Diastereoselective synthesis of C2-symmetric hexacoordinated phosphate anions (HYPHATs) with predetermined chirality from 1,2-diaryl-ethane-1,2-diols

Article abstract of DOI:10.1016/S0022-328X(01)01251-7

C₂-symmetric HYPHAT anion (5) made of a central phosphorus(V) atom, one hydrobenzoin and two tetracholorocatechol ligands can be simply prepared in high yield as its dimethylammonium salt using a one pot-process and simple commercially available or easily prepared starting materials. The presence of the chiral hydrobenzoin ligands (e.g. R,R) leads to the formation of diastereomeric anions (Δ,R,R/Δ,R,R). A partial control over the configuration of the adduct by the chiral ligands is observed (diastereomeric ratio (d.r.) 75:25 in 5% DMSO-CHCl₃). This asymmetric induction can be improved (d.r. up to 90:10) by the introduction of ortho bromo substituents on the phenyl rings of the 1,2-diaryl-ethane-1,2-diol ligands.

Full text of DOI:10.1016/S0022-328X(01)01251-7

Journal of Organometallic Chemistry 643–644 (2002) 392–403
www.elsevier.com/locate/jorganchem
Diastereoselective synthesis of C2-symmetric hexacoordinated
phosphate anions (HYPHATs) with predetermined chirality from
1,2-diaryl-ethane-1,2-diols
Je´roˆme Lacour *, Anne Londez
De´partement de Chimie Organique, Uni6ersite´ de Gene`6e, quai Ernest-Ansermet 30, CH-1211 Gene6a 4, Switzerland
Received 20 July 2001; accepted 11 September 2001
Dedicated to Professor Franc¸ois Mathey
Abstract
C2-symmetric HYPHAT anion (5) made of a central phosphorus(V) atom, one hydrobenzoin and two tetracholorocatechol
ligands can be simply prepared in high yield as its dimethylammonium salt using a one pot-process and simple commercially
available or easily prepared starting materials. The presence of the chiral hydrobenzoin ligands (e.g. R,R) leads to the formation
of diastereomeric anions (D,R,R/L,R,R). A partial control over the configuration of the adduct by the chiral ligands is observed
(diastereomeric ratio (d.r.) 75:25 in 5% DMSO–CHCl3). This asymmetric induction can be improved (d.r. up to 90:10) by the
introduction of ortho bromo substituents on the phenyl rings of the 1,2-diaryl-ethane-1,2-diol ligands. © 2002 Elsevier Science
B.V. All rights reserved.
Keywords: Predetermination of configuration; Hexacoordinated phosphorus; A(1,3) strain; Asymmetric induction
resonance (NMR) chiral shift, resolving and asymmet-
ric inducing reagent onto organic and organometallic
1. Introduction
derivatives—with a predilection for octahedral metallo-
organic complexes [5].
However, with some chiral C2-symmetric cations, low
NMR shifts and asymmetric inducing properties were
central phosphorus atom with three identical symmetric
The octahedral geometry of pentavalent hexacoordi-
nated phosphorus allows the formation of chiral an-
ions—D and L enantiomers—by complexation of the
recently observed using anion 2 [6]. Assuming that its
D3-symmetry was not adapted for the chiral recognition
of such cations, we decided to investigate the synthesis
of C2-symmetric hexacoordinated phosphate anions;
our interest being motivated by the overall efficiency of
such symmetry in asymmetric reactions or molecular
recognition processes [7]. Furthermore, the introduction
of a well-chosen C2-symmetric chiral ligand could per-
mit a diastereoselective synthesis of the resulting anion
by predetermination of the configuration [8] around the
phosphorus atom (Fig. 1).
bidentate ligands [1,2]. Tris(benzenediolato)phosphate
anion 1, of particular interest for its easy preparation
from catechol, PCl5 and an amine, is unfortunately
configurationally labile in solution as an ammonium
salt due to an acid-induced racemization mechanism [3].
Recently, we reported that the introduction of electron-
withdrawing groups (chlorine atoms) on the aromatic
nuclei increases the configurational stability of the re-
sulting
tris(tetrachlorobenzenediolato)phosphate(V)
derivative or TRISPHAT 2 [4]. This D3-symmetric an-
ion can be resolved by association with a chiral ammo-
nium cation. It is an efficient nuclear magnetic
However, chiral induction from ligands is, as a gen-
eral rule, not very efficient for didentate ligands,
whereas with ligands of higher denticity, especially
those of four, five, and six donor atoms, a complete
* Corresponding author. Fax: +41-22-3287396.
E-mail address: jerome.lacour@chiorg.unige.ch (J. Lacour).
0022-328X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)01251-7

J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
393
chiral induction can often be observed. In tris(diden-
tate) octahedral complexes, most successful examples of
predetermination of chirality have thus been made us-
ing three chiral didentate ligands [8,9]. When only one
chiral ligand and two achiral didentate ligands are
coordinated to a metal center, only sterically demand-
ing ligands lead to measurable effects of chiral induc-
tion upon complex formation. Diastereoselectivity
remains, however, low in most cases and often depends
upon non-covalent interactions [10]. This was also ob-
served in the only example known in the literature—
prior to our own studies—of a hexacoordinated
phosphate anion bearing a single chiral ligand. Anion
3, containing (−)-mandelic acid along with two pyro-
catechol rings, is configurationally labile and epimerizes
in solution to give a 55:45 mixture of diastereomers
[3c,11] (Fig. 2).
Recently, we have shown that chiral C2-symmetric
ligand BINOL ([1,1%]binaphthalenyl-2,2%-diol) can con-
trol single-handedly the configuration of the resulting
anion (BINPHAT 4, diastereomeric ratio d.r.\39:1)
[6,12]. This successful example of high diastereoselectiv-
ity in the stereocontrolled synthesis of hexacoordinated
phosphate anion 4 prompted us to test the generality of
this chiral auxiliary approach with other chiral diols.
Herein, we report on the synthesis of C2-symmetric
phosphate anions containing 1,2-diaryl-ethane-1,2-diols
as chiral auxiliaries and on the resulting asymmetric
induction (diastereoselectivity) from the chosen ligands.
A modest selectivity was observed using 1,2-diphenyl-
ethane-1,2-diol (anion 5a, d.r.575:25). It can be im-
proved by the introduction of ortho bromo substituents
on the phenyl rings (5b, d.r.590:10).
2. Results and discussion
2.1. General considerations and choice of a class of
chiral ligands
Fig. 1. D3-symmetric hexacoordinated phosphate anions 1 and 2
(TRISPHAT).
Our goal was thus to synthesize new C2-symmetric
hexacoordinated phosphate anions of configuration (D
or L) controlled by a chiral ligand. Keeping in mind the
overall efficiency of anion 2 in asymmetric applications
[5], we chose to introduce two tetrachlorocatechol (6)
moieties along with the chiral ligand. 1,2-Diphenyl-eth-
ane-1,2-diol (7a) and its derivatives are chiral C2-sym-
metric compounds which have been studied to a large
extent in asymmetric chemistry as chiral ligands or
auxiliaries [13]. Their general use is most probably due
to their easy preparation in high enantiomeric purity by
asymmetric dihydroxylation (AD) of (E)-stilbenes [14].
This AD procedure developed by Sharpless and
coworkers is general and can be applied to a large
variety of substituted stilbenes [15]. This gives a possi-
bility of facile ligand optimization after the feedback
from initial experiments. For these reasons, we selected
1,2-diaryl-ethane-1,2-diols as chiral ligands and decided
to test their efficiency as chiral inducers for the determi-
nation of configuration of hexacoordinated phosphate
anions.
Fig. 2. Hexacoordinated phosphate anions 3 and 4 (BINPHAT) with
predetermined configuration.
However, the desired anions of general structure 5
(Fig. 3) were unknown in the literature. It was therefore
necessary to develop an efficient and general synthetic
procedure for their preparation. Several routes were
considered before achieving this goal. They are de-
scribed—along with the synthesis of the chiral lig-
ands—in the following paragraphs.
Fig. 3. Phosphate anions 5a–5d containing 1,2-diaryl-ethane-1,2-diol
7a–7d ligands.

