The design of second generation MOP-phosphonites: Efficient chiral hydrosilylation of functionalised styrenes

Article abstract of DOI:10.1039/c5dt04475h

A series of enantiopure MOP-phosphonite ligands, with tailored steric profiles, have been synthesised and are proven to be very successful in high-yielding, regio- and enantioselective catalytic hydrosilylation reactions of substituted styrenes, affording important chiral secondary alcohols.

Full text of DOI:10.1039/c5dt04475h

Dalton
Transactions
View Article Online
View Journal
COMMUNICATION
The design of second generation
MOP-phosphonites: efficient chiral
Cite this: DOI: 10.1039/c5dt04475h
hydrosilylation of functionalised styrenes†
Received 4th August 2015,
Accepted 13th November 2015
James T. Fleming, Arne Ficks, Paul G. Waddell, Ross W. Harrington and
Lee J. Higham*
DOI: 10.1039/c5dt04475h
www.rsc.org/dalton
A series of enantiopure MOP-phosphonite ligands, with tailored parent (S)-H-MOP and (R)-MeO–MOP ligands, and we wished
steric profiles, have been synthesised and are proven to be very to examine this effect for phosphonite derivatives. Importantly,
successful in high-yielding, regio- and enantioselective catalytic we also wanted to examine the effect of incorporating a second
hydrosilylation reactions of substituted styrenes, affording impor- chiral moiety, which we achieved by appending either (R)- or
tant chiral secondary alcohols.
(S)-binol to isolate the diastereomeric pairs (S,Rb)-Ia/(S,Sb)-Ia
and (R,Rb)-Ib/(R,Sb)-Ib. By investigating the coordination chem-
istry and catalytic performance of the four ligand pairs, we
were able to demonstrate that subtle differences in the posi-
tion of the palladium atom relative to the lower naphthyl ring
in these complexes appears to have important consequences
for chiral induction in the catalytic hydrosilylation of styrene.4
These results led us to question whether the large chiral binol
group was in fact necessary for effective asymmetric catalysis.
In this study we sought to investigate the effect of replacing
the binol moiety of Ia and Ib with achiral ancillary aryloxides,
in order to establish if the enantioselectivities remain high by
virtue of the phosphorus-bonded binaphthyl backbone, and
thus eliminate the requirement for a second chiral centre. To
the best of our knowledge this has not been demonstrated for
MOP phosphonites, which are likely to effectively catalyse
alternate substrates than their phosphine cousins, by virtue of
their differing electronic properties. Herein we describe the
synthesis of the phenoxy derivatives (S)-3a and (R)-3b, which
allowed us to probe the effect of removing the chelating
element of the added aryloxide, before describing how the sen-
sitivity of these ligands to hydrolysis demanded the design of
next generation biphenoxy-based MOP-phosphonites (S)-4a
and (R)-4b, which were further improved upon to yield the
user-friendly, chiral phosphonites (S)-5a and (R)-5b, that
perform better than the corresponding MOP phosphines in
the hydrosilylation of functionalised styrenes (Chart 2 and
Table 2).
Chiral mono- and bidentate phosphorus ligands have achieved
considerable acclaim in a variety of transition metal-catalysed
asymmetric transformations.1 Atropisomeric monophosphines
with a binaphthyl backbone2 have proven to be particularly
successful in chiral hydrosilylations of alkenes.3
We recently reported a new class of MOP-phosphonite
ligands (Ia and Ib), based on a MOP/XuPhos-type hybrid
(Chart 1).4 Chiral phosphonites are ligands with good π-accep-
tor properties and are effective in a range of asymmetric hydro-
cyanations,5 hydroformylations6 and hydrogenations.7 Ligands
Ia and Ib possess a phosphorus donor bound directly to a
binaphthyl backbone – the a series has a hydrogen in the 2′
position of the lower naphthyl ring whilst a methoxy group
occupies the site in the b series (Chart 1). This difference has
profound implications for the catalytic performance of the
Chart 1 A selection of binaphthyl-containing monophosphorus-donor
ligands.
