Diastereoisomer interconversion in chiral biphepPtX2 complexes

Article abstract of DOI:10.1021/om000629a

Reaction of biphepPt(COa) (biphep = 2,2'-bis(diphenylphosphino)-l,l/-biphenyl) with BINOL or HN(TCHPhCHPhOH (TfNO) yielded square-planar biphepPtX2 (X2 = BINOL, N(Tf)CHPhCHPhO) complexes as a mixture of diastereomers (-1:1). BiphepPtCk also reacted with Na2BINOL to generate biphepPt(BINOL) as a 1:1 mixture of diastereomers. With racemic BINOL or TfNO ligands, the mixtures were prone to isomerize to thermodynamic diastereomer mixtures (BINOL, 95:5; TfNO, +- 97:3) by an X2-X2 ligand-ligand exchange mechanism that was rapid at room temperature. With enantiopure ligands the X2-X2 ligandligand exchange process was degenerate and nonproductive. However, thermolysis of 1:1 mixtures of enantiopure biphepPt(BINOL) diastereomers (92-122 °C) cleanly established thermodynamic equilibrium by a process that involves biphenyl atropisomerism (AH= 27(2) kcal mol 1, AS= -5(5) eu). Two mechanisms for this process were considered, concerted stereoinversion via a planar seven-membered metallacycle, and one-arm-off prior to a biphenyl isomerization (anti disposed PPh2 units). In pyridine, a third mechanism for atropisomerism was identified and proposed to involve a five-coordinate pyridine intermediate (not observed) with an enhanced phosphine dissociation rate. Pyridine lowered the ; isomerization temperature of enantiopure complexes by -50 °C. X-ray structures of the thermodynamically favored biphepPt(TfNO) ((±)-4a) and the thermodynamically less favored biphepPt(BINOL) (A(S)-5b) diastereomers were obtained, and a stereochemical model to explain the diastereoselectivity was formulated.

Full text of DOI:10.1021/om000629a

4376
Organometallics 2000, 19, 4376-4384
Dia ster eoisom er In ter con ver sion in Ch ir a l Bip h ep P tX2
Com p lexes
Melanie D. Tudor, J ennifer J . Becker, Peter S. White, and Michel R. Gagne´*
Department of Chemistry CB#3290, University of North Carolina,
Chapel Hill, North Carolina 27599-3290
Received J uly 20, 2000
Reaction of biphepPt(CO3) (biphep ) 2,2′-bis(diphenylphosphino)-1,1′-biphenyl) with
BINOL or HN(Tf)CHPhCHPhOH (TfNO) yielded square-planar biphepPtX2 (X2 ) BINOL,
N(Tf)CHPhCHPhO) complexes as a mixture of diastereomers (∼1:1). BiphepPtCl2 also reacted
with Na2BINOL to generate biphepPt(BINOL) as a 1:1 mixture of diastereomers. With
racemic BINOL or TfNO ligands, the mixtures were prone to isomerize to thermodynamic
diastereomer mixtures (BINOL, 95:5; TfNO, >97:3) by an X2-X2 ligand-ligand exchange
mechanism that was rapid at room temperature. With enantiopure ligands the X2-X2 ligand-
ligand exchange process was degenerate and nonproductive. However, thermolysis of 1:1
mixtures of enantiopure biphepPt(BINOL) diastereomers (92-122 °C) cleanly established
thermodynamic equilibrium by a process that involves biphenyl atropisomerism (∆Hq ) 27(2)
kcal mol-1, ∆Sq ) -5(5) eu). Two mechanisms for this process were considered, concerted
stereoinversion via a planar seven-membered metallacycle, and one-arm-off prior to a
biphenyl isomerization (anti disposed PPh2 units). In pyridine, a third mechanism for
atropisomerism was identified and proposed to involve a five-coordinate pyridine intermediate
(not observed) with an enhanced phosphine dissociation rate. Pyridine lowered the
isomerization temperature of enantiopure complexes by ∼50 °C. X-ray structures of the
thermodynamically favored biphepPt(TfNO) ((()-4a ) and the thermodynamically less favored
biphepPt(BINOL) (λ(S)-5b) diastereomers were obtained, and a stereochemical model to
explain the diastereoselectivity was formulated.
In tr od u ction
What is the mechanism by which 2,2′-bis(diphenyl-
phosphino)biphenyl (biphep) interconverts its chiral
skew conformations when coordinated to a transition
metal fragment? We pose this question in the context
of recent clever experiments that take advantage of a
chiral metal fragment’s ability to affect the population
of biphep conformations and thereby access selective
catalysts for the asymmetric transfer hydrogenation of
ketones.1 These experiments in effect trick biphep into
acting like BINAP, the well-known and highly selective
diphosphine ligand. This despite the fact that biaryl
rotation in free biphep is too fast for enantiomer
resolution (∆Gq ) 22 ( 1 kcal mol-1, 125 °C).2
In the Mikami/Noyori experiment, addition of enan-
tiopure 1,2-diphenylethylenediamine to biphepRuCl2-
(dmf)n quickly generates a 1:1 mixture of diastereomers
that is stable in CDCl3, but that relaxes to a 3:1 mixture
with a half-life in the 10’s of seconds in 1:2 CDCl3/(CD3)2-
CDOD (eq 1). All data point to biaryl rotation being the
key step in the atrop-interconversion and at rates
consistent with the rotational barrier measured in the
free ligand. Simple on-metal atrop-inversion and one-
arm-off prior to biaryl rotation were proposed as reason-
able mechanisms for the equilibration of the diastere-
omers.
To gain insight into the mechanism of diastereomer
interconversion and to specifically address the issue of
on-metal ligand twisting versus arm-off (free ligand-like)
mechanisms for atrop isomer inversion, we have exam-
ined the isomerization of substitution-inert chiral bi-
phepPtX2 complexes, where X2 is BINOL or a chelating
O,NTf ligand. Unlike octahedral Ru(II) complexes,
square-planar Pt(II) complexes generally react via as-
sociative mechanisms. We rationalized that this distinc-
tion would enable one-arm-off and on-metal inversion
processes to be distinguished (both intramolecular
processes). Moreover, chiral P2PtX2 complexes would
also distinguish and rank the relative importance of
competing ligand exchange (no biaryl rotation, an
intermolecular isomerization) and atrop-inversion mech-
anisms for diastereomer equilibration (e.g., Scheme 1).
Though the products from the intra- and intermolecular
isomerization are enantiomeric, we loosely refer to both
processes as “isomerization” mechanisms.
Resu lts a n d Discu ssion
(1) Mikami, K.; Korenaga, T.; Terada, M.; Ohkuma, T.; Pham, T.;
Noyori, R. Angew. Chem., Int. Ed. 1999, 38, 495-497.
(2) Desponds, O.; Schlosser, M. Tetrahedron Lett. 1996, 37, 47-48.