394
J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
2.2. Synthesis of the 1,2-diaryl-ethane-1,2-diols (7a–7d)
albeit low yields (Eq. (2), 26 and 15%, respectively) [19].
With these 1,2-diaryl-ethane-1,2-diols in our hands, we
turned our attention to the synthesis of the correspond-
ing phosphate anions with a strong emphasis on the
preparation of anion 5a derived from diol 7a.
For the preparation of 1,2-diphenyl-ethane-1,2-diol
7a, we reproduced the Sharpless procedure with com-
mercially available AD-mix b and (E)-stilbene (Eq. (1)).
(R,R)-7a was obtained in 80% yield in high enan-
tiomeric purity (e.e.\99%) [14].
2.3. Synthesis of phosphate anion 5a. Preliminary
studies
For the synthesis of HYPHAT anion 5a, we first
attempted to adapt the synthetic procedure developed
for TRISPHAT 2—that is the one-time addition of
three equivalents of tetrachlorocatechol 6 to one equiv-
alent of PCl5 followed by the addition of an amine.
Ligands 6 (two equivalents) were added to a solution of
PCl5 in toluene at 70 °C. After 16–24 h, concentration
in vacuo and additions of CH2Cl2, (R,R)-7a (one equiv-
alent) and Bu3N (two equivalents), the spontaneous
formation of a precipitate was observed. This white
powder corresponded to the desired salt [nBu3NH][5a],
albeit in low chemical yield (13%) along with minor
amounts of [nBu3NH][2] (Eq. (3)).
(1)
For reasons that will be explained later (cf. Section
2.6),
(R,R)-1,2-bis-(2-bromo-phenyl)-ethane-1,2-diol
(7b), rac-1,2-bis-(4-trifluoromethyl-phenyl)-ethane-1,2-
diol (7c) and rac-1,2-bis-(3,5-bis-trifluoromethyl-
phenyl)-ethane-1,2-diol (7d) were also prepared for this
project. Compound 7b was synthesized in three steps:
Wittig olefination of 2-bromo-benzaldehyde and (2-
bromo-benzyl)-triphenyl-phosphonium under the con-
ditions reported by Kelly et al. afforded a mixture of
(Z)- and (E)-a,a%-dibromo-stilbene (77%) [16], which
was treated with TeCl4 at reflux in CHCl3 to give pure
(E) isomer in 84% yield [17]. Asymmetric dihydroxyla-
tion of this compound, following the conditions re-
ported by Kagan and Girard, afforded 7b (Eq. (1), 86%
yield, e.e.\98%) [18].
(3)
(2)
1H- and 31P-NMR analyses (DMSO-d6, Fig. 4) re-
vealed essentially one set of signals indicating the pres-
ence of one of the two diastereomeric ion pairs (D,R,R
or L,R,R, d.r. ꢀ96:4) [20]. Although, the high
diastereoselectivity was satisfactory, the low yield was
disappointing and many reactions were performed to
improve this synthetic route. In all cases, selective
precipitation of [ammonium][5a] salts without traces of
[ammonium][2] was difficult. We also observed that
prolonged reaction times were detrimental to the yield.
Larger amounts of degradation products and of
TRISPHAT 2 were isolated indicating possible recom-
binations of the ligands among the intermediates and
products when enough time was given. We thus turned
our attention to the design of a more efficient synthetic
route.
Racemic derivatives 7c and 7d were prepared follow-
ing the procedure of Cuvinot and Alexakis by pinacol
coupling of the corresponding aldehydes in one step
2.4. Optimized syntheses of anion 5a
It was our analysis that the principal problem of the
previous route was the presence of two different biden-
tate ligands, which could react simultaneously and lead
Fig. 4. 31P-NMR spectrum (162 MHz, DMSO-d6) of [nBu3NH][5a].

J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
395
to four different anionic compounds. This is repre-
sented in Fig. 5 considering the theoretical case of two
different ligands aa and bb (2:1 ratio) which can bind to
PCl5. A statistical 1:6:8:12 repartition among the four
possible products is then expected; the desired com-
pound [P(aa)2bb] being synthesized with a maximum
44% theoretical yield.
To obtain a mixed [P(aa)2bb] adduct in high yield
and in the exclusive of all others, we considered that the
best synthetic route would be one allowing the sequen-
tial introduction of each of the three ligands in three
different and orthogonal chemical steps [21]. Undeni-
ably, if only one ligand was to be present at each
chelation step, all chances of mutual competition be-
tween ligands would be nullified and the resulting mix-
ture of compounds avoided. A synthetic route using
three separate and high-yielding chelation steps should
therefore lead to the formation of a single phosphate
anion in high combined yield.
Fig. 5. Statistical distribution among possible phosphates from the
reaction of two different ligands (aa and bb, 2:1 ratio) and PCl5.
For the implementation of this approach, we consid-
ered the known facts that (i) PX3 derivatives (X=halo-
gens, OR, NR2) can react with diols to form
monochelated adducts [22] and (ii) monocyclized P (III)
compounds are readily oxidized to bicyclic spirophos-
phoranes using ortho-chloranil 8 as an oxidant (Fig. 6)
[23]. These two chelation steps were ideal for our
projected route, as they could be performed sequen-
tially in high yields. To perform the final chelation step
and obtain the desired anion, it would then be sufficient
to add a third diol ligand (ligand bb in Fig. 6) to the
spirophosphorane intermediate. After displacement of
the last substituent X, chelation would occur in the
presence of a base forming the final ring. In a final set
of beneficial twists, we realized that if substituent X
were to be a base—such as an amino group—then the
two protons delivered by the last diol would be directly
scavenged in solution with no need to add an extra
base. Salts [XH2][P(aa)2bb] would directly result from
the procedure. We also recognized that most diols are
compatible with ortho-chloranil 8 and chelation steps 2
and 3 could therefore happen in a one-pot procedure.
In our initial attempts following these guidelines,
chiral ligand 7a was introduced at the start of the
synthesis (as ligand aa in Fig. 6). We assumed that its
early chelation around the phosphorus atom would
maximize the chances of asymmetric induction at each
of the following steps. Phosphoramidite 9a was thus
synthesized in a decent yield (60%) by reaction of
(R,R)-7a with tris(dimethylamino)phosphine 10 [22c].
Then, reaction of 9a with a 1:1 mixture of 6 and 8
Fig. 6. Outline of the envisioned synthetic strategy for making of
[XH2][P(aa)2bb] salts.
afforded the desired [Me2NH2][5a] as
a
single
diastereomer (Fig. 7). Decent chemical yields—higher
than those from the previous method—were obtained
(54, 61 and 63% using CH2Cl2, CH2Cl2–hexane 4:1 or
Et2O as solvents respectively).
Fig. 7. Two-steps synthesis of [Me2NH2][5a] via hydrobenzoin phos-
phoramidite 9a.

