General Remarks: All diols and (di)amino alcohols other than
those described below were obtained from commercial sources and
used without further purification. Triisopropyl borate was distilled
and kept under argon. Ϫ Melting points were determined using
apparatus from Gallenkamp Inc., UK. Values given are uncor-
rected. Ϫ Elemental analyses were determined by Mikroanaly-
tisches Laboratorium Kolbe, Mülheim an der Ruhr, Germany.
N,N,NЈ,NЈ-Tetraethyl-1,3-diamino-2-propanol (2f): 1,3-dibromo-2-
propanol (436 mg, 2.00 mmol, purity: 95%) and diethyl amine
(439 mg, 6.00 mmol) were dissolved in chloroform (5 mL). Potas-
sium carbonate (829 mg, 6.00 mmol) was added and the mixture
was heated at 50 °C overnight. After cooling and evaporation of
the solvent, water (25 mL) was added and the pH of the water layer
was adjusted to 1 using concentrated aqueous HCl. The water layer
was washed with dichloromethane (3 ϫ 25 mL). Hereafter, the pH
was adjusted to 12 with 4  NaOH (phase separation visible), and
the water layer was extracted with dichloromethane (3 ϫ 25 mL).
NMR Measurements were performed at
AMX300, and DMX600. Samples measured at low temperature
were filled under argon and sealed. Commercial CDCl and D
were used; CD Cl was distilled before use. No TMS was employed;
for H-NMR spectra, residual CHCl (δ ϭ 7.24) or CHDCl (δ ϭ
.30) were used for internal reference. For 1 C-NMR spectra,
CDCl (δ ϭ 77.0) or CD Cl (δ ϭ 53.8) were used. For spectra in
O, tert-butyl alcohol (CH
a Bruker AC200,
3
2
O
The latter organic layers were combined and dried with MgSO
After filtration and evaporation in vacuo, 2f was obtained (316 mg,
8%) as a clear liquid. The purity, according to NMR, was judged
to be 94%. The contaminant was found to be N,N-diethyl-3-amino-
,2-propanediol and was attributed to the presence of 3-bromo-1,2-
propanediol in the starting dibromide. Further purification was not
4
.
2
2
1
3
2
7
3
5
3
2
2
1
1
13
D
2
3
: δ ϭ 1.20 and 31.2 for H and C,
1
1
respectively) was added. The B spectra were recorded relative to
external BF ϫ OEt
1
attempted. Ϫ H NMR (200 MHz, CDCl
.63 (m, 1 H, CH), 2.48 (dq, 8 H, CH CH
2.34 (m, 4 H, CH CHOH), 0.92 (t, 12 H, CH ). Ϫ C NMR
3
): 4.28 (br s, 1 H, OH),
3
, JCH2CH3 ϭ 7.1 Hz),
3
2
.
3
3
2
1
3
Several complexes were measured at both Ϫ78 °C and room tem-
perature. For B(1a)(2e), series were measured from Ϫ78 to 22 °C
at 20Ϫ30 °C intervals to investigate their dynamic behavior. This
showed fast exchange at temperatures of Ϫ50 °C (very broad) or
higher and slow exchange at Ϫ78 °C, with coalescence probably at
about Ϫ60 °C. In another experiment, the crystalline form 3 was
2
3
(50.3 MHz, CDCl3): 65.2 (CHOH), 57.7 (CH CHOH), 47.2
2
(CH CH ), 11.6 (CH ).
2
3
3
N,N,NЈ,NЈ-Tetrabenzyl-1,3-diamino-2-propanol (2g): Sodium car-
bonate (1.09 g, 10.3 mmol) and 1,3-diamino-2-propanol (201 mg,
2
.23 mmol) were dissolved in water (10 mL) and heated to reflux.
Benzyl bromide (1.57 g, 9.16 mmol) was added dropwise over
0 min. After additional refluxing for 1 h, the mixture was cooled
to room temperature and extracted with diethyl ether (3 ϫ 20 mL).
The ether layers were washed with brine and dried with MgSO
dissolved in CD
2
Cl
2
at Ϫ95 °C and measured within two minutes
1
13
1
(
H) at Ϫ80 °C. After 30 min ( C spectrum), the H spectrum was
3
1
remeasured. The H spectra did not differ from each other and the
1
13
H and C spectra were both identical to the spectra obtained for
4
.
a sample which was prepared at room temperature.
After filtration and evaporation, 2g (632 mg, 1.40 mmol, 63%) was
obtained as a yellowish oil, contaminated with about 7 wt-% benzyl
alcohol. This was used directly for the formation of B(1a)(2g) (see
below). An acid/base extraction as described above for 2f gave 2g
The ee measurements described here employed commercial (S)-1-
(
9-anthryl)-2,2,2-trifluoroethanol as the (diamagnetic) chiral shift
reagent. Samples generally contained 5Ϫ10 mg of the substance of
which the ee was to be determined, dissolved in 0.75 mL CDCl
1
3
.
(218 mg, 0.484 mmol) with a purity > 97 w%. Ϫ H NMR
The shift reagent was added in portions of 5Ϫ10 mg until one or
more H-NMR resonances of the substance (usually an easily ob-
served singlet) were completely resolved into their respective en-
3
(200 MHz, CDCl ): 7.40Ϫ7.25 (m, 20 H, Ph), 3.84 (qn, 1 H, CH,
1
3
2
J
CHCH2 ϭ 6.1 Hz), 3.59 (dd, 8 H, CH
2
Ph, Jgem ϭ Ϫ13.5 Hz), 2.46
1
3
(d, 4 H, CH
2
CH). Ϫ C NMR (50.3 MHz, CDCl ): 139.0 (α-Ph),
3
1782
Eur. J. Org. Chem. 1999, 1775Ϫ1786

Observation of Stereotopic Group Recognition in Chiral Borate Complexes in Solution
FULL PAPER
1
5
29.0, 128.2 (o/m-Ph), 127.0 (p-Ph), 65.7 (CHOH), 58.9 (CH
7.7 (CH CH).
2
Ph), CDCl
Ph), 128.0, 126.5 (o/m-Ph), 127.3 (p-Ph), 85.2 (CHϪPh), 65.9, 65.6
3
): 10.7. Ϫ 13C NMR (50.3 MHz, CDCl
3
): 141.3, 141.0 (α-
2
(
(
CH
CHCH
2
N), 64.7, 64.4 (CHCH
).