The six new MOP-phosphonite ligands were synthesised fol-
lowing a procedure we have developed, starting with the air-
stable, chiral primary phosphine precursors (S)-1a and (R)-1b.8
In a two-step, one-pot reaction, the primary phosphine was
treated with phosphorus pentachloride, quantitatively yielding
the dichlorophosphines (S)-2a and (R)-2b (Scheme 1).4 Our
initial targets (S)-3a and (R)-3b were synthesised by employing
School of Chemistry, Bedson Building, Newcastle University, Newcastle upon Tyne,
NE1 7RU, UK. E-mail: lee.higham@ncl.ac.uk
†Electronic supplementary information (ESI) available: Full experimental and
crystallographic data. CCDC 1416469 (R)-3b, 1416470 (R)-(4b), 1416471 (S)-(6a)
and 1416472 (S)-(7a). For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c5dt04475h
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Communication
Dalton Transactions
Chart 2 Phosphonite ligands (S)-3a, (R)-3b, (S)-4a, (R)-4b, (S)-5a and
(R)-5b.
Fig. 1 Molecular structure of (R)-3b. Hydrogen atoms have been
omitted for clarity. Selected average bond distances (Å) and angles (°):
P2–C19 1.830(2), P2–O4 1.6461(17), P2–O5 1.6619(18), O4–C29
1.402(3), O5–C29’ 1.391(3); O4–P2–O5 99.13(9), O4–P2–C19 93.63(9),
O5–P2–C19 97.85(10), C19–C18–C18’–C19’ –86.5(3).
Scheme 1 The synthesis of the phosphonite ligands from (S)-1a and
(R)-1b.
two equivalents of phenol, however both syntheses proved to
be temperamental as a result of their sensitivity to hydrolysis,
with (S)-3a particularly low-yielding. Therefore we sought to
replace the two phenol groups with the achiral, chelating 2,2′-
biphenol, in order to ascertain whether we could inhibit this
decomposition. This approach led to the successful prepa-
ration of the enantiopure compounds (S)-4a and (R)-4b; the
ligands can be stored under nitrogen for a period of weeks
with only minimal decomposition. However, during our initial
investigations into the coordination chemistry of these
ligands, we noticed slow decomposition in solution and their
performance in catalytic hydrosilylation required improvement
(vide infra). Thus, in order to push the design concept still
further and optimise both ligand stability and catalytic capa-
Fig. 2 Molecular structure of (R)-4b. Hydrogen atoms have been
omitted for clarity. Selected average bond distances (Å) and angles (°):
P1–C2 1.812(7), P1–O1 1.658(4), P1–O2 1.657(4), O1–C12 1.411(7),
O2–C12’ 1.396(7), C11–C11’ 1.482(9); O1–P1–O2 99.5(2), O1–P1–C2
96.6(2), O2–P1–C2 100.3(3), C2–C1–C1’–C2’ –86.5(7), C12–C11–C11’–
C12’ –48.6(8).
bility, we synthesised the bulky ortho-methyl substituted start- lower HOMO (Highest Occupied Molecular Orbital) and LUMO
ing material 2,2′-dihydroxyl-3,3′-dimethyl-1,1′-biphenyl (ESI†), (Lowest Unoccupied Molecular Orbital) energies than their
and used it to prepare the ligands (S)-5a and (R)-5b which aryl counterparts, and have a lower phosphorus contribution
possess a more imposing steric profile. Both phosphonites to the HOMO, implying the phosphonites are weaker P-ligand
were isolated in high yield, and are purified by passing σ-donors. To study the coordination chemistry of our ligands
through a plug of alumina in air. They can be stored under with Pd(II), we reacted them with [Pd(η3-C4H7)Cl]2 in a 1 : 1,
nitrogen without decomposition over several months, and P : Pd ratio, to quantitatively synthesise the complexes [(η3-
their coordination complexes also exhibit better stability.
C4H7)PdCl(LP)].