396
J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
This procedure was advantageous because phosphate
salt [Me2NH2][5a] precipitated as soon as it was formed
giving no chances for the anion to decompose or equili-
brate to other compounds. Most probably, the in-situ
formation of HNMe2 is favorable as it traps immedi-
ately the spirophosphorane intermediate to form the
phosphate anion. However, although useful, this proce-
dure was difficult to scale up due to the delicate synthe-
sis of 9a. We turned our attention to a final approach.
We considered that it would be more practical to
start with phosphoramidite 11 derived from tetra-
chlorocatechol 6 and introduce the chiral ligand at the
last step (as ligand bb in Fig. 6). This approach would
be both economical in chiral ligand 7a and elegant in
terms of strategy as it could be generalized to all diols
and not just hydrobenzoins [6]. Compound 11, simply
prepared by heating anhydrous 6 and freshly distilled
10 in refluxing toluene, was obtained in good to excel-
lent yields (80–97%); purification being effected by
sublimation of the crude product. Then, treatment of
11 with 8 and ligand 7a in an appropriate solvent
combination afforded the desired [Me2NH2][5a] salt in
decent to good yields (45–70%, Scheme 1, Table 1,
entry 1–4).
Scheme 1. (a) 6 (1.7 equivalents for 1.8 equivalents of 10), NH4Cl (2
mol%), toluene, reflux; (b) o-chloranil 8 (1.7 equivalents), 7a (1.0
equivalents), solvent, 20 °C.
Table 1
Optimized syntheses of [Me2NH2][5a] salts starting from tetrachloro-
catechol 6
a
Entry
1 c
Solvent
Time
Yield
d.r. b
[2]% f
3:2 e
18
2
2
18
4
68
48–55
45
70
70
96:4
6
6
5
5
8
The crude mixture from the first reaction containing
essentially pure 11, we realized that this synthetic se-
quence could be performed in a ‘one-pot’ procedure.
This rendered the synthesis of salts [Me2NH2][5a] even
easier; reaction times were reduced from 18 to 4 h with
no loss of chemical yield (Table 1, entry 5). In all these
2 c
3 c
4 c
5 d
CH2Cl2
Et2O
Et2O
4:1 e
\96:4
\96:4
\96:4
\96:4
a Reaction time in hours.
b 31P-NMR, DMSO-d6, signals at l −79.2 and −79.8 ppm, respec-
reactions using ligand 7a, essentially
a
single
1
diastereomer precipitated as observed in H and 31P-
tively.
c Using isolated 11.
NMR (DMSO-d6, d.r. ]96:4).
d One-pot protocol.
With decent amounts of [ammonium][5a] salts in our
hands, we turned our attention to the configurational
stability of anion 5a.
e CH2Cl2–hexane mixtures.
f Contamination of [Me2NH2][2] in the precipitate.
2.5. Configurational stability of 5a and asymmetric
induction properties of ligand 7a
As just mentioned, [ammonium][5a] salts were always
obtained as precipitates in high diastereomeric ratios
(d.r. ]96.4). For the analyzes of the precipitates, care
was taken to use DMSO-d6 as a NMR solvent as
Koenig and Klaebe had observed that this solvent
promoted only little epimerization of the chiral hexaco-
ordinated phosphate he had prepared [3c].
However, upon dissolution of the [ammonium][5a]
salts in MeOH-d4 a slow but definite equilibration
between the two diastereomers was observed as the
ratio changed from \96:4 to 61:39 after 29 h (Fig. 8).
Using CDCl3—with minor amounts of DMSO-d6 to
solubilize the substrate—the change was more dramatic
as a 75:25 ratio was obtained immediately after dissolu-
tion (Fig. 9, spectrum c). All these results demonstrat-
ing a poor configurational stability in solution for
anion 5a in [ammonium][5a] salts.
Fig. 8. Epimerization of [Me2NH2][5a] as a function of time. 31P-
NMR (162 MHz, MeOH-d4) (a) 1.5 h., d.r. 88:11; (b) 2.3 h.; d.r.
86:14; (c) 9.7 h, d.r. 67:32; (d) 17.8 h., 64:36; (e) 28.4 h., d.r. 61:39.
Major (M) and minor (m) diastereomers.

J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
397
The existence of an equilibrium between diastereo-
meric [ammonium][D-5a] and [ammonium][L-5a] salts
was further proven while attempting the purification of
salt [nBu3NH][5a] (d.r. \96:4) from its [nBu3NH][2]
contaminant. Following a protocol developed in our
group for the isolation and purification of TRISPHAT
salts [24], a cation exchange of Bu3NH+ to the purple
cation of crystal violet 12—or [tris(para-dimethyl-
aminophenyl)methylium][chloride]—was realized by
chromatography (basic Al2O3, CH2Cl2) and resulted in
the isolation of the desired ion pair [12][5a]. However, a
65:35 ratio of two diastereomeric salts (1H- and 31P-
NMR, DMSO-d6) was observed demonstrating that a
partial epimerization had occurred during the chro-
matography. This diastereoselectivity (d.r. 65:35)—dif-
ferent from the one observed for [nBu3NH][5a] in CDCl3
(d.r. 75:25)—demonstrates that the cationic counter-ion
plays also a role on the overall position of the thermo-
dynamic equilibrium between the diastereomers.
In conclusion, the high diastereoselectivity observed
for the [ammonium][5a] salts isolated by precipitation
was fortunate. When dissolved, an equilibration hap-
pened between the diastereomers; the rate and position
of the equilibrium depend on the solvent and on the
counter-ion. In all cases, a low diastereoselectivity
(d.r.B75:25) resulted from the epimerization.
2.6. Impro6ed asymmetric induction and configuration
stability from ligands 7b–7d
We reasoned that the low diastereoselectivity might
result from a lack of steric interactions between the
phenyl groups of 7a and the tetracatecholate ligands
due, possibly, to the high conformational flexibility of
the phenyl rings. In that case, introduction of ortho
substituents on the phenyl rings would increase the size
of the aromatic groups and rigidify the skeleton of the
ligand by creation of allylic A(1,3) strains between the
substituents and the stereogenic centers [25]. A higher
diastereoselectivity should then result from the modifi-
cation. Good literature precedents and ease of synthesis
led us to consider and prepare ligand 7b having ortho-
bromo substituents.
Anion 5b was then prepared—following the opti-
mized procedure—by reaction of 11 and diol 7b to give
the [Me2NH2][5b] salt in 68% yield and—in the precip-
itate—a high diastereoselectivity (Eq. (4), d.r. ꢀ98:2,
Fig. 10, spectrum a). Dissolution of salt [Me2NH2][5b]
led to an equilibration. In 15% DMSO-d6–CDCl3, a
90:10 diastereomeric ratio between the diastereomers of
5b (Fig. 10, spectrum b) was obtained (vs. 75:25 for 5a).
In MeOH, only a 68:32 ratio (Fig. 10, spectrum c) was
observed (vs. 61:39 for 5a).
Fig. 9. 1H-NMR (400 MHz) of (a) [Me2NH2][5a], DMSO-d6, d.r.
\96:4 and (b) [12][5a], DMSO-d6, d.r. 65:35 and (c) [Me2NH2][5a],
15% DMSO-d6–CDCl3, d.r. 75:25. Major (M) and minor (m)
diastereomers.
(4)
Fig. 10. 31P-NMR (162 MHz) of [Me2NH2][5b] in (a) DMSO-d6, 80
h., d.r. \96:4; (b) 15% DMSO-d6–CDCl3; 6 days, d.r. 90:10;
(c) MeOH-d4, 11 days, d.r. 68:32. Major (M) and minor (m)
diastereomers. TRISPHAT is indicated by the number 2.