3 3
), 46.6, 45.4 (NCH ), 21.7, 21.5
(
S)-N,N-Dimethyl-2-hydroxy-4,4-dimethyl-1-aminopentane (8): This
[10]
3
amino alcohol derivative was prepared by Jacobsen-epoxidation
2
of 4,4-dimethyl-1-pentene, followed by ring-opening using Me NH,
as follows. The catalyst (S,S)-(ϩ)-N,NЈ-bis(3,5-di-tert-butylsalicyli-
dene)-1,2-cyclohexanediaminomanganese(III) chloride, 68 mg,
Borate Ester of (R,R)-Hydrobenzoin and N,N,NЈ,NЈ-Tetramethyl-
,3-diamino-2-propanol B(1a)(2e) (mixture of 3/4): Crystals suitable
for single-crystal X-ray analysis were grown by slow evaporation of
1
0.107 mmol) and 4,4-dimethyl-1-pentene (196 mg, 2.00 mmol) were
the solvent of a sample containing 720 mg of B(1a)(2e) in about
dissolved in dichloromethane (2 mL). A buffered sodium hypo-
chlorite solution (2.5 mL, containing 1.43  NaOCl and 14 m
phosphate buffer at pH 11.5) was added and the two-phase mixture
was stirred vigorously at room temperature overnight. Hexane
5
mL acetone under an argon stream. Crystals (3, 352 mg, 49%)
were isolated while about half of the solvent was still present. These
crystals were used for the X-ray and the elemental analyses. For a
second batch, evaporation of all solvent was brought about re-
sulting in quantitative crystallization. Washing of the crystals with
ice-cold acetone resulted in pure B(1a)(2e) (3, > 85%) with the same
melting point as found for the first batch. Melting point: 152 °C.
(20 mL) and water (20 mL) were added and the organic layer was
washed with brine (2 ϫ 20 mL). The combined water layers were
extracted with dichloromethane (3 ϫ 20 mL) and the combined or-
ganic layers were dried with MgSO
4
. After filtration and evapo-
Ϫ MS (EI): 369 [M ϩ 1]. Elemental analysis calculated (found) for
ration of the solvent the crude epoxide was obtained as a brownish
liquid, still containing manganese. In order to remove the catalyst
remainder, the product was dissolved in dichloromethane (1 mL)
and put through a small silica column (2 mL volume), eluted with
dichloromethane. After evaporation of the solvent a (less-colored)
liquid (160 mg, < 70%) was obtained, which still showed consider-
1
C
21
H
29BN
2
O
3
: C 68.49 (68.36), H 7.94 (8.05), N 7.61 (7.51). Ϫ H
NMR (200 MHz, CDCl
CHϪPh), 4.26 (qn, 1 H, CHCH
3
): 7.26 (m, 10 H, Ph), 4.74 (m, 2 H,
3
2
, JCHCH2 ϭ 6.9 Hz), 2.90Ϫ2.70
11
(
(
(
(
(
(
m, 4 H, CH
64.2 MHz, CDCl
α-Ph), 128.0, 126.4 (o/m-Ph), 127.3 (p-Ph), 85.2 (CHϪPh), 66.7
2
N), 2.61, 2.49 (2 s, 6 Hϩ6 H, NCH
3
). Ϫ B NMR
13
3
): 10.6. Ϫ C NMR (50.3 MHz, CDCl ): 141.4
3
1
able line-broadening with H NMR. This was used directly in the
1
CHCH
300 MHz, CD
m, 2 H, CHϪPh), 4.18 (br qn, 1 H, CHCH
NϪB), 2.90Ϫ2.40 [m, 2 H, CH N (uncoord.)], 2.82, 2.73, 2.70,
.65 (4 s: about 1:3:3:1, 6 H, BϪNCH ), 2.20, 2.17 [2 s: about 3:1, 6
(uncoord.)]. Ϫ 13C NMR (75.5 MHz, CD
H, NCH Cl , Ϫ78 °C):
40.3, 139.7, 139.5, 139.2 (about 3:3:1:1, α-Ph), 127.4Ϫ124.6 (> 10
resonances, o/m/p-Ph), 84.4, 84.2, 84.0, 83.5 (about 1:1:3:3,
CHϪPh), 65.2, 65.0 (about 3:1, CHCH ), 63.8, 63.7 (about 1:3,
CH NϪB), 62.2, 61.6 [about 1:3, CH N (uncoord.)], 46.2, 45.2
about 1:3, NCH (uncoord.)], 44.7, 44.0 (about 1:3, BϪNCH ). A
2
), 64.4, 63.8 (CH
Cl , Ϫ78°C): 7.34Ϫ7.00 (m, 10 H, Ph), 4.70Ϫ4.40
), 3.62Ϫ3.15 (m, 2 H,
2 3
N), 46.5, 45.7 (NCH ). Ϫ H NMR
second step, the ring-opening with dimethylamine. After adding
2
2
water (0.50 mL), 2-propanol (1.00 mL), and aqueous Me
0.50 mL 40%), the mixture was stirred at room temperature for
8 h. After evaporation of the solvent and the excess Me NH, the
2
NH
2
(
CH
2
2
2
4
2
3
residue was taken up in water (20 mL) and acidified to pH 1 using
concentrated aqueous HCl. The solution was washed with di-
chloromethane (3 ϫ 20 mL), which was colored dark-brown. The
pH of the water layer was raised to 12 using NaOH (300 mg), and
extracted with dichloromethane (5 ϫ 20 mL). The latter layers were
3
2
2
1
2
2
2
4
combined, dried with MgSO , and the solvent was evaporated in
[
3
3
vacuo resulting in > 90% pure N,N-dimethyl-2-hydroxy-4,4-di-
methyl-1-aminopentane (8, 51 mg, 23%). An ee of 10Ϫ15% was
sample prepared from (±)-hydrobenzoin gave identical NMR spec-
tra.
1
determined by H NMR using the diamagnetic chiral shift reagent
S)-1-(9-anthryl)-2,2,2-trifluoroethanol. Ϫ 1H NMR (300 MHz, Borate Ester of (R,R)-Hydrobenzoin and N,N,NЈ,NЈ-Tetraethyl-1,3-
(
1
CDCl
ϩ dd, 2 H, NCH
1
1
6
3
): 3.68 (m, 1 H, CHOH), 2.23 (s, 6 H, NCH
3
), 2.21, 2.04 (m
, for dd: JCHCH2 ϭ 3.3 Hz, Jgem ϭ Ϫ12.1 Hz),
.29, 1.09 (2 dd, 2 H, CH
tBu, 3 CHCH2 ϭ 7.9, 2.6 Hz, 2
4.4 Hz), 0.90 (s, 9 H, C(CH
6.8 (NCH ), 64.2 (CHOH), 48.6 (CH
], 30.1 [C(CH ].
diamino-2-propanol, B(1a)(2f):
7.37Ϫ7.08 (m, 10 H, Ph), 4.65 (m, 2 H, CHϪPh), 4.17 (m, 1 H,
CHCH ), 3.10Ϫ2.50 (m, 12 H, CHCH and CH CH ), 1.06 (br t,
12 H, CH , JCH2CH3 ϭ 6.8 Hz). Ϫ B NMR (64.2 MHz, CDCl ):
): 141.3 (α-Ph), 127.9, 126.4
(o/m-Ph), 127.2 (p-Ph), 84.8 (CHϪPh), 66.0 (CHCH ), 58.2, 57.6
CHCH ), 46.8 (CH CH ), 9.8 (CH ).
