Liquid coordination complexes of Lewis acidic metal chlorides: Lewis acidity and insights into speciation

Article abstract of DOI:10.1039/c7dt02408h

Coordination complexes of Lewis acidic metal chlorides AlCl₃, GaCl₃, InCl₃, SbCl₃, SnCl₄, SnCl₂, ZnCl₂ and TiCl₄ with trioctylphosphine (P₈₈₈) and trioctylphosphine oxide (P₈₈₈O) were synthesised. All compounds formed liquid coordination complexes (LCCs) at ambient temperature, although decomposition via a redox mechanism was detected in some cases. The Lewis acidity of the metal chlorides (measured in 1,2-dichloroethane solutions) and the LCCs (measured neat) was quantified by using the Gutmann acceptor number (AN) approach. In general, LCCs were equally or more Lewis acidic than the corresponding metal chlorides. The AN values were compared with the catalytic activity of selected LCCs in a model Diels-Alder reaction. Insight into speciation of LCCs was gained using multinuclear NMR spectroscopy, revealing that most LCCs comprised charge-neutral complexes rather than ionic ones. The relationship between the speciation, Lewis acidity (AN scale) and catalytic activity is discussed in detail. This approach reveals several new, promising catalytic systems, such as P₈₈₈O-InCl₃, with Lewis acidity enhanced compared to chloroindate ionic liquids, and P₈₈₈O-TiCl₄, with hydrolytic stability enhanced with respect to neat TiCl₄.

Full text of DOI:10.1039/c7dt02408h

View Article Online
View Journal
Dalton
Transactions
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Hogg, L. C.
Brown, K. Matuszek, P. Latos, A. Chrobok and M. Swadzba-Kwasny, Dalton Trans., 2017, DOI:
10.1039/C7DT02408H.
This is an Accepted Manuscript, which has been through the
Volume 45 Number
1 7 January 2016 Pages 1–398
Royal Society of Chemistry peer review process and has been
accepted for publication.
Dalton
Transactions
Accepted Manuscripts are published online shortly after
An international journal of inorganic chemistry
www.rsc.org/dalton
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-9226
PAPER
Joseph T. Hupp, Omar K. Farha et al.
Efficient extraction of sulfate from water using
framework
a Zr-metal–organic
rsc.li/dalton

Please do not adjust margins
Dalton Transactions
Page 1 of 15
View Article Online
DOI: 10.1039/C7DT02408H
Journal Name
ARTICLE
Liquid coordination complexes of Lewis acidic metal chlorides:
Lewis acidity and insights into speciation
Received 00th January 20xx,
Accepted 00th January 20xx
James M. Hogg,a Lucy C. Brown,a Karolina Matuszek,b Piotr Latos,b Anna Chrobokb and Małgorzata
Swadźba-Kwaśnya*
DOI: 10.1039/x0xx00000x
Coordination complexes of Lewis acidic metal chlorides: AlCl3, GaCl3, InCl3, SbCl3, SnCl4, SnCl2, ZnCl2 and TiCl4 with
trioctylphosphine (P888) and with trioctylphosphine oxide (P888O) were synthesised. All compounds formed liquid
coordination complexes (LCCs) at ambient temperature, although decomposition via redox mechanism was detected in
some cases. Lewis acidity of the metal chlorides (measured as 1,2-dichloroethane solutions) and the LCCs (measured neat)
was quantified by Gutmann acceptor number (AN). In general, LCCs were equaly or more Lewis acidic than the
corresponding metal chlorides. The AN values were compared with the catalytic activity of selected LCCs in a model Diels-
Alder reaction. Insights into speciation of LCCs was gained using multinuclear NMR spectroscopy, revealing that most LCCs
comprised charge-neutral complexes, rather than ionic ones. The relationship between the speciation, Lewis acidity (AN
scale) and catalytic activity was discussed in detail. This approach revealed several new, promissing catalytic systems, such
as P888O-InCl3, with Lewis acidity enhanced compared to chloroindate ionic liquids, and P888O-TiCl4, with hydrolytic stability
enhanced with respect to neat TiCl4.
www.rsc.org/
which leaves little room for modification of the catalytic
activity. These shortcomings are driving the development of
new generations of ionic liquid-like Lewis acids, such as liquid
Introduction
Lewis acidic ionic liquids, in particular those with
chloroaluminate(III) anions, are the oldest and are the most
thoroughly studied group of ionic liquids to gain widespread
interest,.1-3 Their catalytic applications (Friedel-Crafts
chemistry, olefin isomerisation and oligomerisation, activation
of transition metal catalysts) have been extensively researched
in the academia, leading to the emergence of several full-scale
industrial processes. In Difasol (IFP) chloroaluminate(III) ionic
liquids were used as solvent and co-catalyst for nickel-
catalysed dimerisation of alkenes.4 IONIKYLATION by
Petrochina,5 CILA by Shell/Petrochina,6 and ISOALKYL by
Chevron (licenced by Honeywell/UOP)7 are all recently-
demonstrated refinery alkylations (alkylation of isobutene with
butenes, to produce high-octane gasoline), and all are
catalysed with chloroaluminate(III) ionic liquids.
coordination complexes (LCCs). LCCs is a recently-coined term
for eutectic mixtures of metal halides and organic donors.8 In
fact, combinations of AlCl3 with donor molecules have long
been used as Lewis acidic reactants in organic chemistry -
always in solutions of organic solvents.9,10 Physical chemists
have suggested that such combinations form ionic mixtures of
coordination complexes, rather than solvated salts.11,12
However, only recently these systems were researched in the
context of ionic liquid applications.
Abbott et al. reported liquid mixtures of urea/amides and
AlCl3, postulating that they form ionic liquids.13 Coleman et al.
studied the speciation of eutectics formed by AlCl3 or GaCl3
with a range of donors (urea, thiourea, acetamide, phosphine
oxide, or phosphine), and reported them to contain
equilibrated cationic, anionic and neutral coordination
complexes of the corresponding metals. A detailed study on
amide-AlCl3 by Liu and co-workers demonstrated that urea
(Ur) and acetamide (AcA) coordinate to AlCl3 through oxygen,
whereas methylacetamide (NMA) and dimethylacetamide
(DMA) have a tendency for bidentate coordination through
both O- and N-donation, increasing as a function of degree of
methylation.14 LCCs of other metal chlorides were merely
touched upon: Abbott et al. reported that urea and ZnCl2,
SnCl2 or FeCl3 form liquids at ambient temperature.15 Coleman
et al. stated the same for trioctylphosphine or
trioctylphosphine oxide combined with SnCl4, SnCl2, ZnCl2 or
FeCl3.8 However, no speciation studies were provided.