398
J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
Fig. 11. Mechanistic rational for equilibration of the diastereomers of [Me2NH2][5a] salt.
A definite increase in diastereoselectivity was ob-
meric purity for the precipitates (d.r. \96:4 and 95:5
for 5c and 5d, respectively).
tained using ligand 7b. However, it was not sufficient to
lead to the preferred formation of a single diastereomer.
This led us to try to stabilize configurationally the
anions hopping that the good diastereomeric ratios
obtained at the precipitation stage would then remain
constant upon dissolution of isolated salts.
It was our analysis that the epimerization between
D-5a and L-5a (or D-5b–L-5b) resulted from the elec-
tron-rich nature of the oxygen atoms of ligand 7a (or
7b). They can be protonated by the acidic Me2NH2+
counter-ion. This labilize the PꢀO bonds and provokes
a ring opening to a configurationally labile spirophos-
phorane (Fig. 11) [26].
Upon dissolution in 5% DMSO-d6–CDCl3, a fast
epimerization of salts [Me2NH2][5c] and [Me2NH2][5d]
occurred. Equilibria were reached in less than 30 min
and a 71:29 diastereomeric ratio was obtained for both
salts. Clearly, the electron-poor aromatic nuclei of lig-
ands 7c–7d—not being directly connected to the oxy-
gen atoms—have
a reduced electron-withdrawing
ability and do not exert any concrete influence on the
rate of racemization.
If one considers that the opening of the ring carrying
the protonated oxygen is the rate-determining step [4],
then stopping the protonation of the electron-rich oxy-
gen atoms would be sufficient to obtain a configura-
tionally stable anion. The introduction of strong
electron-withdrawing trifluoromethyl groups on the
aromatic nuclei of the chiral ligands might thus help
stabilize the phosphates. Ligands 7c and 7d were thus
prepared for that goal [27].
Salts [Me2NH2][5c] and [Me2NH2][5d], derived from
ligands 7c and 7d respectively, were prepared using the
one-pot protocol in non-optimized 48 and 38% yields
respectively (Scheme 2). H-NMR analyses of the salts
in DMSO-d6 revealed—as usual—a high diastereo-
Scheme 2. (a) 6 (1.7 equivalents for 1.8 equivalents of 10), NH4Cl (2
mol%), toluene, reflux; (b) o-chloranil 8 (1.7 equivalents), 7c or 7d
(1.0 equivalents), solvent, 20 °C.
1

J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
399
3. Conclusion
NMR spectra are recorded at 295 K unless otherwise
stated. H-NMR spectra were recorded on Varian XL-
1
In this article, we have shown that C2-symmetric
hexacoordinated phosphate anions can be easily synthe-
sized in a ‘one-pot’ process from simple-to-make or
commercially available starting materials. These anions
can be obtained as solids in high diastereomeric purity.
However, once dissolved, they epimerize rapidly. Using
1,2-diphenyl-ethane-1,2-diol, a modest asymmetric in-
duction is observed (d.r. 75:25 in 5% DMSO–CHCl3),
which can be improved (d.r. up to 90:10) by the intro-
duction of ortho bromo substituents on the phenyl
rings. Despite all our efforts, we have not been able to
grow crystals suitable for X-ray analysis. The absolute
configuration of salt (−)-[Me2NH2][5a] derived from
(R,R)-7a remains so far unknown. Further studies are
conducted in our group to find a chiral ligand, which
would lead to high asymmetric induction and chemical
stability in the resulting phosphate anion.
200 (200 MHz) and Bru¨ker AMX-400 (400 MHz) spec-
trometers. Chemical shifts are given in ppm relative to
Me4Si with the solvent resonance used as the internal
standard. 31P-NMR spectra were recorded on a Bru¨ker
AMX-400 (162 MHz). Chemical shifts are reported in
ppm relative to H3PO4. 13C-NMR spectra were
recorded on Varian XL-200 (50 MHz), Bru¨ker AMX-
400 (100 MHz) spectrometers. Chemical shifts are given
in ppm relative to Me4Si, with the solvent resonance
used as the internal standard (CDCl3 77.0 ppm,
CD3OD 49.0 ppm, Me2SO-d6 39.5 ppm, acetone-d6 29.8
ppm). Assignments may have been achieved using
COSY, HETCOR and/or NOESY experimental data.
19F-NMR spectra were recorded on a Bru¨ker AMX-400
(376 MHz) spectrometer. Chemical shifts are reported
in ppm relative to C6F6. Optical rotations were mea-
sured on a Perkin Elmer 241 polarimeter in a ther-
mostated (20 °C) 10 cm long microcell with high
pressure sodium (u=589 nm) or mercury (u=578 nm)
lamps and are reported as follows: [h]2D0 (c g/100 ml,
solvent). Melting points (m.p.) were measured in open
capillary tubes on a Stuart Scientific SMP3 melting
point apparatus and are uncorrected. Electronspray
mass spectra (ES-MS) were obtained on a Finnigan
SSQ 7000 spectrometer by the Department of Mass
Spectroscopy. Mass spectra (MS) were obtained on a
Varian CH4 or SM1 spectrometer and relative intensi-
ties are given in parentheses (m/z, %). Elemental analy-
ses were performed at the ‘Institut de Chimie
Pharmaceutique de l’Universite´ de Gene`ve’ by Dr H.-J.
Eder.
4. Experimental
4.1. General considerations
All reactions were carried out under an atmosphere
of dry nitrogen or argon using an inert gas–vacuum
double manifold and standard Schlenk techniques with
magnetic stirring, unless otherwise stated. Solvents were
dried and distilled prior to use. Chloroform (Fluka)
and CDCl3 (SDS) were filtered on basic alumina before
use. Diethylether was distilled from sodium metal and
benzophenone ketyl. Toluene was distilled from sodium
metal. Methanol, hexane and dichloromethane were
distilled from calcium hydride. Deionized water was
used for aqueous solutions. Purifications of reaction
products were carried out by chromatography using
J.T. Baker silica gel (30–60 mm) or basic aluminum
oxide (Fluka, type 5016A). Analytical thin layer chro-
matography was performed on Macherey–Nagel 0.25
mm silica gel plates. Visualization was accomplished
with UV light. (R,R)-1,2-Diphenyl-ethane-1,2-diol (7a)
4.2. Preparation of [nBu3NH][5a] by reaction with PCl5
To a solution of PCl5 (208 mg, 1.00 mmol) in toluene
(4.0 ml) at 50 °C, tetrachlorocatechol 6 (496 mg, 2.00
mmol) was slowly added, and the mixture heated to
70 °C overnight. (R,R)-1,2-Diphenyl-ethane-1,2-diol
(7a, 214 mg, 1.00 mmol) and tri-n-butylamine (476 ml,
2.00 mmol) were then added and the mixture was
heated to 70 °C for a further 12 h. The precipitate was
then filtered, washed with CH2Cl2, and dried to afford
[nBu3NH][5a] as a white solid (150 mg, ꢀ13%) in
]96:4 mixture of diastereomers. 31P{1H}-NMR (162
MHz, Me2SO-d6, major (M) and minor (m)
and
(R,R)-1,2-bis-(2-bromo-phenyl)-ethane-1,2-diol
(7b) were synthesized according to literature procedures
[14,18].
In all the phosphate salts syntheses, we define the
stereoselectivity at the precipitation step by the NMR
analysis of the precipitate in Me2SO-d6, because that
this solvent seems to prevent fast epimerization between
the diastereomers, thus giving us a good image of the
real diastereomeric composition of the precipitate. Usu-
ally, only traces of a second diastereomer were observ-
able in Me2SO-d6, whereas, both diastereomers were
clearly seen upon dissolution of the precipitate in an-
other solvent or solvent mixture. In such cases, both
NMR interpretations are given.
1
diastereomers): l −79.3 (M); −79.8 (m, trace). H-
NMR (400 MHz, Me2SO-d6, major diasteromer): l 0.89
(t, J=7.5 Hz, 9H), 1.30 (m, 6H), 1.55 (m, 6H), 2.99 (m,
6H), 4.76 (s, 2H), 7.02 (m, 4H), 7.23 (m, 6H), 8.84 (s
broad, 1H); minor diastereomer not sufficiently visible
in this solvent to be reliably described. ES–MS; m/z:
(−) 735.0; (+) 186.1.