The phosphonites were fully characterised by multinuclear
Single crystals of [(η3-C4H7)PdCl(S)-(4a)] ((S)-6a) and [(η3-
NMR spectroscopy and High Resolution Mass Spectrometry. C4H7)PdCl((S)-(5a))] ((S)-7a) suitable for X-ray analysis were
Their 31P{1H} spectra show characteristic singlet peaks; grown (Fig. 3 and 4); the asymmetric unit of (S)-7a comprises
δ (ppm) = (S)-3a (154.8), (R)-3b (155.6), (S)-4a (177.7), (R)-4b two molecules in different conformations (the second struc-
(180.0), (S)-5a (172.4) and (R)-5b (174.2). The resonances for ture and an overlay is provided in the ESI†) and there is
the phenoxyl-derived phosphonites have a notable upfield twofold disorder of the methallyl ligand and the palladium
shift compared to the biphenoxyl analogues. Single crystals of atom in one of these independents. The phosphonites show
(R)-3b and (R)-4b suitable for X-ray crystallographic analysis monodentate coordination through the phosphorus, and there
were obtained (Fig. 1 and 2); there are two crystallographically is a pseudo-square-planar geometry around the palladium.
independent molecules in the asymmetric unit of (R)-3b.
The allyl carbons trans to phosphorus exhibit a longer Pd–C
Density Functional Theory calculations at the B3LYP/6-31G* bond length (2.195, 2.206 Å) than the allyl carbons cis to the
level of theory (ESI†) revealed that the phosphonites have phosphorus (2.096, 2.095 Å), which is consistent with the
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Communication
protons cis to the phosphorus, by a η3–η1–η3 mechanism, due
to the stronger trans effect of the phosphonite donor ligand.4,9
The rapid exchange process resulted in broadened peaks in
the NMR spectra at room temperature; cooling of CD2Cl2 solu-
tions of the complexes to −20 °C sharpened the resonances
and allowed for full characterisation of the methallyl ligands
in both isomers (ESI†). Reaction of (S)-7a with NaBArF resulted
in loss of NaCl and the formation of [(η3-C4H7)Pd(S)-(5a)]BArF
((S)-8a), where (S)-5a acts as a chelating P,C-π-donor. The
upfield 13C{1H} NMR coordination chemical shifts for both C1′
and C2′ suggest a η2-binding mode (Fig. 5), which is in agree-
ment with the results of an NMR study reported by Pregosin
and co-workers.9
As discussed, MOP phosphine ligands are known to give
high enantioselectivity in the palladium-catalysed asymmetric
hydrosilylation of alkenes, particularly styrene, to give chiral
secondary alcohols (Scheme 2).3 To gain an insight into how
the different stereoelectronic profiles of our phosphonites
impacts upon their catalytic activity in the same transform-
ation, we prepared catalysts by reacting each phosphonite with
[Pd(η3-C3H5)Cl]2 (Table 1). We chose to test our phosphonites
against the well-known H-MOP and MeO–MOP phosphines
(the latter is a commercial compound), employing P : Pd ratios
of 1 : 1 and 2 : 1 at room temperature – full conversion was
obtained in all cases. We noted a general increase in enantio-
selectivity and reaction rate in the order (S)-5a/(R)-5b>(S)-4a/
(R)-4b>(S)-3a/(R)-3b, with the introduction of the biphenyl
moiety, and subsequently the methyl groups, markedly
improving the ligand performance. Phosphonite (S)-5a gave
excellent enantioselectivity for (R)-1-phenylethanol (Table 1,
95%, entry 18), which is slightly higher than that for the
(S)-H-MOP phosphine (Table 1, 94%, entry 4), although the
latter reaction reached conversion more quickly. We also
tested the H-MOP phosphonite ligands (S,Rb)-Ia and (S,Sb)-Ia
at 0 °C, these ligands are far less selective than (S)-5a.4
Fig. 3 Molecular structure of (S)-6a. Hydrogen atoms have been
omitted for clarity. Selected average bond distances (Å) and angles (°):
Pd1–P1 2.2478(10), Pd1–Cl1 2.3583(9), Pd1–C17 2.162(4), Pd1–C18
2.195(4), Pd1–C19 2.096(4), P1–C2 1.815(4), P1–O1 1.619(3), P1–O2
1.610(3), C11–C11’ 1.477(5); P1–Pd1–Cl1 91.55(4), P1–Pd1–C18 165.51(11),
P1–Pd1–C19 98.27(13), C18–Pd1–C19 67.46(16), C18–C17–C19 114.6(4),
C2–C1–C1’–C2’ –80.8(5), C12–C11–C11’–C12’ 42.3(6).
