Palladium-catalysed Grignard cross-coupling using highly concentrated Grignards in methyl-tetrahydrofuran

Article abstract of DOI:10.1039/b921026a

The cross-coupling of Grignard reagents at 5 molar concentration in methyl-tetrahydrofuran with no added reaction solvents can be achieved with the appropriate matching of catalyst to substrate, significantly reducing solvent use when compared to a typical protocol in tetrahydrofuran. Me-THF was also used in an improved microwave accelerated synthesis of the [Pd(L)Cl₂] pre-catalysts from sodium tetrachloropalladate in very high yield.

Full text of DOI:10.1039/b921026a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/greenchem | Green Chemistry
Palladium-catalysed Grignard cross-coupling using highly concentrated
Grignards in methyl-tetrahydrofuran†
Edward J. Milton and Matthew L. Clarke*
Received 13th October 2009, Accepted 25th November 2009
First published as an Advance Article on the web 12th January 2010
DOI: 10.1039/b921026a
The cross-coupling of Grignard reagents at 5 molar concen-
tration in methyl-tetrahydrofuran with no added reaction
solvents can be achieved with the appropriate matching
of catalyst to substrate, significantly reducing solvent use
when compared to a typical protocol in tetrahydrofuran.
Me-THF was also used in an improved microwave acceler-
ated synthesis of the [Pd(L)Cl2] pre-catalysts from sodium
tetrachloropalladate in very high yield.
stage and energy-intensive clean-up of aqueous waste streams
can be reduced. Me-THF is more expensive than THF, in part
due to not being produced in such large quantities, but we felt
that if solvent usage during cross-coupling could be reduced
significantly then the protocols would have very significant
environmental and economic advantages relative to traditional
procedures. Grignard synthesis or cross-coupling reactions in
Me-THF are not a new invention4 since this solvent has been
slowly gaining popularity in industrial research, but to the best
of our knowledge, no report has evaluated a range of commonly
employed Pd catalysts in Grignard cross-coupling under the
rather challenging conditions that enable reduced solvent usage.
In this communication, we report that Grignard cross-coupling
is possible at 5 molar concentration with no added reaction
solvents and using reasonably low catalyst loadings.
We predominantly focussed on ferrocene-type ligands, since
dppf has a strong track record in Kumada and Negishi cross-
coupling.5 More recently, 1,1¢-ditert-butylphosphinoferrocene
(dtbpf) has been found to be a very good ligand in difficult
Pd-catalysed Suzuki reactions,5b giving superior activity relative
to dppf and 1,1¢-bis-diisopropylphosphino-ferrocene (dippf).
Johnson Matthey have developed this chemistry to a synthet-
ically useful level and sell [PdCl2(L)] complexes of the above lig-
ands. 1-Ditert-butyl-1¢diphenylphosphino-ferrocene (dtbdppf)
has barely been studied in cross-coupling or organometallic
chemistry, something that we also wished to rectify. An improved
synthesis of [Pd(L)Cl2] complexes was first devised (Scheme 1)
that benefits from simplicity and reduced solvent usage, elim-
inates the need to convert simple Pd salts to [PdCl2(PhCN)2]
using excess PhCN, and the subsequent ligand displacement
encountered in the most common method for synthesising
[PdCl2(P∧P)] complexes. Equimolar amounts of Na2[PdCl4] and
diphosphine are simply treated with Me-THF–MeCN solvent
(Me-THF–MeCN = 30 : 1) and heated using microwave heating
at 140 ◦C for 10 min. The desired complexes precipitate on
cooling and are filtered and then washed with water. Yield and
purity are very high.
Palladium-catalysed cross-couplings are now very widely used
in both academic and industrial synthesis. There are many
studies where important breakthroughs in the cross-coupling
of low-reactivity electrophilic partners such as aryl chlorides
have become possible after optimisation of ligand structure.1
On the other hand, our interest has been the investigation of
problematic nucleophiles in cross-coupling catalysis,2 and as part
of this effort, we turned our attention to the Grignard cross-
coupling (Kumada coupling).3 There are certainly examples
where the use of Grignard reagents, one of the most economical
nucleophiles, creates difficulties. An obvious potential problem
is reduced functional group tolerance, but there are also reports
that suggest problems with low yields and increased homo-
coupling for these nucleophiles on a given substrate relative to
organo-zinc or organoborane nucleophiles.
