P-alkene bidentate ligands: an unusual ligand effect in Pd-catalysed Suzuki reactions

Article abstract of DOI:10.1016/j.tet.2006.12.012

The versatile reagent bis(diphenylphosphino)benzaldehyde was used to prepare a variety of electronically and sterically varied ligands using Wittig or aldol methodology in high yields. Pd-catalysed Suzuki reactions were evaluated using these ligands. The influence of the functional group on the alkene, which is in direct conjugation with the phosphorus centre of the ligand, was visible in the activity profiles of the reactions. In general, the activity and stability of the Pd-ligand system increased as electron deficiency and steric bulk increased at the alkene. Furthermore, the ligands could be used at decreased catalyst loadings with improved yields and did not themselves participate in the reactions as substrates.

Full text of DOI:10.1016/j.tet.2006.12.012

Tetrahedron 63 (2007) 1624–1629
P–alkene bidentate ligands: an unusual ligand effect
in Pd-catalysed Suzuki reactions
*
D. Bradley G. Williams and Megan L. Shaw
Department of Chemistry, University of Johannesburg, PO Box 524, Auckland Park, Johannesburg 2006, South Africa
Received 10 August 2006; revised 21 November 2006; accepted 7 December 2006
Abstract—The versatile reagent bis(diphenylphosphino)benzaldehyde was used to prepare a variety of electronically and sterically varied
ligands using Wittig or aldol methodology in high yields. Pd-catalysed Suzuki reactions were evaluated using these ligands. The influence
of the functional group on the alkene, which is in direct conjugation with the phosphorus centre of the ligand, was visible in the activity profiles
of the reactions. In general, the activity and stability of the Pd–ligand system increased as electron deficiency and steric bulk increased at the
alkene. Furthermore, the ligands could be used at decreased catalyst loadings with improved yields and did not themselves participate in the
reactions as substrates.
Ó 2007 Elsevier Ltd. All rights reserved.
7
1
. Introduction
to be carried out. Scrivanti has recently reported on the sta-
bilisation of catalyst systems in the presence of electron-de-
ficient alkenes. On the other hand, earlier work by Amatore
8
Despite the existence of several ‘ligandless’ transition metal-
mediated reactions, ligands play pivotal roles in directing
the outcome of the overwhelming majority of such catalysed
transformations. For example, unactivated aryl halides such
as chlorides or electron-rich bromides may usually only
be used in Pd-catalysed reactions when a very electron-rich
1
9
et al. has indicated that dba (as originating from Pd(dba)
catalyst precursors) strongly competes with PPh for com-
plexation to co-ordinatively unsaturated Pd(PPh ) , slowing
3
the overall rate of the oxidative addition step (using PhI) in
the catalytic cycle and exerting a controlling effect on the
apparent rate of the overall reaction. Furthermore, the al-
kene reactant present in Heck reactions has been shown to
decelerate the oxidative addition step, with methyl acrylate
retarding that step more than styrene does. Fairlamb has
broadened the scope of the prior work by employing a range
of functionalised dba ligands, indicating that the more elec-
tron-rich the dba ligand in the catalyst precursor (by virtue of
aryl substitution on the dba ligand), the higher the overall
rate of the oxidative addition reaction. This is because
the functionalised dba ligand competes less efficiently for
the catalytically active Pd(PPh ) species when it is elec-
tron-rich, allowing higher concentrations of this species to
exist. With this as a background, we were motivated to report
our work on the synthesis and evaluation of ‘bidentate’
phosphine–alkene ligands derived from the readily available
bis(diphenylphosphino)benzaldehyde.
2
3
2
9
a
2
9b
ligand is used to modify the metal. Similarly, workers at
Sasol have recently very elegantly shown the ligand-depen-
dent effect of the co-catalysed oligomerisation of ethylene
to form 1-hexene or 1-octene, depending on the type of ligand
employed. Furthermore, the hydroformylation reaction has
been shown to be highly dependent upon the presence of a
ligand and veryspecifically its bite angle. It is therefore clear
that various characteristics ofligands modify the outcomes of
reactions.
9
c
3
4
1
0
In the instance of Pd-mediated chemistry, electron-deficient
alkenes have been shown to be useful as reagents in Heck
reactions. Conversely, several workers have employed suc-
cinic anhydride and similar electron-deficient alkenes not as
substrates but as added ligands to prove the intermediacy
of L Pd(0) type materials, the complexes of which provide
3
2
5
2
not only characteristic NMR data arising from substantial
p-back donation from the electron-rich Pd to the p-acidic
alkene, but also allow X-ray structural determinations
6
2. Results and discussion
Bis(diphenylphosphino)benzaldehyde 1 was converted in
high yields into several alkenes with varying electronic
properties at the alkene moiety, using Wittig or aldol-type
technology (Scheme 1).
Keywords: Suzuki; Sterics; Ligand effects; Palladium.
