Synthesis of bisquinolone-based mono- and diphosphine ligands of the aza-BINAP type

Article abstract of DOI:10.1021/jo800665t

(Chemical Equation Presented) Mono- and bisphosphine ligands based on the 4,4′-bisquinolone structural framework (BIQUIP ligands) were generated by direct microwave-assisted palladium-catalyzed carbon-phosphorus cross-coupling reactions employing the corr

Full text of DOI:10.1021/jo800665t

Synthesis of Bisquinolone-Based Mono- and
Diphosphine Ligands of the Aza-BINAP Type
Nuzhat Arshad, Jamshed Hashim, and C. Oliver Kappe*
Christian Doppler Laboratory for MicrowaVe Chemistry
(CDLMC) and Institute of Chemistry, Karl-Franzens
UniVersity Graz, Heinrichstrasse 28, A-8010 Graz, Austria
oliVer.kappe@uni-graz.at
ReceiVed March 25, 2008
FIGURE 1. Bis- and monophosphine bi(hetero)aryl ligands.
separation from the reaction mixture or to modify other ligand
properties.1,3
Of more recent interest among these ligands are bisphosphines
based on biheteroaryl backbones.1,4,5 The main advantage of
these systems is the possibility to synthesize ligands with a
variety of electronic properties, as the electronic properties of
the heteroaryl rings impact directly on the electronic properties
of the phosphine ligators. In addition, a much wider variety of
potential frameworks offering more flexible synthetic routes are
available as compared to standard biaryl ligands. It has been
demonstrated, for example, that the biheteroaryl ligands BI-
TIANP (2)4 and BINAPFu (3)5 sometimes display higher
catalytic activities and enantioselectivities as compared to
BINAP in asymmetric hydrogenations and/or Heck reactions.
In addition to bisphosphines, monodentate chiral phosphine
ligands (for example MeO-MOP 4) and biaryl-based mono-
phosphines in general (for example, XPhos 5) are becoming
increasingly important as ligands for those transition-metal-
catalyzed reactions where bisphosphine-metal complexes can-
not be used because of their low selectivity toward a desired
reaction pathway.6
Inspired by the success of biheteroaryl phosphine ligands
in asymmetric catalysis,1,4,5 we became interested in the
design and generation of both mono- and bisphosphine
ligands based on the recently disclosed 4,4′-bisquinolone
framework (Figure 2).7
The putative bisquinolone phosphine ligands of the BIQUIP
type would offer several potential advantages compared to their
binaphthyl counterparts (BINAP) as the electronic properties
of the cyclic enamide system in the quinolone moiety are
expected to exert a unique impact on the electronic properties
of the phosphine ligators. In addition, the presence and proximity
of soft (phosphine) and hard (carbonyl) donor groups in the
ligand molecule provides the potential of hemilabile coordinating
abilities in this type of hybrid ligand (Figure 2).8 The 2(1H)-
quinolone system itself offers several readily available diver-
Mono- and bisphosphine ligands based on the 4,4′-bisqui-
nolone structural framework (BIQUIP ligands) were gener-
ated by direct microwave-assisted palladium-catalyzed
carbon-phosphorus cross-coupling reactions employing the
corresponding heteroaryl bromides and diphenylphosphine
as substrates.
Ligand design for asymmetric catalysis has been and con-
tinues to be an important area of synthetic organic chemistry.
Novel classes of ligands that offer additional synthetic op-
portunities or provide new insights into fundamental chemical
processes are constantly being pursued. In this context, atropi-
someric, C2-symmetric phosphines have played a crucial role
in the development of asymmetric catalysis and are among the
most effective ligand systems known today.1–5 The most
successful ligand in this class is the well-known binaphthyl
bisphosphine-based BINAP ligand (Figure 1, 1). BINAP induces
very high ee’s in several asymmetric transition-metal-catalyzed
processes, including hydrogenations, hydrosilylations, hydro-
cyanations, Heck reactions, and enamine isomerizations.2 Until
now, more than 60 frameworks based on the biaryl bisphosphine
theme have been developed in order to effectively tune
efficiency (turnover) and selectivity (ee) but also to facilitate
(1) (a) Shimizu, H.; Nagasaki, I.; Saito, T. Tetrahedron 2005, 61, 5405. (b)
ˇ
Kocˇovsky, P.; Vyskocil, S.; Smrcina, M. Chem. ReV. 2003, 103, 3213.