Grignard cross-couplings typically use solutions of Grignard
reagents in the increasingly environmentally unacceptable sol-
vent, tetrahydrofuran (THF) at concentrations of between 1
and 2.5 molar, often with an added reaction solvent. Most
Grignard reagents are not soluble at higher concentrations.
At the end of a typical cross-coupling procedure, additional
organic solvents (generally diethyl ether) have to be added prior
to quenching with water, since THF is miscible with water.
The products are then isolated from the Mg salts by washing
with organic solvent, and then generally further purified. There
has been increasing interest in the THF surrogate, methyl-
tetrahydrofuran (Me-THF) as it can be produced from waste
biomass, and has some desirable features relative to THF.4 Me-
THF has been reported to be more stable to acids, and to be a
better solvent for Grignard reagents allowing more concentrated
solutions to be used. In addition, Me-THF is not miscible with
water which suggests less solvent can be used in the work-up
School of Chemistry, University of St Andrews, EaStCHEM, St
Andrews, Fife, UK. E-mail: mc28@st-andrews.ac.uk; Fax: (44) 1334
463808; Tel: (44) 1334 463850
† Electronic supplementary information (ESI) available: Experimental
procedures. See DOI: 10.1039/b921026a
Scheme 1 Microwave accelerated synthesis of [PdCl2(L)] pre-catalysts
direct from Na2PdCl4 in Me-THF.
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 381–383 | 381
©

View Article Online
Table 1 Grignard cross-coupling at high concentration: ligand effectsa
Entry
Aryl halide
Grignard reagent
L in [PdCl2/Ln]
T/◦C
Time/h
Starting material (%)
Product (%)
1
2
3
4
5
6
7
8
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
4-bromo-anisole
4-bromo-anisole
4-bromo-anisole
4-bromo-anisole
4-bromo-anisole
4-(Me2N)C6H4Br
4-(Me2N)C6H4Br
4-(Me2N)C6H4Br
4-(Me2N)C6H4Br
4-(Me2N)C6H4Br
4-(CF3)C6H4Cl
4-(CF3)C6H4Cl
4-(CF3)C6H4Cl
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
PhMgBr
dppf
50
50
50
50
50
50
50
50
50
50
50
50
55
55
55
55
55
75
75
75
75
50
50
50
50
50
50
50
50
2
2
2
2
2
2
2
2
2
2
2
2
6
6
6
6
6
6
6
6
6
2
2
2
2
2
2
2
2
19
60
61
28
21
19
1
78
33
76
54
40
39
51
59
37
77
15
74
25
7
76
20
30
56
71
75
96
22
67
24
46
60
61
49
41
63
23
82
13
93
75
21
25
56
21
4
dippf
dtbpf
dtbdppf
dppe
dppb
dppp
dppp
dippf
dtbpf
dppf
dtbppf
dppp
dippf
dtbpf
dppf
dtbppf
dippf
dtbpf
dppf
dtbppf
2 ¥ PPh3
dppe
dppp
dppb
dippf
dtbpf
dppf
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
PhMgBr
PhMgBr
PhMgBr
PhMgBr
4-(CF3)C6H4Cl
PhMgBr
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
2-bromo-4-fluoro-anisole
4-(CF3)C6H4MgBr
4-(CF3)C6H4MgBr
4-(CF3)C6H4MgBr
4-(CF3)C6H4MgBr
4-(CF3)C6H4MgBr
4-(CF3)C6H4MgBr
4-(CF3)C6H4MgBr
4-(CF3)C6H4MgBr
79
75
44
79
96
98
34
33
2
66
67
dtbppf
a Reactions conducted at 1 mmol scale using 1.2 mmol of 5M Grignard reagent in 0.24 mL Me-THF for time given. Conversions were determined
by 19F NMR spectroscopy, and checked using H NMR spectroscopy against an internal standard with product identity and side products further
examined by GCMS (entries 8–17 by H NMR using methyl naphthalene as internal standard and GCMS only). In the majority of the reactions,
GCMS indicates that homo-coupled aryl halide is only present in trace amounts, but exceptions are entry 2 (20% homocoupled), entry 3 (14%
homocoupled), entry 4 (16% h) and entries 5 and 6 (6%H).