400
J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
4.4. Preparation of [tris(para-dimethylamino-
phenyl)methylium][5a] or [12][5a]
4.3. One-pot preparation of (−)-dimethylammonium
bis(tetrachlorobenzenediolato)mono((R,R)-1,2-diphenyl-
ethane-1,2-diolato)phosphate(V) or (−)-[Me2NH2][5a]
[nBu3NH][5a] (23.0 mg, 2.46 mmol) and crystal violet
(10.2 mg, 2.50 mmol) were dissolved in CH2Cl2 (3.0 ml)
and the resulting solution stirred for 5 min and then
concentrated in vacuo. Purification by chromatography
(basic Al2O3, 0.5×2 cm, CH2Cl2) afforded 7 mg (0.65
mmol, 25%) of [12][5a] as a dark purple solid containing
a 65:35 mixture of diastereomers: 31P{1H}-NMR (162
MHz, Me2SO-d6, major (M) and minor (m)
diastereomers): l −79.3 (M), −79.8 (m). 1H-NMR
(400 MHz, Me2SO-d6, major (M) and minor (m)
diastereomers): l 3.07 (s, 18H), 4.46 (s, 2H, m), 4.64 (s,
2H, M), 6.85 (m, 9H), 7.12 (m, 13H). ES–MS; m/z:
(−) 735.0; (+) 372.2.
Tetrachlorocatechol 6 (211 mg, 0.85 mmol) and a
catalytic amount of NH4Cl (one crystal) were mixed
together in dry toluene (2.0 ml). Freshly distilled tris-
(dimethylamino)phosphine 10 (163 ml, 0.90 mmol) was
then slowly added. Dimethylamine evolved and the
mixture was refluxed for 15 min. The remaining solvent
and excess of 10 were removed under reduced pressure
and the residue carefully dried in vacuo. (R,R)-1,2-
Diphenyl-ethane-1,2-diol 7a (107 mg, 0.50 mmol) and
ortho-chloranil 8 (209 mg, 0.85 mmol) were added and
the mixture dissolved in a mixture of CH2Cl2–hexane
4:1 (5 ml). The dark red solution progressively lost its
color within 4 h and a precipitate was formed. When
the color did not change anymore, the precipitate was
filtered, washed carefully successively with the CH2Cl2–
hexane 4:1, toluene and CH2Cl2–hexane 4:1 again. It
was then dried to afford [Me2NH2][5] as a single
diastereomer (1H,31P-NMR) however along with 8% of
inseparable [Me2NH2][2] salt (304 mg, 70%). The
product could be further purified by trituration in
CH2Cl2. [h]2D0= −101 (c 0.99; Me2SO). M.p. \210 °C
(decomposition). 31P{1H}-NMR (162 MHz, Me2SO-d6,
4.5. Preparation of tetrachlorobenzoŽ1,3,2-
dioxaphosphol-2-yl-dimethyl-amine (11)
In a sublimating flask and under a nitrogen atmo-
sphere, tetrachlorocatechol (6, 1.24 g, 5.0 mmol) and a
catalytic amount of NH4Cl (ꢀ10 mol%) were mixed
together in dry toluene (5.0 ml). Freshly distilled tris-
(dimethylamino)phosphine 10 (1.09 ml, 6.0 mmol) was
slowly added. Dimethylamine evolved. The mixture was
refluxed for 1 h and 30 min. The remaining solvent and
excess of 10 were removed under reduced pressure. The
gray residue was sublimated (170 °C, 0.06 mbar) to
afford 11 (1.56 g, 97%) as a white crystalline solid. m.p.
1
major diastereomer): l −79.3; H-NMR (400 MHz,
Me2SO-d6, major diastereomer): l 2.53 (s, 6H, Me),
4.76 (s, 2H), 7.01 (m, 4H), 7.23 (s broad, 6H), 8.15
(s broad, 2H, NH2). 13C{1H}-NMR (100 MHz,
Me2SO-d6, major diastereomer): l 34.3 (NCH3), 80.4
(CH), 111.9 (Cq, d, JCP=19 Hz), 112.1 (Cq, d, JCP=18
Hz), 119.8 (Cq), 119.9 (Cq), 126.7 (CH), 128.1 (CH),
128.2 (CH), 139.1 (Cq, d, JCP=14 Hz), 142.7 (Cq, d,
1
125–130 °C. H-NMR (400 MHz, C6D6): l 1.85 (d,
J
HP=9.2 Hz, 6H). 13C{1H}-NMR (100 MHz, C6D6): l
35.0 (NCH3), 115.3 (Cq), 125.0 (Cq), 143.5 (Cq, d,
J
CP=8.0 Hz). 31P{1H}-NMR (162 MHz, C6D6): l
156.9.
J
CP=5.4 Hz), 143.0 (Cq, d, JCP=5.3 Hz). ES-MS; m/z:
4.6. Preparation of dimethylammonium
bis(tetrachlorobenzenediolato)mono((R,R)-1,2-
bis(2-bromo-phenyl))-ethane-1,2-diolato)phosphate(V)
or (−)-[Me2NH2][5b]
(−) 735.0. Anal. Found: C, 41.72; H, 2.93; N, 1.85.
Calc. for C28H20Cl8NO6P (92%) and C20H8Cl12NO6P
(8%, [Me2NH2][TRISPHAT]): C, 41.97; H, 2.45; N,
1.78%.
In 15% Me2SO-d6–CDCl3, both diastereomers can be
observed after epimerization: 31P{1H}-NMR (162
MHz, major (M) and minor (m) diastereomers): l
−82.1 (m), −82.3 (M). 13C{1H}-NMR (100 MHz,
major (M) and minor (m) diastereomers): l 35.3
(NCH3), 80.2 (CH, m), 81.2 (CH, M), 113.1 (Cq, d,
(−)-(R,R)-1,2-Bis-(2-bromo-phenyl)-ethane-1,2-diol
(7b, 186 mg, 0.50 mmol), phosphoramidite 11 (161 mg,
0.50 mmol), and ortho-chloranil (123 mg, 0.50 mmol)
were mixed together under a nitrogen atmosphere. The
mixture was dissolved in Et2O (5 ml). The dark red
solution progressively lost its color within 2 h and a
precipitate was formed. When the color did not change
anymore, the precipitate was filtered, washed succes-
sively with Et2O, toluene and Et2O again, and dried to
afford the desired dimethylammonium phosphate salt
as a white solid (350 mg, 68%) a 98:2 mixture of
diastereomers and 2% of inseparable [Me2NH2][2].
JCP=18 Hz, M), 113.2 (Cq, d, JCP=19 Hz, m), 113.3
(Cq, d, JCP=19 Hz, M), 113.9 (Cq, d, JCP=20 Hz, m),
121.6 (Cq, m), 121.7 (Cq, M), 121.8 (Cq, M), 123.0 (Cq,
m), 126.7 (CH, M), 127.6 (CH, m), 127.6 (CH, M),
127.9 (CH, M), 128.2 (CH, m), 128.3 (CH, m), 138.3
(Cq, d, JCP=13 Hz, M), 138.4 (Cq, d, JCP=11 Hz, m),
141.1 (Cq, d, JCP=6.6 Hz, m), 142.0 (Cq, d, JCP=5.8
Hz, M), 142.2 (Cq, d, JCP=4.9 Hz, M), 143.0 (Cq, d,
31P{1H}-NMR
(162
MHz,
Me2SO-d6,
major
1
diastereomer): l −79.5. H-NMR (400 MHz, Me2SO-
d6, major diastereomer): l 2.53 (s, 6H), 5.23 (s, d,
JCP=5.3 Hz, m).