(1) Liga n d a n d P r ecu r sor Syn th esis. Given the
capricious nature of the known methods for synthesizing
10.1021/om000629a CCC: $19.00 © 2000 American Chemical Society
Publication on Web 09/22/2000

Chiral BiphepPtX2 Complexes
Sch em e 1
Organometallics, Vol. 19, No. 21, 2000 4377
Ta ble 1. 31P Ch em ica l Sh ift a n d Cou p lin g
Con sta n ts for Selected Com p lexes (CD2Cl2)
31P chemical shift
(ppm)
J P-Pt (J P-P)
complex
(Hz)
1
2
7.9
2.2
7.2
3.5
8.9
4.8
2.9
2.6
3610
3570
4a (major)
3760 (31.3)
3210 (31.3)
3725 (32.3)
3210 (32.3)
3670
4b (minor)
5a (major)
5b (minor)
3643
the biphep ligand,2,3 we have examined several alterna-
tive syntheses of this ligand. First explored was the
oxidative coupling of the monoanion obtained by the
lithiation of Ph3PO with sec-BuLi in THF (-78 °C) (eq
2). Adding this deep orange solution to a bright blue
shifts and J P-Pt values are collected in Table 1 for all
reported complexes.
(2) Syn th esis of Bip h ep P tX2 Com p lexes. The
reaction of 2 with 0.9 equiv of (()-3 in CD2Cl2 yielded
the desired product after 1 h as a ∼1:1 ratio of diaster-
eomers, (()-4a and (()-4b (eq 4), along with traces of 1
and 2. The isomer ratio remained constant in solution
solution of Cu(OPiv)2 (0 °C) yielded the coupled diphos-
phine oxide ligand. This Cu(II)-based methodology is
superior to the copper bronze-mediated Ullman coupling
of 2-iodotriphenylphosphine oxide, as it reduces the
number of synthetic steps from 2 to 1 and the time to
multigram quantities of ligand from Ph3PO to 2 days
from 1 week. The crude biphep oxide was then reduced
with Cl3SiH in refluxing xylene to yield the desired
ligand contaminated with unreacted PPh3. Biphep was
obtained in 30% overall yield after a single recrystalli-
zation from toluene/ethanol. Two alternative procedures
4
based on the (dppe)NiCl2-catalyzed coupling of ClPPh2
or HPPh25 with the ditriflate of 2,2′-biphenol were found
to directly yield biphep, but unidentified byproducts and
long reaction times made these approaches less ame-
nable to scale-up. During the preparation of this article
Hayashi reported a Pd(OAc)2/dppb-catalyzed procedure
for coupling the ditriflate of 2,2′-biphenol with HPPh2O
in high yields that is amenable to scale-up.6
The reaction of biphep with (COD)PtCl2 was straight-
forward and provided biphepPtCl2, 1, which was recrys-
tallized from CH2Cl2/MeOH (eq 3). The desired biphepPt-
(CO3), 2, could be readily generated from the dichloride
and Ag2CO3 in wet CH2Cl2 by the procedure of An-
drews.7 Like 1, the carbonate was soluble in CH2Cl2 and
was completely characterized. Phosphorus chemical
over 3 days at ambient temperature provided that traces
of 2 were present. The reduced rate of λ- and δ-skew
interconversion in biphep (cf. dppe)8 is consistent with
the observation of two diastereomers in the 31P NMR;
at a minimum the biphep δ/λ interconversion is slow
on the NMR time scale. Consistent with our previous
characterization of dppePtNO complexes,8 we assign the
upfield set of doublets (3.5 and 4.8 ppm) to the phos-
phines in (()-4a and (()-4b that are trans to the
alkoxide (J Pt-P typically 3200-3250 Hz) and the down-
field resonances (7.2 and 8.9 ppm) to those trans to the
triflamide (J Pt-P typically 3600-3700 Hz), Table 1. The
1:1 ratio of enantiopure diastereomers, λ(1R,2S)-4a and
δ(1R,2S)-4b, was invariant and the mixture could be
isolated using 1 equiv of (1R,2S)-3 (eq 5) in 40% yield
after recrystallizing from CH2Cl2/hexanes.
(3) Desponds, O.; Schlosser, M. J . Organomet. Chem. 1996, 507,
257-261.
(4) Ager, D. J .; East, M. B.; Eisenstadt, A.; Laneman, S. A. Chem.
Commun. 1997, 2359-2360.
(5) Cai, D.; Payack, J . F.; Bender, D. R.; Hughes, D. L.; Verhoeven,
T. R.; Reider, P. J . J . Org. Chem. 1994, 59, 7180-7181.
(6) Ogasawara, M.; Yoshida, K.; Hayashi, T. Organometallics 2000,
19, 1567-1571.
(7) Andrews, M. A.; Gould, G. L.; Klooster, W. T.; Koenig, K. S.; Voss,
E. J . Inorg. Chem. 1996, 35, 5478-5483.
(8) Becker, J . J .; White, P. S.; Gagne´, M. R. Inorg. Chem. 1999, 38,
798-801.

4378 Organometallics, Vol. 19, No. 21, 2000
Tudor et al.
toluene/hexane. Recrystallization from CH2Cl2 provided
a single crystal of λ(S)-5b, which is described in detail
in section 5. Dissolving a portion of the X-ray crystals
confirmed (31P NMR) that the crystals contained only
the minor isomer, and thus the structure corresponded
to λ(S)-5b.
(3) In flu en ce of Ad d ed X2 Liga n d s on Dia ster eo-
m er In ter con ver sion . Since one mechanism for dia-
stereomer interconversion in chiral biphep complexes
is exchange of the chiral X2 ligand for one of opposite
absolute stereochemistry (no biaryl rotation, Scheme 1),
we undertook a series of NMR experiments wherein like
and unlike ligands were added to the ∼1:1 mixture of
racemic and enantiopure diastereomers obtained in
section 2.
Addition of 0.2 equiv of (()-3 to a mixture of (()-4a
and (()-4b that was otherwise stable to isomerization
in an NMR tube at room temperature caused one of the
diastereomers to slowly increase at the expense of the
other. After 12 h, the diastereomer that was character-
ized by the upfield set of two resonances was undetect-
able by 31P NMR; the major isomer (4a , vide infra) is
highly favored (>97:3) at equilibrium. Thus, the ratio
of the racemates 4a /4b is only constant in the absence
of excess ligand, and 0.9 equiv (or less) of 3 was utilized
to make a 1:1 mixture of diastereomers. A single
diastereomer of (()-4a was obtained as indicated in eq
4, except with 1.1 equiv of (()-3 and overnight gentle
heating. X-ray quality crystals of (()-4a were obtained
by slow diffusion of Et2O into a CH2Cl2 solution (see
section 5).10 Similarly, addition of 20 mol % of (()-
BINOL to (()-5a /5b led to the rapid conversion of the
mixture to one major isomer in a ∼95:5 ratio, as
determined by 31P NMR. The major isomer of biphepPt-
(BINOL) was normally obtained by combining (()-
BINOL and 2 in CH2Cl2, which leads to the slow
precipitation of the major isomer from solution. Since
the kinetic diastereoselectivity is known to be low from
experiments with (S)-BINOL (section 2), the diastere-
omers must isomerize rapidly upon formation. These
experiments demonstrate that excess ligand catalyzes
diastereomer exchange by enabling ligand-ligand ex-
change to match the biphep and X2 ligand stereochem-
istries.