With these and several other common pre-catalysts in hand,
we turned our attention to the synthesis of the Grignards and
evaluating the catalysts in cross-coupling.‡ Most of our studies
were carried out with PhMgBr. This can be made as a rather
viscous solution at concentrations as high as 5 M. (SAFETY
NOTE: We have not experienced any difficulties making small
scale solutions (~30 mmol) of highly concentrated Grignard
reagents at the temperatures described, but it is essential that
great care is taken in synthesising new Grignard reagents at
this concentration, in particular monitoring reaction exotherms.
While this is the case for any Grignard reagent, it can be
envisaged that extra caution is needed at higher concentration.
Scale up of these reactions, in common with all Grignard
reactions, should first be evaluated using calorimetry.) In order
to facilitate transfer of the Grignard reagents, they are heated
to 50–75 ◦C to give less viscous solutions and transferred to the
reaction vessel. Alternatively, a very concentrated solution of
catalyst and substrate can be added to ArMgBr.
The catalyst screening studies have been carried out using no
extra reaction solvents other than the minimal solvent solvating
the Grignard reagent (around 800 mg (~4 mmol) of aryl halide
per mL solvent), at low catalyst loadings (0.2 mol%) and at
moderate temperatures (50–75 ◦C) since we envisaged these sorts
of conditions would be needed in a real application. Table 1
and Scheme 2 show these results. In the case of 2-bromo-4-
fluoro-anisole, palladium catalysts derived from dppp stand out
Scheme 2 Protocol for Grignard cross-coupling reactions in Me-THF.
as being able to operate well under this challenging regime. It is
interesting to note that dtbpf, that has been shown to give superb
Pd catalysts for the Suzuki coupling of boronic acids with both
aryl bromides and deactivated chlorides under conditions where
dppf fails, is a poor catalyst in these reactions with significant
amounts of starting material and homo-coupling product. Given
that the Suzuki results in reference 5b infer high activity in
oxidative addition, and that this ligand has been shown to
promote reductive elimination over a hundred times faster than
dppf complexes,6 this suggests inefficient transmetalation, a
reaction known to be incredibly sensitive to steric effects.7 We
have not optimised these reactions, but the dppp/Pd catalysed
reaction gives essentially quantitative yields of a crude material
(~95% purity) after work up in which only a further 0.6 ml
MeTHF added to assist work-up. While this material that will
be contaminated with traces of Pd might be sufficiently pure
for further application and downstream purification, it can be
further purified by column chromatography (86% isolated yield)
in a standard purification procedure. A procedure that is greener
382 | Green Chem., 2010, 12, 381–383
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
still would use a recyclable catalyst that could be separated prior
to the work up; we are optimistic that such a catalyst could be
developed given the considerable advances that have been made
in catalyst recycling. In any case, this procedure used < 1 ml
of Me-THF per mmol of Grignard reagent for the Grignard
synthesis, catalytic reaction and aqueous extraction of magne-
sium salts. Hardly any reports in the literature give amounts of
solvent used, but one of best Pd-catalysed procedures reported
that does give full details uses over 30 mL of THF and Et2O
per mmol reaction.3b Although this amount could probably be
reduced, we would argue that this shows that the saving in solvent
could be considerable and make this approach practical. A few
other aryl halides were investigated, and one aspect of this study
worth highlighting is that the ligand must be matched to the
aryl halide, and in this case, this is not easily predicted based
on the rather limited range of reactions we have examined thus
far; dppf and the previously unstudied ligand dtbdppf appear
to be the best catalysts in general. Significant optimisation of
Grignard synthesis and cross-coupling would need to be carried
out as this type of protocol is transferred to other Grignard
reagents or scaled up, but to check that this type of approach
would work on another Grignard reagent, p-CF3C6H5MgBr was
synthesised and coupled with 2-bromo-4-fluoro-anisole at 5M
concentration. This Grignard did not store well, needing to be
prepared the same day as the coupling reaction in order to obtain
reproducible results, but none-the-less good conversions can be
obtained in just 2 h.