(2) (a) Noyori, R.; Takaya, H. Acc. Chem. Res. 1990, 23, 345. (b) Noyori,
R. AdV. Synth. Catal. 2003, 345, 15.
(3) (a) Berthold, M.; Mignani, G.; Woodward, G.; Lemaire, M. Chem. ReV.
2005, 105, 1801. (b) Tanaka, K. Synlett 2007, 1977.
(4) (a) Benincori, T.; Brenna, E.; Sannicolo´, F.; Trimarco, L.; Antognazza,
P.; Cesarotti, E.; Demartin, F.; Pilati, T. J. Org. Chem. 1996, 61, 6244. (b)
Benincori, T.; Brenna, E.; Sannicolo´, F.; Trimarco, L.; Antognazza, P.; Cesarotti,
E. J. Chem. Soc., Chem. Commun. 1995, 685. (c) Spino, C.; Mayes, N.; Desfosse´s,
H.; Sotheeswaran, S. Tetrahedron Lett. 1996, 37, 6503.
(6) (a) Hayashi, T. Acc. Chem. Res. 2000, 33, 354. (b) Gibson, S. E.; Rudd,
M. AdV. Synth. Catal. 2007, 349, 781.
(7) (a) Hashim, J.; Glasnov, T. N.; Kremsner, J. M.; Kappe, C. O. J. Org.
Chem. 2006, 71, 1707. (b) Hashim, J.; Kappe, C. O. AdV. Synth. Catal. 2007,
349, 2353.
(5) (a) Andersen, N. G.; Parvez, M.; Keay, B. A. Org. Lett. 2000, 2, 2817.
(b) Andersen, N. G.; Parvez, M.; McDonald, R.; Keay, B. A. Can. J. Chem.
2004, 82, 145.
(8) Braunstein, P.; Naud, F. Angew. Chem., Int. Ed. 2001, 40, 680.
10.1021/jo800665t CCC: $40.75
Published on Web 05/29/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 4755–4758 4755

TABLE 1. Reaction Optimization for Transition-Metal-Catalyzed
C-P Cross Couplingsa
product
entry solvent catalyst (mol %) T (°C) time (min) distributionb (%)
1
2
3
4
5
6
7
8
9
DMF
palladacyclec (3)
180
180
180
130
125
125
125
125
125
125
35
35
30
45
45
45
45
45
45
45
30
0/73/27
20/51/29
0/72/28
0/88/12
0/91/9d
18/64/18
6/70/24
42/36/13
20/57/23
88/0/12
25/12/63
FIGURE 2. Bisquinolone-based bisphosphine ligands.
dioxane palladacyclec (3)
DMSO palladacyclec (3)
sification sites9,10 and therefore the basicity at phosphorus can
be readily adjusted by incorporating substituents with differing
predefined electronic properties onto the bisquinolone core. This
considerably expands the tunability of electronic properties on
the ligating phosphorus and also allows some influence on the
dihedral angle by variation of the Z-substituent (Figure 2).1
Herein, we outline a short synthetic route toward bisquino-
lones that utilizes Pd-catalyzed C-P cross-coupling protocols
involving the corresponding 3-bromo- and 3,3′-dibromobisqui-
nolones as starting materials.