3
H NMR (200 MHz, CDCl ):
3
2
2
2
J
J
gem
ϭ
2
2
2
3
13
3
11
3
)
3
). Ϫ C NMR (75.5 MHz, CDCl
3
):
3
3
3
1
), 30.2 11.0. Ϫ C NMR (50.3 MHz, CDCl
3
2
2
tBu), 45.3 (NCH
3
[C(CH
)
3 3
)
3 3
2
(
2
2
3
3
General Borate Esterification Procedure: Unless stated otherwise,
each borate ester derivative described below was prepared as fol-
lows. Equimolar amounts of triisopropyl borate, diol, and amino
alcohol were dissolved in toluene (1 mmol in 10 mL) and concen-
trated in vacuo at 50 °C. After twice repeated coevaporation of 2-
propanol with toluene, the compound was dried in vacuo and used
as such.
Borate Ester of (R,R)-Hydrobenzoin and N,N,NЈ,NЈ-Tetrabenzyl-
1
1
7
1
Ϫ
,3-diamino-2-propanol, B(1a)(2g): H NMR (200 MHz, CDCl
.46Ϫ7.20 (m, 30 H, Ph and Bz), 5.13 (s, 2 H, CHϪPh), 4.40 (m,
H, CHCH ), 3.65 (m, 8 H, CH -Bz), 2.57 (m, 4 H, CHCH ).
B NMR (64.2 MHz, CDCl ): 23.1. Ϫ C NMR (50.3 MHz,
): 140.1 (α-Ph), 139.2, 139.0 (α-Bz), 130.0Ϫ125.9 (o/m/p-Ph
and -Bz), 85.5 (CHϪPh), 70.3 (br, CHCH ), 58.9, 58.8 (CH -Bz),
57.7, 57.4 (CHCH ).
3
):
2
2
2
11
13
3
CDCl
3
Borate Ester of (R,R)-Hydrobenzoin and (Dimethylamino)ethanol,
2
2
1
B(1a)(2c): H NMR (200 MHz, CDCl
(
3
): 7.25 (m, 10 H, Ph), 4.76
OB), 3.25Ϫ2.95 (m, 2 H,
). Ϫ 11B NMR (64.2 MHz, CDCl
):
0.8. Ϫ 13C NMR (50.3 MHz, CDCl
): 141.0 (α-Ph), 127.9, 126.3
o/m-Ph), 127.2 (p-Ph), 85.0 (CHϪPh), 58.7, 57.9 (CH OB and
CH N), 45.0 (NCH ).
2
br s, 2 H, CHϪPh), 3.98 (m, 2 H, CH
CH N), 2.74 (br s, 6 H, NCH
2
Borate Ester of (S)-1-Phenyl-1,2-ethanediol and (±)-N,N-Dimethyl-
2
3
3
1
1
-amino-2-propanol: B(1b)(2d) (1:1 mixture of diastereomers):
NMR (200 MHz, CDCl ): 7.30Ϫ7.00 (m, 5 H, Ph), 5.03 (m, 1 H,
CHϪPh), 4.20Ϫ4.00, 3.59 (2 m, 2Hϩ1 H, CHCH and CH OB),
.90Ϫ2.60 (m, 2 H, CH N), 2.56 (s, 6 H, NCH ), 1.22, 1.21 (2 d,
CHCH3 ϭ 6.1 Hz). Ϫ B NMR (64.2 MHz,
H
1
3
3
(
2
3
2
2
3
2
2
3
3
11
Borate Ester of (R,R)-Hydrobenzoin and (±)-N,N-Dimethyl-1-am-
3 H, CHCH
ino-2-propanol, B(1a)(2d) (1:1 mixture of diastereomers): H NMR CDCl ): 11.0. Ϫ C NMR (50.3 MHz, CDCl
200 MHz, CDCl ): 7.26 (m, 10 H, Ph), 4.77 (m, 2 H, CHϪPh), 128.1, 125.4 (o/m-Ph), 127.0 (p-Ph), 76.7 (CHϪPh), 71.5 (CH
), 65.6, 65.4 (CH N), 64.6, 64.5 (CHCH ), 45.9, 45.1 (NCH
21.5 (CHCH ).
3
,
J
1
13
3
3
): 142.7 (α-Ph),
OB),
(
3
2
4
1
.24 (m, 1 H, CHCH
.33 (d, 3 H, CHCH
3
3
), 2.85Ϫ2.65 (m, 8 H, CH
, JCHCH3 ϭ 5.9 Hz). Ϫ B NMR (64.2 MHz,
2
N and NCH
3
2
3
3
), 21.6,
11
3
3
Eur. J. Org. Chem. 1999, 1775Ϫ1786
1783

J. Huskens, M. T. Reetz
FULL PAPER
Borate Ester of (S)-1-Phenyl-1,2-ethanediol and N,N,NЈ,NЈ-Tetra-
methyl-1,3-diamino-2-propanol, B(1b)(2e): H NMR (200 MHz, (CH
Ph), 76.4 (CHOB), 73.1 (CH
2
Ph), 72.8 (CH
2
OBz), 58.7, 58.0
1
2
N and CH OB), 44.6 (NCH
2
3
).
CDCl
.28Ϫ4.04, 3.63 (2 m, 2Hϩ1 H, CHOB and CH
m, 4 H, CH N), 2.55, 2.45 (2 s, 6 H ϩ 6 H, NCH
64.2 MHz, CDCl
): 10.9. Ϫ 13C NMR (50.3 MHz, CDCl
α-Ph), 128.1, 125.7 (o/m-Ph), 127.0 (p-Ph), 76.4 (CHϪPh), 71.0
CH OB), 66.4 (CHOB), 63.8, 63.3 (CH N), 45.8, 45.4 (NCH ).