In summary, we have used a convenient microwave-assisted
synthesis of some typical Pd catalysts used for cross-coupling,
and investigated the forgotten ferrocene ligand, 1-ditert-butyl-
1¢diphenylphosphino-ferrocene, and studied how ligand struc-
ture alters reactivity in the Grignard cross-coupling under chal-
lenging conditions in Me-THF, a new solvent that is beginning to
attract attention as a greener replacement to THF if the process
economics can overcome its higher cost.4c These studies show
a significant saving in solvent use in both Grignard synthesis
(expected from the literature), but also the Grignard cross-
coupling and the work-up; good conversions were obtained
in short times at S/C of 500 in the few examples we have
examined thus far. The reaction is strongly dependent on ligand
structure, but 1-ditert-butyl-1¢diphenylphosphino-ferrocene and
1,1¢-bis-diphenylphosphino-ferrocene seem to be the preferred
ligands at present. A more thorough investigation into the full
scope of this reaction will be carried out and reported in full in
due course. It is hoped that this study suggests that Grignard
cross-coupling in Me-THF might be ecologically and econom-
ically attractive and worth developing further in industrial
synthesis.
turnings (0.789 g, 1.02 eq., 32.5 mmol) via an addition funnel. NO
INITIATOR NEEDED. CAUTION, EXOTHERM! (Heat generated
sufficient to maintain gentle reflux.) After addition, the reaction mixture
was warmed to 90 ◦C for 4 h (until the Mg was consumed) to give a bright
orange solution. To a solution of the Grignard (PhMgBr) in Me-THF
(0.24 mL, 1.2 eq., 1.2 mmol), 2-bromo-4-fluoroanisole (0.205 g, 0.13 mL,
1 mmol) was added followed by the catalyst [Pd(DPPP)Cl2] (0.2 mol%,
1.2 mg) and the reaction mixture heated to 50 ◦C for 2 h. 1H NMR
spectroscopy and GCMS showed the desired product. Excess Grignard
reagent was quenched with solid NH4Cl, organics filtered through a
plug of MgSO4 (washed with 0.6 mL meTHF) and concentrated in
vacuo to give a yellow oil in essentially quantitative yield. GCMS:
Retention time, MS (EI) m/z (%): 18.706 mins (96%), [M+] = 202.1;
C14H10F4O requires 202.08. Although this material is of high purity,
column chromatography (petroleum ether 40–60) yielded the product as
a colourless liquid (0.173 g, 0.86 mmol, 86%). This compound has been
reported previously.5b 1H NMR (400 MHz, CDCl3) d = 3.82 (3H, s,
OCH3), 6.81 (1H, dd, J = 4.62, 8.8 Hz, ArCH), 6.87–6.94 (1H, m,
ArCH), 6.96 (1H, dd, J = 3.2, 9.2 Hz, ArCH), 7.25 (1H, tt, J = 7.54,
1.4 Hz, ArCH), 7.34 (2H, t, J = 6.96 Hz, ArCH), 7.4–7.45 (2H, m,
1
ArCH); 19F{ H} NMR (376.5 MHz, CDCl3) d = -124.51; 13C NMR
(100 MHz, CDCl3) d = 56.2, 112.3, 112.4, 114.2, 114.4, 117.3, 117.6,
127.4, 128.2, 129.4, 132.0, 132.1, 137.5, 152.7, 156.7 (d, J = 236.4 Hz);
MS (EI) m/z (%): 202.1 (100.0) [M+]. Similar protocols were used for the
other screening experiments, with product identities, and conversions
determined by 1H NMR spectroscopy against an internal standard
and GCMS analysis. The data obtained matched literature values and
samples previously prepared in our laboratory by other methods.