In order to test the practicability of the planned C-P cross-
coupling reactions, we initially performed optimization studies
on the readily available 3-bromo-1-methyl-4-phenylquinolin-
2(1H)-one (6).9 For the transition-metal-catalyzed couplings of
bromide 6 with diphenylphosphine, the reaction parameters were
optimized using controlled microwave heating in sealed reaction
vessels.11 Based on our previous experience with metal-
catalyzed C-P coupling protcols,12 we were able to quickly
optimize the conditions for the specific heteroaryl bromide
substrate 6 (Table 1). Optimum conversion and selectivity for
cross-coupling was obtained by employing equimolar amounts
of the bromide substrate (6) and KOAc base, in combination
with 3 mol % of Herrmann’s palladacycle catalyst and a slight
excess of diphenylphosphine (1.2 equiv). At a reaction temper-
ature of 125-130 °C, complete conversions were achieved
within 45 min using either dry THF (entry 4) or n-BuCl (entry
5) as solvent. From the experiment described in entry 5, a 70%
yield of 3-diphenylphosphino-2-quinolone 7 was isolated by
flash chromatography. The use of other solvents (entries 1-3)
or different Pd or Ni catalysts (entries 6-11) provided significant
amounts of dehalogenated byproduct or led to very low
conversions.
THF
palladacyclec (3)
n-BuCl palladacyclec (3)
n-BuCl Pd2(dba)3 (3.5)
n-BuCl PdCl2(dppf) (10)
n-BuCl Pd(OAc)2 (5)
n-BuCl Pd(PPh3)4 (8)
10 n-BuCl Pd/C (10)
11 DMF
NiCl2(PPh3)2 (20) 180
a Reaction conditions: 0.25 mmol of 3-bromoquinolone 6, catalyst
(3-20 mol %), 0.25 mmol of KOAc, 1.5 mL of dry solvent, 1.2 equiv
of Ph2PH, Ar atmosphere, single-mode microwave irradiation. b Product
distribution refers to relative peak area (%) ratios of crude HPLC-UV
(215 nm) traces: starting material/product/dehalogenated product. c Herrmann’s
palladacycle,
[trans-di(µ-acetato)bis[o-di-o-tolylphosphino)benzyl]dipalla-
dium(II). d 70% isolated yield by flash chromatography.
SCHEME 1
For the generation of the desired BIQUIP ligands (Figure 2) it
was necessary to first prepare the corresponding mono- and
dibromobisquinolone derivatives 9 and 10. In order to have ligands
available with differing electronic properties in the aromatic ring
we have chosen the 6,7-dimethoxy- and the 6-trifluoromethyl
analogues as our initial synthetic targets.13 The required
bisquinolones 8a,b were obtained following our recently
reported Pd- or Ni-catalyzed methods.7 Subsequent dibromina-
tion of dimethoxybisquinolone 8a in DMF with N-bromosuc-
cinimide (NBS) at room temperature (5 equiv, 50 min), provided
the desired bisbromo derivative 10a in 82% isolated yield
(Scheme 1). Controlled monobromination using only 2.5 equiv
of NBS at rt for 35 min furnished a 62% isolated yield of
monobromo analogue 9a after separation from the bisbromo
derivative 10a by flash chromatography. In sharp contrast, the
electron deficient bisquinolone 8b proved exceedingly difficult
to brominate and forcing conditions had to be employed.
Optimum isolated yields (68%) of the mono- bromo derivative
9b were obtained by applying microwave-assisted NBS bro-
mination conditions in MeCN (7 equiv NBS, 150 °C, 2.5 h).9
Subsequent addition of an additional quantity of NBS (7 equiv)
(9) Glasnov, T. N.; Stadlbauer, W.; Kappe, C. O. J. Org. Chem. 2005, 70,
3864.
(10) (a) Kappe, T. Farmaco 1999, 54, 309. (b) Kappe, T.; Karem, A. S.;
Stadlbauer, W. J. Heterocycl. Chem. 1988, 25, 857. (c) Kreimeyer, A.; Laube,
B.; Sturgess, M.; Goeldner, M.; Foucaud, B. J. Med. Chem. 1999, 42, 4394. (d)
Jung, J.-C.; Jung, Y.-Jo.; Park, O.-S. J. Heterocycl. Chem. 2001, 38, 61.