Despite this widespread interest, chlorometallate(III) ionic
liquids have some key drawbacks: (i) the organic cation, which
plays a spectator role in Lewis acid chemistry, is also the cost-
incurring component, and (ii) the comprises
surrounded exclusively by halide ligands (typically chlorides),
a metal
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

PleaDsaeltdoonnTortaandsjaucsttimonasrgins
Page 2 of 15
ARTICLE
In catalytic applications of LCCs, their speciation was found to Gutmann AN measurements
Journal Name
View Article Online
DOI: 10.1039/C7DT02408H
have a crucial impact on the catalytic performance. In The Gutmann acceptor numbers of metal chlorides were
oligomerisation of 1-decane, DMA-MCl3 (M = Al or Ga) yielded measured as 1:1 adducts of MClx and triethylphosphine oxide
significantly higher proportion of heavy oligomer products (P222O) as described by Gutmann.19 Solutions of accurately
compared to Ur-MCl3 or AcA-MCl3.16 In Friedel-Crafts alkylation weighed metal chlorides (ca. 0.58 mmol) in accurately known
of benzene with 1-decene, DMA-GaCl3 gave much higher volumes of 1,2-dichloroethane (ca. 1.5 ml) were prepared, to
reaction rate than Ur-GaCl3.17 In refinery alkylation catalysed achieve concentrations of ca. 0.35 mol dm-3. To each solution,
with L-AlCl3 (L = AcA, NMA, DMA, among others) the catalytic 1 mol eq. of P222O (ca. 0.58 mmol) was added, and the mixture
performance changed as a function of the methylation in the was stirred for 30 min. All solids dissolved readily upon the
amide ligand.18 All these can be explained in terms of differing addition of P222O. The 31P NMR spectra of the solutions were
coordination modes, as demonstrated by Liu and co-workers.14 recorded, and then the solutions were diluted to ca. 0.25 mol
Despite increasing interest in LCCs as Lewis acidic catalysts, dm-3, and subsequently to ca. 0.20 mol dm-3, with 31P NMR
only systems based on AlCl3 or GaCl3 have been wider spectra recorded for each concentration. Recorded 31P NMR
explored.
chemical shifts were plotted as a function of concentration and
In this scoping study, we screened a wide range of Lewis acidic extrapolated to infinite dilution of P222O in the studied sample
metal chlorides for their LCC-forming potential, and (δinf). The acceptor number was calculated according to Eq 1,
investigated their Lewis acidity in order to identify promising where δhex inf is the 31P NMR chemical shift for infinite dilution
catalytic systems.
of P222O in hexane.
AN = 2.348·(δinf - δhex inf
)
Eq 1
Experimental
Materials
For each LCC, three samples (ca. 1 g each) were weighed out
accurately into samples vials. P222O was weighed accurately
into each sample (ca. 1, 2 and 3 wt %). 31P NMR spectra were
recorded for the three concentrations, and the chemical shift
value was extrapolated to the value of infinite dilution. The AN
values were calculated from Eq 1.
Aluminium(III) chloride (99.999%) and gallium(III) chloride
(99.999%) were purchased in sealed ampules, under argon
from Alfa Aesar. Titanium tetrachloride (99%), tin tetrachloride
(99.999%), tin dichloride (99.99%), indium trichloride (99%),
zinc dichloride (98 %) and antimony trichloride (99.95%) were General procedure for Diels-Alder reaction
all purchased from Sigma Aldrich and used as received.
Trioctylphosphine oxide (99.5%) was provided by Cytec and
dried under reduced pressure (80 °C, < 1 mbar, 48 h) before
use. Triocytlphosphine (99.5%) was provided by Cytec in a
sealed canister under an inert atmosphere and used as
received. Triethylphosphine oxide (99%) and ethyl acrylate
were purchased from Sigma Aldrich and was used as received.
Cyclopentadiene (ex. Sigma Aldrich) was supplied as a dimer
and was distilled immediately before use.
All reactions were carried out under dry argon. In a typical
procedure, the LCC catalyst (0.1 - 1.0 mol% per 1.0 mole of
dienophile) was placed in a two-neck round-bottomed flask
equipped with stirring bar. Then, a mixture of dienophile (16
mmol) and diene (24 mmol) was added drop-wise to the
vigorously stirred (1500 rpm) reaction mixture in an ice bath.
The reaction was carried out for 5 - 120 min, and its progress
monitored by gas chromatography. Afterwards, the reaction
was quenched with few drops of water, and 1 cm3 of
dichloromethane was added to homogenise the mixture prior
to GC analysis.
Synthesis of liquid coordination complexes
LCCs were synthesised on a 2 g scale, following the literature
procedure.8 All syntheses were carried out in a argon-filled
glovebox (MBraun LabMaster dp, <0.6 ppm O2 and H2O). To a
neutral donor ligand (0.25 - 1.00 mol eq.) a metal chloride was
added in small aliquots. Heat was typically evolved. The
reaction mixture was stirred until the flask was cool to touch
before additional aliquots were added. This was repeated until
GC analysis was performed using a Perkin Elmer Clarus 500 gas
chromatograph equipped with an SPB™-5 column (30 m x 0.2
mm x 0.2 μm) with n-decane as internal standard.
Results and discussion
the desired molar ratio of metal chloride of (ꢀMClx = 0.50 - 0.75) Synthetic strategy
was reached. Heat was typically evolved. Afterwards, the
Liquid coordination complexes were prepared from a range of
reaction mixture was stirred (0.5 - 24 h, 30 - 80 °C), until
homogeneous liquid was obtained. All LCCs were stored in the
glovebox until used.
Lewis acidic metal chlorides: AlCl3, GaCl3, InCl3, SbCl3, SnCl4,
SnCl2, ZnCl2 and TiCl4, combined with either of the two ligands:
trioctylphosphine (P888) or trioctylphosphine oxide (P888O). The
6 - 8 carbon alkyl chain length is considered optimal for
generating liquids, rather than solids (shorter chains promote
crystallisation, whereas longer ones - wax formation).20
Indeed, in our earlier work on the AlCl3 and GaCl3-LCCs, these
two ligands yielded LCCs that were liquid at ambient
conditions for the widest range of compositions.8
Multinuclear NMR spectroscopy
31P, 67Zn, 115In and 121Sb NMR spectra were recorded on a
Bruker Avance DPX 400 MHz spectrometer, at 162, 25, 88 and
96 MHz, respectively. 49Ti and 119Sn NMR spectra were
recorded on a Bruker Avance DPX 600 MHz spectrometer, at
34 and 224 MHz, respectively., All samples were studied neat,
using a DMSO-filled, sealed capillary as an external lock.