A nearly 1:1 ratio of biphepPt(BINOL) diastereomers
was obtained by a salt elimination protocol that relied
on a soluble source of Na2BINOL obtained from in situ
deprotonation of BINOL with NaOtBu9 (eq 6). In
toluene, NaCl elimination from (()-Na2BINOL‚tBuOHx
and biphepPtCl2 is complete within 30 min, and the
BINOLate diastereomers ((()-5a and (()-5b) are ob-
tained in 82% yield after filtration in CH2Cl2; 31P
chemical shifts and J Pt-P coupling constants are com-
piled in Table 1. Enantiopure mixtures of these diaster-
eomers were obtained by an analogous procedure using
(S)-Na2BINOL and biphepPtCl2, or from 2 and (S)-
BINOL (eq 7). The thermodynamically less preferred
Unlike the above examples, addition of (1R,2S)-3 and
(S)-BINOL to the enantiopure mixtures of δ(1R,2S)-4a ,
λ(1R,2S)-4b, and δ(S)-5a , λ(S)-5b, respectively, did not
change the ratio of diastereomers, even after overnight
heating at 70 °C in toluene. On the other hand, 1 equiv
of the opposite ligand enantiomers, (1S,2R)-3 (eq 8) and
(R)-BINOL (eq 9), respectively, led to isomerization and
thermodynamic product ratios. Figure 1 shows a revers-
ible first-order treatment11 of the conversion of λ(S)-5b
to the major diastereomer (λ(R)-5a ) by the action of 1
equiv of (R)-BINOL at 45 °C, as monitored by 31P NMR12
(10) Dissolving the X-ray crystals confirmed that it was the major
diastereomer that crystallized.
(11) The integrated rate expression for a standard reversible first-
order process (A T B) is
isomer (λ(S)-5b, vide infra) can be obtained in pure form
(33% yield, 50% max) by fractional precipitation from
(9) The salt can be generated in either THF or toluene with a trace
of THF. The presence of tBuOH is apparently important to solubility,
as the pure salt is only sparingly soluble in THF once the tBuOH has
been removed in vacuo.
Plotting the left-hand side of the expression versus time yields a line
with a slope of k, the forward rate constant. Laidler, K. J . Chemical
Kinetics, 3rd ed.; Harper & Row: New York, 1987.

Chiral BiphepPtX2 Complexes
Organometallics, Vol. 19, No. 21, 2000 4379
Ta ble 2. Ra te a n d Equ ilibr iu m Da ta for th e
Th er m a l Con ver sion of λ(S)-5b to δ(S)-5a
equilibrium
(major:minor)
temp (°C)
k (s-1
)
Keq
45a
92.6
108.9
113.6
122.8
94.4:5.6
94.2:5.8
93.7:6.3
93.1:6.9
92.7:7.3
16.9(5)
16.2(4)
14.9(4)
13.5(5)
12.7(5)
2.3 × 10-5
1.2 × 10-4
1.8 × 10-4
4.6 × 10-4
a
Data obtained from the equilibration experiment described in
Figure 1. It does not fit on the van’t Hoff plot since excess BINOL
H-bonds to the product and perturbs the equilibrium.33
F igu r e 1. Reversible first-order treatment for the conver-
sion of λ(S)-5b to λ(R)-5a (eq 9, see ref 11) with 1 equiv of
(R)-BINOL at 45 °C in chlorobenzene. The flat lines are
controls with 1 equiv of (S)-BINOL and no additive.
[minor]0 ) 15.0 mM; [major]t)0 ) 8.2 mM; [minor]eq ) 0.9
mM; and [major]eq ) 14.1 mM.
F igu r e 2. Eyring analysis for the thermal conversion of
λ(S)-5b to δ(S)-5a .
isomerization mechanisms were operative.13 Heating a
∼1:1 mixture of δ(1R,2S)-4a and λ(1R,2S)-4b in toluene-
d8 to 100 °C, with or without added (1R,2S)-3, caused
the minor isomer to disappear, though extensive de-
composition masked whether the minor isomer was
converting to the major or it was decomposing. Since
the major diastereomer, (()-4a (>97:3), did not decom-
pose at 100 °C, this suggested that the minor diastere-
omer either decomposed before or competitive with
isomerization.
On the other hand, heating a chlorobenzene solution
of λ(S)-5b to elevated temperature (92-122 °C) cleanly
converted it to the major isomer (∼94:6). Rate data
obtained by 31P NMR indicated that the process obeyed
standard first-order relaxation to equilibrium kinetics.
First-order rate constants for the minor to major
conversion and equilibrium ratios are collected in Table
2. Eyring analysis of rate data obtained over a ∼30 °C
temperature range yielded ∆Hq and ∆Sq of 27(2) kcal
mol-1 and -5(5) eu, respectively (Figure 2).14 A van’t
Hoff analysis of the equilibrium data obtained from the
thermolysis experiments provided ∆H° ) -2.2(4) kcal
mol-1 and ∆S° ) -0.6(10) eu for the reaction.
(k ) 6.6 × 10-4 s-1, ∆Gq ) 23 kcal mol-1, 94.4:5.6 at
equilibrium). Since the minor triflamide diastereomer
is not observed at equilibrium, an irreversible first-order
rate treatment was utilized; k ) 3.4 × 10-4 s-1 (35 °C,
∆Gq ) 23 kcal mol-1) for minor to major (eq 8).
When 1 equiv of (R)-BINOL is used to isomerize the
enantiopure diastereomers, the equilibrium ratio of (R)-
and (S)-BINOL is 1, and the equilibrium constant for
eq 9 can be directly determined from the ratio of
diastereomers. Keq is thus 16.9 at 45 °C for biphepPt-
(BINOL) and >32 for biphepPt(ONTf) at 35 °C.
These experiments indicate that BINOL and 3 pro-
vide a strong thermodynamic preference for a single
biphep skew conformation. Furthermore, these data are
most consistent with a diastereomer interconversion
mechanism that is not biphep inversion, but simple
ligand-ligand exchange.
(4) Th er m olysis of Dia ster eom er Mixtu r es. Al-
though ligand-ligand exchange dominates when both
enantiomers of the chiral ligands are available, the
question of biaryl inversion in the absence of added
ligand remained unanswered. The enantiopure diaste-
reomer mixtures were thermolyzed to determine if other
(5) Str u ctu r a l Ra tion a le for th e Th er m od yn a m ic
Selectivity. To provide a structural rationale for the
(13) Since adventitious traces of ligand in racemic complexes could
catalyze the isomerization through ligand exchange, only enantiopure
diastereomers were studied.
(12) Since biphenyl atropisomerism does not occur under these
conditions, δ(S)-5a , which is already the favored isomer, does not
appreciably react with (R)-BINOL. Equations 8 and 9 thus only
describe one set of reacting enantiomers.
(14) Errors are reported at the 95% confidence level (3σ), assuming
a 10% margin of error on each rate constant, and were calculated by
a nonlinear least-squares method; computer program kindly provided
by Prof. Barry Carpenter of Cornell.

4380 Organometallics, Vol. 19, No. 21, 2000
Tudor et al.