3
): 7.42Ϫ7.18 (m, 5 H, Ph), 5.02 (m, 1 H, CHϪPh),
Borate Ester of (R,R)-1,4-Dibenzylthreitol and N,N,NЈ,NЈ-Tetra-
methyl-1,3-diamino-2-propanol, B(1f)(2e): H NMR (200 MHz,
4
2
OB), 2.78Ϫ2.65
1
11
(
(
(
(
2
3
). B NMR
): 142.8
CDCl
4
3
): 7.20 (m, 10 H, Ph), 4.45, 4.44 (2 s, 2 H ϩ 2 H, CH
.02 (qn, 1 H, CHCH N), 3.92 (br, 2 H, CHOB), 3.48 (br, 4 H,
OBz), 2.62Ϫ2.50 (m, 4 H, CH N), 2.31, 2.26 (2 s, 6 H ϩ 6
2
Ph),
3
3
2
CH
2
2
2
2
3
1
1
13
H, NCH
(50.3 MHz, CDCl
Ph), 76.0 (CHOB), 73.0 (CH
): 7.84Ϫ6.98 (m, 15 H, Ph), 5.86 (s, 1 H, (CHCH N), 63.7, 63.5 (CH N), 45.8, 45.2 (NCH
CHϪPh), 4.36 (m, 1 H, CHCH ), 3.00Ϫ2.65 (m, 4 H, CH N), 2.42
3
). Ϫ B NMR (64.2 MHz, CDCl
): 138.1 (α-Ph), 127.9, 127.3 (o/m-Ph), 127.1 (p-
Ph), 72.8 (CH OBz), 66.2
).
3
): 10.7. Ϫ C NMR
Borate Ester of (S)-1,1,2-Triphenyl-1,2-ethanediol and N,N,NЈ,NЈ-
3
1
Tetramethyl-1,3-diamino-2-propanol,
200 MHz, CDCl
B(1c)(2e):
H
NMR
2
2
(
3
2
2
3
2
2
Borate Ester of 1,2:5,6-Diisopropylidene- -mannitol and (Dimeth-
). Ϫ B NMR (64.2 MHz, CDCl
): 11.4. Ϫ 13
NMR (200 MHz, CDCl ):
NMR (50.3 MHz, CDCl ): 148.3, 144.6, 141.3 (α-Ph), 128.8Ϫ125.0
OB and CHϪ and CH 1g), 3.05Ϫ2.76
), 86.0 (CHϪPh), 66.8 (CHCH ), 64.0, 63.4
N), 2.55 (s, 6 H, NCH ), 1.35, 1.26 [2 s, 6 H ϩ 6 H,
N), 46.0, 45.3 (NCH ).
D
1
1
(
s, 12 H, NCH
3
3
C
1
ylamino)ethanol, B(1g)(2c):
.11Ϫ3.75 (m, 10 H, CH
m, 2 H, CH
H
3
3
4
(
2
2
(
(
o/m/p-Ph), 87.6 (CPh
CH
2
2
2
3
2
3
1
1
13
C(CH
3
)
2
]. Ϫ B NMR (64.2 MHz, CDCl
): 108.9 [C(CH ], 78.0, 77.6 (CH 1g), 66.4
1g), 59.4, 57.8 (CH N and CH OB), 45.1 (NCH ), 26.3,
), 2.63 (m, 25.1 [C(CH ].
3
): 10.8. Ϫ C NMR
Borate Ester of 1,1,2,2-Tetraphenyl-1,2-ethanediol and (±)-N,N-Di-
(50.3 MHz, CDCl
NMR (200 MHz, (CH
): 7.28Ϫ7.00 (m, 20 H, Ph), 4.29 (m, 1 H, CHCH
3
3 2
)
1
methyl-1-amino-2-propanol, B(1d)(2d):
CDCl
H
2
2
2
3
3
3
3
3 2
)
2
5
(
H, CH
.6 Hz). Ϫ 11B NMR (64.2 MHz, CDCl
): 145.9, 144.1 (α-Ph), 128.6Ϫ126.3 (o/m/p-Ph),
COB), 67.6 (CH N), 65.3 (br, CHCH ), 46.0 (NCH ).
2
N), 2.14 (s, 6 H, NCH
3
), 1.35 (d, 3 H, CHCH
3
, JCHCH3 ϭ
Borate Ester of 1,2:5,6-Diisopropylidene-
Tetramethyl-1,3-diamino-2-propanol,
D
-mannitol and N,N,NЈ,NЈ-
): 14.0. Ϫ 13C NMR
3
1
B(1g)(2e):
H
NMR
50.3 MHz, CDCl
3
(
200 MHz, CDCl
CH 1g), 2.61 (m, 4 H, CH
3 2
1.34, 1.27 [2 s, 6 H ϩ 6 H, C(CH )
3
): 4.07Ϫ3.65 (m, 9 H, CHCH
2
N and CHϪ and
),
]. Ϫ B NMR (64.2 MHz,
): 10.5. Ϫ C NMR (50.3 MHz, CDCl ): 108.9 [C(CH ],
): 7.37Ϫ6.98 (m, 20 H, Ph), 4.30 (m, 1 H, 78.1, 77.6 (CH 1g), 66.5 (CHCH N), 66.3 (CH 1g), 64.0, 63.4
), 26.3, 25.2 [C(CH ].
8
3.0 (Ph
2
2
3
3
2
2
N), 2.45, 2.35 (2 s, 6 H ϩ 6 H, NCH
3
1
1
Borate Ester of 1,1,2,2-Tetraphenyl-1,2-ethanediol and N,N,NЈ,NЈ-
1
13
Tetramethyl-1,3-diamino-2-propanol,
200 MHz, CDCl
CHOB), 2.75 (m, 4 H, CH
64.2 MHz, CDCl
): 11.9. Ϫ 13C NMR (50.3 MHz, CDCl
44.2 (α-Ph), 129.0Ϫ125.9 (o/m/p-Ph), 91.7, 82.9 (Ph COB), 67.2
CHOB), 64.8 (CH N), 46.1 (NCH ).
