Development of the first P-stereogenic PCP pincer ligands, their metallation by palladium and platinum, and preliminary catalysis

Article abstract of DOI:10.1002/1522-2675(20011114)84:11<3519::AID-HLCA3519>3.0.CO;2-9

The potentially tridentate P-stereogenic [P*CP*] ligands 1,3-{(bis[(tert-butyl)(phenyl)phosphino]methyl)-benzene and 1,3-{(bis[(tert-butyl)(phenyl) phosphino]methyl})1-2-bromobenzene have been synthesized as the protected phosphine-borane adducts. Deprotection with a secondary amine affords the free phosphine ligand which can be metallated by Pd and Pt with standard metal synthons. Two of the resultant [P*CP*] metal complexes have been characterized by X-ray crystallography. The complexes exhibit a C₂ symmetric environment about the remaining binding site of the square-planar center, with t-Bu groups filling two quadrants of the open site. The Pd complexes can be converted by use of a Ag salt to the analogous aquo complex, which is catalytically active in the aldol condensation of methyl 2-isocyanoacetate and benzaldehyde. Preliminary results and comparisons with previously reported catalysts with more distal C-stereogenicity are presented.

Full text of DOI:10.1002/1522-2675(20011114)84:11<3519::AID-HLCA3519>3.0.CO;2-9

Helvetica Chimica Acta Vol. 84 (2001)
3519
Development of the First P-Stereogenic PCP Pincer Ligands, Their
Metallation by Palladium and Platinum, and Preliminary Catalysis1)
a
b
by B.Scott Williams ), Paulo Dania)2), Martin Lutzb), Anthony L.Spek* ), and Gerard van Koten*a)3)
a) Department of Metal-Mediated Synthesis, Debye Institute, Universiteit Utrecht, Padualaan 8, 3584 CH,
Utrecht, The Netherlands
b) Bijvoet Center for Biomolecular Research, Crystal and Structural Chemistry, Universiteit Utrecht,
Padualaan 8, 3584 CH, Utrecht, The Netherlands
In memoriam Professor Luigi M. Venanzi for his pioneering work in organometallic chemistry, and in PCP-
pincer chemistry in particular
The potentially tridentate P-stereogenic [P*CP*] ligands 1,3-{bis[(tert-butyl)(phenyl)phosphino]methyl}-
benzene and 1,3-{bis[(tert-butyl)(phenyl) phosphino]methyl}-2-bromobenzene have been synthesized as the
protected phosphine-borane adducts. Deprotection with a secondary amine affords the free phosphine ligand
which can be metallated by Pd and Pt with standard metal synthons. Two of the resultant [P*CP*] metal
complexes have been characterized by X-ray crystallography. The complexes exhibit a C2 symmetric
environment about the remaining binding site of the square-planar center, with t-Bu groups filling two
quadrants of the open site. The Pd complexes can be converted by use of a Ag salt to the analogous aquo
complex, which is catalytically active in the aldol condensation of methyl 2-isocyanoacetate and benzaldehyde.
Preliminary results and comparisons with previously reported catalysts with more distal C-stereogenicity are
presented.
Introduction. Transition metal complexes of so-called −pincer ligands× have found
wide application in a variety of fields since this ligand type was first reported by
Moulton and Shaw in 1976 [1]. The ligands can be abbreviated as −[ECE]× ligands,
where E is the donating heteroatom. A number of early reactivity studies showed that
the chemistry of square-planar late-transition-metal pincer complexes of this type is
particularly rich ([2] and references therein). Subsequent studies have demonstrated
the utility of such complexes in Lewis acid catalysis [2a][3], Heck reactions [4],
Kharasch additions [5], catalytic dehydrogenation [6], transfer hydrogenation [7], and
investigations of CÀH and CÀC oxidative addition reactions ([8] and references in
[8a]).
Recently, many efforts have been concentrated upon the immobilization of the
catalytic systems upon recoverable supports, such as dendrimers [5c][9], highly
symmetric organic frameworks [10], polymers [11], fullerenes [12], and solid supports
[13]. Because of these advances in attachment of pincer ligands to recoverable
supports, any new modification of the pincer framework or discovery of new patterns of
reactivity can be applied in a straightforward manner to a practical catalytic process, in
which the catalyst can be easily separated from the reaction medium, and thus the
1
)
)
)
This work was presented in part at OMCOS 11, Taipei, Republic of China, in July 2001.
Present Address: Avantium Technologies B. V., P. O. Box 2915, 1000 CX, Amsterdam, The Netherlands.
e-mail: g.vankoten@chem.uu.nl.
2
3

3520
Helvetica Chimica Acta Vol. 84 (2001)
products of the reaction. For this reason, a major focus of the research in our laboratory
is the development of the basic chemistry of pincer-supported organometallic catalysts.
Despite the wide variety of catalytically active pincer complexes that have been
reported, relatively few attempts have been made to explore enantioselective reactions
with chiral versions of the pincer system [14]4). We report herein the synthesis and
procedures for the palladation and platination of a new type of [PCP] pincer ligand, as
well as some initial results of the use of these complexes in a catalytic test reaction,
namely the asymmetric aldol condensation of methyl 2-isocyanoacetate and benzalde-
hyde to form chiral dihydrooxazoles (Scheme 1).
Scheme 1. The Aldol Condensation of Methyl 2-Isocyanoacetate and Benzaldehyde
This asymmetric catalytic reaction was initially reported by Hayashi, Ito, and co-
workers, who used a Au catalyst [15], and was also the subject of detailed further study
by Togni and Pastor under similar conditions [16]. Pregosin and co-workers showed this
reaction to be catalysed by PdII centres as well [17]. This reaction is interesting both
because it is a potential route to b-hydroxy-amino acids, and because it involves the
formation of a CÀC bond with simultaneous creation of two chiral centres, resulting in
4
)
While not true pincer ligands in the strict sense as defined above, the tridentate ligands developed by
Richards et al. ([14c] and references therein) as well as those of Bergman, Tilley, and co-workers [14d] are
closely related, and should be kept in mind in discussions of this area.