J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
401
4.8. Dimethylammonium bis(tetrachlorobenzenediolato)-
JHP=0.8 Hz, 2H), 7.16 (m, 2H), 7.34 (m, 4H), 7.41 (d,
mono(1,2-bis(para-trifluoromethylphenyl))-ethane-
1,2-diolato)phosphate(V) or [Me2NH2][5c]
J=7.9 Hz, 2H), 8.15 (s broad, 2H); minor diastereomer
not sufficiently visible to be reliably described. 13C{1H}-
NMR (100 MHz, Me2SO-d6, major diastereomer): l
This compound was prepared from 7c in a similar
way to (−)-[Me2NH2][5a] using the one-pot protocol
and 6 (126 mg, 0.51 mmol), 10 (98 ml, 0.54 mmol), 8
(125 mg, 0.51 mmol) and 7c (105 mg, 0.30 mmol).
CH2Cl2 was used as solvent and a reaction time of 3 h
needed. [Me2NH2][5c] was obtained as a white solid
(168 mg, ꢀ48%) and contained 20% of inseparable
[Me2NH2][2]. 31P{1H}-NMR (162 MHz, Me2SO-d6, ma-
jor diastereomer): l −79.1. 19F-NMR (376 MHz,
Me2SO-d6, major diastereomer): l 102.5. 1H-NMR (400
MHz, Me2SO-d6, major diastereomer): l 2.53 (s, 6H),
4.86 (d, J=0.8 Hz, 2H), 7.24 (d, J=8.0 Hz, 4H), 7.65
(d, J=8.0 Hz, 4H), 8.15 (s broad, 2H). ES-MS; m/z:
(−) 870.4.
In 5% Me2SO-d6–CDCl3, both diastereomers can be
observed after epimerization: 31P{1H}-NMR (162 MHz,
major (M) and minor (m) diastereomers): l −82.4
(M), −82.0 (m). 19F-NMR (376 MHz, major (M) and
minor (m) diastereomers): l 101.2 (M), 101.3 (m).
1H-NMR (400 MHz, major (M) and minor (m)
diastereomers): l 2.53 (s, 6H, Me), 4.86 (d, J=0.8 Hz,
2H, m), 4.91 (m, 2H, M), 7.14 (d, J=8.0 Hz, 4H,
M+m), 7.50 (d, J=8.0 Hz, 4H, M), 7.56 (d, J=7.8
Hz, 4H, m), 8.25 (s broad, 2H, NH2).
34.3 (NCH3), 78.3 (CH), 81.5 (Cq), 112.0 (Cq, d, JCP
=
19 Hz), 112.3 (Cq, d, JCP=18 Hz), 120.1 (Cq), 120.2
(Cq), 122.1 (Cq), 128.0 (CH), 128.9 (CH), 129.8 (CH),
132.2 (CH), 137.9 (Cq, d, JCP=14 Hz), 142.6 (Cq, d,
J
CP=28 Hz). ES-MS; m/z: (−) 892.3. Anal. Found: C,
35.90; H, 2.21; N, 1.54. Calc. for C28H18Br2Cl8NO6P:
C, 35.82; H, 1.93; N, 1.49%.
In 15% Me2SO-d6–CDCl3, both diastereomers can be
observed after epimerization: 31P{1H}-NMR (162 MHz,
major (M) and minor (m) diastereomers): l −82.1 (m),
−82.3 (M). 1H-NMR (400 MHz, major (M) and minor
(m) diastereomers): l 2.63 (s, 6H, Me), 5.33 (s, d,
JHP=1.7 Hz, 2H, m), 5.38 (s, 2H, M), 7.03 (m, 2H, M),
7.06 (m, 2H, m), 7.28 (m, 4H, M), 7.32 (m, 4H, m),
7.43 (dd, J=8.0, 1.3 Hz, 2H, M), 8.07 (d, J=6.6 Hz,
2H, m), 8.23 (s broad, 2H, NH2).
4.7. Preparation of rac-1,2-bis(para-trifluoromethyl-
phenyl)-ethane-1,2-diol] (7c)
Preparation of the copper reagent: 7.5 g (30.0 mmol)
of CuSO4·5H2O and 1.96 g (30.0 mmol) of Zn were
stirred in 100 ml of water until complete precipitation
of copper(0) was observed. The precipitate was filtered,
washed with water, acetone, dry Et2O, and dried in
vacuo to yield 1.90 g of copper.
4.9. Preparation of rac-1,2-bis(3,5-bistrifluoromethyl)-
phenyl-ethane-1,2-diol (7d)
TiCl4 (1.64 ml, 14.9 mmol) was added dropwise to a
suspension of Cu (1.90 g, 30.0 mmol) in dry THF (25
ml) under a nitrogen atmosphere. The temperature of
the reaction medium was regulated with a water bath.
A brown precipitate was observed and the reaction
mixture refluxed for 2 h. To the resulting green precip-
itate, para-trifluoromethylbenzaldehyde (683 ml, 5.0
mmol) was added dropwise and the mixture refluxed
for another 2 h. It was left to cool overnight, then
hydrolyzed by an aqueous solution of 4.50 g of
trisodium citrate. The solid residue was filtered off and
most of the solvent was removed under reduced pres-
sure. The resulting oil was extracted with Et2O, washed
with brine, dried over anhydrous Na2CO3, filtered and
concentrated in vacuo to afford 750 mg of a white
solid. Purification by chromatography (SiO2, hexane–
AcOEt 70:30 up to 50:50) afforded 225 mg of 7c (26%)
Compound 7d was prepared in a similar way as for
7c using TiCl4 (3.28 ml, 30.0 mmol), Cu (3.60 g, 60.0
mmol), 3,5-bis-trifluoromethyl-benzaldehyde (2.42 g,
10.0 mmol) to afford, after recrystallization (toluene) of
the crude product, 379 mg of 7d (15% yield). M.p.
1
135–137 °C. H-NMR (400 MHz, CDCl3): l 3.06 (s,
2H), 4.87 (s, 2H), 7.52 (s, 4H), 7.82 (s, 2H). 19F-NMR
(376 MHz, CDCl3): l 100.6. 13C{1H}-NMR (100 MHz,
CDCl3): l 77.7 (CH), 122.3 (CH, q, J=3.0 Hz), 123.0
(Cq, q, J=273 Hz), 127.0 (CH), 131.8 (Cq, q, J=33
Hz), 141.5 (Cq). EI–MS; m/z: 467, 449, 243, 195, 145,
127, 69.
4.10. Dimethylammonium bis(tetrachlorobenzene-
diolato)mono(1,2-bis(3,5-trifluoromethylphenyl))-ethane-
1,2-diolato)phosphate(V) or [Me2NH2][5d]
1
with a DL-meso ratio of 93:7. M.p. 139–143 °C. H-
This compound was prepared from 7d in a similar
way to (−)-[Me2NH2][5a] using the one-pot protocol
and 6 (211 mg, 0.85 mmol), 10 (165 ml, 0.91 mmol), 8
(209 mg, 0.85 mmol) and 7d (243 mg, 0.50 mmol).
CH2Cl2 was used as solvent and a reaction time of 4.25
h needed. [Me2NH2][5d] was obtained as a white solid
(200 mg, 38%) as essentially a single diastereomer.
NMR (400 MHz, CDCl3): l 3.03 (s, 2H), 4.76 (s,
2H), 7.23 (d, J=8.3 Hz,, 4H), 7.53 (d, J=8.3 Hz,
4H). 19F-NMR (376 MHz, CDCl3): l 101.1. 13C{1H}-
NMR (100 MHz, CDCl3): l 78.4, 123.9 (q, J=259
Hz), 125.3 (q, J=3.8 Hz), 127.3, 130.5 (q, J=32 Hz),
144.4.

