DOI: 10.1002/chem.201101467
Improved Syntheses of Phosphine Ligands by Direct Coupling of
Diarylbromophosphine with Organometallic Reagents
AHCTUNGTRENNUNG
Lei Liu,[a] Hai-Chen Wu,*[a] and Jin-Quan Yu[b]
In memory of Jonathan B. Spencer
Ligand synthesis lies at the heart of modern organometal-
lic chemistry. The design and synthesis of new ligands plays
a key role in the development of highly efficient asymmetric
catalysis. Among those thousands of various ligands that
have been synthesized so far, phosphine ligands are un-
doubtedly the most used ligands in both asymmetric cataly-
sis and Pd catalysis.[1]
During our continued mechanistic studies of homogene-
ous hydrogenation,[2] the preparation of chiral phosphine li-
gands with systematically tuned electronic properties is es-
sential. For instance, we synthesized a series of BINAP-type
phosphine ligands 1–5 (Scheme 1; BINAP=2,2’-bis(diphe-
nylphosphino)-1,1’-binaphthyl) to investigate the interplay
between the electronic properties of ligands and substrates
and how this affects the enantioselection of rhodium-cata-
lyzed asymmetric hydrogenation reactions.[2c] BINAP is one
of the most efficient ligands for chiral induction and has
been most extensively studied in asymmetric catalysis since
it was developed by Noyori et al in 1980.[3] We therefore
chose BINAP as the model ligand and began to tune the
electronic properties of the four phenyl rings.[4]
The first practical synthesis of BINAP was reported by
Noyori et al. in 1986 with an overall yield of 14%.[5] An im-
proved method was reported by Cai et al. (Merck Inc.) in
1994 involving a nickel-catalyzed coupling reaction between
the chiral ditriflate of BINOL and diphenylphosphine. The
yield of this significantly shorter route was as high as ap-
proximately 75%.[6] A modified version of Caiꢀs method
was later developed by Laneman et al (Monsanto) in 1997,[7]
which also involves chiral BINOL ditriflate (BINOL=1,1’-
binaphthalene-2,2’-diol) but has the advantage of using in-
dustrially available diphenylchlorophosphine as a starting
material. We initially chose the Merck method for the syn-
theses of ligands 1–5 due to its short sequence (Scheme 1).
When the phenyl rings were substituted by electron-donat-
ing groups, such as p-tolyl (ligand 4) or m-xylyl (ligand 5)
groups, the substituted BINAPs were obtained in moderate
yields (52% for ligand 4 and 46% for ligand 5). However,
when electron-withdrawing groups (such as CF3) were intro-
duced to the phenyl ring, this method gave either only a
trace amount of the desired products or none at all. The
Monsanto method was also tested for the syntheses of li-
gands 1 and 2 but no desired products were obtained.
We then turned to Noyoriꢀs method for the syntheses of
the oxides of ligands 1 and 2. Binaphthylmagnesium bro-
mide was coupled with bisACHTUNTGRNEUNG[3,5-bis(trifluoromethyl)-phenyl]-
phosphinyl chlorides to afford phosphine oxide 1a
(Scheme 2). The complex product mixture was subjected to
silica gel chromatography and all the products were isolated
and characterized. The products were, in order of elution,
Scheme 1. Syntheses of ligands 1–5 using the Merck method. Conditions:
triflate (1.0 equiv), XPAr2 (1.05 equiv), [NiACHTNUTRGNE(NUG dppe)Cl2] (0.2 equiv),
DABCO (8.0 equiv) in DMF (10 mLmmolÀ1 triflate); heating at 1008C
for 2–3 days.
[a] Dr. L. Liu, Prof. Dr. H.-C. Wu
1,1’-binaphthylene (7; 6%), tris
nyl]phosphine oxide (8; 17%), 2,2’-dibromo-1,1’-binaphthyl
(6; 3%), 2,2’-bis(bis[3,5-bis(trifluoromethyl)phenyl])phos-
phinyl-1,1’-binaphthyl (1a; 18%), bis[3,5-bis(trifluorome-
ACHTUNGTRNE[NUNG 3,5-bis(trifluoromethyl)phe-
Key Laboratory for Biomedical Effects
of Nanomaterials & Nanosafety, Institute of High Energy Physics
Chinese Academy of Sciences, 19B Yuquan Road
Beijing 100049 (P. R. China)
Fax : (+86)10-88235745
A
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
thyl)phenyl]phosphinic acid (9; 5%) and bis(binaphthyl)di-
phosphine oxide 10 (40%; see the Supporting Information
for characterization data) (Scheme 2). Interestingly, when
the reaction was carried out at room temperature, 8 and 10
became the main products and 1a was not found in the mix-
ture. The formation of 8 indicated the presence of a nucleo-
[b] Prof. Dr. J.-Q. Yu
Department of Chemistry, The Scripps Research Institute
10550 N. Torrey Pines Rd., La Jolla, California 92037 (USA)
Supporting information for this article is available on the WWW
10828
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10828 – 10831