Furthermore, comparison between P888 and P888
O was
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 15
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
potentially offering insight into the influence of the donor upon exposure to air is a significant advantage in terms of ease
View Article Online
DOI: 10.1039/C7DT02408H
hardness on speciation and Lewis acidity.
and safety of handling.
The compositional range (metal chloride to ligand ratio, Lewis acidity of the LCCs, similarly to that of halometallate
reported as the molar ratio of the former, χMClx) was also ionic liquids, depends on the metal centre, and on speciation.21
varied. In our preliminary work, we reported the L-AlCl3 (L = In LCCs however, added complexity arose from he presence of
P888O or P888) systems to be liquid for χAlCl3 = 0.50 - 0.60, and a ligand (soft P888 vs. hard P888O), expected to influence
the L-GaCl3 (L = P888O or P888) LCCs to be liquid at χGaCl3 = 0.50 - electrophilicity on the metal.
0.75.8 In this work, the equimolar composition (χMClx = 0.50)
Insights into speciation
was the starting point; whenever homogenous liquids had
To enable meaningful interpretation of the AN studies,
been obtained, a composition with the excess of metal
knowledge of the LCCs speciation was indispensible. At the
chloride (χMClx = 0.60, corresponding to MClx:L ratio of 1.5:1)
same time, a full speciation study of every system (comprising
minimum three techniques, such as multinuclear NMR
spectroscopy, Raman and FT-IR spectroscopies, XPS or EXAFS,
and supporting studies, e.g. conductivity measurements) was
not feasible for a single publication. Therefore, in this work,
preliminary speciation of LCCs was proposed, based on
multinuclear NMR studies combined with a thorough literature
research. Systems of particular interest were identified, and
their detailed speciation will be published in separate
contributions.
was synthesised.
All metal chlorides, combined with either of the two ligands at
χMClx = 0.50, formed homogenous liquids (Table 1). At χMClx
=
0.60, homogenous LCCs were generated by combining GaCl3,
SbCl3, SnCl4, TiCl4 or ZnCl2 with at least one of the two ligands.
As previously reported, L-GaCl3 LCCs were homogenous liquids
up to χMClx = 0.75.8
Table 1. Combinations of metal chlorides and P888 or P888O ligands that formed LCCs, at
given χMClx values.
Amongst the LCCs in this study here, AlCl3 and GaCl3-LCCs have
already been characterised by 27Al NMR and Raman
spectroscopies (71Ga NMR signals were too broad to be
interpretable).8 In this work, it was attempted to record 115In,
121Sb, 119Sn, 49Ti and 67Zn NMR spectra of the relevant LCCs,
but only the 119Sn NMR spectra contained well-defined signals.
The remaining nuclei are quadrupolar (in contrast to spin 1/2
119Sn), thus generally give wider signals, which broaden further
with decreasing symmetry around the studied nuclei.
Fortunately, the 31P{1H} NMR spectra for all LCCs gave
interpretable signals (δ 31P, Table 2), with signals shifted
upfield in comparison to uncoordinated ligand (Δδ 31P, Table
2). The corresponding 31P NMR spectra are reported in
Supplementary Information, Figure 1 SI - 4 SI.
MClx
χMClx
P888
O
P888
AlCl3
0.50
0.60
0.67
Colourless liquid$
Yellow liquid$
Yellow liquid with
white precipitate$
Yellow liquid$
Yellow liquid$
Yellow liquid with
white precipitate$
GaCl3
InCl3
0.50
0.60
0.67
0.75
Colourless liquid$
Colourless liquid$
Colourless liquid$
Colourless liquid$
Colourless liquid$
Colourless liquid$
Colourless liquid$
Colourless liquid$
0.50
0.60
Colourless liquid
Colourless liquid with
white precipitate$$
Colourless liquid with
white precipitate
Colourless liquid with
white precipitate
SbCl3
SnCl2
0.50
0.60
Colourless liquid
Colourless liquid
Liquid, redox reaction
-
0.50
0.60
Colourless liquid
Colourless liquid with
white precipitate
Colourless liquid
Colourless liquid with
white precipitate
SnCl4
TiCl4
0.50
0.60
Colourless liquid
Colourless liquid
Liquid, redox reaction
-
0.50
0.60
Yellow liquid
Yellow solid with white
powder
Red liquid
Red liquid
ZnCl2
0.50
0.60
Colourless liquid
Colourless liquid
Colourless liquid
Colourless liquid with
white precipitate
$ - previously reported; $$ - a very small amount of precipitate
The presence of a hydrophobic ligand (P888O or P888) increased
the hydrolytic stability of halides such as SnCl4 or TiCl4; the
corresponding LCCs did not fume/hydrolyse violently when
exposed to air. Although they hydrolysed over prolonged
exposure to the atmosphere, the lack of rapid HCl release
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

PleaDsaeltdoonnTortaandsjaucsttimonasrgins
Page 4 of 15
ARTICLE
Journal Name
Table 2. 31P NMR chemical shifts (δ 31P / ppm) for free P888 and P888O, and for ligands in
the corresponding LCCs with different metal halides and χMClx values. Changes in 31P
NMR chemical shift upon ligand complexation to a metal (Δδ 31P / ppm) are also listed.
View Article Online
DOI: 10.1039/C7DT02408H
P888
O
P888
MClx
χMClx
δ 31
P
Δδ 31
P
δ 31
P
Δδ 31
P
/ ppm
-
0.00
0.50
49.91
-
-29.44
-
AlCl3
75.20
77.38
25.29
27.47
-26.65
-22.16
2.79
7.28
0.60
76.57
77.37
78.07
26.66
27.46
28.16
-23.29
6.15
GaCl3
0.50
0.60
0.67
0.75
76.50
80.15
26.59
30.24
-9.87
-5.99
-2.61
-0.86
19.57
23.45
26.83
28.58
79.97
81.23
30.06
31.32
81.31
83.41
31.40
33.50
84.37
85.21
86.02
34.46
35.30
36.11
Figure 1. 31P NMR spectra of a) P888O-AlCl3 χAlCl3 = 0.50, b) P888O-AlCl3 χAlCl3 = 0.60,
c) P888-AlCl3 χAlCl3 = 0.50, d) P888-AlCl3 χAlCl3 = 0.60.