402
J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
31P{1H}-NMR (162 MHz, Me2SO-d6, major diastereo-
mer): l −79.8. 19F-NMR (376 MHz, Me2SO-d6, major
diastereomer): l 101.8. H-NMR (400 MHz, Me2SO-d6,
major diastereomer): l 2.53 (s, 6H), 5.02 (d, JHP=2.8
Hz, 2H), 7.67 (s, 2H), 8.01 (s, 4H), 8.15 (s broad, 2H).
ES–MS; m/z: (−) 1006.4.
In 5% Me2SO-d6–CDCl3, both diastereomers can be
observed after epimerization: 31P{1H}-NMR (162 MHz,
major (M) and minor (m) diastereomers): l −81.6 (m),
−82.3 (M). 19F-NMR (376 MHz, major (M) and mi-
nor (m) diastereomers): l 100.6 (M), 100.9 (m). 1H-
NMR (400 MHz, major (M) and minor (m)
diastereomers): l 2.53 (s, 6H, Me), 4.85 (d, JHP=2.4
Hz, 2H, m), 4.94 (d, JHP=1.5 Hz, 2H, M), 7.46 (s, 4H,
M+m), 7.83 (s, 2H, M), 8.00 (s, 4H, m), 8.14 (s broad,
2H, NH2).
(c) J.J. Jodry, J. Lacour, Chem. Eur. J. 6 (2000) 4297;
(d) D. Monchaud, J. Lacour, C. Coudret, S. Fraysse, J.
Organomet. Chem. 624 (2001) 388.
1
[6] J. Lacour, A. Londez, C. Goujon-Ginglinger, V. Buß, G.
Bernardinelli, Org. Lett. 2 (2000) 4185.
[7] (a) J.K. Whitesell, Chem. Rev. 89 (1989) 1581;
(b) A. Alexakis, P. Mangeney, Tetrahedron Asymm. 1 (1990)
477;
(c) C. Rosini, L. Franzini, A. Raffaelli, P. Salvadori, Synthesis
(1992) 503.
[8] (a) A. von Zelewsky, Stereochemistry of Coordination Com-
pounds, Wiley, Chichester, 1996;
(b) U. Knof, A. von Zelewsky, Angew. Chem. Int. Ed. Engl. 38
(1999) 302.
[9] J.L. Pierre, Coord. Chem. Rev. 178–180 (1998) 1183.
[10] (a) M. Yashiro, T. Murata, S. Yoshikawa, S. Yano, J. Chem.
Soc. Dalton Trans. (1994) 1073;
(b) M. Yashiro, T. Murata, M. Sato, M. Kosaka, K. Kobayashi,
T. Sakurai, S. Yoshikawa, M. Komiyama, S. Yano, J. Chem.
Soc. Dalton Trans. (1994) 1511;
(c) A. Tatehata, Inorg. Chem. 21 (1982) 2496;
(d) R.S. Vagg, P.A. Williams, Inorg. Chim. Acta 51 (1981) 61;
(e) T.J. Goodwin, P.A. Williams, R.S. Vagg, Inorg. Chim. Acta
63 (1982) 133.
Acknowledgements
[11] (a) A. Munoz, G. Gence, M. Koenig, R. Wolf, Bull. Soc. Chim.
Fr. (1975) 1433;
We thank the Swiss National Foundation (AL, JL),
the Federal Office for Education and Science (COST
D11, JL), and the ‘Socie´te´ Acade´mique de Gene`ve’ for
financial support.
(b) M. Koenig, A. Klaebe, A. Munoz, R. Wolf, J. Chem. Soc.
Perkin Trans. II (1976) 955.
[12] BINPHAT 4—synthesized in
a one-pot process and good
yield—is however sensitive to acids and decompose rapidly in
CHCl3 or CH2Cl2 when associated with acidic ammonium
counterions.
References
[13] For recent applications of 7a and its derivatives, see (a) E.P.
Kundig, C.M. Saudan, F. Viton, Adv. Synth. Catal. 343 (2001)
51;
[1] (a) M.J. Gallagher, I.D. Jenkins, in: E.L. Eliel, S.H. Wilen
(Eds.), Topics in Stereochemistry, vol. 3, Wiley, New York,
1968, p. 76;
(b) K. Ding, A. Ishii, K. Mikami, Angew. Chem. Int. Ed. 38
(1999) 497;
(c) M. Hasegawa, D. Taniyama, K. Tomioka, Tetrahedron 56
(2000) 10153;
(d) C.A. Ray, T.W. Wallace, R.A. Ward, Tetrahedron Lett. 41
(2000) 3501;
(e) H. Fujioka, N. Kotoku, Y. Nagatomi, Y. Kita, Tetrahedron
Lett. 41 (2000) 1829;
(b) D. Hellwinkel, in: G.M. Kosolapoff, L. Maier (Eds.), Or-
ganic Phosphorus Compounds, vol. 3, Wiley, New York, 1972,
p. 251;
(c) R. Luckenbach, in: M. Regitz (Ed.), Methoden der Organis-
chen Chemie. Band E2, Georg Thieme Verlag, Stuttgart, 1982, p.
897;
(d) R.A. Cherkasov, N.A. Polezhaeva, Russ. Chem. Rev. 56
(1987) 163;
(e) R. Burgada, R. Setton, in: F.R. Hartley (Ed.), The Chemistry
of Organophosphorus Compounds, vol. 3, Wiley, New York,
1994, p. 185;
(f) M. Shindo, K. Koga, K. Tomioka, J. Org. Chem. 63 (1998)
9351;
(g) E.P. Ku¨ndig, D. Amurrio, G. Anderson, D. Beruben, K.
Khan, A. Ripa, L. Ronggang, Pure Appl. Chem. 69 (1997) 543.
[14] (a) Z.-M. Wang, K.B. Sharpless, J. Org. Chem. 59 (1994) 8302;
(b) K.B. Sharpless, W. Amberg, Y.L. Bennani, G.A. Crispino, J.
Hartung, K.-S. Jeong, H.-L. Kwong, K. Morikawa, Z.-M.
Wang, D. Xu, X.-L. Zhang, J. Org. Chem. 57 (1992) 2768.
[15] (a) K. Ishimaru, K. Monda, Y. Yamamoto, K.-Y. Akiba, Synth.
Commun. 30 (2000) 575;
(f) R.R. Holmes, Chem. Rev. 96 (1996) 927;
(g) R. Holmes, Acc. Chem. Res. 31 (1998) 535.
[2] (a) D. Hellwinkel, Angew. Chem. (1965) 378;
(b) D. Hellwinkel, S.F. Mason, J. Chem. Soc. (B) (1970) 640;
(c) D. Hellwinkel, W. Krapp, Phosphorus 6 (1976) 91.
[3] (a) M. Koenig, A. Klaebe, A. Munoz, R. Wolf, J. Chem. Soc.
Perkin Trans. II (1979) 40;
(b) B.M. Kim, J.K. Park, Bull. Korean Chem. Soc. 20 (1999)
744.
(b) A. Klaebe, M. Koenig, R. Wolf, P. Ahlberg, J. Chem. Soc.
Dalton Trans. (1977) 570;
(c) J. Cavezzan, G. Etemad-Moghadam, M. Koenig, A. Klaebe,
Tetrahedron Lett. (1979) 795.
[16] T.R. Kelly, Q. Li, V. Bhushan, Tetrahedron Lett. 31 (1990) 161.
[17] T.W. Wallace, I. Wardell, K.-D. Li, P. Leeming, A.D. Redhouse,
S.R. Challand, J. Chem. Soc. Perkin Trans I (1995) 2293.
[18] S.Y. Zhang, C. Girard, H.B. Kagan, Tetrahedron Asymmetry 6
(1995) 2637.
[4] (a) J. Lacour, C. Ginglinger, C. Grivet, G. Bernardinelli, Angew.
Chem. Int. Ed. Engl. 36 (1997) 608;
[19] D. Cuvinot, A. Alexakis, private communication.
[20] The two diastereomers—when they occur—can be observed in
DMSO-d6: 31P-NMR (162 MHz, l −79.3 and −79.8, e.g. Fig.
9).
[21] R.B. Merrifield, G. Barany, W.L. Cosand, M. Engelhard, S.
Mojsov, Pept. Proc. Am. Pept. Symp. 5 (1977) 488.
(b) J. Lacour, C. Ginglinger, F. Favarger, Tetrahedron Lett.
(1998) 4825.
[5] (a) J. Lacour, J.J. Jodry, C. Ginglinger, S. Torche-Haldimann,
Angew. Chem. Int. Ed. Engl. 37 (1998) 2379;
(b) J. Lacour, C. Goujon-Ginglinger, S. Torche-Haldimann, J.J.
Jodry, Angew. Chem. Int. Ed. 39 (2000) 3695;