Ta ble 3. Cr ysta llogr a p h ic Da ta for (()-4a a n d
λ(S)-5b
is distorted 13° relative to the P-Pt-P plane and
toward the pseudoaxial P-Phs. The triflamide portion
of 4a is similar to this fragment coordinated to dppePt8
and is characterized by an axial phenyl substituent R
to the nitrogen ligand and an equatorial phenyl group
R to oxygen. These orientations are apparently also
(()-4a ‚H2O‚(C2H5)2O
λ(S)-5b‚3CH2Cl2
empirical
formula
fw
space
group
a
PtP2SC51H40F3NO3‚H2O‚ PtP2C56H40O2‚3CH2Cl2
(C2H5)2O
1083.99
orthorhombic
Pcab (61)
19.1768(8)
24.0196(10)
41.0509(16)
18908.8(13)
16
1255.75
orthorhombic
P212121 (19)
15.6986(5)
15.7021(5)
21.2205(6)
5230.9(3)
4
3
maintained in solution, as J Pt-H coupling to the pseudo-
3
axial H on the carbinol carbon is small, while J Pt-H for
the pseudoequatorial H on the center R to N is large
b
c
3
(36 Hz). A Karplus-like relationship correlating J Pt-H
V, Å3
Z
to the Pt-X-C-H dihedral has been noted.17
The sterically dominant group in the O,NTf chiral
ligand is the SO2 moiety, as one oxygen protrudes into
the open quadrant created by one of the back pointing
pseudoaxial P-phenyl substituents. The thermodynamic
selectivity for the major isomer is readily understood
by considering that the biphep atropisomer would
position a sterically dominant pseudoequatorial P-Ph
into this same quadrant. The major isomer thus ef-
ficiently gears the SO2 portion and the protruding
pseudoequatorial P-Ph.
T, °C
Dc, g/cm3
λ, Å
-100
-100
1.523
1.595
Mo KR (0.71073)
3.12
0.042
0.062
Mo KR (0.71073)
3.08
0.023
0.027
µ, nm-1
a
Rf
b
Rw
GOFc
2.6596
1.1986
a
b
2 1/2
] .
c GOF )
Rf ) ∑(Fo - Fc)/∑Fo. Rw ) [∑w(Fo - Fc)2/∑wFo
[∑w(Fo - Fc)2/(n - p)]1/2, where n ) number of reflections and p )
number of parameters.
Unlike 4a , λ(S)-5b is the thermodynamically less
favored diastereomer and represents the “mismatched”
combination of biphep and BINOL stereochemistries. In
addition to confirming the relative stereochemistry of
the biphep and BINOL fragments, the mismatched Pt-
BINOL complex λ(S)-5b also has an unusual solid-state
structure (Figure 4A).18 In particular, the structure does
not have C2-symmetry, as the biphep ligand is highly
distorted above the square plane in an effort to minimize
the interaction between its forward pointing pseu-
doequatorial P-Ph groups and the 3,3′-C-H’s of BINOL;
that is, both attempt to occupy the same quadrants. To
minimize this interaction, the structure skews the
biphenyl backbone so that one-half of biphep has two
P-Ph substituents roughly evenly disposed about the
square plane (neutral axial/equatorial designations)
while the other half has overemphasized the axial/
equatorial positions to such an extent that one P-Ph
almost lies in the square plane while the other is pulled
well back. Most illustrative of this P-Ph skewing are
the P-Pt-P-Cipso dihedral angles expressed as rota-
tions above and below the P-Pt-P square plane (Figure
4B). In the more “symmetric” half of the molecule,
P-Ph(23) is rotated 73.8° above the plane while P-Ph-
(29) is rotated 44° below. In contrast, the more “dis-
torted” phosphine has P-Ph(53) and P-Ph(47) rotated
6.2° above and 113° below the plane, respectively. The
sum of these dihedral angles is similar for each phos-
phorus, 117.8 and 119.2°, respectively, indicating that
the axial/equatorial distortions are primarily torsional
in nature. In addition to twisting, there is also signifi-
cant bond distance and angle variability in the struc-
ture. In particular, P(1)-Pt-O(22) (85.60(5)°) is signifi-
cantly more acute than P(2)-Pt-O(1) (95.70(5)°), though
F igu r e 3. ORTEP representation of the X-ray structure
of (()-4a , (λ(1S,2R) shown). Thermal ellipsoids are at the
50% probability level. Selected bond distances (Å) and bond
angles (deg): Pt-P1 ) 2.2505(18), Pt-P2 ) 2.2380(19),
Pt-N1 ) 2.094(6), Pt-O4 ) 2.024(5), P1-Pt-P2 ) 93.03-
(7), O1-Pt-N1 ) 81.25(21), P2-Pt-O4 ) 87.45(15), P1-
Pt-N4 ) 99.55(17), C51-C56-C61-C66 ) 65.1(10).
high thermodynamic selectivity observed for the com-
bination of a single chiral ligand and a single biphep
skew conformation, X-ray structures of (()-4a and λ(S)-
5b were obtained. Acquisition and structure refinement
parameters are collected in Table 3; complete tables of
metrical parameters are available in the Supporting
Information. (()-4a crystallizes into a centrosymmetric
space group with two independent, but nearly identical
molecules in the asymmetric unit. As indicated in Figure
3 for an arbitrary molecule, the structure of the major
isomer corresponds to 4a . The biphep portion of the
molecule is unexceptional and looks like many BINAP15
16
(17) (a) Cerasino, L.; Williams, K. M.; Intini, F. P.; Cini, R.; Marzilli,
L. G.; Natile, G. Inorg. Chem. 1997, 36, 6070-6079. (b) Appleton, T.
G.; Hall, J . R. Inorg. Chem. 1971, 10, 1717-1725. (c) Erickson, L. E.;
McDonald, J . W.; Howie, J . K.; Clow, R. P. J . Am. Chem. Soc. 1968,
90, 6371-6382. (d) Yano, S.; Tukada, T.; Saburi, M.; Yoshikawa, S.
Inorg. Chem. 1978, 17, 2520-2526. (e) Erickson, L. E.; Sarneski, J .
E.; Reilley, C. N. Inorg. Chem. 1975, 14, 3007-3017.
and a recent biphepPdCl2 structure with a biphenyl
dihedral angle of 65.1(10)°. Torsional strain in the
seven-membered chelate causes the biphep ligand to
skew relative to the P-Pt-P plane (C61-C56 vector is
angled 63° to the P-Pt-P plane). The N-Pt-O plane
(18) For other examples of unusual BINOL structures as a conse-
quence of stereochemical mismatching, see: (a) Brunkan, N. M.; White,
P. S.; Gagne´, M. R. Angew. Chem., Int. Ed. 1998, 37, 1579-1582. (b)
Brunkan, N. M.; White, P. S.; Gagne´, M. R. J . Am. Chem. Soc. 1998,
120, 11002-11003.
(15) Ohta, T.; Takaya, H.; Noyori, R. Inorg. Chem. 1988, 27, 566-
569.
(16) Ogasawara, M.; Yoshida, K.; Hayashi, T. Organometallics 2000,
19, 1567-1571.