*
Corresponding author. Tel.: +2711 489 3431; fax: +2711 489 2819;
e-mail: dbgw@rau.ac.za
0
040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.12.012

D. B. G. Williams, M. L. Shaw / Tetrahedron 63 (2007) 1624–1629
1625
2
. R = H
3. R = CO CH
3
complexes of this study, it is the ability of the alkene to ac-
cept this electron density that allows such p-back donation
to take place. ‘Unactivated’ alkenes are electron-rich, while
enoate esters are electron-deficient, the latter being better
able to accommodate p-back donation into their lower
Ph P
2
2
4. R = CO C(CH )
2 3 3
Ca
R
Ph P
Ha
2
Wittig
Cb
Hb
or aldol
O
1
1
lying p* (LUMO) orbitals. Alkene complexation is also
witnessed by changes in the C]O IR stretching frequencies.
H
or
1
5
. R = CO CH CH
Application of the complexes to a Pd-catalysed Suzuki reac-
tion (Scheme 2) making use of bromobenzonitrile (6) and
phenylboronic acid (7) yielded exciting and intriguing
results, which are summarised in Tables 2–4. The results
2
2
3
Ph P
2
Ha
Ca
Cb
R
ꢀ
of reactions performed at 110 C show a clear trend of
R
enhanced catalyst activity and stability that mimics the
series of ligands from most electron-rich to most electron-
poor, the latter providing the best results (rate, yield and
catalyst stability). The stability of the catalyst was gauged
by the formation or not of Pd black, the formation of which
corresponded with a dramatic slowing in the progress of the
reaction if the reaction was not already essentially complete,
also indicative of a less stable catalyst.
Scheme 1.
Accordingly, reaction of 1 with methyltriphenylphospho-
nium bromide in the presence of sodium amide afforded
the styrene derivative 2 in 81% yield, while the enoate
derivatives 3 and 4 were prepared making use of (methoxy-
carbonylmethylene) triphenylphosphorane and (tert-butoxy-
carbonylmethylene) triphenylphosphorane in yields of 92%
and 80%, respectively. Diester 5 was prepared using a base-
promoted Knoevenagel reaction in a yield of 66%.
OH
x% Pd(OAc)
2
Br
B
OH
2x% Ligand
+
Potassium
phosphate
DMF
NC
NC
Each of the new ligands (2 equiv) was complexed in situ
with Pd(dba) by simply mixing the ligand and Pd(dba) in
6
7
8
2
2
deuterated chloroform. NMR spectroscopy unambiguously
showed complexation of the Pd metal to the P(III) atom by
a substantial (ca. 40 ppm) down-field shift as detected in
Scheme 2.
It is clear from the results in Table 2 that all of the reactions
are fast, with the exception of those involving styrene ligand
2 and, to a lesser extent, PPh . It is also evident that the
ligands containing electron-withdrawing groups provide an
enhanced rate of reaction, which is further increased with
additional steric bulk. Saturated versions of ligands 3 and
4 showed poor activity in these reactions, highlighting the
need for and the effect of the unsaturation in enhancing
the activity of catalysts derived from our ligands.
3
1
the P NMR spectrum for each of the Pd complexes of
ligands 2, 3, 4 and 5 (Table 1), which is consistent with pub-
lished data. Here, the signals manifested as one sharp
3
6
singlet for each ligand with only very little or no free ligand
being detected. This indicated that the complex, whatever its
actual constitution and geometry be, gave rise to equivalent
phosphines. This is in contradistinction to prior observations
on the complexation behaviour of Pd(dba) towards PPh ,
2
3
9
a
where non-equivalent phosphines were observed. Addi-
tionally, the NMR spectra also clearly proved that the alkene
forms a chelate structure to the Pd centre by strong up-field
shifts of the signals characteristic of the unsaturated moiety
a
Table 2. Yields of biphenyl 8 with varying ligands
Time
(h)
PPh
(%)
3
Ligand 2
(%)
Ligand 3
(%)
Ligand 4
(%)
Ligand 5
(%)
1
13
in both the H and C NMR spectra for Pd complexes of 3, 4
and 5, also consistent with the literature. In the case of sty-
6
1
2
4
8
1
2
50
56
61
67
69
71
14
16
17
21
22
26
64
70
74
83
84
85
63
71
76
83
86
87
79
82
86
88
91
93
3
1
rene derivative 2, only the shift in the P NMR spectrum
was observed; no concomitant shifts in the other spectra were
seen. This situation is readily explained by the fact that the
extent of p-back donation from a metal onto a ligand is
strongly dependent on the ability of that ligand to accept
additional electron density. In the cases of the ligands and
2
4
a
ꢀ
Pd(OAc)
2
(5%), 10% ligand, 110 C, DMF, isolated yield.