InCl3
0.50
0.50
72.04
70.91
22.13
21.00
11.49
-2.77
40.93
26.67
For χAlCl3 = 0.60, binuclear complexes were detected (Eq 3).8
The [Al3Cl10]- anion is not stable at ambient conditions and
decomposes to precipitate AlCl3,1,22 therefore χAlCl3 = 0.60 was
the upper compositional limit for AlCl3-LCCs.
SbCl3
0.60
0.50
69.60
70.96
19.69
21.05
-
-
SnCl2
SnCl4
-3.89
25.55
0.50
68.74
71.40
18.83
21.49
102.7
(redox)
132.14
(redox)
3AlCl3 + 2L → [AlCl3L] + [Al2Cl6L] ⇄ [AlCl2L2][Al2Cl7]
Eq 3
In agreement with this assignment, 31P NMR spectrum of
P888O-AlCl3, χAlCl3 = 0.60 (Figure 1 b) contained three signals,
26.66, 27.46 and 28.16 ppm, corresponding to [AlCl3(P888O)],
[AlCl2(P888O)2]+, and [Al2Cl6(P888O)], respectively. This supports
the speciation proposed currently for all O-donor-AlCl3 LCCs
(with monodentate ligands) studied to date.8,14
0.60
0.50
68.88
71.38
87.27
18.97
21.47
37.36
-
-
TiCl4
27.68
57.12
0.60
0.50
0.60
-
-
34.50
63.94
9.41
ZnCl2
65.56
68.23
15.65
18.32
-20.03
The 31P NMR spectrum of the P888-AlCl3, χAlCl3 = 0.50, featured
two broad overlapping singlets at -26.65 and -22.16 ppm
(Figure 1 c). It could be mistaken for a doublet because the
AlCl3-LCCs. Both χMClx = 0.50 and 0.60 compositions were
liquids, irrespective of the ligand (Table 1). Speciation of L-AlCl3
(L = P888 or P888O), has been already studied as a function of
composition, using 27Al NMR and Raman spectroscopies.8 It
was proposed that, at χMCl3 = 0.50, these LCCs contain
equilibrated mononuclear complexes (Eq 2).
1
two signals are of similar intensity, but the coupling of JP-Al
=
467 Hz is exceedingly high for Al-phosphine complex, with
1
reported values always JP-Al < 300 Hz.23-25 The change in 31P
NMR chemical shifts, compared to that of neat P888 (Δδ 31P =
2.79 - 7.28 ppm), is quite small, when compared to other data
in Table 2, and to Δδ 31P = 19.4 ppm reported for [AlCl3(P111)].25
This however is in agreement with the literature: in Al-
phosphine adducts, Δδ 31P decreases with increasing bulkiness
of the phosphine, and even may assume negative values.26
Finally, Raman spectra of P888-AlCl3, χAlCl3 = 0.50, featured
bands characteristic of the [AlCl4]- anion.8 This points towards
speciation consistent with Eq 2, with equilibrated [AlCl3(P888)],
[AlCl2(P888)2]+, and [AlCl4]-.
AlCl3 + L → [AlCl3L] ⇄ [AlCl2L2][AlCl4]
Eq 2
The 31P NMR spectrum of P888O-AlCl3, χAlCl3 = 0.50 (Figure 1 a,
Table 2) featured two 31P NMR signals, corresponding to two
phosphorus-containing complexes proposed in Eq 2, 25.29
ppm - [AlCl3(P888O)] and 27.47 ppm - [AlCl2(P888O)2]+.
The 31P NMR spectrum of the P888-AlCl3, χAlCl3 = 0.60, contains a
singlet at -23.29 ppm (Figure 1 d). With the reported Raman
spectrum being inconclusive,8 the 27Al NMR spectrum features
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 15
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
2GaCl3 + P888O → [Ga2Cl6(P888O)] ⇄ [GaCl2(P888O)2]V[iGewa3ACrtilc1l0e]Online
a broad signal with a maximum at 107.8 ppm, consistent with
AlCl3-phosphine adducts.23,25 This may be interpreted as the
equilibrium in Eq 3 being strongly left-shifted (towards
molecular adduct), or as a very rapid exchange between
charged and neutral complexes. AlX3-phosphine complexes are
rare,24 and systems with aluminium halide in excess is quite
unique, which offers no literature backing of either hypothesis.
DOI: 10.1039/C7DT02408H
Eq 4
Finally, the spectrum for χGaCl3 = 0.75 is complex (Figure 2 d,
left; Table 2), with three major signals and some low-intensity
features, suggesting a number of equilibrated oligomeric
chlorogallate species.
Speciation of P888-GaCl3 system appears to deviate from that
proposed in our earlier work (Coleman et al.).8 All
compositions, χGaCl3 = 0.50 - 0.75, gave singlets in their 31P
GaCl3-LCCs. L-GaCl3 LCCs were liquid for χMClx = 0.50 - 0.75,
irrespective of the ligand (Table 1).8 This derives from
propensity of chlorogallate(III) ions to form oligomeric chains,
as reported for chlorogallate(III) ionic liquids.27-29 Raman
studies of L-GaCl3, χMClx = 0.50 - 0.75, have been reported in
our earlier work.8
NMR spectra (Figure 2, right; Table 2). The signal at χGaCl3
=
0.50 is relatively broad, which could indicate several species in
an equilibrium (Figure 2 a, right), however, the peaks for χGaCl3
= 0.60 - 0.75 are very narrow, suggesting belonging to a single
species (Figure 2 b-d, right).
The 31P NMR spectrum of P888O-GaCl3, χGaCl3 = 0.50 (Figure 2 a,
left; Table 2) featured two 31P NMR signals at 76.50 and 80.15
ppm, corresponding to two phosphorus-containing complexes:
[GaCl3(P888O)] and [GaCl2(P888O)2]+, respectively, in analogy to
Eq 2. Compared to AlCl3-based analogue (Figure 1 a), the signal
for the neutral adduct was much more prominent.