Fig. 4 Molecular structure of (S)-7a. Hydrogen atoms have been
omitted for clarity. Selected average bond distances (Å) and angles (°):
Pd1–P1 2.2368(11), Pd1–Cl1 2.3739(11), Pd1–C18 2.152(4), Pd1–C19
2.095(4), Pd1–C20 2.206(4), P1–C2 1.804(4), P1–O1 1.610(3), P1–O2
1.622(3), C11–C11’ 1.485(6); P1–Pd1–Cl1 94.59(4), P1–Pd1–C19 96.50(15),
P1–Pd1–C20 162.95(15), C19–Pd1–C20 66.8(2), C19–C18–C20 117.4(5),
C2–C1–C1’–C2’ –81.9(6), C12–C11–C11’–C12’ –46.0(7).
Interestingly, although the methoxy-substituted ligands
(R)-MeO–MOP phosphine and (R)-5b performed poorly compared
stronger trans influence of the phosphonite compared to the
chloride ligand. In (S)-6a the lower naphthyl fragment of the
binaphthyl backbone is face-to-face with the palladium centre
(Fig. 3). Neither of the two independents of (S)-7a display this
feature, with the lower naphthyl group being orientated away
from the palladium centre and facing the biphenyl moiety
(Fig. 4 and ESI†). In the two independents of (S)-7a the torsion
angle of the dimethyl substituted biphenyl moiety is of oppo-
site sign, and also when comparing the biphenyl moiety in
(R)-4b and (S)-6a (Fig. 2 and 3), implying no restriction to
rotation about the C11–C11′ bond.
Fig. 5 Proposed structure of (S)-8a (left) and a fragment showing the
13C{1H} NMR coordination chemical shift (ppm) between the major
isomers of (S)-7a and (S)-8a (right).
The 31P{1H} NMR peaks of (S)-6a and (S)-7a experience
slight shifts compared to the free ligands, and show the pres-
ence of two independent resonances due to the two isomers
formed; δ (ppm) = (S)-6a (172.0 and 173.6) and (S)-7a (172.9
and 177.6). The isomers are a result of the rotation of the
allyl moiety, via a selective syn/anti exchange of the allyl
Scheme 2 The synthesis of chiral secondary alcohols via the palladium
catalysed asymmetric hydrosilylation and subsequent oxidation of
styrene derivatives.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Communication
Dalton Transactions
Table 1 Catalytic results from the asymmetric hydrosilylation of
styrenea
Table 2 Asymmetric
styrenesa
hydrosilylation
of
heteroatom-substituted
L
P : Pd
T
Timeb
Conv.c
eed
L
Ar
4-ClC6H4
4-ClC6H4
4-ClC6H4
T
Timeb
Conv.c
eed
1
2
3
4
5
6
7
8
(S)-H-MOP
(S)-H-MOP
(S)-H-MOP
(S)-H-MOP
(R)-MeO-MOP
(R)-MeO-MOP
(S)-3a
(S)-3a
(R)-3b
(R)-3b
(S)-4a
(S)-4a
(R)-4b
(R)-4b
(S)-5a
(S)-5a
(S)-5a
(S)-5a
(R)-5b
1 : 1
1 : 1
2 : 1
2 : 1
1 : 1
2 : 1
1 : 1
2 : 1
1 : 1
2 : 1
1 : 1
2 : 1
1 : 1
2 : 1
1 : 1
1 : 1
2 : 1
2 : 1
1 : 1
2 : 1
1 : 1
2 : 1
1 : 1
2 : 1
rt
0 °C
rt
0 °C
rt
rt
rt
rt
rt
rt
rt
rt
rt
rt
rt
0 °C
rt
0 °C
rt
rt
1 min
4 h
1 h
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
88%e
>99%
>99%
22%e
74 (R)
93 (R)
75 (R)
94 (R)
22 (R)
20 (R)
63 (R)
73 (R)
7 (S)
1
2
3
4
5
6
7
8
(S)-H-MOP
(S)-H-MOP
(S)-5a
rt
0 °C
rt
0 °C
rt
0 °C
rt
1 h
24 h
18 h
336 h
10 min
12 h
70 min
240 h
>99%
>99%
>99%
<5%e
>99%
>99%
>99%
>99%
78 (R)
94 (R)
86 (R)
—
35 (R)
58 (R)
78 (R)
85 (R)
12 h
10 min
1 h
24 h
24 h
24 h
24 h
6 min
1 h
10 min
1 h
2 min
5 h
2 h
72 h
1 h
70 min
168 h
168 h
72 h
168 h
(S)-5a
4-ClC6H4
(S)-H-MOP
(S)-H-MOP
(S)-5a
4-MeOC6H4
4-MeOC6H4
4-MeOC6H4
4-MeOC6H4
(S)-5a
0 °C
9
a The catalyst was generated in situ from the ligand (0.50 mol%) and
[Pd(η3-C3H5)Cl]2 (0.125 mol%) and reacted with styrene (10 mmol) and
trichlorosilane (12 mmol). b Time taken for reaction to reach
completion. c Determined by 1H NMR spectroscopy. d % ee determined
by chiral GC; absolute configuration assigned by comparison of the