Helvetica Chimica Acta Vol. 84 (2001)
3521
four possible stereoisomers. It is thus a useful test reaction for exploring the chiral
induction provided by new, chiral Lewis acid catalysts.
In 1994, Venanzi and co-workers prepared a Pt complex of a [PCP] ligand, in which
the benzylic positions of the ligand had been modified to form stereogenic centres
[14a]. This complex was found to be an active catalyst for the aforementioned aldol
condensation. While the enantiomeric excesses and diastereoisomeric selectivities
obtained were moderate for this [PCP]-Pt complex, it was clearly demonstrated that
this sort of species was capable of chiral induction. More recently, Zhang and co-
workers have prepared Pd and Pt complexes of a similar chiral [PCP] ligand, and tested
their activities in the same aldol condensation [14b]5). While selectivities remained
modest, the synthetic approach used to synthesize the chiral ligand was considerably more
straightforward than that of Venanzi and co-workers, thus increasing the potential use-
¬
fulness of the system. In related studies, Gimenez and Swager have prepared chiral [SCS]-
Pd complexes in which the chirality is located on a tail attached to C(4) of the [SCS]
aryl ring, although no significant chiral induction was observed with this catalyst [19].
Since previous attempts to catalyse this reaction with pincer-supported metal
complexes had centered upon complexes containing the chiral information at the
benzylic C-atom, it seemed a logical approach to synthesize catalysts with the chiral
information to be found at the P-atom itself, since the binding pocket of square-planar
[PCP] complexes is quite deep, and thus the R-groups bound to the P-atom are in close
proximity to a potential substrate. We therefore set out to prepare the C2-symmetric
complexes with sterically demanding R-groups.
Results and Discussion. Preparation of P-Stereogenic PCP Ligands [P*CP*]. The
racemic monophosphorus compound tert-butyl(phenyl)phosphane-borane 1 was
prepared in reasonable (67%) yield by a modification of the procedure of Imamoto
et al. (Scheme 2) [20]. This secondary phosphine-borane can be deprotonated, as was
shown by Wolfe and Livinghouse, by BuLi in Et2O, in the presence of (À)-sparteine
[21]. Warming this mixture to 308 for 1 h, followed by cooling and reaction with either
1,3-bis(bromomethyl)benzene or 1,3-bis(bromomethyl)-2-bromobenzene resulted in
the formation of the terdentate [P*CP*] ligands 2a and 2b, respectively, in
diastereoisomerically pure form (yields 53% purified and 99% crude, resp.).
5
)
This same ligand has also been used for catalytic dehydrogenation with a Ru catalyst (see [18]).

3522
Helvetica Chimica Acta Vol. 84 (2001)
Scheme 2 Synthesis of [P*CP*] Ligands 2a,b
i) 1) t-BuLi, Et2O, À788; 2) LiAlH4, Filtration; 3) Me2SBH3. ii) 1) BuLi, (À)-sparteine, Et2O, À788; 2) 308,
90 min; 3) di- or tribromoxylene, À788.
The most striking spectral feature of these species is the sharp doublet (3J(P,H) ˆ
14 Hz in both cases) assigned to the t-Bu group. If the alkylation of the bromoxylenes is
carried out in the absence of (À)-sparteine or in THF in place of Et2O, a mixture of both
diastereoisomers is obtained, in which the t-Bu doublets of both diastereoisomers are
1
clearly separated in the H-NMR spectrum. Presumably in the latter case, the THF
‡
coordinates strongly to the Li cation, preventing sparteine complexation, thereby
rendering the reaction completely nonstereoselective. The Me analogue of 2a, 1,3-
(PMePh(BH3)CH2)2C6H4, can be prepared from PHMePh(BH3) and 1,3-bis(bromo-
methyl)benzene, yet surprisingly, even when the reaction has been performed in Et2O
and in the presence of (À)-sparteine, an equimolar mixture of diastereoisomers has
been obtained. This is presumably a result of the difference in steric congestion
between the lithium methyl-phosphide-sparteine complex and the lithium tert-butyl-
phosphide-sparteine complex.
Epimerization of Free Phosphine Centres. The phosphine boranes 2a,b can be
deprotected with an excess of various types of amines [20][21]6) to form the related
phosphines 3a,b, which are potential [P*CP*] ligand precursors (Scheme 3). Since the
usefulness of these ligands is dependant upon their barrier to epimerization, a sample of
3a was deprotected with Et2NH, the volatiles were removed, and C6D6 was added via
1
vacuum transfer. The epimerization of 3a at 808 was monitored by H-NMR in a J.
Young NMR tube. A barrier of 28.1 Æ 0.1 kcal/mol was determined for the loss of chiral
purity7). Such an inversion barrier is actually rather low for phenyldialkyl phosphines
[22], but sufficiently high that epimerization of metal catalysts derived from this ligand
would not be a significant consideration, as quaternerization of the P-atom makes it
configurationally stable8).
6
)
The most convenient methods have been the use of pyrrolidine (neat, 408, ca. 12 h) or Et2NH (50/50
mixture with THF, 558, 1.5 h). The amine can be subsequently removed by evaporation. No sign of
inversion has been observed under these relatively mild conditions, though epimerization becomes rapid at
higher temperatures and over longer times (vide infra). In contrast to the report in [21], we have not been
able to remove the amine-boranes by sublimation.
7
8
)
This barrier is for the rac to meso epimerization, and was determined by use of the Eyring equation.
Keq(375 K) ˆ 0.33, thus the rac form is slightly more stable, so the barrier for meso-to-rac epimerization is
27.3 kcal/mol.
Assuming a very minimal metalÀP bond energy of 17 kcal/mol (kdissoc > 1 sÀ1 at 258), this would provide a
barrier > 40 kcal/mol to inversion (kepim < 5 ¥ 10À10 sÀ1 at 1208, t1/2 > 40 years; assuming entropy effects upon
the epimerization rate are relatively minor).
)