Previously-reported Raman spectra feature less bands in the
Ga-Cl stretching frequency region, compared to other L-GaCl3
LCCs, which suggests lower number of complexes present.8
The band at 349 cm-1 is dominant for χGaCl3 = 0.50, and the
band at 372 cm-1 dominates for χGaCl3 = 0.60 - 0.75. The Ga-Cl
vibrational frequency of 349 cm-1 is in close agreement with
the band at 349 cm-1, reported for a neutral adduct,
[GaCl3(PPh3)], by Reid and co-workers.30 The same group
reported several failed attempts to synthesise [GaCl2L2]+
cations (L = phosphine), with GaCl3 having a strong preference
towards molecular adduct with monodentate phosphines,31 in
contrast to findings for AlCl3.25 The band at 372 cm-1 was
recorded also in other L-GaCl3 LCCs, and may thus be assigned
to Ga-Cl stretching frequency in oligomeric neutral GaIII
complex, [GanCl3n(P888)], forming in the excess of GaCl3.
Combining the 31P NMR spectra recorded in this work with the
literature Raman spectra,8 and with the literature data,25,30,31
the P888-GaCl3 LCCs appear to be molecular adducts (Eq 5 and
6).
GaCl3 + P888 → [GaCl3(P888)]
Eq 5
Eq 6
nGaCl3 + P888 → [GanCl3n(P888)]
The difference in the action of P888O and P888 on the dimeric
‘Ga2Cl6’ unit can possibly be interpreted in analogy to work of
Gandon and co-workers.32 In their study of GaX3 adducts with
NHC-donors, they found that strong donors (such as P888O)
induce heterolytic splitting, whereas homogenous splitting is
caused by weaker donors (here - P888).
Figure 2. 31P NMR spectra of L-GaCl3, where L = P888O (left) or P888 (right) a) χGaCl3 = 0.50,
b) χGaCl3 = 0.60, c) χGaCl3 = 0.67, d) χGaCl3 = 0.75 (*= artefact)
The 31P NMR spectrum of P888O-GaCl3, χGaCl3 = 0.60 (Figure 2 b,
left; Table 2), is similar to that of the AlCl3 analogue (Figure 1
b), but only two signals are visible, with the lowest intensity
signal either absent, or merged with the baseline. However, in
combination with Raman studies,8 speciation analogous to
that in Eq 3 can be confirmed.
The presence of two signals at 81.31 and 83.41 ppm for P888O-
GaCl3, χGaCl3 = 0.67 (Figure 2 c, left; Table 2) corresponds well
to [GaCl2(P888O)2]+ and [Ga2Cl6(P888O)] - see Eq 4. This
assignment is in agreement with the reported Raman spectra,
which (in the Ga-Cl stretching frequency region) are very
similar of all O-donor-GaCl3 LCCs.8
InCl3-LCCs. LCCs with equimolar composition, L-InCl3, χInCl3
=
0.50 (L = P888O or P888), contained very small amounts of
precipitate, which was filtered off before further studies. For
χInCl3 > 0.50, InCl3 was found to precipitate in significant
quantities (Table 1). Similarly, in chloroindate(III) ionic liquids,
dimeric or oligomeric chloroindate(III) anions were not
observed, with excess of InCl3 precipitating.29,33,34
31P NMR signals recorded for the two InCl3-LCCs studied here
revealed broad singlets, at 72.04 ppm for P888O-InCl3, χInCl3
0.50 and at 11.49 ppm for P888-InCl3, χInCl3 = 0.50 (Table 2).
Whereas chloroindate(III) ionic liquids gave broad but well-
defined signals for the tetrahedral [InCl4]- anion in their 115In
=
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

PleaDsaeltdoonnTortaandsjaucsttimonasrgins
Page 6 of 15
ARTICLE
Journal Name
NMR spectra,34 no 115In NMR signals could be distinguished for
either LCC. Single peaks in 31P NMR spectra and no evidence
for [InCl4]- in 115In NMR spectra suggest that simple molecular
adducts, liquid at ambient temperatures, are formed in both
cases (Eq 7).
View Article Online
DOI: 10.1039/C7DT02408H
InCl3 + L → [InCl3L] (L = P888 or P888O)
Eq 7
This is supported by the literature on coordination chemistry
of InIII. Indium has significantly larger covalent radius,
compared to gallium and aluminium, easily adopting 4-, 5- or
6-coordinate environment.24 Hence, chloroindate(III) ionic
liquids contain equilibrated [InCl6]3-, [InCl5]2- and [InCl4]-
29,33,34
anions, depending on the composition, χInCl3
.
In adducts
with phosphine oxides, the formation of a molecular or ionic
compound depended on the ligand bulkiness, e.g.
35
Figure 3. Crystal structures for a) [SbCl3(P111)] (a polymeric array), and b) [SbCl3(PPh3)]
(dimeric structure).36
[InCl3(P111O)3] vs. [InCl2(PPh3O)4][InCl4].
Nevertheless, the
ligand size did not decrease the coordination number of the
cationic indium, and always excess of ligand was required to
induce heterolytic splitting (which is never the case in this
study). In adducts of InX3 with phosphines, 4- and 5-coordinate
molecular adducts are common, with some 6-coordinate
species known, but heterolytic splitting is rare, only with
extremely bulky ligands.24
The 31P NMR spectra of the P888O-SbCl3 system featured a
singlet at 70.91 ppm (Δδ 31P = 21.0 ppm) for χSbCl3 = 0.50, and a
singlet at 69.60 ppm (Δδ 31P = 19.69 ppm) for χSbCl3 = 0.60
(Table 2). These are comparable with the 31P NMR spectrum of
the [SbCl3(PCy3O)] adduct, which also contained a singlet at
65.3 ppm (Δδ 31P = 15.6 ppm).38 The crystal structure for this
adduct, published by Burford and co-workers, revealed a
dimer with pentacoordinate SbIII, which was remarkably similar
to the structure of [SbCl3(PPh3)], Figure 3 b.38
The liquid-phase speciation of the P888O-SbCl3 system is
proposed to contain neutral adducts, possibly in the dimeric
form, for χSbCl3 = 0.50 (Eq 8). The P888O-SbCl3 χSbCl3 = 0.60
composition likely contain charge-neutral Cl-bridged
oligomers, with bridging modes like these in Figure 3, but
further studies would be required to confirm this.