(11) (a) Kappe, C. O. Angew. Chem., Int. Ed. 2004, 43, 6250. (b)
Appukkuttan, P.; Van der Eycken, E. Eur. J. Org. Chem. 2008, 1133.
(12) Stadler, A.; Kappe, C. O. Org. Lett. 2002, 4, 3541.
(13) (a) Electron-donating substituents such as methoxy groups on the
aromatic ring have in some cases resulted in higher catalytic activity in
asymmetric hydrogenations. Takahashi, H.; Achiwa, K. Chem. Lett. 1989, 18,
305. (b) Morimoto, T.; Chiba, M.; Achiwa, K. Tetrahedron Lett. 1989, 30, 735.
(c) Yamamoto, N.; Murata, M.; Morimoto, T.; Achiwa, K. Chem. Pharm. Bull.
1991, 39, 1085.
4756 J. Org. Chem. Vol. 73, No. 12, 2008

SCHEME 2
SCHEME 3
and resubjection to the microwave conditions (150 °C, 2.5 h)
led to the bisbromo derivative 10b in 60% overall yield.
at 160 °C for 10 min cleanly provided monophosphine 11a in
75% isolated yield.
With all the required four bromo precursors in hand, we
continued to investigate the key mono and double C-P cross-
coupling transformations leading to the desired bisquinolone
phosphine ligands. Starting with dibromo bisquinolone 10a, it
was quickly realized that the initially optimized reaction
conditions for the model substrate 6 (Table 1) had to be adjusted
for bisquinolone 10a mainly because of solubility issues. An
extensive reoptimization revealed that for this particular sub-
strate, DMF proved to be a very suitable solvent for bisphos-
phination. Thus, treatment of dibromide 10a with 2.2 equiv of
Ph2PH, 3.5 equiv of KOAc, and 13 mol% of Herrmann’s
palladacycle in dry DMF at 160 °C for 10 min (MW) produced
a 73% isolated product yield of bisphosphine bisquinolone
ligand 12a (Scheme 2). Under these conditions, the highest
selectivities for bisphosphination could be achieved. Using lower
amounts of catalyst and base resulted in incomplete conversions
(monophosphination), while at higher reaction temperatures (180
°C) more dehalogenated byproducts could be observed.
Optimum results were obtained at relatively high substrate
concentrations which kept oxidation to the corresponding
phosphine oxides to a minimum. In order to isolate the
bisphosphine ligand 12a, an aqueous workup followed by
extraction into chloroform and precipitation with methanol was
performed. Standard chromatographic workup using silica gel
chromatography caused partial decomposition of the material
on the column. The isolated bisphosphine bisquinolone 12a
proved to be quite air stable but could be oxidized to the
corresponding diphosphine oxide 14a by stirring with an excess
of aqueous H2O2 in THF.14
The same synthetic principles as outlined in Scheme 2 for
the electron-rich dimethoxyquinolone scaffolds were then ap-
plied to the electron-deficient trifluoromethyl-substituted qui-
nolones (Scheme 3). The corresponding mono- and dibromide
precursors 9b and 10b, respectively, were therefore subjected
to the Pd-catalyzed coupling conditions. For these particular
substrates, the use of THF as solvent and 130 °C as reaction
temperature under otherwise identical reaction conditions proved
optimal. In contrast to the electron-rich methoxy-functionalized
bisquinolones 11a and 12a (Scheme 2), however, phosphino
bisquinolones 11b and 12b bearing electron-withdrawing CF3
groups in the 6,6′ positions proved to be somewhat oxidation
sensitive during isolation. Typically, mixtures of the phosphines
and their mono- and dioxides (for 12b) were obtained after
extractive workup. Therefore, for characterization purposes the
crude reaction products were rapidly oxidized (30 min) with
aqueous H2O2 in THF to provide the corresponding oxides 13b
and 14b in good overall yields (Scheme 3).