Helvetica Chimica Acta Vol. 84 (2001)
3523
Scheme 3. Transition Metal Complexes of [P*CP*] Ligands
i) HNR2. ii) For 4a: 3a ‡ [Pd(MeCN)4][BF4]2 in MeCN, 608 followed by LiCl; for 4b: 3b ‡ Pd2(dba)3 in C6H6,
228, 12 h; for 5: 3b ‡ Pt2(p-tolyl)4(m-SEt2)2, C6H6, 558, 5 min [23].
Transition Metal Complexes of [P*CP*] Ligands. Complexes of the form
[P*CP*]MY (M ˆ Pd, Yˆ Cl, (4a); M ˆ Pd, Yˆ Br, (4b); M ˆ Pt, Yˆ Br, (5)) can
be formed by metallation of the deprotected ligands 3a (for 4a) or 3b (for 4b and 5),
from the appropriate organometallic precursors (Scheme 3). All of the complexes have
been characterized by multinuclear (1H, 13C, 31P) NMR and polarimetry, while 4a and 5
were additionally characterized by X-ray crystallography.
Crystals of 4a suitable for X-ray crystallography were obtained by crystallization
from CH2Cl2/hexane, while those of 5 were obtained from toluene/pentane. Molecular
plots of 4a and 5 are shown in Figs. 1 and 2, and selected bond lengths, angles, and
dihedral angles are given in the Table. Compound 4a has approximate C2 symmetry,
while compound 5 crystallizes with two molecules in the asymmetric unit, which both
have an exact, crystallographic C2 symmetry. The dihedral angle of the aryl plane and
the coordination plane is 14.46(10)8 in 4a, and 9.81(6)8 and 10.01(6)8 in 5. The most
interesting feature of these crystal structures is the fact that the t-Bu groups of the
phosphine are oriented out and away from the ligand in the direction of the halide,
whereas the Ph groups of the phosphine are held nearly perpendicular to the square
plane of the complex (88.18(11)8 and 93.01(11)8 for the two ClÀPÀPdÀC (ipso)tor-
sion angles in 4a). This effect is less pronounced in the Pt structure, in which the same
torsion angles of the Ph groups are reduced to 78.07(7)8 and 84.89(6)8.
When the complexes (especially that of Pd, 4a) are viewed along the halogen-metal
axis, two of the four quadrants are occupied by t-Bu groups, while two are relatively
empty due to the distance of the Ph groups. This creates a deep chiral pocket around the
halogen, which could potentially have a strong chiral induction in reactions in which a
substrate is bound in the position occupied by the halogen. Cross et al. have named this

3524
Helvetica Chimica Acta Vol. 84 (2001)
Fig. 1. Displacement ellipsoidplot of 4a (50% probability). H-Atoms omitted for clarity.
a −staggered× configuration in their analysis of pincer and pincer-type ligand structures
[24]. The contrasting situation, or −eclipsed× form, would place all four phosphine
substituents evenly distanced from the square plane of the metal, thus placing large
bulky groups in all four quadrants, which would be expected to give much less
selectivity, and to dramatically slow the rate of catalysis.
Reaction of complexes 4 and 5 with AgZ in wet acetone affords the corresponding
cationic aqua complexes 6 and 7 (M ˆ Pd, Z ˆ BF4, 6a; M ˆ Pd, Z ˆ OTf, 6b; M ˆ Pt,
Z ˆ OTf, 7). Complex 6b has been synthesized on a preparative scale (452 mg, 95%)
and fully characterized by elemental analysis, multinuclear NMR (1H, 31P), and
polarimetry. Since we were interested in the application of these catalysts to the aldol
condensation of CNCH2CO2Me with aldehydes, CNCH2CO2Me was added to 6b to
form the isocyanate adduct 8, which was spectroscopically (1H-, 13C-NMR; IR)
characterized. The change in nCN (Dn ˆ 56 cmÀ1) between free CNCH2CO2Me and 8
indicates that the isocyanate is bound through the isocyanate C-atom.
Chiral Aldol Condensation Reaction. The complexes 6a,b are active catalysts in the
aldol condensation of methyl 2-isocyanoacetate with benzaldehyde. The product of this
reaction, methyl 4,5-dihydro-5-phenyl-1,3-oxazoline-4-carboxylate exists in four iso-
meric forms: two (E)-enantiomers and two (Z)-enantiomers. Previous work by the
groups of Venanzi and Zhang has demonstrated that the reaction has some
enantioselectivity when performed with catalysts in which the chiral information is
located at the benzyl C-atom of the [PCP]-Pt and -Pd catalysts (see Introduction)

Helvetica Chimica Acta Vol. 84 (2001)
3525
Fig. 2. Displacement ellipsoid plot of the first independent molecule of 5 (50% probability). H-Atoms omitted
for clarity. Symmetry operation: 1 À x, y, À z.
Table. SelectedBondLengths, Angles, andTorsion Angles for 4a and 5
4a
5
Bond lengths [ä]
PdÀCl
2.3571(12)
2.011(3)
2.2975(8)
2.2947(8)
PtÀBr
2.5430(4)
2.5442(4)
2.021(3)
PdÀC(23)
PdÀP(1)
PdÀP(2)
Pt'ÀBr'
PtÀC(121)
Pt'ÀC(122)
PtÀP
2.016(3)
2.2926(5)
2.2928(6)
96.73(1)
97.16(1)
83.27(1)
Pt'ÀP'
Bond Angles [8]
ClÀPdÀP(1)
99.55(3)
96.34(3)
82.22(8)
BrÀPtÀP
ClÀPdÀP(2)
Br'ÀPt'ÀP'
C(121)ÀPtÀP
C(122)ÀPt'ÀP'
BrÀPtÀPÀC(11)
Br'ÀPt'ÀP'ÀC(12)
BrÀPtÀPÀC(51)
Br'ÀPt'ÀP'ÀC(52)
C(23)ÀPtÀP(1)
C(23)ÀPtÀP(2)
ClÀPdÀP(1)ÀC(1)
ClÀPdÀP(2)ÀC(11)
ClÀPdÀP(1)ÀC(5)
ClÀPdÀP(2)ÀC(15)
81.86(8)
82.84(1)
Dihedral Angles [8]
À 40.21(10)
À 36.00(14)
88.19(13)
93.01(13)
À 48.94(9)
À 43.84(9)
78.07(7)
84.90(7)
[14a,b]. (E/Z)-Ratios of 70/30 and 78/22, and e.e. values of 65% and 24% were
reported for the (E)-isomers, respectively. In both cases, the chirality of these
complexes was located on the benzylic C-atom of the [PCP] ligand.
We anticipated that moving the chiral site closer to the substrate binding pocket by
the synthesis of a P-chiral version of the [PCP] ligand would increase the enantio- and