SbCl3-LCCs. Both ligands gave homogenous liquids for χSbCl3
0.50. The P888O-SbCl3, χSbCl3 = 0.60, LCC was also a liquid.
=
The 31P NMR spectrum of P888-SbCl3, χSbCl3 = 0.50, featured a
singlet at -2.77 ppm (Table 2). Burford and co-workers
reported that reactions of SbCl3 with three different
phosphines, at χSbCl3 = 0.50, yield the corresponding neutral
adducts, [SbCl3(phosphine)], characterised by 31P NMR singlets
at 5.6 - 25.1 ppm.36 In the crystal structure, [SbCl3(P111)]
formed a Cl-bridged polymeric chain (Figure 3 a), whereas
complexes with bulkier phosphines, [SbCl3(PPh3)] and
[SbCl3(PCy3)], formed dimers with two Cl-bridges (Figure 3 b).
In all cases, the complex adopted square based pyramidal
geometry around the SbIII centre. In analogy, the formation of
a molecular adduct was assumed, with the caveat that it may
be assuming a dimeric structure also in solution (Eq 8).
SnCl2-LCCs. The L-SnCl2 LCCs (L = P888O or P888) were liquid for
χSnCl2 = 0.50, and contained unreacted SnCl2 for χSnCl2 > 0.50
(Table 1). This was in contrast to chlorostannate(II) ionic
liquids, which were homogenous up to χSnCl2 = 0.67 due to the
formation of [Sn2Cl5]-.39 This was the first indication that
molecular adducts, rather than ionic species, should be
expected in L-SnCl2 LCCs.
2SbCl3 + 2L → 2[SbCl3L] ⇄ [Sb2Cl6L2]
Eq 8
The 119Sn NMR spectrum of P888-SnCl2, χSnCl2 = 0.50, featured a
broad singlet at -70.9 ppm (Figure 4 a), and the corresponding
31P NMR spectrum had a sharp singlet at -3.89 ppm (Figure 4 c,
Table 2). These are in a good agreement with spectroscopic
data reported for the neutral adduct, [SnCl2(P222)], which a
broad 119Sn NMR signal at -82.5 ppm, and a 31P NMR singlet at
-3.7 ppm.40 In its crystal structure, the bond angles around tin
at ca. 90° angle due to very strong inert pair effect.
However, after two days (sample stored in the glovebox under
ambient conditions, argon atmosphere), weak signal
a
corresponding to [PV888Cl]+ appeared at 103 ppm (Figure 5 SI
b).37 In an older sample, a dark precipitate (likely antimony
black) appeared, indicating that the LCC is unstable, and
undergoes a complex redox reaction, which was further
confirmed by 31P NMR spectroscopy of the remaining liquid
(Figure 5 SI c). A fresh sample of the LCC was synthesised for
acceptor number and Diels-Alder studies, but the P888-SbCl3,
χSbCl3 = 0.60, composition was not studied, owing to this
decomposition.
In conclusion, tricoordinate [SnCl2(P888)] adduct, liquid at
ambient temperature, was formed (Eq 9).
SnCl2 + P888 → [SnCl2(P888)]
Eq 9
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 15
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
View Article Online
DOI: 10.1039/C7DT02408H
Figure 4. NMR spectra of L-SnCl2 LCCs at χSnCl2 = 0.50 a) 119Sn NMR spectrum of P888
-
SnCl2, b) 119Sn NMR spectrum of P888O-SnCl2, c) 31P NMR spectrum of P888-SnCl2, d) 31P
NMR spectrum of P888O-SnCl2
Figure 5. NMR spectra of SnCl4-LCCs a) 119Sn NMR spectrum of P888O-SnCl4, χSnCl4 = 0.50,
b) 119Sn NMR spectrum of P888O-SnCl4, χSnCl4 = 0.60 c) 31P NMR spectrum of P888O-SnCl4,
χSnCl4 = 0.50, d) 31P NMR spectrum of P888O-SnCl4, χSnCl4 = 0.60.
Crystal structures of a general formula [SnCl2L], where L =
trialkylphosphine oxide, or triarylphosphine oxide, are not
known,‡ in contrast to [SnCl2L2] adducts. For example, Reid and
co-workers reported on crystal structures of [SnCl2(P111O)2]
and [SnCl2(PPh3O)2], crystallised from solutions with SnCl2:L
ratio of χSnCl2 = 0.33.41 The 31P NMR spectra featured broad
singlets at 51.2 and 36.3, respectively. The 119Sn NMR
spectrum of [SnCl2(P111O)2] had a singlet at -204 ppm, but no
signal was recorded for [SnCl2(PPh3O)2].
This highly shielded signal at -710 ppm can be identified as
belonging to the hexacoordinate [SnCl4(P888O)2],42 with the two
multiplets likely corresponding to cis- and trans-conformations
of the P888O ligand in the octahedron. Coulpling constants are
slightly lower than these reported in the literature (2J119Sn-31P
=
144 - 216 Hz). The less shielded peak at -449.2 ppm is assigned
to [SnCl4(P888O)], which is the stoichiometric composition for
χSnCl4 = 0.50. The chemical shift for the least shielded signal
changed from -336.3 ppm for χSnCl4 = 0.50, to -211.1 ppm for
χSnCl4 = 0.60, compared to -150 ppm for neat SnCl4. Also, its
area visibly increased with increasing χSnCl4 value. This signal is
assigned to tin(IV) chloride, in a dynamic equilibrium with its
complexes (Eq 11).
The 119Sn NMR spectrum of P888O-SnCl2, χSnCl2 = 0.50, reported
in this work, featured a broad, low intensity singlet at -179.5
ppm (Figure 4 b), with the 31P NMR spectrum featuring a sharp
singlet at 70.96 ppm (Figure 4 d, Table 2). The 119Sn NMR signal
corresponds to a tetracoordinate tinII complex, with no signals
either in the tricoordinate region (downfield from -100 ppm),
nor for SnCl2 (>0 ppm).23 Combining the requirement for
tetracoordinate environment, equimolar ligand to tin ratio and
single peak in 31P NMR spectrum, a charge-neutral dimer,
[Sn2Cl4(P888)2], with two bridging chlorides, is proposed as the
main species (Eq 10).
[SnCl4(P888O)2] + SnCl4 ⇄ 2[SnCl4(P888O)]
Eq 11
The P888-SnCl4 χSnCl2 = 0.50 system was a homogenous liquid at
ambient temperature, but both 119Sn and 31P NMR
spectroscopy of the fresh sample revealed a redox reaction (Eq
12). Consequently, higher loadings of SnCl4 were not tested.