Chiral BiphepPtX2 Complexes
Organometallics, Vol. 19, No. 21, 2000 4381
Sch em e 2
Sch em e 3
F igu r e 4. (A) ORTEP representation of λ(S)-5b (thermal
ellipsoids at the 50% probability level). Selected bond
distance (Å) and angles (deg) not discussed in the text:
P(1)-Pt-P(2) ) 91.456(24), O(1)-Pt-O(22) ) 87.67(7). (B)
Chem3D picture highlighting the biphep distortion. The
non-oxygen-containing fragment of BINOL is removed for
clarity.
the P(1)-Pt (2.2368(6)) and O(22)-Pt (2.0706(17)) bond
distances (Å) are longer than those on the “distorted”
side of the molecule (2.2180(7) and 2.0353(17) Å, re-
spectively). The biphep and BINOL dihedral angles
describing the atrop-isomerism are 63.4(4)° and 57.8-
(4)°, respectively, and the O-Pt-O plane is distorted
10° from the P-Pt-P plane and toward the pseudoaxial
P-Ph substituents (counterclockwise). Since a singlet
is observed in the 31P NMR, the structural distortions
must be either time averaged or unique to the solid
state. The δ(S)-5a diastereomer is presumably thermo-
dynamically favored because the BINOL projects its
3,3′-positions toward the less congested P-Ph pseudo-
axial positions.19,20
(6) Mech a n istic P ossibilities. Chiral ligand addi-
tive studies (section 3) showed that ligand-ligand
exchange with enantiomeric ligands was the most rapid
mechanism for diastereomer interconversion and that
biphep stereoinversion was not competitive. However,
since enantiopure X2 ligand exchange processes cannot
be responsible for the thermal isomerization of λ(S)-5b
to δ(S)-5a (∼94:6) (section 4), this requires net biphep
atrop-inversion by a second mechanism. As noted by
Mikami and Noyori for octahedral Ru complexes,1 two
reasonable but kinetically indistinguishable mecha-
nisms are possible, both of which also apply to biphep-
PtX2: dissociation induced biaryl rotation (Scheme 2),
and on-metal concerted inversion (Scheme 3). Both
mechanisms have weak points, though, as the dissocia-
tive pathway breaks a strong P-Pt bond to generate a
three-coordinate PPtX2 intermediate (with or without
solvent coordination), and the on-metal skew inversion
proceeds through a strained planar seven-membered
metallacycle containing two long P-Pt bonds and four
sp2 centers.
The large difference in rate between biphepRu(NN)-
Cl2 (t1/2 10’s of min at 25 °C) and the biphepPtX2
complexes (t1/2 of hours at 100 °C) is most easily
reconciled with a dissociative mechanism that is facile
with 18-electron octahedral Ru(II) complexes21 and
attenuated with substitution-inert biphepPtX2 com-
plexes (Scheme 2). Since preequilibrium dissociation of
a chelated diphosphine arm is documented in dppePt-
Me4 reductive elimination (165-200 °C),22 similar phos-
phine dissociation-induced reactivity may be operative
in biphepPtX2 complexes; the elevated isomerization
(21) (PPh3)4RuCl2 is known to be extensively dissociated in solution,
giving rise to the 14- and 16-electron complexes, (PPh3)2RuCl2 and
(PPh3)3RuCl2, respectively. (a) Caulton, K. G. J . Am. Chem. Soc. 1974,
96, 3005-3006. (b) J ames, B. R.; Markham, L. D. Inorg. Chem. 1974,
13, 97-100. Substitution of water from [RuII(OH2)(bpy)2(PR3)]2+ by
incoming ligands is also known to be dissociation initiated; see:
Leising, R. A.; Ohman, J . S.; Takeuchi, K. J . Inorg. Chem. 1988, 27,
3804-3809. For a general discourse on substitution mechanisms of
octahedral organometallic complexes, see: (c) Atwood, J . D. Inorganic
and Organometallic Reaction Mechanisms, 2nd ed.; VCH Publishers:
New York, 1997.
(19) BINOL also strongly affects the relative energetics of the
biphenyl atropisomers in Brintzinger’s biphenylCp2Ti(BINOL) com-
plexes; see ref 20.
(20) Ringwald, M.; Stu¨rmer, R.; Brintzinger, H. H. J . Am. Chem.
Soc. 1999, 121, 1524-1527.
(22) Crumpton, D. M.; Goldberg, K. I. J . Am. Chem. Soc. 2000, 122,
962-963.

4382 Organometallics, Vol. 19, No. 21, 2000
Tudor et al.
Sch em e 4
temperatures are consistent with an expectedly slow
initiating P-Pt bond breakage.
A dissociative radical mechanism has been proposed
in a conceptually similar thermal isomerization (60-
110 °C) of biphenyl-linked metallocene BINOL com-
plexes (A, M ) Ti, Zr). Stable free radical additives such
as TEMPO were observed to accelerate the isomeri-
zation rate, presumably via a dechelated cyclopentadi-
enyl radical.20
relative to the diamine. Moreover, if ring strain in the
planar biphep transition state was attenuated by length-
ening the P-Ru and P-Pt bonds (with concomitant
compression of P-M-P bond angles), then the energetic
cost for such a distortion would depend on the absolute
ground-state bond strengths (P-Ru < P-Pt)26 and the
plasticity of the P-M-P bond angle. In this scenario
partial breakage of the P-Ru and P-Pt bonds occurs
en route to the transition state, the energetic cost of
which is regained on the return to products. Arguing
against any significant breakage in at least the diamine
case, however, is the modest rate decrease (∼2×) on
substituting Ru for Os in B. Another possibility for
explaining the Ru > Pt rate ordering points to the axial
chloride ligands in biphepRu(NN)Cl2. If these ligands
buttress the pseudoaxial P-Ph groups sufficiently to
raise the ground state energy, then reduction of this
interaction in the transition state would lead to an
overall rate enhancement relative to a square-planar
coordination environment, all things being equal. Steric
effects on the rate of biphep atrop-inversion however
must be interpreted cautiously, as the barrier is known
to be variable and sensitive to substitution in simple
constrained biaryl compounds such as C-F , more rigid,
substituted backbones tending toward higher activation
energies.27 Since the available data do not distinguish
between the two proposed mechanisms for biaryl atrop-
inversion (the activation parameters are not helpful), a
mechanism for the thermal isomerization of diastereo-
mers cannot be assigned at this time. Future studies
The on-metal atrop-inversion process, however, can-
not be dismissed (Scheme 3). In particular, the rapid
atrop-isomerism observed in 2,2′-diamino-1,1′-biphenyl
complexes of Ru and Os (millisecond time scale), known
to rigorously proceed by a concerted mechanism, B,
suggests that the planar biphep transition state might
be feasible.23,24 Most problematic in assessing the
importance of this mechanism is rationalizing the large
difference in rate observed for biphenyldiamine com-
plexes, B (t1/2 ∼100 ms, 25 °C), the diphosphine com-
plexes biphepRu(NN)Cl2 (t1/2 ∼20 min, 25 °C), and
biphepPt(BINOL) (t1/2 8.5 h, 90 °C). Since substituting
NH2 for PPh2 increases the M-X bond lengths (X ) N,
P),25 this change might add additional strain to the
cyclic transition state and slow the biphep systems
(23) (a) Alguindigue, S. S.; Khan, M. A.; Ashby, M. T. Organome-
tallics 1999, 18, 5112-5119. Similarly, 1,1′-biisoquinoline derivatives
of Ru and Os are known to rigorously undergo atrop-inversion by a
chelated transition state. (b) Ashby, M. T.; Govindan, G. N.; Grafton,
A. K. J . Am. Chem. Soc. 1994, 116, 4801-4809. (c) Ashby, M. T. J .
Am. Chem. Soc. 1995, 117, 2000-2007. (d) Ashby, M. T.; Alguindigue,
S. S.; Khan, M. A. Organometallics 2000, 19, 547-552.
(24) Phosphite ligands based on 2,2′-biphenol are also known to
rapidly invert their atropisomers while coordinated to Mo and Pt; see:
Hariharasarma, M.; Lake, C. M.; Watkins, C. L.; Gray, G. M.