3526
Helvetica Chimica Acta Vol. 84 (2001)
diastereoselectivity of the reaction. Indeed, for the Pd complexes 6, the diastereoiso-
meric ratio was found to be higher, yielding 94 98% (E)-product. However, the
enantiomeric excess was never found to be higher than 11%. Variations of the
concentration of catalyst, aldehyde, and H¸nig×s base did not appreciably affect the
enantioselectivity, nor did changing the reaction solvent from CH2Cl2 to THF or
toluene. Use of the isocyanate adduct 8, under anhydrous conditions also showed no
appreciable enantioselectivity. In contrast to the experience of Venanzi and co-workers
[14a], we found the Pt analogue 7 of our catalyst to be slower than the blank reaction
that occurs in the absence of transition metal catalyst. This is presumably due to the
greater steric bulk about the open site of our catalyst relative to that of Venanzi and co-
workers.
It is perhaps not surprising that our [P*CP*] catalyst displays a lower enantiose-
lectivity than that of the Au catalyst of Hayashi and co-workers [15], as the selectivity
of the latter has been ascribed to a H-bonding interaction with a pendant amino group,
which is lacking in the Pd/Pt system. However, it is striking that the enantioselectivity is
not higher than that of the other [PCP]-Pd/Pt-pincer systems discussed above in which
the chirality is more distant from the active site of the catalyst. Further investigation of
other reactions of these [P*CP*] catalysts will be necessary before nonspeculative
explanations can be advanced, and these are underway in our laboratory.
Conclusions. The wide use and versatility of pincer ligands in catalysis, and
especially advances in macromolecular pincer catalysts that can be recycled with high
efficiency by current and developing techniques, makes the development of new pincer-
based systems a very attractive field of catalyst development. To date, pincer-based
systems have shown poor to moderate enantioselectivities in catalytic reactions that
result in chiral products. We have shown that it is possible to prepare enantiomerically
pure transition metal pincer complexes with chirality at the P-atoms themselves which
are directly bonded to the transition metal. An initial test reaction between methyl 2-
isocyanoacetate and benzaldehyde did not generate high enantiomeric excesses, though
this does not preclude the possibility of enantioselectivity in other reactions. We are
currently involved in the use of these ligands with other metals and catalytic reactions,
as well as the synthesis of related P-stereogenic pincer ligands in order to find systems
in which our techniques for recoverable catalyst synthesis can be applied to useful
asymmetric transformations.
This material is based upon work supported by the North Atlantic Treaty Organization (NATO) under a
grant awarded in the year 2000 (to B. S. W.). Any opinions, findings, and conclusions or recommendations
expressed in this publication are those of the authors and do not necessarily reflect the view of NATO. A. L. S.
and M. L. acknowledge funding from the Council for the Chemical Sciences of the Dutch Organization for
Scientific Research (CW-NWO) and P. D. acknowledges funding from the Conselho Nacional de Desenvolvi-
¬
mento Cientifico e Tecnologico of Brazil. B. S. W. would like to thank H. P. Dijkstra and Dr. S. Medici for helpful
discussions.
Experimental Part
General. Unless otherwise noted, all transformations were carried out under N2, by standard Schlenk
techniques, or under vacuum. Recrystallizations were performed under ambient atmosphere, either at r.t. or
À308. Both protio- and deutero-solvents were purified by standard methods in reactions run under N2. Et2NH
and pyrrolidine were degassed before use. 1,3-Bis(bromomethyl)-2-bromobenzene [25] and Pt2(p-tolyl)4(m-