Table 1. Selected physical data for free versus Pd(0)-complexed ligands 2–5
31
a
H (ppm)
b
H (ppm)
a
C (ppm)
b
C (ppm)
ꢁ1
CO (cm
Ligand
P (ppm)
n
)
Free ligand 2
Pd complex of 2
ꢁ12.9
7.34
7.35
5.40
5.36
129.0
129.7
116.0
116.9
n/a
n/a
32.9
Free ligand 3
Pd complex of 3
ꢁ14.0
8.44
5.95
6.27
4.44
139.1
95.5
119.7
72.8
1713
1695
29.3
Free ligand 4
Pd complex of 4
ꢁ13.1
8.28
4.40
6.17
4.40
139.1
94.8
121.9
73.2
1700
1683
31.0
Free ligand 5
Pd complex of 5
ꢁ12.5
8.25
7.34
n/a
n/a
142.0
112.7
127.8
97.4
1727
1697
28.6

1626
D. B. G. Williams, M. L. Shaw / Tetrahedron 63 (2007) 1624–1629
ꢀ
a
Table 3. Yields of biphenyl 8 in reactions performed at 82 C
scale reactions. This negated the possibility of the ligands
being subject to such Pd-mediated reactions, the products
of which might have been argued to be the actual and active
ligands for the Pd metal.
Time (h) Ligand 2 (%) Ligand 3 (%) Ligand 4 (%) Ligand 5 (%)
1
2
4
8
1
2
14
21
25
28
29
29
13
31
43
57
63
68
16
43
56
66
69
72
14
35
59
73
81
83
A second series of ligands was synthesised using the Knoe-
venagel reaction, producing malonate-derived ligands 9, 10
and 11 that were structurally similar to 5 (Scheme 3), in an
effort to probe the stereoelectronic influence of the ligands
more fully.
2
4
a
ꢀ
2
Pd(OAc) (5%), 10% ligand, 82 C, DMF, isolated yield.
Table 4. Yields of biphenyl 8 at lower catalyst loading (1% Pd) in reactions
ꢀ
Di-alkyl malonate
a
performed at 82 C
Ph P
Ph P
2
2
Piperidine
Time (h) Ligand 2 (%) Ligand 3 (%) Ligand 4 (%) Ligand 5 (%)
Toluene
O
1
40 °C
H
CO R
2
1
2
4
8
1
2
13
18
22
31
36
51
25
28
34
44
53
73
25
35
43
54
62
75
29
43
52
62
72
84
RO C
2
1
9. R = CH
3
10. R = CH(CH )
3
2
2
4
11. R = C(CH )
3
3
a
Scheme 3.
2
Pd(OAc) (1%), 2% ligand, DMF, isolated yield.
The Knoevenagel products were evaluated as ligands in the
Suzuki reaction (Scheme 2) making use of 5% catalyst and
1% catalyst (Table 6). As before, the lower catalyst loading
afforded improved reactions, and at both catalyst loadings
there was a distinct trend of increasing yield in response to
increased steric hindrance of the ligand. Catalysts derived
from the latter set of ligands proved more efficient and robust
than the analogous monoester ligands 2–4.
Table 5. Yields of biphenyl 8 at lower catalyst loading (1% Pd) in reactions
ꢀ
a
performed at 110 C
Time (h) Ligand 2 (%) Ligand 3 (%) Ligand 4 (%) Ligand 5 (%)
1
2
4
8
1
56
65
74
80
87
84
87
89
91
92
82
86
88
90
94
85
87
88
93
97
2
Rate enhancements in otherwise identical catalysed reac-
tions usually imply that the particular ligand speeds up the
rate-limiting step in the reaction. The now-accepted three-
step mechanism of the Suzuki reaction involves oxidative
addition of the aryl halide to the active Pd(0) catalyst,
transmetallation from the boronic acid or ester to form an
Ar–Pd–R species and reductive elimination as the final
a
2
Pd(OAc) (1%), 2% ligand, DMF, isolated yield.
1
2
A similar reactivity trend was obtained when carrying out
ꢀ
the reactions at 82 C (Table 3). Here too, the activity
increased with increasing electron deficiency. For PPh3
and ligand 2 rapid deactivation of the catalyst and concom-
itant precipitation of Pd black were observed at both temper-
(
C–C bond-forming) step. Electron-rich ligands tend to
2
enhance the rate of the oxidative addition step, since these
ligands destabilise the Pd(0) oxidation state and render the
metal centre more nucleophilic and susceptible to oxidation.
In contrast, electron-deficient ligands tend to destabilise the
higher Pd(II) oxidation state, thereby promoting the reduc-
tive elimination step. For the present set of ligands, we
believe that it is the rate of the transmetallation step or the
reductive elimination step that is being enhanced, and not
the oxidative addition step, argued as follows.
ꢀ
ꢀ
atures (110 C and 82 C).
Having experienced success at the relatively high catalyst
loading of 5% Pd, we reduced the loading to 1% Pd (Tables
4
slight rate enhancements were observed over the first hour
in otherwise similar yielding reactions. Most of the benefit
was seen for ligand 2 (Table 3 vs Table 4).