1 (a) A. Zapf and M. Beller, Top. Catal., 2002, 19, 101; (b) C. E.
Tucker and J. G. DeVries, Top. Catal., 2002, 19, 111; (c) C. Barnard,
Platinum Met. Rev., 2008, 52, 38; (d) M. L. Clarke and J. J. R. Frew,
Specialist Periodical Reports: Organometallic Chemistry, Royal Society
of Chemistry, Ed. I. J. S. Fairlamb and J. Lynam, 2009, 35, 19.
2 (a) M. L. Clarke, Adv. Synth. Catal., 2005, 347, 303; (b) J. A. Fuentes,
M. E. France, G. J. Roff, E. J. Milton and M. L. Clarke, Beilstein J. Org.
Chem., 2007, 3, art.18; (c) E. J. Milton, J. A. Fuentes and M. L. Clarke,
Org. Biomol. Chem., 2009, 7, 2645; (d) K. Damian, M. L. Clarke and
C. J. Cobley, Appl. Organomet. Chem., 2009, 23, 272.
3 (a) K. Tamao, K. Sumitani and M. Kumada, J. Am. Chem. Soc., 1972,
94, 4374; (b) G. Y. Li, J. Organomet. Chem., 2002, 653, 63; (c) A.
Alimardanov, L. Schmieder-van de Vondervoort, A. H. M. de Vries
and J. G. de Vries, Adv. Synth. Catal., 2004, 346, 1812; (d) J. A. Miller
and R. P. Farrell, Tetrahedron Lett., 1998, 39, 7275; (e) R. Martin
and S. L. Buchwald, J. Am. Chem. Soc., 2007, 129, 3844; (f) C. E.
Hartmann, S. P. Nolan and C. J. Cazin, Organometallics, 2009, 28,
2915; (g) Iron catalysed Grignard cross-coupling has developed further
recently, but further studies are needed to lower catalysts loadings
and determine mechanism (no reports use concentrated Me-THF
solutions). For some examples, see: T. Hatekeyama and M. Nakamura,
J. Am. Chem. Soc., 2007, 129, 9844; R. B. Bedford, M. Betham, D. W.
Bruce, S. A. Davis, R. M. Frost and M. Hird, Chem. Commun., 2006,
1398; R. B. Bedford, D. W. Bruce, R. M. Frost and M. Hird, Chem.
Commun., 2005, 4161.
4 (a) D. F. Aycock, Org. Process Res. Dev., 2007, 11, 156; (b) C.
Werner, F. Platz, A. Kanschlik-Conradsen, US Patent (to Honeywell
International) US 7205,414 (2007); (c) Our interest to follow up
our initial finding (that dtbdppf was a better ligand than dtbpf for
this reaction) was crystallised by an interesting debate, where the
issue of overcoming the extra cost of this green solvent was high-
lighted. See: http://gaussling.wordpress.com/2006/10/18/fabulous-
2-methyltetrahydrofuran/.
5 (a) T. Hayashi, M. Konishi, Y. Kobori, M. Kumada, T. Higuchi and
K. Kirotsu, J. Am. Chem. Soc., 1984, 106, 158; (b) T. J. Colacot and
H. A. Shea, Org. Lett., 2004, 6, 3731; (c) A. L. Boyles, I. R. Butler
and S. C. Quayle, Tetrahedron Lett., 1998, 39, 7766; I. R. Butler, W. R.
Cullen, T. J. Kim, S. J. Rettig and J. Trotter, Organometallics, 1985, 4,
972.
We acknowledge the EPSRC for funding, Johnson-Matthey
for the loan of some of the Pd salts, and IMCD for a donation
of Me-THF.
Notes and references
6 G. Mann, Q. Shelby, A. H. Roy and J. F. Hartwig, Organometallics,
‡ Example experimental procedure: Bromobenzene (5 g, 3.36 mL,
32 mmol) in Me-THF (6.4 mL) was added DROPWISE to magnesium
2003, 22, 2775.
7 M. L. Clarke and M. Heydt, Organometallics, 2005, 24, 6475.
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 381–383 | 383
©