Helvetica Chimica Acta Vol. 84 (2001)
3527
SEt2)2 [23] were prepared according to published literature procedures. Other reagents, unless specified, were
used as received from commercial suppliers or prepared by inorganic synthesis procedures. Enantiomeric
excesses in catalytic reactions were determined by chiral HPLC, while diastereoisomeric ratios were determined
by NMR.
Preparation of (tert-Butyl)(phenyl)phosphine borane (1). This preparation is a variant upon that published
by Imamoto and co-workers [20]. In a typical experiment, a flame-dried Schlenk flask was charged with a stir
bar, PPhCl2 (5.80 g, 32.4 mmol), and dry ether (100 ml). The reaction vessel was cooled to À788 and t-BuLi
(1.5m soln. in pentane, 28ml, 42 mmol) was added dropwise via syringe over the course of 0.5 h. The mixture
was allowed to stir at À788 for 1.5 h, after which the temp. was raised to r.t. over the course of another 1.5 h. The
temp. was subsequently reduced to À418, and LiAlH4 (1.43 g, 37.6 mmol) was added in a single portion. The
cooling bath was removed, and the mixture filtered through Celite on sintered glass under N2. The solids were
washed with dry ether (2 Â 20 ml) and borane dimethyl sulphide complex (16.2 ml, 94%, 158mmol) was added
via syringe at 08. This was allowed to stir for 0.5 h at r.t. before being carefully poured onto a mixture of ice
(100 g) and 4m HCl (50 ml). The org. layer was separated and the aq. layer washed twice with ether (2 Â 50 ml).
The combined org. layers were washed with H2O (50 ml) and brine (50 ml) before being dried (MgSO4) and
filtered. The crude product was dissolved in CH2Cl2 (100 ml) and filtered through silica to give a heavy oil upon
evaporation of the CH2Cl2 (3.90 g, 67%).
Synthesis of (S,S) 1,3-(P(BH3)(tBu)(Ph)CH2)2(C6H4) (2a) and (S,S) 1,3-(P(BH3)(tBu)(Ph)CH2)2(2-
Br)C6H3 (2b). This preparation was derived from the method of sparteine resolution described by Wolfe and
Livinghouse [21]. 2a and 2b were prepared by identical methods from 1 and either 1,3-bis(bromomethyl)
benzene (2a) or 1,3-bis(bromomethyl)-2-bromobenzene (2b). The synthesis and characterization of 2a is given
here. Phosphine borane 1 (3.90 g, 21.7 mmol) was placed in a Schlenk flask with a stirbar, (À)-sparteine (7 ml,
29 mmol) and ether (80 ml). The temp. was lowered to À788, and BuLi (1.6m soln. in pentane, 16.7 ml,
8.8 mmol) was added dropwise via syringe. We have noticedthat it is essential to have more ( À)-sparteine than
BuLi present in order to ensure the formation of only one isomer. Within minutes, a voluminous white precipitate
had formed. After 0.5 h, the cooling bath was removed, and the reaction allowed to come to r.t. over the course
of 0.5 h. The reaction was placed in a warm water bath (ca. 308) for 1.5 h in order to allow the sparteine-lithium-
phosphide complex to form the desired isomer. Subsequently, the temp. was reduced to À788 and 1,3-
bis(bromomethyl)benzene was added, and the reaction warmed to À208 over the course of 2 h before being
placed in the freezer (À 308) overnight. The reaction was poured into toluene (100 ml), CH2Cl2 (200 ml), and
3% H2SO4 (200 ml), the org. layer separated, and the aq. layer extracted thrice with CH2Cl2 (100, 100, and
50 ml). The combined org. fractions were reduced to 150 ml, and washed (aq. Na2CO3, 50 ml) and brine (50 ml).
The solvents were evaporated, the product passed through silica in CH2Cl2, and subsequently precipitated with
hexane (2.09 g, 42%, 2 crops). [a]2D3 ˆ ‡106 (c ˆ 0.21, CHCl3). 1H-NMR (CDCl3): 1.11 (d, 3J(P,H) ˆ 13.8, 22 H
(includes some intensity from broad BH3 signal), PÀt-Bu); 3.2 3.5 (8line ABX pattern, 4 H, diastereoisotopic
PÀCH2); 6.9 7.0 (m, HÀC(3), HÀC(4), and HÀC(5) of central aryl ring); 7.18( s, HÀC(1) of central aryl
ring); 7.3 7.5 (m, 6 H, m- and p-signals of PÀPh); 7.75 (t, 3J(P,H)(H,H) ˆ 8.1, 4 H, o-signals of PÀPh).
13C-NMR (CDCl3): 25.8(C Me3); 27.0, 30.0 (2d, J ˆ 30.0, PÀCMe3 and PÀCH2); 126.0 (d, 1J(P,C) ˆ 47,
PÀC(ipso)); 127.9 (arom.); 128.2 (d, 3J(P,C) ˆ 9, PÀC(meta)); 128.8 (arom.); 131.2 (arom.); 132.7 (arom.);
133.9 (d, 2J(P,C) ˆ 7, P ÀC(ortho)). 31P-NMR (CDCl3): 34 (br. q). Anal. calc. for C28H42B2P2 (462.21): C 72.76,
H 9.16, P 13.40; found: C 72.52, H 9.08, P 13.46.
Compound 2b was isolated only in crude form (3.22 g, 99%) and thus characterized spectroscopically:
[a]2D3 ˆ ‡136 (c ˆ 1.01, CH2Cl2). 1H-NMR (CDCl3): 1.11 (d, 3J(P,H) ˆ 13.8, 22 H (includes some intensity from
broad BH3 signal), PÀt-Bu); 3.2 3.5 (8line ABX pattern, 4 H, diastereoisotopic P ÀCH2); 6.9 7.0 (m, 3 H,
HÀC(3), HÀC(4), and HÀC(5) of central aryl ring); 7.3 7.5 (m, 6 H, m- and p-signals of PÀPh); 7.75
(t, 3J(P,H)(H,H) ˆ 8.1, 4 H, o-signals of PÀPh). 13C-NMR (CDCl3): 25.9 (CMe3); 27.5, 30.5 (2d, J ˆ 30.0,
PÀCMe3 and PÀCH2); 126.0 (d, 1J(P,C) ˆ 47, PÀC(ipso)); 126.6 (arom.); 128.6 (d, 3J(P,C) ˆ 9, PÀC(meta);
129.1 (t, J ˆ 6.1, CÀBr); 129.7 (arom.); 131.6 (d, 4J(P,C) ˆꢀ 2, PÀC(para); 134.2 (d, 2J(P,C) ˆ 7, P ÀC(ortho));
134.6 (arom). 31P-NMR (CDCl3): 34 (br. q).
Deprotection of Phosphine Boranes to form 1,3-{Bis[(tert-butyl)(phenyl)phosphino]methyl}benzene and
1,3-{Bis[tert-butyl(phenyl)phosphino]methyl}-2-bromobenzene (3a,b): The phosphine boranes 2a,b were
deprotected by heating in neat pyrrolidine (408, 12 h) or 50/50 (v/v) mixture of THF and Et2NH (558, 1.5 h)
to afford the free phosphines 3a,b, respectively (31P-NMR: 8.5 (for 3a), 8.1 (for 3b)). The volatiles were removed
under vacuum and the free phosphines directly used in metallation reactions without further purification.
Preparation of {(S,S)-1,3-{Bis[(tert-butyl)(phenyl)phosphino]methyl}phenyl}chloropalladium (4a): Fresh-
ly deprotected 3a (from 585 mg, 1.35 mmol of 2a) was dissolved in MeCN, and was allowed to react with