ꢀ
and 5). At the lower catalyst loading of 1% Pd at 82 C,
ꢀ
If one assumes a P-only co-ordination mode of the ligand
onto the metal during the catalytic cycle, then the complex
of ligand 2 would oxidatively add fastest, followed by the
catalyst generated from enoate esters 3 and 4, with those
of diesters 5 and 9–11 being slowest, since the alkene is in
conjugation with the P-atom. This conjugation would
When repeating the reactions at 110 C, extremely fast
reactions were observed for all of the ligands employed
and very high yields were obtained (Table 5). The reactions
were essentially complete within the first hour, with slow
increases in yields being observed up to 12 h. The higher
activity at the lower catalyst loading may be ascribable to
lower levels of aggregation of the homogeneous catalyst.
Notwithstanding the overall high rates and yields of these
reactions, the overall relative catalyst activity followed the
same pattern previously observed.
Table 6. Yields of 8 using Knoevenagel products as ligands
i
t
9
(R¼Me)
5 (R¼Et)
81
85
10 (R¼ Pr)
86
94
11 (R¼ Bu)
88
97
a
a
a
a
8
0
b
b
b
b
8
4
The ligands themselves did not participate in Pd-mediated
reactions of the Heck-type (to which they are activated),
and could be isolated intact in yields of >80% from large-
a
ꢀ
Pd(OAc)
Pd(OAc)
2
(5%), 10% ligand, 82 C, 12 h, isolated yield of 8.
ꢀ
b
2
(1%), 2% ligand, 82 C, 12 h, isolated yield of 8.

D. B. G. Williams, M. L. Shaw / Tetrahedron 63 (2007) 1624–1629
1627
decrease the electron density of the P-atom along this gen-
eral series, leading to slower oxidative addition reactions.
If one assumes a chelated binding mode, as has been demon-
strated to be the case for these ligands, then a scenario in
which the oxidative addition step is rate-limiting would
also follow the order 2>3,4>5,9,10,11 since the extent of
p-back donation is expected to increase along this series,
thereby decreasing the electron density on the Pd atom.
That this is not the observed order of activity eliminates
the oxidative addition step as rate-limiting in these reactions
QP2010 instrument fitted with quadrupole mass detector.
High resolution mass spectra were recorded on a VG
70-SEQ. DMF was pre-dried over phosphorus pentoxide
and distilled over calcium hydride; THF was distilled from
sodium/benzophenone and DCM was dried by passing it
through alumina followed by distillation over calcium
hydride prior to use. All other materials were commercially
available. All reactions were carried out under an inert
atmosphere. Column chromatography was performed on
silica gel (230–400 mesh) using eluants as indicated.
(
in fact, the ligands may well slow the rate of this step). By
default, therefore, it is either the transmetallation or the re-
ductive elimination step that is rate-limiting and is enhanced
by our P–alkene ligands. The p-accepting ligands of this
study would tend to destabilise all Pd(II) reaction intermedi-
ates, and should therefore promote all steps involving Pd in
this oxidation state. At this stage we cannot ascertain which
of these steps is rate-limiting and shows a rate enhancement
due to the presence of our ligands (it is possible that both
steps benefit from the ligands). What is certain, though,
is that the more electron-deficient the alkene, within this
series of ligands, and the more sterically encumbered the
ligand, the greater the rate of the reaction and the higher
the stability of the catalyst. (Such steric effects have recently
4.2. (2-Vinylphenyl)diphenylphosphine 2
Methyltriphenylphosphonium bromide/sodium amide
(398 mg, 1.85 equiv) was stirred in 4 mL THF for 30 min
at room temperature. 2-(Diphenylphosphino)benzaldehyde
(150 mg, 1 equiv) was dissolved in 2 mL of THF and added
dropwise to the phosphonium solution over 10 min. The
reaction mixture was allowed to stir for 4 h or until it was
judged to be complete via TLC monitoring, whereafter it
was quenched with a saturated aqueous NaHCO solution.
3
The reaction mixture was then extracted with diethyl ether
and washed with brine. The organic layer was dried over
MgSO and the excess solvent removed in vacuo. Flash
4
1
3
been observed elsewhere. ) This work provides an inter-
chromatography in 10:1 hexane/EtOAc afforded product
ꢀ
1
4
1
esting contrast to earlier work, which indicates that the
transmetallation of Pt–phosphine complexes is somewhat
inhibited by bulkier phosphines. Additionally, what is appar-
ent is that the influence of the alkene is important in the
rate-limiting step (which, we contend, is not the oxidative
addition step). This again provides an interesting contrast
to the prior work in which the electron-deficient alkene
of the dba ligand determines the concentration of the active
2 as a white solid (81%). Mp 102–103 C; H NMR
(300 MHz, CDCl ) d 7.58 (dd, 1H, J¼7.7 and 4.4 Hz),
3
7.32–7.23 (m, 12H), 7.14 (t, 1H, J¼7.2 Hz), 6.81 (dd with
unresolved fine coupling, 1H), 5.40 (dd with unresolved
fine coupling, 2H, J¼17.4 and 10.8 Hz);
1
3
C NMR
(75 MHz, CDCl ) d 142.3 (d, 1C, J¼22.3 Hz), 136.2 (d,
3
2C, J¼10.0 Hz), 135.3 (d, 1C, J¼24.0 Hz), 135.0 (d, 1C,
J¼14.0 Hz), 133.9 (d, 4C, J¼19.4 Hz), 133.1 (1C), 128.9
(1C), 128.6 (2C), 128.4 (d, 4C, J¼6.9 Hz), 127.8 (1C),
125.4 (d, 1C, J¼4.3 Hz), 115.9 (d, 1C, J¼2.3 Hz);
ꢁ1
PdL species, and therefrom the overall activity of the
2
9
,10
system.