2SnCl2 + 2P888O → [SnCl(P888O)(μ-Cl)2SnCl(P888O)]
Eq 10
SnIVCl4 + PIII888 → [PV888Cl][SnIICl3]
Eq 12
SnCl4-LCCs. The P888O-SnCl4, χSnCl4 = 0.50 and 60, formed
homogenous liquids (Table 1). Their 119Sn NMR spectra both
contained three groups of signals: -336.3, -449.2 and -710 ppm
for χSnCl4 = 0.50 (Figure 5 a), and -211.1, -449.2 and -710 ppm
for χSnCl4 = 0.50 (Figure 5 b). In the multiplet at -710 ppm, two
poorly resolved doublets (or multiplet fragments) could be
This contrasts with reports of Burford and co-workers, who
synthesised [SnCl4(P111)2] with no evidence of redox reaction,
43
even despite pyrophoric nature of P111
.
This might be
attributed to very different reaction conditions: after 1 h
reaction in dichloromethane, the solid product was isolated.
identified, at -709.5 ppm (2J119
= 149.4 Hz) and at
Sn-31P
TiCl4-LCCs. Both P888-TiCl4 compositions were bright red liquids
(Table 1). The colour is characteristic of phosphine complexes
of TiIV and arises due to phosphine⟶TiIV LMCT absorption; for
example, both [TiCl4(PPh3)] and [TiCl4(PPh3)2] absorb ca. 480
nm.44
-711.7 ppm (2J119 Sn-31P = 109.2 Hz) - see insert in Figure 5.
Otherwise, all signals were broad singlets.
The corresponding 31P NMR spectra both contained two
groups of signals: at 68.74 and 71.40 ppm for χSnCl4 = 0.50, and
at 68.88 and 71.38 ppm for χSnCl4 = 0.6 (Table 2, Figure 5 c and
d).
In 31P NMR spectra of P888-TiCl4, χTiCl4 = 0.50 and 0.60, singlets
at 27.68 and 34.50 ppm, respectively, were found (Table 2).
Very strong deshielding of the nuclei compared to the neat
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

PleaDsaeltdoonnTortaandsjaucsttimonasrgins
Page 8 of 15
ARTICLE
ligand (Δδ 31P = 57.12 and 63.94, respectively) is in agreement
Journal Name
View Article Online
with the literature reports for phosphine complexes of TiIV ZnCl2-LCCs. The L-ZnCl2 LCCs, χZnCl2
=
0.50, formed
DOI: 10.1039/C7DT02408H
(Δδ 31P = 26 ppm for PPh3, 71.2 ppm for PMe3).45
homogenous liquids for both L = P888O and P888 (Table 1). With
The TiCl4-phosphine adducts in solution are characterised by increasing zinc chloride loading to χZnCl2 = 0.60, P888O-ZnCl2
complex equilibria between molecular complexes: [TiCl4(Pnnn)] formed a homogenous liquid, whereas P888-ZnCl2 contained
and [TiCl4(Pnnn)2] - see Eq 13 and 14, in addition to a dimer with white precipitate.
two bridging chlorides, [TiCl2(Pnnn)(μ-Cl)2TiCl2], (Eq 15), in The 31P NMR spectrum of P888-ZnCl2, χZnCl2 = 0.50, gave a
excess of TiCl4.45-47
singlet at -20.03 ppm. In the crystal structures of adduct of
ZnX2 with neutral ligands, the most common are [ZnX2L2]
motifs (χZnCl2 = 0.33), but with higher χZnCl2 values bridged
structures are reported, without decreasing the coordination
TiCl4 + Pnnn ⇄ [TiCl4(Pnnn)]
[TiCl4(Pnnn)] + Pnnn ⇄ [TiCl4(Pnnn)2]
[TiCl4(Pnnn)] + TiCl4 ⇄ [Ti2Cl8(Pnnn)]
Eq 13
Eq 14
Eq 15
24
number. Indeed, equimolar adducts of zinc(II) halides with
phosphines are known to form [ZnX(PR3)(μ-X)2ZnX(PR3)],24 for
example recently-reported [ZnCl(PtBu)(μ-Cl)2ZnCl(PtBu)], Figure
6.51
Depending on the solvent, temperature and phosphine, 31P
NMR spectra of the solutions may feature either a single peak,
45
or separate signals for each of the three proposed species
47
are seen. For example, studies of TiCl4 adducts with P222 or
P444 in C6D6 afforded three distinguishable 31P NMR signals for
[TiCl4(Pnnn)2], [TiCl4(Pnnn)] and [Ti2Cl8(Pnnn)]. Irrespective of the
phosphine, and for a wide range of TiCl4 loadings (from χTiCl4
=
0.50 to χTiCl4 > 0.67), these three signals were present,
indicating the presence of all three species in all
compositions.47 The same was suggested for other phosphines,
even if singlets were found in 31P NMR spectra. 45,46
The 31P NMR signals recorded in this work are within the same
range as in the literature, and liquids are formed even with
excess of TiCl4, thus analogous speciation is concluded, i.e.
equilibrated [TiCl4(P888)2], [TiCl4(P888)] and [Ti2Cl8(P888)] in both
studied compositions (Eq 16).
Figure 6. The structure of [ZnCl(PtBu)(μ-Cl)2ZnCl(PtBu)], with ellipsoids at 50%
probability level and arbitrary size for hydrogens.51
[TiCl4(P888)2] + [Ti2Cl8(P888)] ⇄ 3[TiCl4(P888)]
Eq 16
The P888O-TiCl4 χTiCl4 = 0.50 formed a yellow liquid, but the
χTiCl4 = 0.60 composition solidified, and was not studied further
(Table 1). A single 31P NMR signal at 87.27 ppm was recorded
for the χTiCl4 = 0.50 composition. With Δδ 31P = 37.36 ppm, it is
the most deshielded signal for P888O recorded in this study
(Table 2), giving the promise of high Lewis acidity.
The existence of large, Cl-bridged structures in the liquid phase
is feasible; for example, chlorozincate(II) ionic liquids may
contain [ZnCl4]2-, [Zn2Cl6]2-, and larger anionic clusters such as
[Zn4Cl10]2-.52,53 In consistence with a singlet in 31P NMR
spectrum, and known chemistry of zinc(II) chloride with
phosphines, the structure of P888-ZnCl2, χZnCl2 = 0.50 is
postulated (Eq 17).