3528
Helvetica Chimica Acta Vol. 84 (2001)
[Pd(MeCN)4][BF4]2 (680 mg, 1.5 mmol) over the course of 12 h, after which a large quantity of palladium black
had formed. LiCl (200 mg, 7.5 mmol) in wet acetone was added to the soln., and after 1 h, the solvents were
removed in vacuo and the product extracted from the solids with CH2Cl2. A brown impurity was removed by
THF/ether precipitation, and the product was isolated by precipitation from ether/EtOH in two crops (118mg,
15% (unoptimized)). [a]2D3 ˆ À129 (c ˆ 0.21, CHCl3). 1H-NMR (CDCl3): 1.30 (t, 3J(P,H) ˆ 7.7, 2 t-Bu); 3.62
(br. m, 2 PÀCH2); 6.88 (t, 7.2, HÀC(4) of central aryl ring); 7.05 (d, J ˆ 7.5, HÀC(3), HÀC(5) of central aryl
ring); 7.42 (br. m, 6 H, m- and p-signals of PÀPh); 8.07 (br. m, 4 H, o-signal of PÀPh). 13C-NMR (CDCl3): 26.9
(CMe); 33.3, 35.6 (2t, J ˆ 11, 5, PÀCMe3, PÀCH2); 123.3 (t, J ˆ 11, arom.); 125.6 (arom.); 128.5 (arom.); 129.9
(t, J ˆ 16, arom.); 130.7 (arom.); 134.3 (t, J ˆ 3, arom.); 148.1 (t, J ˆ 0, arom.); 158.2 (arom.). 31P-NMR
(CDCl3): 53.2. Anal. calc. for C28H35ClP2Pd (575.41): C 58.45, H 6.16; found: C 58.39, H 6.07.
Preparation of {(S,S)-1,3-{Bis[(tert-butyl)(phenyl)phosphino]methyl}phenyl}bromopalladium (4b): Fresh-
ly deprotected 3a (from 3.06 g, 5.66 mmol of 2a) was dissolved in benzene, and was allowed to react with
Pd2(dba)3 ¥ CHCl3 (2.75 g, 2.66 mmol) over the course of 12 h. The product was filtered through silica in ether.
Subsequently, the product was purified by column chromatography (silica, CH2Cl2) and crystallization from
acetone/hexane at À308 (680 mg, 19% (unoptimized) as 4b ¥ 1/3(C2H6O)). The crystals were washed thrice with
hexane. 1H-NMR (CDCl3): 1.35 (t, 3J(P,H) ˆ 7.6, 2, t-Bu); 3.63 (br. m, 2 PÀCH2); 6.85 (t, J ˆ 7.4, HÀC(4) of
central aryl ring); 7.03 (d, J ˆ 7.4, HÀC(3), HÀC(5) of central aryl ring); 7.41 (br. m, 6 H, m- and p-signals for
PÀPh); 8.05 (br. m, 4 H, o-signal for PÀPh). 13C-NMR (CDCl3): 27.0 (CMe3); 33.6, 35.6 (2t, J ˆ 12, PÀCMe3
and PÀCH2); 123.3 (t, J ˆ 10, arom.); 125.6 (arom.); 128.5 (arom.); 129.7 (t, J ˆ 20, arom.); 130.6 (arom.); 134.3
(arom.); 147.6 (3t, J ˆ 10, arom.). 31P-NMR (CDCl3): 54.0. Anal. calc. for C28H35BrP2Pd (619.76): C 54.43,
H 5.49, P 10.03; found: C 54.26, H 5.38, P 10.03.
Preparation of {(S,S)-1,3-{Bis[(tert-butyl)(phenyl)phosphino]methyl}phenyl}bromoplatinum (5): Freshly
deprotected [P*C(Br)P*] (3b, from 930 mg of 2b, 1.7 mmol) was suspended in benzene with Pt2(p-tolyl)4(m-
SEt2)2 (800 mg, 0.86 mmol). The soln. was placed in a water bath at 558. After 1 min, the 31P-NMR spectrum of
1
the crude mixture (unlocked, unreferenced) showed primarily signals for the product, 7, (d ˆ 50.0, J(Pt,P) ˆ
2895 Hz), a small amount of starting material (d ˆ 8.2), and a small amount of an intermediate, putatively
assigned as the mono-phosphine-coordinated species cis-Pt(p-tolyl)2(SEt2)(k1-3b) (d ˆ 21.7, 1J(Pt,P) ꢀ 1700 Hz;
d ˆ 8.8). After 5 min at 558, only the product was visible by 31P-NMR spectroscopy. The volatiles were removed
under vacuum. The product was passed through a column of silica gel (40 cm) in CH2Cl2 and crystallized from
toluene/pentane (0.599 g, 49% in two crops). [a]2D3 ˆ À51 (c ˆ 1.04, CH2Cl2). 1H-NMR (CDCl3): 1.42
(t, 3J(P,H) ˆ 7.7, 2 t-Bu); 3.65 (br. m, 2 PÀCH2); 6.98( t, J ˆ 7.2, HÀC(4) of central aryl ring); 7.09 (d, with Pt
satellites, 3J(H,H) ˆ 7.5, 4J(H,Pt) ˆ 21, HÀC(3), HÀC(5) of central aryl ring); 7.43 (br. m, 6 H, m- and p-signals
for PÀPh); 8.14 (br. m, 4 H, o-signal or PÀPh). 13C-NMR (CDCl3): 27.3 (CMe3); 33.6, 36.2 (2t, with Pt satellites,
J(P,C) ˆ 15, 2J(Pt,C) ˆ 59 (33.6); J(P,C) ˆ 16, 2J(Pt,C) ˆ 105 (36.2); PÀCMe3 and PÀCH2); 123.6 (t with Pt
satellites, J(P,C) ˆ 8, J(Pt,C) ˆ 39, arom.); 124.9 (arom.); 128.5 (t, J ˆ 5, arom.); 130.8(arom.); 134.4 ( t, J ˆ 11,
arom.); 146.4 (arom.); 148.1 (arom.). 31P-NMR (CDCl3): 49.9 (s, with Pt satellites, 1J(Pt,P) ˆ 2900).
Crystal-Structure Determinations: X-Ray intensities were measured on a Nonius KappaCCD diffractom-
eter with rotating anode (l ˆ 0.71073 ä) at a temp. of 150(2) K. The structures were solved with automated
Patterson methods (DIRDIF 97) [26] and refined with SHELXL97 [27] against F2 of all reflections. Molecular
illustration, structure checking, and calculations were performed with the PLATON [28] package.
Crystal-Structure Determination of 4a: X-Ray-quality crystals of 4a were grown from a mixture of CH2Cl2/
hexane at r.t. by slow evaporation. C28H35ClP2Pd, Fw ˆ 575.35, colorless needle, 0.45  0.06  0.06 mm3,
hexagonal, P63 (No. 173), a ˆ b ˆ 20.3046(2), c ˆ 11.6707(1) ä, Vˆ 4166.93(7) ä3, Z ˆ 6, 1 ˆ 1.376 g/cm3. 81498
reflections were measured. 6380 reflections were unique (Rint ˆ 0.066). 5431 reflections were observed [I >
2s(I)]. The q range was 1.16 27.478 with indices hkl À26/26, À26/26, À15/15. The absorption correction was
based on multiple measured reflections (program PLATON [28] routine MULABS, m ˆ 0.89 mmÀ1, 0.91 0.96
transmission). Non-H-atoms were refined freely with anisotropic displacement parameters. H-atoms were
refined as rigid groups. 295 refined parameters, 1 restraint. R-values [obs. refl.]: R1 ˆ 0.0292, wR2 ˆ 0.0566. R-
values [all refl.]: R1 ˆ 0.0429, wR2 ˆ 0.0605. GoF ˆ 1.022. Flack x-parameter À0.05(2) [29]. Residual electron
density between À0.49 and 0.65 e/ä3.
Crystal-Structure Determination of 5: Crystals of 5 suitable for X-ray-diffraction study were grown from
toluene/pentane. C28H35BrP2Pt, Fw ˆ 708.50, colorless block, 0.33  0.15  0.15 mm3, monoclinic, C2 (No. 5),
a ˆ 15.5173(1), b ˆ 13.0548(1), c ˆ 15.3827(1) ä, b ˆ 118.7775(3)8, Vˆ 2731.29(3) ä3, Z ˆ 4, 1 ˆ 1.723 g/cm3.
60576 reflections were measured. 16878 reflections were unique (Rint ˆ 0.033). 15320 reflections were observed
[I > 2s(I)]. The q range was 1.51 40.238 with indices hkl À28/27, À23/23, À27/27. An analytical absorption
correction was applied (program PLATON [28], routine ABST, m ˆ 6.73 mmÀ1, 0.34 0.65 transmission). The