Organometallics 1999, 18, 2593-2600.
(25) Ru-N bond lengths in (C6H6)RuCl(2,2′-diamino-1,1′-biphenyl)
are 2.172(6) Å; cf. the Pt-P bond lengths of ∼2.24 Å observed herein.
(26) (a) Li, C.; Cucullu, M. E.; McIntyre, R. A.; Stevens, E. D.; Nolan,
S. P. Organometallics 1994, 13, 3621-3627. (b) Haar, C. M.; Nolan,
S. P.; Marshall, W. J .; Moloy, K. G. Organometallics 1999, 18, 474-
479.
(27) This argument was originally put forth by Ashby and co-
workers; see footnote 23a and references therein.

Chiral BiphepPtX2 Complexes
Organometallics, Vol. 19, No. 21, 2000 4383
were recorded at ambient temperatures on a Bruker AMX300
examining Pd analogues might shed additional light on
the subject.
1
or AMX400 spectrometer (300 or 400 MHz, H, respectively).
31P NMR spectra were obtained using 85% H3PO4 as an
external reference. All solvents but THF were purified by
passing over activated alumina columns,30 and THF was dried
over Na-benzophenone ketyl. Optical rotation measurements
were obtained on
a J asco DIP-1000 digital polarimeter.
Elemental analyses were performed by E+R Microanalytical
Laboratory, Inc. (Parsippany, NJ ).
Bip h ep . To a flame-dried flask was added 4.0 g (14.4 mmol)
of triphenylphosphine oxide under nitrogen. The solid was
dissolved in 140 mL of THF and cooled to -78 °C, and then
1.05 equiv (15.1 mmol) of freshly titrated sec-butyllithium was
slowly added. The deep red homogeneous solution was stirred
at -78 °C for 3 h and then warmed to 0 °C. Copper(II) pivalate
(43.1 mmol, 11.5 g) in 50 mL of THF was transferred via
cannula to the triphenylphosphine oxide solution. After stir-
ring for 4 h at 0 °C, the solvent was removed by rotary
evaporation and the blue residue dissolved in a 3:1 solution
of ethyl acetate/THF. The residual copper and lithium were
extracted with a saturated ammonium chloride solution (2 ×
100 mL) and a saturated sodium bicarbonate solution (4 × 100
mL) until the organic layer was colorless. The organic layer
was washed with brine (2 × 100 mL), dried over magnesium
sulfate, and filtered. After solvent was removed by rotary
evaporation, the crude biphep oxide was dried in vacuo and
dissolved in 90 mL of xylenes. To the stirring solution was
added 3.4 mL of triethylamine (2.38 g, 23.5 mmol) and 2.86
mL of trichlorosilane (3.23 g, 28.7 mmol, 4 equiv). The solution
was heated at 100, 120, and 140 °C, each for 1 h. After cooling
to 50 °C, 45 mL of 30% NaOH was added slowly. The layers
were separated, and the NaOH layer was extracted with 30
mL of toluene. To the combined organic fractions was added
45 mL of 30% NaOH, and the mixture was vigorously shaken.
The aqueous layer was removed and back extracted with
toluene (2 × 60 mL). The organic layers were combined,
washed with brine (2 × 100 mL), and dried over MgSO4, and
the solvent was removed by rotary evaporation. The crude
mixture was recrystallized from 1:3 toluene/ethanol to afford
crystals of biphep in an overall 30% yield. 31P NMR (121.5
MHz, CD2Cl2): δ -13.7. 1H NMR (300.1 MHz, CD2Cl2): δ 7.24
(m, 24 H), 7.07 (br d, 2H, J ) 9 Hz), 6.89 (br d, 2H, J ) 9 Hz).
13C{31P} NMR (75.5 MHz, CD2Cl2): δ 147.7, 138.4, 137.7,
137.0, 134.2, 133.8, 131.2, 128.8, 128.6, 128.5, 128.4, 128.0.31
Anal. Calcd for C36H28P2; C, 82.74; H, 5.40. Found: C, 82.77;
H, 5.71.
Since square-planar Pt(II) complexes tend to react via
associative mechanisms, we reasoned that donor ligands
might catalyze the isomerization by increasing the
dissociation rate of a phosphine arm from a five-
coordinate intermediate (Scheme 4). Indeed, λ(S)-5b is
gradually converted to the thermodynamically preferred
δ(S)-5a isomer (95:5) with gentle warming in pyridine
(24 h at 40 °C). This corresponds to a reduction in the
isomerization temperature of 60-70 °C.28 Moreover, this
decrease in temperature allowed a ∼1:1 mixture of
δ(1R,2S)-4a and λ(1R,2S)-4b to be converted to
δ(1R,2S)-4a (>97:3) at 52 °C; thermolysis in toluene
(100 °C) leads to extensive decomposition. While the
arm-off mechanism is reasonable for the pyridine-
catalyzed reaction, it is not clear whether this repre-
sents an accelerated arm-off mechanism or a new
mechanism that operates at low temperatures; that is,
pyridine catalysis does not necessarily shed light on the
noncatalyzed process.
(7) Su m m a r y. Examination of diastereomer inter-
conversion processes in biphepPtX2 complexes revealed
that at least three mechanisms are available. In the
presence of enantiomeric chiral X2 ligands, the data
clearly indicate that X2-X2 ligand exchange processes
dominate the isomerization kinetics and that biphep
stereoinversion is not competitive. On the other hand,
when the X2 ligand is enantiopure, X2-X2 ligand
exchange is unproductive and higher energy processes
are detected. Thermolysis of the minor biphepPtBINOL
diastereomer (92-122 °C) in chlorobenzene leads to a
thermodynamic mixture (∼94:6) by a mechanism that
requires biphep biaryl rotation and atrop-inversion of
the diphosphine stereochemistry. Two mechanisms are
consistent with the data: on-metal inversion through
a planar seven-membered transition state, and one-arm-
off prior to rotation. The isomerization kinetics of the
enantiopure ligand systems could be substantially ac-
celerated with pyridine in a process that is proposed to
involve a five-coordinate pyridine adduct (undetected),
which has a higher phosphine dissociation rate than in
the square-planar complex (or the on-metal stereoin-
version rate).
