Palladium and rhodium ureaphosphine complexes: Exploring structural and catalytic consequences of anion binding

Article abstract of DOI:10.1021/ic051110f

The addition of a chloride ion to Pd and Rh complexes supported by the ureaphosphine ligand L results in the formation of chelating diphosphine complexes that retain some catalytic activity.

Full text of DOI:10.1021/ic051110f

Inorg. Chem. 2005, 44, 7708−7710
Palladium and Rhodium Ureaphosphine Complexes: Exploring
Structural and Catalytic Consequences of Anion Binding
Paul A. Duckmanton, Alexander J. Blake, and Jason B. Love*
School of Chemistry, UniVersity Park, UniVersity of Nottingham, Nottingham NG7 2RD, U.K.
Received July 4, 2005
The addition of a chloride ion to Pd and Rh complexes supported
by the ureaphosphine ligand L results in the formation of chelating
diphosphine complexes that retain some catalytic activity.
a chloride ion and the apparent retention of this supramo-
lecular interaction during the Rh-catalyzed hydroformylation
of octene.
Reaction between the ureaphosphine L, generated by the
addition of PhNCO to m-H2NC6H4PPh2, and PdCl2(PhCN)2
in CH2Cl2 results in the rapid formation of the bis-
(phosphine)palladium complex PdCl2(L)2 (1) in good yield
(Scheme 1). If the reaction mixture is rapidly worked up
Chelating diphosphines are ubiquitous as ligands that
stabilize and define transition-metal centers during important
catalytic processes. Interwoven variables such as bite angle,
steric demand, and electronic effects are key to catalyst
efficacy and have led to the generation of a myriad of
elegantly designed, chelating diphosphine ligands for use in
catalytic processes.1 Recently, two new and potentially
versatile supramolecular approaches to chelating diphosphine
ligands were reported in which (i) direct hydrogen-bonding
interactions between 2-pyridone- and 2-hydroxypyridine-
phosphine tautomers2 and (ii) the assembly of monomeric,
pyridine-functionalized phosphines on a bis[zinc(II) porphy-
rin] template3 both resulted in chelate formation and unique
catalytic activity.
Scheme 1. Synthesis and Anion-Binding Reactions of Ureaphosphine
Complexes 1 and 2a
a Reagents and conditions: (i) [PdCl2(PhCN)2], CH2Cl2, 5 min; (ii)
reaction time 1 h; (iii) Bun N‚Cl, CH2Cl2; (iv) [{Rh(CO)2Cl}2], CH2Cl2;
4
As a new supramolecular approach to chelating diphos-
phine catalysts, we envisaged that anion sequestration by a
suitably designed phosphine ligand would result in chelate
formation and subsequent control of catalytic activity. While
ligands able to act as ionophores for both metal and anion
are well-known,4 to our knowledge, the use of anions as
supramolecular constructs in catalyst ligand design has yet
to be explored. Significantly, it has been shown that anion-
binding ureas can affect the outcome of Pd-catalyzed alkene
hydrocarbonylation,5 and, furthermore, an anion-assisted
trans/cis isomerization was observed in acetanilide-function-
alized phosphine complexes of Pd.6 These observations
suggest that hydrogen-bonding interactions are sufficiently
robust to direct the outcome of catalytic reactions.
(v) [{Rh(CO)2Cl}2], Bu4NCl, CH2Cl2.
(after ca. 5 min), pure cis-1 is isolated (δP 34.3 ppm7).
However, prolonged reaction at room temperature results in
the precipitation of 1 as a pale-yellow solid in a 1:3 cis/
trans isomeric ratio (δP 25.5 ppm [trans]). The insolubility
of 1 in common organic solvents can be attributed to polymer
(4) Kovbasyuk, L.; Kra¨mer, R. Chem. ReV. 2004, 104, 3161. Baer, A. J.;
Koivisto, B. D.; Coˆte´, A. P.; Taylor, N. J.; Hanan, G. S.; Nierengarten,
H.; Dorsselaer, A. V. Inorg. Chem. 2002, 41, 4987. Bondy, C. R.;
Gale, P. A.; Loeb, S. J. Chem. Commun. 2001, 729. Choi, K.;
Hamilton, A. D. Coord. Chem. ReV. 2003, 240, 101. Coxall, R. A.;
Lindoy, L. F.; Miller, H. A.; Parkin, A.; Parsons, S.; Tasker, P. A.;
White, D. J. Dalton Trans. 2003, 55. Liu, X.; Kilner, C. A.; Halcrow,
M. A. Chem. Commun. 2002, 704. Love, J. B.; Vere, J. M.; Glenny,
M. W.; Blake, A. J.; Schro¨der, M. Chem. Commun. 2001, 2678. Qin,
Z.; Jennings, M. C.; Puddephatt, R. J. Inorg. Chem. 2002, 41, 3967.