B(1d)(2e):
H
NMR
CDCl
3
3
3 2
)
(
3
2
2
1
2
N), 2.23 (s, 12 H, NCH
3
). Ϫ 1 B NMR
(CH
2
N), 46.0, 45.6 (NCH
3
3 2
)
(
1
(
3
3
): 146.8,
Borate Ester of Catechol and (±)-N,N-Dimethyl-1-amino-2-propa-
2
1
nol, B(1h)(2d): H NMR (200 MHz, CDCl
CH 1h), 4.15 (m, 1 H, CHCH ), 2.90, 2.71 (2 dd, 2 H, CH
gem ϭ Ϫ11.3 Hz, JCHCH2 ϭ 5.1, 10.1 Hz), 2.52, 2.38 [2 s, 6 H,
3
): 6.67Ϫ6.55 (m, 4 H,
2
3
3
2
N,
2
3
Borate Ester of (R,R)-Diisopropyltartrate and Aminoethanol,
J
1
3
11
B(1e)(2a): Prepared by coevaporation with chloroform. H NMR N(CH
200 MHz, CD Cl ): 6.79 (br s, 2 H, NH ), 4.93 (m, 2 H, CHCH ), (64.2 MHz, CDCl
.23, 4.10 (2 s, 2 H, CHOB), 3.61 (br m, 2 H, CH OB), 2.93 (br 150.6 (COB 1h), 119.0, 118.5, 109.3, 108.7 (CH 1h), 65.5, 65.3
3 2 3
) ], 1.20 (d, 3 H, CHCH , JCHCH3 ϭ 6.0 Hz). Ϫ B NMR
1
3
(
2
2
2
3
3 3
): 11.8. Ϫ C NMR (50.3 MHz, CDCl ): 150.7,
4
2
3
11
m, 2 H, CH
NMR (64.2 MHz, CD
2
N), 1.13 (d, 12 H, CH
3
, JCHCH3 ϭ 6.2 Hz). Ϫ
B
2 3 2 3
(CHCH ), 45.6, 44.5 [N(CH ) ], 21.2 (CHCH ).
13
2
Cl ): 11.3. Ϫ
2
C NMR (50.3 MHz,
Borate Ester of Catechol and N,N,NЈ,NЈ-Tetramethyl-1,3-diamino-
CD
CH
2
Cl
2
): 174.2 (COO), 76.2 (CHOB), 69.2 (CHCH
3
), 58.8 (br,
1
2
-propanol, B(1h)(2e): H NMR (200 MHz, CDCl
3
): 6.69Ϫ6.47 (m,
N), 2.33
): 11.8. Ϫ 13
): 150.6, 150.5 (COB 1h), 119.0, 118.8,
), 109.2, 108.4 (CH 1h), 67.1 (CHOB), 62.9 (CH ), 45.1 (CH ).
N), 1.23, 1.18
2
OB), 42.5 (br, CH
2
N), 21.9 (CH ).
3
4
H, CH 1h), 4.17 (m, 1 H, CHOB), 2.65, (br, 4 H, CH
2
1
1
Borate Ester of (R,R)-Diisopropyltartrate and 1,3-Diamino-2-propa-
(br s, 12 H, CH
3
). Ϫ B NMR (64.2 MHz, CDCl
3
C
1
nol, B(1e)(2b): H NMR (200 MHz, CDCl
CHCH ), 4.38, 4.25 (2 s, 2 H, CHOB 1e), 4.20 (br s, 4 H, NH
.15 (m, 1 H, CHOB 2b), 3.20Ϫ2.70 (m, 4 H, CH
3 3
): 5.00 (m, 2 H, NMR (50.3 MHz, CDCl
3
2
2
3
4
(
2
Borate Ester of (±)-4-Methyl-2,4-pentanediol and 1,3-Diamino-2-
3
2 d, 12 H, CH , JCHCH3 ϭ 6.2 Hz). Ϫ B NMR (64.2 MHz,
3
11
1
propanol, B(1j)(2b): H NMR (200 MHz, CDCl
3
): 4.06 (m, 1 H,
13
3 3
CDCl ): 11.1. Ϫ C NMR (50.3 MHz, CDCl ): 173.5, 173.1
2
CH 1j), 3.83 (m, 1 H, CH 2b), 2.63, 2.51 (2 dd, 4 H, CH
2
N, Jgem
Ϫ12.9 Hz, JCHCH2 ϭ 4.4, 6.8 Hz), 1.58, 1.34 (2 dd, 2 H, CH
ϭ
(
(
COO), 76.5, 75.7 (CHOB 1e), 73.0 (br, CHOB 2b), 68.8, 68.5
CHCH ), 44.7, 44.6 (br, CH N), 21.7, 21.6 (CH ).
3
2
1j,
3
2
3
2
gem ϭ Ϫ13.9 Hz, 3JCHCH2 ϭ 2.8, 11.6 Hz), 1.50 (br, 4 H, NH
J
2
),
, JCHCH3 ϭ 6.2 Hz).
3
Borate Ester of (R,R)-Diisopropyltartrate and N,N,NЈ,NЈ-Tetra-
1.12 [s, 6 H, C(CH
Ϫ
3
)
2
], 1.08 (d, 3 H, CHCH
3
13
1
11
methyl-1,3-diamino-2-propanol, B(1e)(2e): H NMR (200 MHz,
B NMR (64.2 MHz, CDCl
3
): 17.2. Ϫ C NMR (50.3 MHz,
): 75.0 (br, CH 2b), 71.4 [C(CH ], 65.3 (CH 1j), 45.5 (CH
N), 31.0, 27.5, 22.9 (CH ).
3
CD
CHCOO), 4.04 (m, 1 H, CHCH
.47, 2.31 (2 s, 6 H ϩ 6 H, NCH
NMR (64.2 MHz, CD Cl
CD Cl ): 172.1 (br, COO), 76.9 (CHCOO), 68.5 (CHCH
CHCH N), 64.0, 63.8 (CH N), 46.3, 45.4 (NCH ), 21.7 (CHCH
2
Cl
2
): 4.96 (sp, 2 H, CHCH
3
, JCHCH3 ϭ 6.3 Hz), 4.28 (s, 2 H,
CDCl
3
3
)
2
2
2
N), 2.80Ϫ2.40 (m, 4 H, CH
2
N),
B
1j), 44.9 (CH
2
3
1
1
2
3
), 1.17 (d, 12 H, CHCH
3
). Ϫ
C NMR (50.3 MHz,
), 67.2
).
Borate Ester of (±)-4-Methyl-2,4-pentanediol and (±)-N,N-Di-
methyl-1-amino-2-propanol, B(1j)(2d) (1:1 mixture of dia-
13
2
2
): 11.0. Ϫ
2
2
3
1
stereomers): H NMR (200 MHz, CDCl
3
): 4.32Ϫ4.02 (m, 2 H, CH
N), 2.18 (s, 6 H, NCH ),
1j), 1.20 [s, 6 H, C(CH ], 1.15 (d, 3 H,
CHCH3 ϭ 6.2 Hz), 1.08 (d, 3 H, CHCH 2d,
(
2
2
3
3
1
j and 2d), 2.34Ϫ2.07 (m, 2 H, CH
1.72Ϫ1.28 (m, 2 H, CH
2
3
Borate Ester of (R,R)-1,4-Dibenzylthreitol and (Dimethylamino)-
ethanol, B(1f)(2c): 1H NMR (200 MHz, CDCl
): 7.21 (m, 10 H, CHCH
Ph), 4.45 (s, 4 H, CH Ph), 3.98 (br, 2 H, CHOB), 3.77 (m, CH
.48 (br, 4 H, CH OBz), 3.02Ϫ2.67 (m, 2 H, CH N), 2.43 (s, 6 NMR (50.3 MHz, CDCl
). Ϫ 11B NMR (64.2 MHz, CDCl ): 11.0. Ϫ 13C NMR
H, NCH 66.0, 65.9, 65.3, 65.2 (CH 1j and 2d), 46.0 (NCH
50.3 MHz, CDCl ): 138.1 (α-Ph), 128.1, 127.5 (o/m-Ph), 127.3 (p- 31.2, 31.1, 27.8, 27.7, 23.1 (CH 1j), 20.9, 20.8 (CH
3
2
3 2
)
3
3
3
1j,
CHCH3 ϭ 6.2 Hz). Ϫ B NMR (64.2 MHz, CDCl
): 71.3 [C(CH ], 66.7, 66.6 (CH
), 45.8 (CH
2d).