J. Lacour, A. Londez / Journal of Organometallic Chemistry 643–644 (2002) 392–403
[22] For examples of P(III) derivatives of 1,2-diphenyl-ethane-1,2-di- Chem. Ber. 125 (1992) 1325;
403
ols, see (a) A.A. Bredikhin, Z.A. Bredikhina, L.M. Gaisina, E.I.
Strunskaya, N.M. Azancheev, Russ. Chem. Bull. 49 (2000) 310;
(b) D.J. Wink, T.J. Kwok, A. Yee, Inorg. Chem. 29 (1990) 5006;
(c) J. Nielsen, O. Dahl, J. Chem. Soc. Perkin Trans. II (1984)
553;
(d) D. Bernard, R. Burgada, Tetrahedron 31 (1975) 797;
(e) Y. Charbonnel, J. Barrans, C.R. Acad. Sci. Ser. C 277 (1973)
571.
(c) I.V. Shevchenko, P.G. Jones, A. Fischer, R. Schmutzler,
Heteroatom. Chem. 3 (1992) 177.
[24] J. Lacour, S. Barche´chath, J.J. Jodry, C. Ginglinger, Tetrahe-
dron Lett. 39 (1998) 567.
[25] R.W. Hoffmann, Chem. Rev. 89 (1989) 1841.
[26] There has been no direct observation of phosphorane intermedi-
ates. They can be bipyramidal, square-planar pyramidal or any
structure in between. See Ref. [1f].
[23] (a) J. Krill, I.V. Shevchenko, A. Fischer, P.G. Jones, R. Schmut-
zler, Chem. Ber. 130 (1997) 1479;
[27] Ligands 7c and 7d were prepared in racemic form as it is not
required to use enantiomerically enriched ligands to observed by
NMR the diastereomers of the derived anions 5c and 5d.
(b) I.V. Shevchenko, A. Fischer, P.G. Jones, R. Schmutzler,