Turner, D. R.; Spencer, E. C.; Howard, J. A. K.; Tocher, D. A.; Steed,
J. W. Chem. Commun. 2004, 1352. Uppadine, L. H.; Drew, M. G. B.;
Beer, P. D. Chem. Commun. 2001, 291. White, D. J.; Laing, N.; Miller,
H. A.; Parsons, S.; Coles, S.; Tasker, P. A. Chem. Commun. 1999,
2077. Wu, B.; Yang, X.-J.; Janiak, C.; Lassahn, P. G. Chem. Commun.
2003, 902.
We report herein the ability of ureaphosphine ligands to
form well-defined chelates at Pd and Rh in the presence of
* To whom correspondence should be addressed. E-mail:
jason.love@nottingham.ac.uk.
(1) Freixa, Z.; van Leeuwen, P. W. N. M. Dalton Trans. 2003, 1890.
(2) Breit, B.; Seiche, W. J. Am. Chem. Soc. 2003, 125, 6608.
(3) Slagt, V. F.; Ro¨der, M.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.;
Reek, J. N. H. J. Am. Chem. Soc. 2004, 126, 4056. Slagt, V. F.; van
Leeuwen, P. W. N. M.; Reek, J. N. H. Chem. Commun. 2003, 2474.
(5) Scheele, J.; Timmerman, P.; Reinhoudt, D. N. Chem. Commun. 1998,
2613.
(6) Lu, X.-X.; Tang, H.-S.; Ko, C.-C.; Wong, J. K.-Y.; Zhu, N.; Yam, V.
W.-W. Chem. Commun. 2005, 1572.
(7) Grim, S. O.; Keiter, R. L. Inorg. Chim. Acta 1970, 4, 56.
7708 Inorganic Chemistry, Vol. 44, No. 22, 2005
10.1021/ic051110f CCC: $30.25
© 2005 American Chemical Society
Published on Web 10/06/2005

COMMUNICATION
Figure 2. Job plots from the addition of [Bun ‚Cl] to the ureaphosphine
4
Figure 1. X-ray crystal structure of the ureaphosphine complex [PdCl2-
(L)2]‚2DMSO (1‚2DMSO; 50% displacement ellipsoids). For clarity, all
hydrogens except those of the urea nitrogens N1, N2, N3, and N4 are
omitted. Selected bond lengths (Å) and angles (deg): Pd1-Cl1 2.2978(11),
Pd1-Cl2 2.3036(11), Pd1-P1 2.3509(12), Pd1-P2 2.3268(12); Cl1-Pd1-
P1 87.16(4), Cl1-Pd1-P2 90.58(4), Cl2-Pd1-P1 95.40(4), Cl2-Pd1-
P2 86.83(4), P1-Pd1-P2 176.13(4), Cl1-Pd-Cl2 177.32(4).
ligand L and the Pd complex 1.
and L confirm that 1:1 anion binding occurs in solution in
both cases. Unfortunately, we have been unable to calculate
accurate association constants using these data, possibly as
a consequence of competing inter- and intramolecular
hydrogen-bonding interactions between L, 1, and the chloride
ion.
The inaccuracy of the above calculations and the reality
that many metal-catalyzed reactions are conducted at elevated
temperature led us to assess the thermal stability of 1‚Cl by
1H NMR spectroscopy in BrC6D5. Between 300 and 403 K,
no significant change in the urea chemical shift was observed
[δ(300 K) ) 10.41 ppm; δ(403 K) ) 10.27 ppm], which
suggests that hydrogen bonding to the anion is uncompro-
mised within this temperature range. In contrast, the non-
anion-bound complex 1 is completely insoluble in BrC6D5
at 400 K, and this implies that intermolecular hydrogen-
bonding interactions remain dominant in this case.