J
3
3
11
13
2
2
OB),
J
3
): 17.3. Ϫ
C
N),
1j),
3
2
2
3
3
)
2
2
3
3
3
2
(
3
3
1784
Eur. J. Org. Chem. 1999, 1775Ϫ1786

Observation of Stereotopic Group Recognition in Chiral Borate Complexes in Solution
FULL PAPER
Borate Ester of (±)-4-Methyl-2,4-pentanediol and N,N,NЈ,NЈ-Tetra-
[CH
methyl-1,3-diamino-2-propanol, B(1j)(2e): H NMR (200 MHz, 45.4 [N(CH
CDCl ): 4.11Ϫ3.89 (m, 2 H, CH 1j and 2e), 2.18 (d, 4 H, CH N, formed at Ϫ78 °C, a mixture of diastereomers 5/6 (63:37) was ob-
CHCH2 ϭ 6.2 Hz), 2.08 (s, 12 H, NCH ), 1.47, 1.22 (2 dd, 2 H, served by C NMR at room temperature. After hydrolysis, an ee
CH 1j), 1.06 [s, 6 H, C(CH ], 1.01 (d, 3 H, CHCH
.1 Hz). Ϫ 11B NMR (64.2 MHz, CDCl ): 15.3. Ϫ 13C NMR chiral shift reagent (S)-1-(9-anthryl)-2,2,2-trifluoroethanol. Using
50.3 MHz, CDCl ): 70.5 [C(CH ], 67.3 (CH 2e), 64.5 (CH 1j), (R,R)-hydrobenzoin, with methylation performed at room tempera-
N), 45.7 (NCH and CH
2
N(CH
3
)
3
], 62.5 (CHOH), 61.9 [CH
2 3 2 3 3
N(CH ) ], 54.5 [N(CH ) ],
1
3 2
) ]. Ϫ Using (±)-hydrobenzoin, with methylation per-
3
2
3
13
J
3
3
1
2
)
3 2
3
, JCHCH3
ϭ
of 0% was determined for 7 by H NMR using the diamagnetic
6
3
(
3
3 2
)
1
3
6
3.6 (CH
2
3
2
1j), 31.3, 27.7, 23.0 (CH
3
1j).
ture, a mixture of diastereomers 5/6 (63:37) was observed by
C
NMR at room temperature. After hydrolysis, an ee of 26% was
determined for 7 by H NMR using the diamagnetic chiral shift
reagent (S)-1-(9-anthryl)-2,2,2-trifluoroethanol.
Borate Ester of (±)-1,3-Diphenyl-1,3-propanediol and (Dimethylami-
1
1
no)ethanol, B(1k)(2c): H NMR (200 MHz, CDCl
3
): 7.40Ϫ7.18 (m,
3
1
0 H, Ph), 5.13 (t, 2 H, CHPh, JCHCH2 ϭ 5.2 Hz), 4.06 (t, 2 H,
3
CH
2
OB, JCH2CH2 ϭ 6.0 Hz), 2.72 (t, 2 H, CH
), 2.28 (t, 2 H, CH
CHPh). Ϫ 11B NMR (64.2 MHz, CDCl
7.2. Ϫ 13C NMR (50.3 MHz, CDCl
): 142.5 (α-Ph), 128.2, 125.0
o/m-Ph), 127.1 (p-Ph), 70.5 (CHPh), 60.1, 59.8 (CH OB and
CH N), 45.4 (NCH ), 41.4 (CH CHPh).
2
N), 2.42 (s, 6 H,
Methylation of Solid B(1a)(2e): Crystallized B(1a)(2e) (3: a single
diastereomer, see above) was used and ground in a mortar. (a) The
solid was suspended in pentane, and MeOTf was added at Ϫ78 °C.
The mixture was warmed to room temperature over 4 h, dichloro-
methane was added and the solvent was evaporated. A dia-
NCH
3
2
3
):
1
3
(
2
2
3
2
1
3
stereomeric ratio (63:37) for 5/6 was observed by C NMR. (b)
Borate Ester of (±)-1,3-Diphenyl-1,3-propanediol and N,N,NЈ,NЈ-
1
MeOTf was added to the solid at room temperature without ad-
Tetramethyl-1,3-diamino-2-propanol,
200 MHz, CDCl ): 7.45Ϫ7.02 (m, 10 H, Ph), 5.06 (br t, 2 H,
CHPh, JCHCH2 ϭ 5.3 Hz), 4.30 (m, 1 H, CHCH N), 2.68 (m, 4
H, CH N), 2.43 (s, 12 H, NCH ), 2.28Ϫ2.15 (m, 2 H, CH CHPh).
B NMR (64.2 MHz, CDCl
50.3 MHz, CDCl ): 145.3, 141.7 (α-Ph), 128.2Ϫ124.8 (o/m/p-Ph),
1.6, 70.8 (CHPh), 66.6 (CHCH N), 64.3, 63.8 (CH N), 46.0, 45.7
NCH ), 42.4, 41.3 (CH CHPh).