Helvetica Chimica Acta Vol. 84 (2001)
3529
structure contains two independent molecules in the asymmetric unit, which both are located on a twofold axis.
Non-H-atoms were refined freely with anisotropic displacement parameters. The H-atoms of Me at C(21) and
C(22) were refined freely with isotropic displacement parameters; all other H-atoms were refined as rigid
groups. 321 refined parameters, 49 restraints. R-values [obs. refl.]: R1 ˆ 0.0243, wR2 ˆ 0.0466. R-values [all refl.]:
R1 ˆ 0.0302, wR2 ˆ 0.0481. GoF ˆ 1.031. Flack x-parameter À0.004(3) [29]. Residual electron density was
between À2.16 and 0.90 e/ä3.
Crystallographic data (excluding structure factors) for the structures in this paper have been deposited with
the Cambridge Crystallographic Data Centre as supplementary publication no. CCDC 169317 (4a) and 169318
(5). Copies of the data can be obtained, free of charge, on application to CCDC, 12 Union Road, Cambridge
CB21EZ UK, (fax: ‡441223336033 or e-mail: deposit@ccdc.cam.ac.uk).
Preparation of Aqua Complexes Aqua{(S,S)-1,3-{bis[(tert-butyl)(phenyl)phosphino]methyl}phenyl}palla-
dium Tetrafluoroborate (6a), Trifluoroacetate (6b) andAqua{( S,S)-1,3-{bis[(tert-butyl)(phenyl)phosphino]me-
thyl}phenyl}platinum Bromide (7): As all three aqua complexes were prepared by the same method, that for 6b
is given as typical. On the benchtop, a sample of [PCP]PdBr (4b) (430 mg, 0.67 mmol) was dissolved in wet
acetone. AgOTf (176 mg, 0.68mmol) was added, which resulted in an off-white precipitate. After 1 h, the
sample was filtered through Celite on glass wool and precipitated with hexane as a microcrystalline white
powder to form the product with one acetone molecule of crystallization (452 mg, 95%). [a]2D3 ˆ À13.5 (c ˆ 1.04,
CH2Cl2). 1H-NMR (C6D6): 1.2 ppm (br., 2 t-Bu); 1.54 (6 H, acetone); 3.0 (4 H, PÀCH2); 3.4 4.8(Pd ÀOH2); 7
(aryl signals); 7.1 8.2 (PÀPh signals). Aryl region not integratable due to C6D5H and overlap. Anal. calc. for
C29H37F3O4P2PdS ¥ (C3H6O) (765.11): C 50.23, H 5.66, P 8.10; found: C 49.99, H 5.55, P 8.20.
For 7: [a]2D3 ˆ À57 (c ˆ 0.76, CH2Cl2).
Isolation of PdIsocyanate Complex (8): On the benchtop, [[PCP]Pd(OH2)][OTf] (108.3 mg, 0.14 mmol)
was dissolved in acetone, to which methyl 2-isocyanoacetate (14 ml, 0.15 mmol) was added. The solvent was
evaporated, and the product precipitated from THF/hexane (84.1 mg, 75%). [a]2D3 ˆ À34 (c ˆ 0.23, CH2Cl2). IR:
1758cm À1 (CˆO), 2219 cmÀ1 (CꢁN). 1H-NMR (CDCl3): 1.27 (t, 3J(P,H) ˆ 8.1, 2 t-Bu); 3.8(br. m, 2 PÀCH2);
3.87 (s, OMe); 5.21 (s, CNCH2); 7.03 (t, J ˆ 7, H ÀC(4) of central aryl ring); 7.13 (d, J ˆ 7.5, HÀC(3), HÀC(5) of
central aryl ring); 7.52 (br. m, 6 H, HÀC(3), HÀC(4), HÀC(5) of PÀPh); 7.73 (br. m, 4 H, HÀC(2), HÀC(6)
of PÀPh). 13C-NMR (CDCl3): 26.6 (CMe3); 33.0, 36.5 (2t, J ˆ 12, PÀCMe2, PÀCH2); 47.3 (s, CNCH2); 53.8
(s, OMe); 123.7 (t, J ˆ 10, arom.); 127.3 (t, J ˆ 10, arom.); 127.9 (t, J ˆ 20, arom.); 129.5 (arom.); 131.9 (arom.);
133.9 (arom.); 148.4 (t, J ˆ 10, arom.); 162.8( s, PdÀCN); 164.4 (s, CˆO). 31P-NMR (CDCl3) ˆ 71.0. We have
been unable to obtain a satisfactory elemental analysis of compound 8, presumably due to loss of isonitrile
ligand under high vacuum conditions.
Typical Catalytic Experiment for Aldol Condensation: A 20-ml vial was charged sequentially with
[[PCP]Pd(OH2)[BF4] (6a) (0.016 mmol, 1 mol-%), 5 ml of CH2Cl2 (distilled from CaH2), isonitrile (145 ml,
1.60 mmol, 100 mol-%), benzaldehyde (162 ml, 1.59 mmol, 100 mol-%, H¸nig×s base (28 ml, 0.16 mmol, 10 mol-
%), and mesitylene (223 ml, 1.60 mmol, 100 mol-%). The reaction was monitored by occasionally removing a
100-ml aliquot, evaporation of the volatiles under a slow stream of N2, and dissolution in CDCl3 for NMR
spectroscopy.
REFERENCES
[1] C. L. Moulton, B. L. Shaw, J. Chem. Soc., Dalton Trans. 1976, 1020.
[2] a) M. Rietveld, D. M. Grove, G. van Koten, New J. Chem. 1997, 21, 751; b) G. van Koten, Pure Appl. Chem.
1989, 61, 1681.
[3] F. Gorla, L. M. Venanzi, A. Albinati, Organometallics 1994, 13, 43.
[4] M. Ohff, A. Ohff, M. E. van der Boom, D. Milstein, J. Am. Chem. Soc. 1997, 119, 11697.
[5] a) D. M. Grove, G. van Koten, A. H. M. Verschuuren, J. Mol. Catal. 1988, 45, 169; b) D. M. Grove, A. H. M.
Verschuuren, G. van Koten, J. A. M. van Beek, J. Organomet. Chem. 1989, 372, C1; c) A. Kleij, R. A.
Gossage, R. J. M. Klein Gebbink, N. Brinkmann, E. J. Reijerse, U. Kragl, M. Lutz, A. L. Spek, G. van
Koten, J. Am. Chem. Soc. 2000, 122, 12112.
[6] M. Gupta, C. Hagen, R. J. Flescher, W. C. Kaska, C. M. Jensen, Chem. Commun. 1996, 2083; M. Gupta,
W. C. Kaska, C. M. Jensen, Chem. Commun. 1997, 461; W. W. Xu, G. P. Rosini, M. Gupta, C. M. Jensen,
W. C. Kaska, K. Krogh-Jespersen, A. S. Goldman, Chem. Commun. 1997, 2273; F. Liu, E. B. Pak, B. Singh,
C. M. Jensen, A. S. Goldman, J. Am. Chem. Soc. 1999, 121, 4086.
[7] P. Dani, T. Karlen, R. A. Gossage, S. Gladiali, G. van Koten, Angew. Chem., Int. Ed. 2000, 39, 743.