In addition to our interest in the aza-BINAP analogues
described above, we also wanted to explore the possibility of
synthesizing aza-BINOL derivatives based on the 4,4′-bisqui-
nolone framework. Since 1990 the enantiomeric atropisomers
of 1,1′-binaphthyl-2,2′-diol (BINOL) have become the most
widely used ligands both for stoichiometric and catalytic
asymmetric reactions.15 The extraordinary success of this
molecule in a wide range of asymmetric transformations and
in chiral recognition has led to the development of several
modified BINOL analogues in order to fine-tune steric and
electronic factors and for the provision of additional binding
sites.16 However, very little is known on BINOL analogues that
are derived from heterobiaryl frameworks.17
A byproduct (ca. 10%) in the phosphination 10a f 12a under
certain conditions was the monophosphinated bisquinolone 11a
presumably formed by a monophosphination/monodehalogena-
tion pathway. Since biaryl-type monophosphines are becoming
increasingly important as ligands in transition-metal-catalyzed
organic transformations,6 monophosphinated bisquinolone 11a
was also synthesized directly in high yield from the monobromo
bisquinolone 9a. Thus, treatment of 9a with 1.2 equiv of Ph2PH,
2 equiv of KOAc and 7 mol % of Herrmann’s catalyst in DMF
The Pd-catalyzed formation of a phenol using 3-bromo-4-
phenylquinolinone 6 as a model substrate for 3,3′-dibromobis-
quinolones 10 is demonstrated in Scheme 4. Following a recent
protocol by Buchwald and co-workers,18 introducing the use
of 2-di-tert-butylphosphino-2′,4′,6′-triisopropylbiphenyl (tert-
(15) Brunel, J. M. Chem. ReV. 2005, 105, 857.
(14) Gelpke, A. E. S.; Kooijman, H.; Spek, A. L.; Hiemstra, H. Chem. Eur.
J. 1999, 5, 2472.
(16) Chen, Y.; Yekta, S.; Yudin, A. K. Chem. ReV. 2003, 103, 3155.
(17) Upadhgay, S. P.; Karnik, A. V. Tetrahedron Lett. 2007, 48, 317.
J. Org. Chem. Vol. 73, No. 12, 2008 4757

SCHEME 4
mixture was purified by automated flash chromatography using
EtOAc/CH3CN as eluent to give 122 mg (82%) of 10a as a yellow
solid: mp 319-320 °C (ethanol); IR (KBr) νmax 1642, 1251, 1166,
1
645 cm-1; H NMR (360 MHz, CDCl3) δ 3.63 (s, 6H), 3.96 (s,
6H), 4.05 (s, 6H), 6.45 (s, 2H), 6.91 (s, 2H); 13C NMR (90 MHz,
CDCl3) δ 31.8, 56.3, 56.5, 97.7, 107.5, 111.5, 115.6, 135.2, 145.2,
146.1, 153.0, 158.2; MS (positive APCI) m/z 594 (100, M), 596
(65, M + 2), 592 (55, M - 2), 595 (30, M + 1); HRMS (APCI+)
calcd for C24H23N2O6Br2 [M + H]+ 592.99174, found 592.99168.