(bip h ep )P tCl2 (1). To a stirring solution of 358.5 mg (0.958
mmol, 1 equiv) of (cod)PtCl2 in 35 mL of CH2Cl2 was added
dropwise 499.9 mg (0.956 mmol, 1 equiv) of biphep in 35 mL
of CH2Cl2. After 10 min of stirring, the solvent was removed
and the white powder was recrystallized in CH2Cl2/MeOH,
producing 685 mg of white crystals (90% yield). 31P NMR (121.5
1
MHz, CD2Cl2): δ 7.9 (J P-Pt ) 3610 Hz). H NMR (300.1 MHz,
CD2Cl2): δ 7.74 (4H, q, J ) 9 Hz), 7.6 (4H, m), 7.4 (8H, m),
7.2 (4H, t, J ) 9 Hz), 7.0 (6H, m), 6.65 (2H, dd, J ) 9, 6 Hz).
13C{31P} (75.5 MHz, CD2Cl2): δ 144.0, 136.1, 135.5, 134.0,
133.2, 132.1, 131.9, 131.3, 128.7, 128.2, 128.1, 128.0, 127.8,
124.8. Anal. Calcd for C36H28P2PtCl2‚CH2Cl2: C, 50.90; H, 3.47.
Found: C, 50.82; H, 3.47.
Regardless of the microscopic details of the process
that leads to biaryl rotation and atrop-inversion, it is
clear that the diphosphine stereochemistry in substitu-
tion-inert biphepPtX2 complexes is kinetically more
robust than the octahedral biphepRu(NN)Cl2 catalyst
reported by Mikami and Noyori.
(bip h ep )P tCO3 (2). To a solution of 300 mg (0.380 mmol,
1 equiv) of 1 in 40 mL of wet CH2Cl2 was added silver
carbonate (190 mg, 0.689 mmol, 1.8 equiv) in 20 mL of wet
CH2Cl2. The mixture was stirred in darkness for 3 h and then
filtered through Celite to remove the AgCl. 2 was precipitated
from solution using hexanes and was recrystallized CH2Cl2/
Exp er im en ta l Section
Gen er a l Com m en ts. (cod)PtCl2 was prepared according to
literature protocol.29 1H, 13C, 13C{31P}, and 31P NMR spectra
(29) (a) Slack, D. A.; Baird, M. C. Inorg. Chim. Acta 1977, 24, 277.
(b) McDermott, J . X.; Whitesides, G. M. J . Am. Chem. Soc. 1976, 98,
6521.
(30) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. Organometallics 1996, 15, 1518-1520.
(31) Not all 13C resonances were located.
(28) Monodentate phosphines such as PPh3, P(2-MePh)3, PBu3,
P(OMe)3, P(OPh)3, PMe3, excess biphep, and other donor solvents such
as DMSO, DMF, THF, and CH3CN either did not affect the rate of
BINOLate diastereomer interconversion or led to decomposition.

4384 Organometallics, Vol. 19, No. 21, 2000
Tudor et al.
hexanes to afford 248 mg of the pale yellow solid (84% yield).
31P NMR (162.0 MHz, CD2Cl2): δ 2.2 (J P-Pt ) 3570 Hz). 1H
NMR (400.1 MHz, CD2Cl2): δ 7.7 (8H, m), 7.5 (6H, m), 7.3
(2H, t, J ) 8 Hz), 7.2 (4H, t, J ) 8 Hz), 7.1 (2H, t, J ) 8 Hz),
7.0 (2H, d, J ) 8 Hz), 6.8 (4H, m). 13C{31P} (75.5 MHz, CD2-
Cl2): δ 143.8, 135.2, 135.0, 134.9, 134.0, 132.2, 132.0, 131.9,
131.8, 128.8, 128.6, 128.1, 127.4, 126.0, 125.6. Anal. Calcd for
126.0, 125.8, 125.7, 124.9, 124.7, 122.8, 121.2. [R]D26.1 ) -138.1
(c 0.605 in ClC6H5, contains 5% λ(S)-5b). Anal. Calcd for
C
56H40P2PtO2: C, 67.13; H, 4.02. Found: C, 66.9; H, 4.0.
(m in or , λ(S)-5b ): 31P NMR (162 MHz, CD2Cl2): δ 2.64
(J P-Pt ) 3643 Hz). 1H NMR (400 Hz, CD2Cl2): δ 8.04 (m, 4 H),
7.70 (d, J ) 8.1 Hz, 2 H), 7.53 (m, 12 H), 7.28 (m, 2 H), 7.09
(m, 8 H), 6.98 (t, J ) 8.3 Hz, 2 H), 6.88 (t, J ) 8.6 Hz, 6 H),
6.76 (m, 2 H), 6.23 (d, J ) 8.8 Hz, 2 H). 13C{31P} NMR (75
MHz, CD2Cl2): δ 161.5, 143.5, 135.5, 134.8, 134.5, 133.1, 132.7,
131.6, 131.2, 130.9, 128.9, 128.8, 128.3, 128.2, 127.9, 127.8,
127.4, 125.4, 125.2, 124.7, 124.1, 121.3.31 [R]D26.1 ) 36.7 (c 0.610
in ClC6H5). Anal. Calcd for C56H40P2PtO2‚2CH2Cl2: C, 59.45;
H, 3.78. Found: C, 59.48; H, 4.02.
C
37H28P2PtO3: C, 57.15; H, 3.63. Found: C, 56.92; H, 3.52.
(bip h ep )P t(1R,2S)-TfNO (δ(1R,2S)-4a , λ(1R,2S)-4b). A
100 mg sample of 2 (0.128 mmol) and 44.4 mg (0.128 mmol)
of (1R,2S)-3 were dissolved in 5 mL of CH2Cl2 and stirred at
room temperature for 1 h. The solvent was removed and the
white solid dissolved in a minimum amount of CH2Cl2 and
precipitated using hexanes to produce a ∼1:1 mixture of
diastereomers. The 1:1 ratio of δ(1R,2S)-4a :λ(1R,2S)-4b was
recrystallized from CH2Cl2/hexanes to yield a colorless solid
in 40% yield. A racemate of this mixture that was stable to
isomerization could only be obtained if a deficiency of (()-3
(0.9 equiv) was used. Employing 1.1 equiv of (()-3 led to (()-
4a , which was recrystallized from CH2Cl2/hexanes (47% yield).
Liga n d Exch a n ge Kin etics. A typical procedure for the
isomerization of λ(S)-5b to λ(R)-5a through ligand exchange
is as follows. A 15 mg sample (15 µmol) of a 1:1 mixture of
λ(S)-5b and δ(S)-5a was dissolved in 1.0 mL of chlorobenzene.
One equivalent of (R)-BINOL (4.3 mg, 15 µmols) was added
to the solution, and then it was transferred into an NMR tube
containing a sealed benzene-d6 capillary to provide a lock
signal and immediately cooled (-78 °C). The sample was
warmed and inserted into a preheated NMR probe (45 °C),
and data collection was initiated; no diastereomer overlap
occurred at 45 °C. Control experiments with 1 equiv of (S)-
BINOL and no additive were carried out in parallel.