3530
Helvetica Chimica Acta Vol. 84 (2001)
[8] B. Rybtchinski, D. Milstein, Angew. Chem., Int. Ed. 1999, 38, 870; M. Kanzelberger, B. Singh, M. Czerw, K.
Krogh-Jespersen, A. S. Goldman, J. Am. Chem. Soc. 2000, 122, 11017.
[9] J. W. J. Knapen, A. W. van der Made, J. C. de Wilde, P. W. N. M. van Leeuwen, P. Wijkens, D. M. Grove, G.
van Koten, Nature 1994, 372, 659; R. A. Gossage, L. A. van de Kuil, G. van Koten, Acc. Chem. Res. 1998, 31,
423.
[10] H. P. Dijkstra, P. Steenwinkel, D. M. Grove, M. Lutz, A. L. Spek, G. van Koten, Angew. Chem., Int. Ed.
1999, 38, 2186; H. P. Dijkstra, M. D. Meijer, J. Patel, R. Kreiter, G. P. M. van Klink, M. Lutz, A. L. Spek,
A. J. Canty, G. van Koten, Organometallics 2001, 20, 3159.
[11] L. A. van de Kuil, D. M. Grove, J. W. Zwikker, L. W. Jenneskens, W. Drenth, G. van Koten, Chem. Mater.
1994, 6, 1675.
[12] M. D. Meijer, N. Ronde, D. Vogt, G. P. M. van Klink, G. van Koten, Organometallics 2001, 20, 3993.
[13] C. Patmamanoharan, P. Wijkens, D. M. Grove, A. P. Philipse, Langmuir 1996, 12, 4372; C. Bezemer,
R. J. M. Klein Gebbink, G. van Koten, submitted for publication.
[14] a) F. Gorla, T. Togni, L. M. Venanzi, A. Albertini, F. Lianza, Organometallics 1994, 13, 1607; b) J. M.
Longmire, X. Zhang, M. Shang, Organometallics 1998, 17, 4374; c) M. A. Stark, G. Jones, C. J. Richards,
Organometallics 2000, 19, 1282; d) M. Gerisch, J. R. Krumper, R. G. Bergman, T. D. Tilley, J. Am. Chem.
Soc. 2001, 123, 5818.
[15] Y. Ito, M. Sawamura, T. Hayashi, J. Am. Chem. Soc. 1986, 108, 6405.
[16] A. Togni, S. D. Pastor, J. Org. Chem. 1990, 55, 1649.
[17] R. Nesper, P. S. Pregosin, K. P¸ntener, M. Wˆrle, Helv. Chim. Acta 1993, 76, 2239; R. Nesper, P. S.
Pregosin, K. P¸ntener, M. Wˆrle, A. Albinati, J. Organomet. Chem. 1996, 507, 8 5.
[18] P. Dani, M. Albrecht, G. P. M. van Klink, G. van Koten, Organometallics 2000, 19, 4468.
¬
[19] R. Gimenez, T. M. Swager, J. Mol. Catal., A 2001, 166, 265.
[20] T. Imamoto, T. Oshiki, T. Onozawa, T. Kusomoto, K. Sato, J. Am. Chem. Soc. 1990, 112, 5244.
[21] B. Wolfe, T. Livinghouse, J. Am. Chem. Soc. 1998, 120, 5116.
[22] R. D. Baechler, K. Mislow, J. Am. Chem. Soc. 1970, 92, 3090.
[23] A. J. Canty, J. Patel, B. W. Skelton, A. H. White, J. Organomet. Chem. 2000, 599, 195.
[24] R. J. Cross, A. R. Kennedy, K. W. Muir, J. Organomet. Chem. 1995, 487, 227.
[25] L. A. van de Kuil, H. Luitjes, D. M. Grove, J. W. Zwikker, J. G. M. van der Linden, A. M. Roelofsen, L. W.
Jenneskens, W. Drenth, G. van Koten, Organometallics 1994, 13, 468.
[26] P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, S. Garcia-Granda, R. O. Gould, J. M. M. Smits,
C. Smykalla, DIRDIF 97 program system, −Technical Report of the Crystallography Laboratory×,
University of Nijmegen, The Netherlands, 1997.
[27] G. M. Sheldrick, SHELXL-97 program for crystal-structure refinement, University of Gˆttingen,
Germany, 1997.
[28] A. L. Spek, PLATON, multipurpose crystallographic tool, Utrecht University, The Netherlands, 2000.
[29] H. D. Flack, Acta Crystallogr., Sect. A 1983, 39, 876.
ReceivedAugust 29, 2001