B(1k)(2e):
H
NMR
3
ditional solvent, and kept overnight. After uptake in CDCl , a dia-
(
3
1
3
3
stereomeric ratio (6:37) for 5/6 was observed by C NMR. (c) Me-
2
OTf was added to dichloromethane (1 mL), and the mixture was
cooled to freezing in N (l)/pentane. The borate (70 mg) was added
2
2
3
2
1
1
13
Ϫ
3
): 19.1, 9.1. Ϫ C NMR
and the sample was kept at freezing point for 1 h. After warming
to room temperature and evaporation of the solvent, a dia-
(
3
7
2
2
1
3
stereomeric ratio (3:1) for 5/6 was observed by C NMR. (d) Me-
(
3
2
OTf was added to dichloromethane (0.35 mL), and the borate
General Methylation Procedure: Unless stated otherwise, each
methylation described below was performed as follows. Freshly pre-
pared borate ester was dissolved in dichloromethane (10 mL/mmol)
and, when applicable, cooled to Ϫ78 °C. Methyl trifluoromethane-
sulfonate (MeOTf, 1.00 equivalents) was added, and the mixture
was stirred for 1 h. After warming to room temperature, dichloro-
methane was evaporated in vacuo, water (5 mL) was added to bring
about hydrolysis of the borate ester, and the pH was brought to 1
with concentrated aqueous HCl. After washing with diethyl ether,
the solvent was removed in vacuo and boronic acid was removed
as the trimethyl ester by coevaporation with MeOH. Water (5 mL)
(
65 mg) was added at room temperature under vigorous stirring.
After evaporation of the solvent, a diastereomeric ratio (63:37) for
5
13
/6 was observed by C NMR. (e) The solid was suspended in
pentane, and MeOTf was added at Ϫ78 °C. Small portions of pre-
cooled dichloromethane were added over 2 h until a clear mixture
was obtained. After warming to room temperature and evaporation
of the solvent, a diastereomeric ratio (3:1) for 5/6 was observed by
13
C NMR.
Methylation of B(1a)(2f): (R,R)-Hydrobenzoin was used. The meth-
ylation was performed at Ϫ78 °C. Selected resonances for
was added and the pH was brought to 12 with 4  NaOH. The B(1a)(2f) · MeOTf (mix of diastereomers, only product observed):
1
1
13
solvent was removed in vacuo and the residue was extracted with
3
B NMR (64.2 MHz, CDCl ): 6.2. Ϫ C NMR (50.3 MHz,
dichloromethane, which was dried with MgSO
solvent removed in vacuo.
4
, filtered, and the
3 3
CDCl ): 84.6, 84.3 (2:1, PhCHOB), 56.2 (NCH ). After hydrolysis
and consecutive purification, 2f · MeOTf (32%) was obtained. An
1
ee of about 30% was determined by H NMR using the diamag-
Methylation of B(1a)(2e): (R,R)-Hydrobenzoin was used. The
methylation was performed at Ϫ78 °C. Spectra for
B(1a)(2e) · MeOTf (mix of diastereomers 5/6, only product ob-
netic chiral shift reagent (S)-1-(9-anthryl)-2,2,2-trifluoroethanol.
1
Spectra for 2f · MeOTf: H NMR (200 MHz, CDCl
3
): 4.02 (m, 1
), 3.21Ϫ2.93 (m, 2 H,
), 2.60Ϫ2.24 (m, 2 H, CH NEt ),
), 1.35 (br t, 6 H, NMeCH CH
, JCH2CH3 ϭ 7.1 Hz).
1
H, CHOH), 3.50 (br q, 4 H, NMeCH
CH NMeEt ), 3.10 (s, 3 H, NCH
.56 (q, 4 H, NCH CH
CH2CH3 ϭ 7.1 Hz), 0.98 (t, 6 H, NCH
2 3
CH
served): H NMR (200 MHz, CDCl
4
3
): 7.26Ϫ7.06 (br m, 10 H, Ph),
CHOB),
], 2.73Ϫ2.65 [m, H,
]. Ϫ 11B NMR (64.2 MHz, CDCl ): 10.8. Ϫ 13C NMR
50.3 MHz, CDCl ): 140.9, 140.5 (3:1, α-Ph), 128.6Ϫ125.8 (o/m/p-
Ph), 85.3, 85.1 (PhCHOB), 69.9, 69.8 [3:1, CH N(CH ], 63.6, 63.3
3:1, CH CHOB), 60.3, 59.9 [1:3, CH N(CH ], 54.6 [N(CH ],
5.1 [N(CH ]. After hydrolysis and consecutive purification,
e · MeOTf (7, 76%, contaminated with 6% 1a) was obtained. After
2
2
3
2
2
.75Ϫ4.50 (br m, 2 H, PhCHOB), 3.67, 3.61 (2 m: 1:3, CH
.26Ϫ3.07 [m, 11 H, CH N(CH
N(CH
2
2
J
Ϫ
2
3
2
3
,
3
2
3
)
3
8
3
3
2
CH
): 64.7 (CH
and NMeCH CH
), 11.9, 8.0 (CH CH ).
3
CH
2
3
)
2
3
1
3
C NMR (50.3 MHz, CDCl
CHOH), 57.8, 57.7 (CH NEt
7.0 (NCH CH
3
3
2
NMeEt
2
), 61.8
),
(
3
(
2
2
2
3
), 56.5 (NCH
3
2
3 3
)
4
2
3
2
(
2
2
3
)
2
3 3
)
4
3 2
)
Methylation of B(1j)(2c): Reaction performed at Ϫ78 °C. Selected
2
1
3
resonances for B(1j)(2c) · MeOTf: H NMR (200 MHz, CDCl ):
recrystallization from ethyl acetate, pure 7 (57%) was obtained as
]. Ϫ 11B NMR (64.2 MHz, CDCl
): 53.8 [N(CH ]. Hydrolysis was per-
Spectra for
O): 4.10 (t, 2 H, CH N), 3.56
]. Ϫ C NMR (50.3 MHz,
OH), 55.3 [N(CH ].
13
3
.13 [s, N(CH
NMR (50.3 MHz, CDCl
formed without consecutive purification.
3
)
3
3
): 17.9. Ϫ
C
1
a white solid. An ee of 48% was determined by H NMR using the
3
3 3
)
diamagnetic chiral shift reagent (S)-1-(9-anthryl)-2,2,2-trifluoro-
ethanol. Pure 1a (84%) was recovered from the ether washings. MS
Ϫ
1
2
c · MeOTf: H NMR (200 MHz, D
t, 2 H, CH OH), 3.26 [s, 9 H, N(CH
O): 68.9 (CH N), 57.0 (CH
2
2
ϩ
1
(
ESI ): 161 (7). Ϫ Spectra for 7: H NMR (200 MHz, CDCl
3
):
13
(
D
2
3 3
)
4
.36 (br s, 1 H, OH), 4.10 (m, 1 H, CHOH), 3.47, 3.22 [2 dd, 2 H,
2
2
2
3 3
)
3
2
2 3 3
CH N(CH ) , JCHCH2 ϭ 1.5, 9.8 Hz, Jgem ϭ Ϫ13.3 Hz], 3.25 [s,
3
9
5
H, N(CH
3
2
)
3
], 2.30, 2.21 [2 dd, 2 H, CH
2
N(CH
3 2
) ,
J
CHCH2
ϭ
Methylation of B(1j)(2e): Reaction performed at Ϫ78 °C. Hydroly-
]. 13C NMR sis was performed without consecutive purification. A mixture of
CF
J ϭ 322 Hz), 70.0 mono- and dimethylated product was obtained.