To classify definitively the nature of anion binding, the
solid-state structural determination of trans-1‚Cl was under-
taken (Figure 3). Unlike with 1, intramolecular hydrogen-
bonding interactions between the hydrogens associated with
the urea nitrogen atoms N1 and N2 and their symmetry
equivalents and the single chloride ion Cl3 occur [N1-Cl3
3.341(4) Å, N2-Cl3 3.201(4) Å] and result in the formation
of an unusual, effectively chelating diphosphine bound to
Pd1; this supramolecular interaction between the urea
hydrogens and the chloride ion also results in meso stereo-
chemistry at Pd1. It is clear that trans orientation of the
ureaphosphine ligands results in a hydrogen-bonding cavity
that is suitable for chloride anion binding. This feature is
potentially important in the catalytic chemistry of these
systems because rhodium complexes of chelating diphos-
phines that can adopt a range of bite angles or even span
trans sites are extremely effective hydroformylation cata-
lysts.1,8
formation due to intermolecular hydrogen bonding. These
interactions are disrupted by dissolution in DMSO, so
allowing the above NMR data to be acquired. An EXSY
NMR experiment showed that no isomer exchange occurs
in solution at room temperature, and this contrasts with the
allosteric, anion-controlled, cis/trans isomerization seen for
Pd complexes of similar acetanilide-functionalized phosphine
ligands.6 Crystals grown from DMSO were analyzed by
X-ray crystallography (Figure 1).
It is clear from the solid-state structure that crystallization
of trans-1 is favored over that of cis-1 and that intermolecular
hydrogen bonding occurs between the urea protons and the
DMSO solvent oxygen atoms O1S and O2S [N1-O1S
2.846(5) Å, N2-O1S 2.962(4) Å, N3-O2S 2.819(5) Å, N4-
O2S 2.836(4) Å]; the ureaphosphine backbones adopt an anti
configuration, presumably for steric reasons. Dissolution of
crystalline trans-1 in DMSO-d results in rapid equilibration
to a 1:3 cis/trans mixture.
Unlike the above, the 1:1 reaction between in-situ-prepared
cis/trans-1 and Bun N‚Cl in CH2Cl2 does not result in
4
precipitation due to polymer formation but instead forms the
anion-bound adduct cis/trans-[Bun N][PdCl2(L)2Cl] (1‚Cl) in
4
high yield (Scheme 1); the 1:1 complex/anion ratio was
supported by elemental analysis, and no change in the cis/
trans isomer ratio was observed upon anion binding. Surpris-
ingly, the addition of Bun N‚H2PO4 to 1 resulted in chloride
4
substitution at Pd and isolation as crystals of substoichio-
metric amounts of the Cl--bound adduct 1‚Cl; other Pd
species are also formed in the bulk material during this
reaction and show JPP coupling in the 31P{1H} NMR
spectrum that is consistent with metal phosphate coordina-
tion. The 1H NMR spectrum of 1‚Cl clearly shows that anion
binding has occurred in solution because the resonances
attributed to the urea N-H’s shift significantly downfield
as compared to the resonances of the anion-free complex 1
(∆δ[DMSO-d] 0.46 ppm) and the ureaphosphine ligand L
(∆δ[DMSO-d] 0.58 ppm) (DMSO-d was used because 1 and
L are otherwise insoluble; 1‚Cl is soluble in chlorinated
solvents). This interaction was also probed using standard
NMR titration methodology (Figure 2), and the Job plots
derived from the addition of anion to solutions of both 1
To date, we have been unable to crystallize either cis-1
or cis-1‚Cl and so cannot unequivocally define the structural
effect of anion binding in this isomer. It is clear, however,
that anion binding does occur in solution in both PdII (∆δ
0.46 ppm) and also Pd0 complexes; preliminary investigations
(8) Bessel, C. A.; Aggarwal, P.; Marschilok, A. C.; Takeuchi, K. J. Chem.
ReV. 2001, 101, 1031. Eberhard, M. R.; Heslop, K. M.; Orpen, A. G.;
Pringle, P. G. Organometallics 2005, 24, 335. Freixa, Z.; Beentjes,
M. S.; Batema, G. D.; Dieleman, C. B.; van Strijdonck, G. P. F.; Reek,
J. N. H.; Kamer, P. C. J.; Fraanje, J.; Goubitz, K.; van Leeuwen, P.
W. N. M. Angew. Chem., Int. Ed. 2003, 42, 1284.
Inorganic Chemistry, Vol. 44, No. 22, 2005 7709

COMMUNICATION
-
precursors, and with or without F-, Cl-, Br-, HSO4 , and
lactate were assessed for the ambient-temperature allylic
alkylation reaction between rac-1,3-diphenyl-2-propenyl
acetate and dimethyl malonate. While all catalyst mixtures
were found to be active and achieved 100% conversion after
18 h, no significant difference was seen with anion present,
and no asymmetric induction was observed in the presence
of the chiral lactate anion; as such, and in line with the solid-
state structure of trans-1‚Cl, it would appear that a meso
isomer may be adopted in preference to the rac isomers. The
RhI complexes 2 and 2‚Cl were assessed as preformed
catalysts for the hydroformylation of octene using 1:1 H2/
CO syngas at 80 °C, 20 bar, and compared to the com-
mercially available ROPAC catalyst, [Rh(CO)(acac)(PPh3)].