3
1
P NMR (121 MHz, CDCl ) d ꢁ13.4; IR (CHCl )/cm
3
P
3
+
3
012, 1434, 1200, 703; EIMS m/z (%) 289 (M , 100%) no
3. Conclusions
discernable fragmentation pattern; HRMS found 288.1068,
calcd 288.1070 for C H P.
2
0 17
We have demonstrated, for the first time according to our
knowledge, that electron-deficient alkene–phosphine bi-
dentate ligands enhance the rate of the Suzuki reaction and
also improve the stability of the catalyst. Initial indications
are that this rate enhancement rests in one of the Pd(II) reac-
tion intermediates and not in the oxidative addition step,
which involves a Pd(0) species. The ligands have been found
not to participate in Heck reactions, and are isolated intact
from the reaction mixtures.
4.3. General procedure for the synthesis
of a,b-unsaturated esters
In a typical experiment, 2-(diphenylphosphino)benzalde-
hyde (0.52 mmol) and the triphenylphosphorane (1.21 equiv)
were added together in 5 mL of dry DCM and stirred at
room temperature for 12 h. Excess solvent was removed
in vacuo and the pure product was obtained via flash chroma-
tography.
4. Experimental
4.3.1. Methyl acrylate 3. Flash chromatography in 10:1
hexane/EtOAc afforded product 3 as a yellow solid (92%).
ꢀ
Mp 82–84 C; H NMR (300 MHz, CDCl ) d 8.44 (dd,
3
1
4
.1. General
1
H, J¼15.8 and 4.4 Hz), 7.61 (dd, 1H, J¼7.5 and 4.2 Hz),
Melting points were determined on a Gallenkamp digital
melting point apparatus and are uncorrected. H NMR
7.34–7.23 (m, 12H), 6.94 (dd, 1H, J¼6.7 and 4.3 Hz), 6.27
1
13
(d, 1H, J¼15.9 Hz), 3.71 (s, 3H); C NMR (75 MHz,
1
3
31
(
300 MHz), C NMR (75 MHz) and P NMR (121 MHz)
spectra were recorded on a Varian Gemini 300 MHz spec-
CDCl ) d 166.9 (1C), 143.0 (d, 1C, J¼25.7 Hz), 139.1 (d,
3
1C, J¼22.6 Hz), 137.9 (d, 1C, J¼16.1 Hz), 135.8 (d, 2C,
J¼9.8 Hz), 133.9 (d, 4C, J¼19.8 Hz), 133.7 (1C), 129.8
(1C), 129.1 (1C), 128.8 (2C), 128.5 (d, 4C, J¼7.1 Hz),
126.6 (d, 1C, J¼4.0 Hz), 119.7 (d, 1C, J¼2.6 Hz), 51.6
1
trometer in CDCl (internal references being TMS for H,
3
chloroform-d for C and 85% aqueous H PO for P spec-
4
tra). C NMR data include P–C coupling information. IR
spectra were recorded on a Tensor 27 Bruker IR spectro-
meter. Mass spectra were recorded on a Shimadzu GCMS-
1
3
31
3
1
3
3
1
(1C); P NMR (121 MHz, CDCl ) d ꢁ14.5; IR (CHCl )/
3
P
3
ꢁ1
+
cm 3016, 1709, 1229, 795; EIMS m/z (%) 347 (M ,

1628
D. B. G. Williams, M. L. Shaw / Tetrahedron 63 (2007) 1624–1629
1
3
00), 287 (MꢁC H O , 5%); HRMS found 346.1131, calcd
133.1 (1C), 129.8 (1C), 128.9 (2C), 128.8 (1C), 128.5 (d,
4C, J¼7.1 Hz), 128.0 (d, 1C, J¼4.3 Hz), 127.2 (d, 1C,
2
3 2
46.1123 for C H O P.