.9, 4.6 Hz, Jgem ϭ Ϫ12.3 Hz], 2.24 [s, 6 H, N(CH )
3 2
1
(50.3 MHz, CDCl
3
): 120.5 (q, CF
3
,
Eur. J. Org. Chem. 1999, 1775Ϫ1786
1785

J. Huskens, M. T. Reetz
FULL PAPER
M. Wang, Angew. Chem 1995, 107, 2605Ϫ2606; Angew. Chem.
Double methylation of B(1j)(2e): Reaction performed at Ϫ78 °C.
[3f]
Int. Ed. Engl. 1995, 34, 2395. Ϫ
H. Imado, T. Ishizuka, T.
Hydrolysis was performed without consecutive purification. Spec-
[3g]
Kunieda, Tetrahedron Lett. 1995, 36, 931Ϫ934. Ϫ
K.
tra for 2e · 2 MeOTf: 1H NMR (200 MHz, D
O): 3.25 [s, 18 H,
N and CHOH). Ϫ 13C NMR
N), 65.5 (CHOH), 55.6 [N(CH ].
2
Nordström, C. Moberg, A. Heumann, J. Organomet. Chem.
[3h]
N(CH
3
)
3
], 3.15 (m, 5 H, CH
O): 71.6 (CH
2
1996, 525, 233Ϫ238. Ϫ
E. Vedejs, O. Daugulis, S. T. Diver,
[3i]
J. Org. Chem. 1996, 61, 430Ϫ431. Ϫ
M. Kinugasa, T. Ha-
(
50.3 MHz, D
2
2
3 3
)
rada, A. Oku, J. Am. Chem. Soc. 1997, 119, 9067Ϫ9068. Ϫ
[3j]
K. Matsuki, H. Inoue, M. Takeda, Tetrahedron Lett. 1993,
4, 1167Ϫ1170.
3
[
[
[
[
4]
5]
6]
M. T. Reetz, J. Rudolph, R. Mynott, J. Am. Chem. Soc. 1996,
Acknowledgments
1
18, 4494Ϫ4495.
J. Huskens, R. Goddard, M. T. Reetz, J. Am. Chem. Soc. 1998,
J. H. gratefully acknowledges a Marie Curie fellowship from the
European Commission (contract number ERBFMBICT961429).
1
20, 6617Ϫ6618.
M. Van Duin, J. A. Peters, A. P. G. Kieboom, H. Van Bekkum,
Tetrahedron 1985, 41, 3411Ϫ3421.
7] [7a]
E. E. Turner, M. M. Harris, Q. Rev., Chem. Soc. 1947, 1,
[
[
1]
[7b]
M. T. Reetz, Pure Appl. Chem. 1996, 68, 1279Ϫ1283.
299Ϫ330. Ϫ
M. M. Harris, in Progress in Stereochemistry,
2] [2a]
F. Terradas, M. Teston-Henry, P. A. Fitzpatrick, A. M. Kli-
Vol. 2 (Eds.: W. Klyne, P. B. D. de la Mare), Butterworths, Lon-
[
2b]
[7c]
banov, J. Am. Chem. Soc. 1993, 115, 390Ϫ396. Ϫ
J. B. Jones,
K. Faber, Biotrans-
formations in Organic Chemistry, Springer, Berlin, 1992. Recent
don, 1958, pp. 157Ϫ195. Ϫ
J. Jacques, A. Collet, S. H. Wilen,
[2c]
Aldrichimica Acta 1993, 26, 105Ϫ112. Ϫ
Enantiomers, Racemates, and Resolutions, Krieger, Malabar,
[7d]
1994. Ϫ
E. L. Eliel, S. H. Wilen, L. N. Mander, Stereochem-
[
2d]
examples:
1, 3332Ϫ3341. Ϫ
Monti, A. Galli, Tetrahedron Lett. 1996, 37, 4247Ϫ4250. Ϫ
D. A. Griffith, S. J. Danishefsky, J. Am. Chem. Soc. 1996, 118,
R. Ch eˆ nevert, M. Dickman, J. Org. Chem. 1996,
istry of Organic Compounds, Wiley, New York, 1994; and refer-
ences cited therein.
[2e]
6
G. Speranza, P. Manitto, G. Fontana, D.
[2f]
[8]
[9]
W. H. Pirkle, D. L. Sikkenga, M. S. Pavlin, J. Org. Chem. 1977,
42, 384Ϫ387.
[
2g]
9
526Ϫ9538. Ϫ
Tetrahedron: Asymmetry 1997, 8, 547Ϫ557. Ϫ
Parish, R. B. Herbert, M. Hughes, D. F. Monk, Tetrahedron:
G. Egri, E. Fogassy, L. Nov a´ k, L. Poppe,
W. H. Pirkle, T. G. Burlingame, S. D. Beare, Tetrahedron Lett.
1968, 56, 5849Ϫ5852.
[
2h]
J. Gill, J. H.
[10]
W. Zhang, E. N. Jacobsen, J. Org. Chem. 1991, 56, 2296Ϫ2298.
M. Palucki, P. J. Pospisil, W. Zhang, E. N. Jacobsen, J. Am.
Chem. Soc. 1994, 116, 9333Ϫ9334.
[2i]
[11]
Asymmetry 1997, 8, 2227Ϫ2230. Ϫ
G. Guanti, E. Narisano,
R. Riva, Tetrahedron: Asymmetry 1997, 8, 2175Ϫ2187.
[
3] [3a]
[3b]
[12]
[13]
R. S. Ward, Chem. Soc. Rev. 1990, 19, 1Ϫ19. Ϫ
C. S.
Z.-M. Wang, K. B. Sharpless, J. Org. Chem. 1994, 59,
8302Ϫ8303.
[
3c]
Poss, S. L. Schreiber, Acc. Chem. Res. 1994, 27, 9Ϫ17. Ϫ
S.
D. E.
[
3d]
R. Magnuson, Tetrahedron 1995, 51, 2167Ϫ2213. Ϫ
J. Dale, J. Chem. Soc. 1961, 910Ϫ922.
Received February 12, 1999
Ward, Y. Liu, D. How, J. Am. Chem. Soc. 1996, 118,
3e]
3
025Ϫ3026. Recent examples: [ D. Seebach, G. Jaeschke, Y.
[O99083]
1786
Eur. J. Org. Chem. 1999, 1775Ϫ1786