While complete conversion of octene to linear/branched (2.6:
1) aldehyde was observed after 1 h using the ROPAC
catalyst, both 2 (10 h) and 2‚Cl (24 h) required prolonged
reaction times for complete conversion to an aldehyde
product of a similar linear/branched ratio (ca. 3:1 in both
cases). Reactions carried out for 2, 4, 6, 8, and 10 h using 2
and 2‚Cl were analyzed, and in both cases, a significant
induction period was observed (5 h for 2, >10 h for 2‚Cl).
The increased reactivity seen for 2 as compared to that of
2‚Cl suggests that the abstraction of chloride necessary to
form a reactive, hydridic intermediate is less hindered in 2
because the urea anion-binding pocket is blocked by a
precoordinated chloride ion in 2‚Cl.
Figure 3. X-ray crystal structure of the anion-bound complex [Bun N]-
4
[PdCl2(L)2‚Cl] (1‚Cl; 50% displacement ellipsoids). For clarity, the [Bun N]+
4
cation and all hydrogens except those of the urea nitrogens N1 and N2 are
omitted. Selected bond lengths (Å) and angles (deg): Pd1-Cl1 2.2857(18),
Pd1-Cl2 2.3007(16), Pd1-P1 2.3484(10); Cl1-Pd1-P1 93.88(3), Cl2-
Pd1-P1 87.25(3), Cl1-Pd1-Cl2 169.59(8), P1-Pd1-P1A 165.83(5). The
1
suffix A on atom P1A indicates the symmetry operation (x, /2 - y, z).
We have shown that the new (bis)ureaphosphine com-
plexes 1 and 2 bind a chloride ion to form, primarily, trans-
chelating diphosphine ligands at the metal. These supramo-
lecular interactions are observed both in solution and in the
solid state (for 1‚Cl) and appear to be retained at the elevated
temperatures required for many catalytic reactions. Pd0
complexes derived from Pd0 precursors, L, and additional
anion act as catalysts for allylic alkylation, although no
product discrimination is observed in the presence of added
anion. However, the presence of anion does appear to
decrease the efficacy of 2 in the rhodium-catalyzed hydro-
formylation of octene.
suggest that Pd0 complexes such as [Pd(TCNE)(L)2] (TCNE
) tetracyanoethylene) and its anion-bound congener [Bun N]-
4
[Pd(TCNE)(L)2‚Cl] can be synthesized and that anion
binding is observed in solution [∆δ(CDCl3) 1.48 ppm].
We have also started to explore the synthesis of RhI
complexes of L and have been able to prepare trans-[RhCl-
(CO)(L)2] (2) and trans-[Bun N][RhCl(CO)(L)2.Cl] (2‚Cl)
4
in good yield by ligand substitution from [{RhCl(CO)2}2]
(Scheme 1). The resonances at 30.1 (JRh-P ) 125.7 Hz) and
30.7 ppm (JRh-P ) 126.7 Hz) in the 31P{1H} NMR spectra
of 2 and 2‚Cl, respectively, are consistent with trans ligand
arrangements; no resonances consistent with cis isomers were
observed. The elemental analysis of 2‚Cl supports a 1:1
complex/anion ratio, and anion binding in solution is inferred
Acknowledgment. We thank the University of Notting-
ham and the EPSRC for support and the EPSRC National
Crystallography Service for collecting the diffraction data
for complex 1.
1
from the H NMR spectrum of 2‚Cl, which shows that the
urea N-H’s shift to higher frequency upon anion addition
[∆δ(2‚Cl vs 2, DMSO-d) 0.39 ppm].
The occurrence of well-defined anion binding in the above
PdII, Pd0, and RhI complexes led us to undertake a prelimi-
nary investigation of the effect of anion incorporation on
the catalytic activity of these metal complexes. Catalyst
mixtures derived from L, palladium(0) dibenzylideneacetone
Supporting Information Available: Experimental details for
all complexes and catalytic reactions (PDF) and X-ray crystal-
lographic files for 1 and 1‚Cl (CIF). This material is available free
of charge via the Internet at http://pubs.acs.org.
IC051110F
7710 Inorganic Chemistry, Vol. 44, No. 22, 2005