4,4′-Bis(3-diphenylphosphanyl-6,7-dimethoxy-1-methylquinolin-
2(1H)-one) (12a, BIQUIP) (Scheme 2). A mixture of 61.2 mg (0.103
mmol) of 4,4′-bis-(3-bromo-6,7-dimethoxy-1-methylquinolin-2(1H)-
one) (10a), 12.6 mg (0.013 mmol, 13 mol %) of Herrmann’s
palladacycle, and 35.4 mg (0.36 mmol, 3.5 equiv) of KOAc in a 2
mL Pyrex microwave vial was equipped with a magnetic stirring
bar and sealed. After flushing with Ar, 0.8 mL of anhydrous DMF
and 39.2 µL (0.23 mmol, 2.2 equiv) of Ph2PH were added with
shaking. The resulting mixture was stirred for 5 min at room
temperature under Ar and then heated for 10 min at 160 °C via
single-mode microwave irradiation. After being cooled to ambient
conditions, the mixture was diluted with 5 mL of water and
extracted with 3 × 15 mL of CHCl3. The combined organic extracts
were washed with saturated aqueous KCl solution and dried over
MgSO4 for 2 h. After filtration and evaporation, the remaining crude
mixture was dissolved in a few drops of CHCl3 and after addition
of excess MeOH, the yellowish precipitate was filtered and dried
to give 60.6 mg (73%) of bisphosphine 12a as yellow crystals: mp
278-279 °C (ethanol); IR (KBr) νmax 1633, 1253 cm-1; 1H NMR
(360 MHz, CDCl3) δ 3.21 (s, 6H), 3.78 (s, 6H), 4.07 (s, 6H), 6.39
(s, 2H), 6.86 (s, 2H), 6.96-7.57 (m, 20H); 31P NMR (90 MHZ,
CDCl3) δ -10.06 (1 signal); 13C NMR (90 MHz, CDCl3) δ: 30.2,
56.3, 97.4, 106.2, 109.1, 113.4, 121.0, 126.0, 129.1, 129.6, 136.9,
142.7, 145.8, 148.1, 153.5, 161.4; MS (positive APCI) m/z 804
(100, M), 805 (60, M + 1), 806 (18, M + 2), 620 (35, M - 184);
HRMS (APCI+) calcd for C48H43N2O6P2 [M + H]+ 805.25909,
found 805.25890.
butyl X-phos, 16) as ligand in related Pd-catalyzed couplings,
we were able to synthesize 3-hydroxy-4-phenyl-1-methylquino-
lin-2(1H)-one 15 in good yield from the bromo precursor 6.
Optimum isolated yields/conversions were obtained by subject-
ing aryl bromide 6 in a 1:1 THF/H2O mixture containing 4 equiv
of KOH as base to microwave irradiation for 15 min at 170 °C.
The use of 4 mol % of Pd2(dba)3 in combination with 8 mol %
of ligand 16 resulted in the best selectivity for C-O bond
formation (93%) with only 7% of dehalogenated byproduct
being formed. The use of other ligand systems such as, for
example, diphenylphosphinoferrocene (dppf) was significantly
less effective and produced more dehalogenated byproduct under
all tested reaction conditions.19
In conclusion, we have shown that both mono- and bispho-
sphine ligands derived from the 4,4′-bisquinolone framework
can be readily synthesized in a two-step bromination/phosphi-
nation protocol starting from the basic biheteroaryl core.
Preliminary studies confirm that these novel types of aza-BINAP
phosphines containing an enamide moiety incorporated into the
heteroaryl ring are indeed useful ligands in transition-metal-
catalyzed transformations (see the Supporting Information).
Experimental Section
4,4′-Bis(3-bromo-6,7-dimethoxy-1-methylquinolin-2(1H)-one) (10a)
(Scheme 1). A mixture of 109 mg (0.25 mmol) of 4,4′-bis(6,7-
dimethoxy-1-methylquinolin-2(1H)-one) (8a), 222.4 mg (1.25
mmol, 5.0 equiv) of N-bromosuccinimide (NBS), and 0.6 mL of
DMF in a vial equipped with a magnetic stirring bar was stirred
for 50 min at room temperature. The reaction was quenched by
addition of ice-water and was filtered. The resulting crude product
Acknowledgment. N.A. and J.H. thank the Higher Education
Commission of Pakistan for Ph.D. scholarships.
Supporting Information Available: Experimental proce-
dures and characterization data of compounds 7-11 and 13-15.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(18) Anderson, K. W.; Ikawa, T.; Tundel, R. E.; Buchwald, S. L. J. Am.
Chem. Soc. 2006, 128, 10694.
(19) For aminations of the Buchwald-Hartig type starting from bromide 6,
see: Wang, Z.; Wang, B.; Wu, J. J. Comb. Chem. 2007, 9, 811.
JO800665T
4758 J. Org. Chem. Vol. 73, No. 12, 2008