sign of optical rotation to literature (ESI). e Reaction stopped.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1 (S)
73 (R)
79 (R)
23 (R)
27 (R)
79 (R)
92 (R)
80 (R)
95 (R)
45 (R)
51 (R)
4 (S)
ducing an electron withdrawing chloride group resulted in
much longer reaction times, however full conversion and a
high ee, better than (S)-H-MOP, were again obtained (Table 2,
entries 1 and 3; note entry 2 for improving the ee generated by
the latter). The chloro-substituted chiral secondary alcohol
product was used to synthesise the antihistamine Clemastine
enantioselectively.12 More significantly, when an electron
donating 4-methoxy substituent was incorporated onto the
styrene precursor, (S)-5a was found to be a remarkably superior
ligand, generating a much higher ee than (S)-H-MOP, albeit
more slowly (85% versus 58%, Table 2, entries 8 and 6). It is of
note that the product alcohol from this reaction was used to
prepare antagonists of lysophosphatidic receptors, which may
(R)-5b
(S,Rb)-Ia
(S,Rb)-Ia
(S,Sb)-Ia
(S,Sb)-Ia
0 °C
0 °C
0 °C
0 °C
60 (R)
59 (R)
77 (R)
a The catalyst was generated in situ from the ligand (0.25 or 0.50 mol%)
and [Pd(η3-C3H5)Cl]2 (0.125 mol%), and reacted with styrene
(10 mmol) and trichlorosilane (12 mmol). b Time taken for reaction to
reach completion. c Determined by 1H NMR spectroscopy. d % ee
determined by chiral GC; absolute configuration assigned by
comparison of the sign of optical rotation to literature (ESI). e Reaction
stopped.
to their H-substituted counterparts – an established trend for have implications in the treatment of lung fibrosis, liver
this ligand family10 – the phosphonite ligand gave much higher disease and cancer.13 Our future work will seek to extend the
enantioselectivities than the corresponding phosphine (51% substrate scope to styrenes bearing other important functional
versus 20%, Table 1, entries 20 and 6), demonstrating further groups.
the potential these derivatives have to improve existing toolkits.
In summary, we have synthesised a series of novel MOP-
It is notable that at room temperature, the hydrosilylation phosphonite ligands from air-stable, chiral primary phosphine
occurs as a highly exothermic, quite violent reaction which is precursors, and have demonstrated the ease with which both
accompanied by a spontaneous colour change to deep black; the sterics and the hydrolytic nature of the ligands can be
similar observations with arylmonophosphinoferrocenes have tuned, by modifying the aryl groups of the aryloxy moiety. In
also been made.11 When a 2 : 1, P : Pd ratio was used there was comparison to our previous report on Ia/Ib, a major finding
often a significant delay in the time taken to observe this here is that a second chiral element is not necessary for the
phenomenon (from 1–8 h, Table 1). However these exothermic chiral induction, crucially both the ee and reaction rates can
colour changes were not observed when the external tempera- be increased without it.4 In the palladium-catalysed asym-
ture of the flask was kept at 0 °C. Consequently, the catalysis metric hydrosilylation of functionalised styrenes, our ligands
performed at 0 °C using both (S)-H-MOP and the sterically proved to be high yielding, and both regio- and enantio-
bulky (S)-5a resulted in longer reaction times, however this was selective; phosphonite (S)-5a achieved ees of up to 95%. When
offset by an increased ee of the product – phosphonite (S)-5a the substrate used was 4-methoxystyrene, (S)-5a is significantly
gave the best ee with a value of 95%.