22 19 2
3
1
J¼2.6 Hz), 52.5 (1C), 52.2 (1C); P NMR (121 MHz,
ꢁ
1
4.3.2. tert-Butyl acrylate 4. Flash chromatography in 10:1
hexane/EtOAc afforded product 4 as a light yellow solid
CDCl ) d ꢁ13.2; IR (CHCl )/cm 3620, 2976, 2401,
3
P
3
+
1691, 1477; EIMS m/z (%) 404 (M , 1%), 345 (MꢁC H O ,
2 3 2
ꢀ
1
(
80%). Mp 91–101 C; H NMR (300 MHz, CDCl ) d
3
100%); HRMS found 404.1155, calcd 404.1178 for
C H O P.
H
8
.29 (dd, 1H, J¼15.9 and 4.5 Hz), 7.59 (dd, 1H, J¼7.4
2
4 21 4
and 4.1 Hz), 7.30–7.19 (m, 12H), 6.91 (dd, 1H, J¼7.1 and
1
3
4
.4 Hz), 6.17 (d, 1H, J¼15.9 Hz), 1.45 (s, 9H); C NMR
4.4.3. Benzylidene diester 10. Flash chromatography in
10:1 hexane/EtOAc afforded product 10 as a yellow solid
(
1
75 MHz, CDCl ) d 165.6 (1C), 141.8 (d, 1C, J¼25.3 Hz),
3
ꢀ
1
39.1 (d, 1C, J¼22.0 Hz), 137.8 (d, 1C, J¼16.0 Hz),
35.8 (d, 2C, J¼10.0 Hz), 134.0 (d, 4C, J¼20.0 Hz), 133.4
(72%). Mp 86–87 C; H NMR (300 MHz, CDCl ) d 8.25
3
1
(d, 1H, J¼2.7 Hz), 7.44 (dd with unresolved fine coupling,
(
1C), 129.5 (1C), 129.0 (1C), 128.8 (2C), 128.5 (d, 4C,
1H), 7.31–7.21 (m, 12H), 6.92 (t, 1H, J¼4.9 Hz), 5.06 (m,
13
J¼7.1 Hz), 126.4 (d, 1C, J¼4.0 Hz), 121.9 (d, 1C,
2H), 1.24 (d, 6H, J¼6.3 Hz), 1.13 (d, 6H, J¼6.2 Hz);
3
1
J¼2.3 Hz), 80.2 (1C), 28.1 (3C); P NMR (121 MHz,
C NMR (75 MHz, CDCl ) d 165.4 (1C), 163.1 (1C),
3
ꢁ
1
CDCl ) d ꢁ13.6; IR (CHCl )/cm
2981, 1710, 1321,
151; EIMS m/z (%) 389 (M , 30%), 333 (MꢁC H ,
141.1 (d, 1C, J¼22.3 Hz), 138.3 (d, 1C, J¼22.6 Hz), 138.0
(d, 1C, J¼14.6 Hz), 135.4 (d, 2C, J¼9.2 Hz), 134.0 (d, 4C,
J¼19.7 Hz), 132.8 (1C), 129.5 (1C), 128.9 (2C), 128.5
(d, 6C), 128.3 (d, 1C, J¼3.97 Hz), 68.9 (1C), 68.7 (1C),
3
3
+
1
1
4
9
0%), 289 (MꢁC H O , 100%); HRMS found 388.1590,
5
9 2
calcd 388.1592 for C H O P.
25 25 2
3
1
2
1.6 (2C), 21.4 (2C);
P NMR (121 MHz, CDCl )
3
ꢁ1
4.4. General procedure for the synthesis of Knoevenagel
phosphine ligands
dP ꢁ12.5; IR (CHCl )/cm
3622, 2980, 2402, 1724,
1523; EIMS m/z (%) 460 (M , 1%), 417 (MꢁC H , 40%),
3
+
3
7
3
HRMS found 460.1802, calcd 460.1804 for C H O P.
73 (MꢁC H O , 100%), 331 (MꢁC H O , 98%);
28 29 4
4 7 2 7 13 2
In a typical experiment dialkyl malonate (1 equiv), 2-(di-
phenylphosphino)benzaldehyde (1.1 equiv) and 0.01 mL of
piperidine (0.1 equiv) were added to a flamed out Dean Stark
4.4.4. Benzylidene diester 11. Flash chromatography in
10:1 hexane/EtOAc afforded product 11 as a light orange
apparatus with 2 mL of toluene and heated in an oil bath
ꢀ
ꢀ
1
(
130–150 C) for 6–8 h until all of the water had been
solid (47%). Mp 91–93 C; H NMR (300 MHz, CDCl )
3
azeotropically removed. After cooling, 5 mL of toluene
was added and the reaction mixture was washed with water
d 8.05 (d, 1H, J¼3.0 Hz), 7.51 (dd, 1H, J¼6.1 and
4.3 Hz), 7.32–7.21 (m, 12H), 6.91 (t, 1H, J¼6.0 Hz), 1.44
1
3
(
2ꢂ5 mL), 1 N HCl (2ꢂ5 mL) and saturated aqueous
(s, 9H), 1.39 (s, 9H); C NMR (75 MHz, CDCl ) d 165.3
3
NaHCO (2ꢂ5 mL). The combined aqueous layers were
(1C), 163.0 (1C), 139.3 (d, 1C, J¼22.3 Hz), 138.3 (d, 1C,
J¼22.3 Hz), 137.8 (d, 1C, J¼14.9 Hz), 135.6 (d, 2C,
J¼9.7 Hz), 134.0 (d, 4C, J¼19.8 Hz), 132.9 (1C), 130.4
(d, 1C, J¼2.3 Hz), 129.4 (1C), 128.9 (2C), 128.6 (1C),
3
back-extracted once with 10 mL of toluene. The organic
layers were combined, dried with Na SO , filtered, evapo-
rated and purified via flash chromatography.