(m a jor , (()-4a ): 31P NMR (162.0 MHz, CD2Cl2): δ 7.2 (d,
1
1J Pt-P ) 3760 Hz, J P-P ) 31.3 Hz, trans to N), 3.5 (d, J Pt-P
)
3210 Hz, J P-P ) 31.3 Hz, trans to O). 1H NMR (300.1 MHz,
CD2Cl2): δ 7.8 (m, 7H), 7.5 (m, 16 H), 7.0 (m, 12 H), 6.7 (m, 2
H), 6.4 (m, 1H), 5.4 (d, J ) 6 Hz, CHO), 5.0 (br s, J Pt-H ) 36
3
Hz, CHN). 13C{31P} (75.5 MHz, CD2Cl2): δ 144.9, 144.8, 143.2,
141.1, 135.3, 135.0, 134.0, 133.3, 132.2, 132.1, 131.9, 131.5,
131.2, 131.0, 130.9, 130.7, 130.2, 128.5, 128.0, 127.8, 127.7,
127.6, 127.5, 127.0, 126.7, 126.5, 126.0, 125.3, 125.0, 85.1 (CH),
71.5 (CH).31 Anal. Calcd for C51H40F3NO3P2PtS: C, 57.74; H,
3.80; N, 1.32. Found: C, 57.29; H, 3.84; N, 1.15. (m in or ,
λ(1R,2S)-4b) could not be obtained free of the major isomer;
characteristic resonances include the following. 31P NMR
Th er m olysis Kin etics. A typical procedure for the thermal
isomerization of λ(S)-5b to δ(S)-5a is as follows. A solution of
6.2 mg (5.1 µmol) of λ(S)-5b in 0.8 mL of chlorobenzene was
placed in an NMR tube containing a sealed capillary that was
filled with benzene-d6 to provide a lock signal. The solution
was heated to 92.6 °C in an oil bath. After a known amount of
time, the NMR tube was removed from the oil bath and
immediately cooled to 0 °C to halt the thermal isomerization.
The conversion to product was monitored by 31P NMR at room
temperature; the two diastereomers are overlapped at elevated
temperature.
Cr ysta llogr a p h y. Crystals of (()-4a suitable for X-ray
crystallography were grown at room temperature from a
saturated CH2Cl2 solution with slow diffusion of diethyl ether.
Crystals of λ(S)-5b were grown from CD2Cl2 by slow evapora-
tion. Single crystals were mounted in oil on the end of a fiber.
Intensity data were collected on a Siemens SMART diffracto-
meter with CCD detection using Mo KR radiation of wave-
length 0.710 73 Å (ω scan mode). The structures were solved
by direct methods and refined by least-squares techniques on
F using structure solution programs from the NARCVAX
system.32 All non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were placed in calculated positions (C-H )
0.96 Å) and allowed to ride on the atoms to which they were
bonded. Crystal data, data collection, and refinement param-
eters are listed in Table 3. Absorption corrections were made
using SADABS.
1
(162.0 MHz, CD2Cl2): δ 8.9 (d, J Pt-P ) 3725 Hz, J P-P ) 32.3
Hz, trans to N), 4.8 (d, 1J Pt-P ) 3210 Hz, J P-P ) 32.3 Hz, trans
to O). 1H NMR (300.1 MHz, CD2Cl2): δ 5.6 (d, J ) 6 Hz, CHO);
5.1 (br s, CHN, overlaps major diastereomer).
bip h ep P t(S)-BINOL (δ(S)-5a , λ(S)-5b). Meth od 1. A
solution of sublimed NaOtBu (27.8 mg, 0.289 mmol) in 5 mL
of THF was transferred via cannula to 40.9 mg (0.143 mmol)
of (S)-BINOL in 5 mL of THF. This solution of Na2BINOL was
transferred via cannula to a slurry of 1 (112.9 mg, 0.143 mmol)
in dry toluene. The yellow mixture became clear and yellow
within 30 min, signifying completion of the reaction. The
toluene was removed in vacuo, and the yellow solid triterated
twice with 5 mL of dry toluene to remove residual tert-butyl
alcohol. CH2Cl2 (10 mL) was added and the solution filtered
to remove the NaCl byproduct. The product was precipitated
from the supernatant in a 1:1 ratio of diastereomers using
hexanes. A yellow solid that was ∼1:1 δ(S)-5a /λ(S)-5b was
isolated in 87% yield. Thermolysis of the latter complexes in
toluene (90 °C) over 3 days provided δ(S)-5a (95:5). A mixture
of racemic diastereomers could be obtained using racemic
BINOL, though this material tended to isomerize, at a rate
dependent on the concentration of free BINOL.
Meth od 2. 2 (24.4 mg, 0.031 mmol) and BINOL (either
racemic or enantiomerically pure) (9.0 mg, 0.031 mmol) were
dissolved in 0.5 mL of CD2Cl2, and the reaction was monitored
by 31P NMR. The reaction with S-BINOL took 5 days to
complete at room temperature (1:1 mixture of diastereomers);
the racemic BINOL reaction took 24 h and precipitated a 95:5
mixture of (()5a /5b from solution. This solid was collected and
recrystallized from chlorobenzene and hexanes in 91% yield.
(m a jor , δ(S)-5a ): 31P NMR (162 MHz, CD2Cl2): δ 2.86
(J P-Pt ) 3670 Hz). 1H NMR (400 Hz, CD2Cl2): δ 7.81 (m, 4 H),
7.72 (d, J ) 7.7 Hz, 2 H), 7.59 (m, 4 H), 7.52 (d, J ) 8.7 Hz, 2
H), 7.39 (m, 4 H), 7.26 (m, 8 H), 7.12 (m, 2 H), 7.01 (m, 4 H),
6.92 (t, J ) 7.7 Hz, 2 H), 6.85 (d, J ) 8.7 Hz, 4 H), 6.72 (d,
J ) 8.7 Hz, 2 H), 6.63 (m, 2 H). 13C{31P} NMR (75 MHz, CD2-
Cl2): δ 161.3, 143.8, 135.9, 135.7, 135.3, 134.0, 132.1, 131.8,
131.4, 131.2, 128.9, 128.7, 128.5, 128.1, 127.9, 127.8, 127.3,
Ack n ow led gm en t. This research was partially sup-
ported by the NSF (CAREER, CHE-9624852), NIGMS
(R01 GM60578-01), the Petroleum Research Fund,
DuPont, and 3M. M.R.G. is a Camille Dreyfus Teacher
Scholar (2000).
Su p p or tin g In for m a tion Ava ila ble: Kinetic plots for the
thermal isomerization of λ(S)-5b to δ(S)-5a , van’t Hoff plot for
the λ(S)-5b to δ(S)-5a equilibrium, kinetic plots for the
isomerization of λ(1R,2S)-4b to λ(1S,2R)-4a , and complete
tables of acquisition and metrical parameters for λ(S)-5b and
(()-4a . This material is available free of charge via the
Internet at http://pubs.acs.org.
OM000629A
(32) Gabe, E. J .; Le Page, Y.; Charland, J . P.; Lee, F. L.; White, P.
S. J . Appl. Crystallogr. 1989, 22, 384.
(33) Andrews, M. A.; Cook, G. K.; Shriver, Z. H. Inorg. Chem. 1997,
36, 5832-5844.