more enantioselective than Hayashi’s benchmark phosphine
The final objective in this study was to extend our styrene (S)-H-MOP, demonstrating how MOP-phosphonite ligands
screening regimen to substrates that MOP phosphines struggle have the potential to fill voids left by their phosphine counter-
to catalyse, or confer only low or moderate enantioselectivity in parts in the field of asymmetric catalysis.
the product, with the aim that the unique stereoelectronic pro-
We wish to acknowledge the financial support of the EPSRC
perties of the phosphonites would allow for improvements to for a Fellowship award (LJH: EP/G005206/1) and its National
be made. As such we extended our screening with (S)-5a to Mass Spectrometry Facility at Swansea, and the School of
include the para-substituted styrenes listed in Table 2. Intro- Chemistry, Newcastle University for a Studentship (JTF).
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Communication
6 J. I. van der Vlugt, J. M. J. Paulusse, E. J. Zijp,
J. A. Tijmensen, A. M. Mills, A. L. Spek, C. Claver and
D. Vogt, Eur. J. Inorg. Chem., 2004, 4193.
7 C. Claver, E. Fernandez, A. Gillon, K. Heslop, D. J. Hyett,
A. Martorell, A. G. Orpen and P. G. Pringle, Chem.
Commun., 2000, 961; M. T. Reetz and T. Sell, Tetrahedron
Lett., 2000, 41, 6333.
Notes and references
1 Phosphorus Ligands in Asymmetric Catalysis, ed. A. Börner,
Wiley-VCH, Weinheim, 2008; Phosphorus(III) Ligands
in Homogeneous Catalysis: Design and Synthesis, ed.
P. C. J. Kamer and P. W. N. M. van Leeuwen, John Wiley &
Sons, Chichester, 2012.
2 T. Hayashi, Acc. Chem. Res., 2000, 33, 354; F. Lagasse
and H. B. Kagan, Chem. Pharm. Bull., 2000, 48, 315;
8 R. M. Hiney, L. J. Higham, H. Müller-Bunz and
D. G. Gilheany, Angew. Chem., Int. Ed., 2006, 45, 7248;
A. Ficks, C. Sibbald, S. Ojo, R. W. Harrington, W. Clegg and
L. J. Higham, Synthesis, 2013, 265.
9 P. G. A. Kumar, P. Dotta, R. Hermatschweiler, P. S. Pregosin,
A. Albinati and S. Rizzato, Organometallics, 2005, 24, 1306.
10 K. Kitayama, Y. Uozumi and T. Hayashi, J. Chem. Soc.,
Chem. Commun., 1995, 1533.
R.
M.
Hiney,
A.
Ficks,
H.
Müller-Bunz,
D. G. Gilheany and L. J. Higham, in Organometallic
Chemistry, ed. I. J. S. Fairlamb and J. M. Lynam,
The Royal Society of Chemistry, Cambridge, 2011, vol.
37, pp. 27–45.
3 T. Hayashi, Catal. Surv. Asia, 1999, 3, 127; J. W. Han
and T. Hayashi, Tetrahedron: Asymmetry, 2014, 25, 11 H. L. Pedersen and M. Johannsen, J. Org. Chem., 2002, 67,
479. 7982.
4 A. Ficks, R. M. Hiney, R. W. Harrington, D. G. Gilheany and 12 A. M. Fournier, R. A. Brown, W. Farnaby, H. Miyatake-
L. J. Higham, Dalton Trans., 2012, 41, 3515. Onzozabal and J. Clayden, Org. Lett., 2010, 12, 2222.
5 L. Bini, C. Muller and D. Vogt, Chem. Commun., 2010, 46, 13 J. H. Hutchinson, T. J. Seiders, B. Wang, J. M. Arruda,
8325.
J. R. Roppe and T. Parr, Br. Pat., GB2470833, 2010.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.