2
4
1
28.5 (d, 2C, J¼7.1 Hz), 128.4 (1C), 81.8 (1C), 81.6 (1C),
3
1
4.4.1. Benzylidene diester 5. Flash chromatography in 10:1
hexane/EtOAc afforded product 5 as a light yellow solid
28.0 (3C), 27.8 (3C);
P NMR (121 MHz, CDCl )
3
ꢁ1
d ꢁ12.1; IR (CHCl )/cm 3623, 2979, 2402, 1723, 1479;
3
ꢀ
1
+
(
66%). Mp 89–91 C; H NMR (300 MHz, CDCl ) d 8.34
3
EIMS m/z (%) 488 (M , 1%), 431 (MꢁC H , 15%), 387
4 9
(
1
4
(
(
d, 1H, J¼2.7 Hz), 7.39 (dd with unresolved fine coupling,
(MꢁC H O , 30%), 331 (MꢁC H O , 100%); HRMS
5
9
2
9 17 2
H), 7.32–7.23 (m, 12H), 6.94 (dt, 1H, J¼5.4 and 1.5 Hz),
.21 (q, 2H, J¼7.2 Hz), 4.12 (q, 2H, J¼7.1 Hz), 1.26
found 488.2113, calcd 488.2117 for C H O P.
30 33 4
1
3
t, 3H, J¼6.9 Hz), 1.09 (t, 3H, J¼7.1 Hz); C NMR
4.5. General procedure for the Suzuki reaction
75 MHz, CDCl ) d 165.6 (1C), 163.4 (1C), 141.9 (d, 1C,
3
J¼22.6 Hz), 138.2 (d, 1C, J¼23.1 Hz), 137.8 (d, 1C, J¼
In a typical experiment, the phosphine ligand, e.g. TPP
(0.114 mmol)was dissolved in 5 mL of solvent and Pd(OAc)2
(0.057 mmol) was added. The reaction mixture was allowed
to stir for 20 min. Potassium phosphate (1.71 mmol),
bromobenzonitrile (1.14 mmol) and phenylboronic acid
(1.14 mmol) were added and the reaction mixture was stirred
at the relevant temperature for the relevant time.
1
4.9 Hz), 135.2 (d, 2C, J¼9.4 Hz), 133.8 (d, 4C, J¼
1
9.8 Hz), 132.8 (1C), 129.5 (1C), 128.8 (2C), 128.5
(
1
1
1C), 128.3 (d, 4C, J¼7.1 Hz), 128.0 (d, 1C, J¼4.3 Hz),
27.8 (d, 1C, J¼2.6 Hz), 61.2 (1C), 61.0 (1C), 13.9 (1C),
3 P
3
1
3.6 (1C); P NMR (121 MHz, CDCl ) d ꢁ13.0; IR
ꢁ
1
(
CHCl )/cm
3
2985, 1727, 1255, 1070; EIMS m/z (%)
+
4
4
33 (M , 100%), 387 (MꢁC H O, 40%), 359 (MꢁC H O ,
2
5
3 5 2
1
5
5%); HRMS found 432.1485, calcd 432.1491 for
4.5.1. 4-Cyanobiphenyl 8. Flash chromatography in 10:1
hexane/EtOAc afforded product 8. H NMR (300 MHz,
1
C H O P.
26 25 4
1
3
CDCl ) d 7.72–7.64 (m, 4H), 7.59–7.39 (m, 5H);
C
NMR (75 MHz, CDCl ) d 145.5 (1C), 139.0 (1C), 132.5
3
4.4.2. Benzylidene diester 9. Flash chromatography in 10:1
hexane/EtOAc afforded product 9 as a yellow crystalline
3
(2C), 129.0 (2C), 128.5 (1C), 127.6 (2C), 127.1 (2C),
118.8 (1C), 110.8 (1C).
ꢀ
1
solid (56%). Mp 109–111 C; H NMR (300 MHz, CDCl )
3
d 8.39 (d, 1H, J¼3.3 Hz), 7.39–7.25 (m, 13H), 6.95 (t, 1H,
1
3
J¼5.7 Hz), 3.76 (s, 3H), 3.63 (s, 3H); C NMR (75 MHz,
Acknowledgements
CDCl ) d 166.3 (1C), 164.0 (1C), 142.9 (d, 1C, J¼
3
2
1
2.6 Hz), 138.4 (d, 1C, J¼23.5 Hz), 138.1 (d, 1C, J¼
4.6 Hz), 135.3 (d, 2C, J¼9.1 Hz), 133.9 (d, 4C, J¼19.8 Hz),
We thank Sasol, the NRF, THRIP and the University of
Johannesburg for financial support of this project.

D. B. G. Williams, M. L. Shaw / Tetrahedron 63 (2007) 1624–1629
1629
Supplementary data
C. M.; Landis, C. R.; Stahl, S. S. J. Am. Chem. Soc. 2004,
26, 14832.
1
Supplementary data associated with this article can be found
in the online version, at doi:10.1016/j.tet.2006.12.012.
7. (a) Mason, M. R.; Verkade, J. G. Organometallics 1992, 11,
2212; (b) Rybin, L. V.; Petrovskaya, E. A.; Rubinskaya,
M. I.; Kuz’mina, L. G.; Struchkov, Y.; Kaverin, V. V.;
Koneva, N. J. Organomet. Chem. 1985, 288, 119; (c) de
Pater, J.; Tromp, D. S.; Tooke, D. M.; Spek, A. L.; Deelman,
B.; Koten, G.; Elsevier, C. J. Organometallics 2005, 24, 6411.
8. (a) Scrivanti, A.; Beghetto, V.; Matteoli, U.; Antonaroli, S.;
Marini, A.; Crociani, B. Tetrahedron 2005, 61, 9752; (b)
Scrivanti, A.; Beghetto, V.; Matteoli, U.; Antonaroli, S.;
Marini, A.; Mandoj, F.; Paolesse, R.; Crociani, B. Tetrahedron
Lett. 2004, 45, 5861.
9. (a) Amatore, C.; Jutand, A.; Khalil, F.; M’Barki, M. A.;
Mottier, L. Organometallics 1993, 12, 3168; (b) Amatore, C.;
Broeker, G.; Jutand, A.; Khalil, F. J. Am. Chem. Soc. 1997,
119, 5176; (c) Amatore, C.; Carr ꢀe , E.; Jutand, A.; Medjour,
Y. Organometallics 2002, 21, 4540.
10. (a) Fairlamb, I. J. S.; Kapdi, A. R.; Lee, A. F. Org. Lett. 2004,
24, 4435; (b) Mac ꢀe , Y.; Kapdi, A. R.; Fairlamb, J. S.; Jutand, A.
Organometallics 2006, 25, 1795.
11. Fleming, I. Frontier Orbitals and Organic Chemical Reactions;
Wiley: Chichester, UK, 1976.
12. See Ref. 8 and references cited therein.
13. Clarke, M. L.; Heydt, M. Organometallics 2005, 24, 6475.
14. Hadei, N.; Kantchev, E. A.; O’Brien, C.; Organ, M. Org. Lett.
2005, 7, 199110.
15. Schulman, E. M.; Christensen, K. A.; Grandt, D. M.; Walling,
C. J. J. Org. Chem. 1974, 39, 2686.
References and notes
1
. Alonso, F.; Beletskaya, I. P.; Yus, M. Tetrahedron 2005, 61,
1771.
1
2
. (a) Huang, X.; Anderson, K. W.; Zim, D.; Jiang, L.; Klapars,
A.; Buchwald, S. L. J. Am. Chem. Soc. 2003, 125, 6653; (b)
Hills, I. D.; Netherton, M. R.; Fu, G. C. Angew. Chem., Int.
Ed. 2003, 42, 5749.
3
. Overett, M. J.; Blann, K.; Bollmann, A.; Dixon, J.; Haasbroek,
D.; Killian, E.; Maumela, H.; McGuinness, D. S.; Morgan, D.
J. Am. Chem. Soc. 2005, 127, 10723.
4
5
6
. Kamer, P. C. J.; Van Leeuwen, P. W. N.; Reek, J. N. H. Acc.
Chem. Res. 2001, 34, 895.
. Battace, A.; Zair, T.; Doucet, H.; Santelli, M. Tetrahedron Lett.
2
006, 47, 459.
. (a) Negishi, E.; Takahashi, T.; Baba, S.; Van Horn, D.;
Okukado, N. J. Am. Chem. Soc. 1987, 109, 2393; (b) Ozawa,
F.; Ito, T.; Nakamura, Y.; Yamamoto, A. J. Organomet.
Chem. 1979, 168, 375; (c) McGuinness, D. S.; Cavell, K. J.
Organometallics 1999, 18, 1596; (d) Fitton, F. P.; McKeon,
J. E. J. Chem. Soc., Chem. Commun. 1968, 4; (e) Sprengers,
J. W.; Agerbeek, M. J.; Elsevier, C. J. Organometallics
2
004, 23, 3117; (f) Popp, B. V.; Christine, T.; Morales,