A yellow crystalline [TiCl4(OPPh3)2] complex was reported, but
no crystal structures of equimolar adducts are known.48 P888O,
often abbreviated as TOPO (trioctylphosphine oxide), is
commonly studied extracting agents for TiIV species from acidic
aqueous solutions, with [TiCl4(P888O)2] as the proposed adduct
structure.49 Furthermore, P888O was used as solvent/capping
agent in the synthesis of titania nanoparticles from TiCl4 and
titanium alkoxides.50 Despite many uses, coordination
chemistry of TiIV with phosphine oxides is solution is not well
studied, especially with equimolar or excess amount of TiCl4.
Based on the available 31P NMR spectrum, and evidently
strong bonding of the ligand, an equimolar adduct
[TiCl4(P888O)] is tentatively assumed as the dominant species.
Oligomers similar to those postulated for TiCl4-P888 (Eq 16) are
conceivable, but considering very strong O-Ti interaction, the
ligand is expected to be much less labile. Given interest in this
system, speciation of P888O-TiCl4 as a function of χTiCl4 will be
the subject of a separate study, supported with vibrational
spectroscopy and crystallography.
ZnCl2 + L → [ZnClL(μ-Cl)2ZnClL]
Eq 17
Adducts of zinc halides with phosphine oxides are far less
studied in inorganic chemistry, although in general with both
halides and O-donors zinc forms tetrahedral complexes.54
Again, in analogy to TiIV, P888O and other phosphine oxides are
commonly studied for ZnII extraction from the aqueous layer,
with [ZnX2(PnnnO)2] stoichiometry suggested, but no studies on
equimolar compositions.55 Also in this case, a tentative
structure
postulated
in
Eq
17
is
proposed,
[ZnCl(P888O)(μ-Cl)2ZnCl(P888O)], but this system lends itself
onto further study.
The summary of plausible speciation of the synthesised LCCs is
gathered in Table 3. Compared to the L-AlCl3 and L-GaCl3 LCCs,
in liquid coordination complexes of other metals molecular
species are preferred over the ionic ones.56 A possible
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 15
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
explanation might be derived from the study on the formation to form a dimeric intermediate, and the presence of sufficient
View Article Online
DOI: 10.1039/C7DT02408H
of ammonia borane, [NH3BH3], vs. diammoniate of diborane, weak interactions stabilising the opposing charges. Although,
[BH2(NH3)2][BH4].56 The formation of the ionic product was for many of the newly-reported LCCs, further studies would be
found to be less energetically favourable compared to the required to confirm the proposed speciation, the charge-
molecular adduct. In general terms, several conditions needed neutral coordination complexes, mono- or oligonuclear,
to be met en route to the charged species, including a pathway appear to be the norm.
Table 3. Suggested speciation in LCCs studied in this work, based on the available NMR spectroscopic data and the literature.
MClx
χMClx
P888
O
P888
AlCl3
0.50
2[AlCl3(P888O)] ⇄ [AlCl2(P888O)2][AlCl4]
[AlCl3L] + [Al2Cl6(P888O)] ⇄ [AlCl2(P888O)2][Al2Cl7]
2[GaCl3(P888O)] ⇄ [GaCl2(P888O)2][GaCl4]
[GaCl3(P888O)] + [Ga2Cl6(P888O)] ⇄ [GaCl2(P888O)2][Ga2Cl7]
[Ga2Cl6(P888O)] ⇄ [GaCl2(P888O)2][Ga3Cl10]
complex oligomers
2[AlCl3(P888
)
] ⇄ [AlCl2(P888
)
2][AlCl4]
0.60
0.50
0.60
0.67
0.75
0.50
0.50
0.60
0.50
0.50
0.60
0.50
0.60
0.50
0.60
[AlCl3L] + [Al2Cl6(P888)] ⇄ [AlCl2(P888)2][Al2Cl7]
GaCl3
[GaCl3(P888)]
[Ga2Cl6(P888)]
[Ga3Cl9(P888)]
complex oligomers
[InCl3(P888)]
InCl3
[InCl3(P888O)]
SbCl3
2[SbCl3(P888O)] ⇄ [Sb2Cl6(P888O)2]
oligonuclear CL-bridged complexes
[Sn2Cl4(P888O)2]
2[SbCl3L] ⇄ [Sb2Cl6L2] → redox
-
SnCl2
SnCl4
[SnCl2(P888)]
redox
[SnCl4(P888O)2] + SnCl4 ⇄ 2[SnCl4(P888O)]
[SnCl4(P888O)2] + SnCl4 ⇄ 2[SnCl4(P888O)]a
[TiCl4(P888O)]
-
TiCl4
[TiCl4(P888)2] + [Ti2Cl8(P888)] ⇄ 3[TiCl4(P888)]
[TiCl4(P888)2] + [Ti2Cl8(P888)] ⇄ 3[TiCl4(P888)]a
-
ZnCl2
[Zn2Cl4(P888O)2]
[Zn2Cl4(P888)2]
-
oligonuclear Cl-bridged complexes
a - the equilibrium reaction does not reflect the actual stoichiometry at the given composition
samples where multiple 31P NMR signals occurred marked with
asterisk.
Gutmann acceptor number of metal chlorides
AN scale for metal chlorides. Acceptor numbers of metal
chlorides followed the order: TiCl4 > SbCl5 > AlCl3 > SnCl4 >
GaCl3 ≈ InCl3 ≈ SnCl2 > ZnCl2 > SbCl3 (Figure 7).
Methodology. The literature data on ANs of metal chlorides
used in this work were incomplete. Moreover, even if the AN
values were reported, they were measured in different
solvents, for example Gutmann and co-workers used 1,2-
dichloroethane (DCE),19 but Beckett et al. selected diethyl
ether (Et2O).57 For the sake of reliable comparison, we
measured AN values for all metal chlorides used in this work,
adhering to the Gutmann’s procedure,19 whereby 31P NMR
spectra of equimolar mixtures of P222O and a metal chloride in
DCE were recorded.
Besides this work, the most comprehensive classification of
Lewis acidic metal chlorides in the literature appears to have
been provided by Kobayashi and co-workers.58 Metal chlorides
were divided into categories with respect to their catalytic
strength: A (active), B (weak) and C (inactive), and with respect
to the aldehyde/aldimine selectivity: 1 (aldehyde-selective), 2
(aldimine-selective) and 3 (neutral).58 ANs measured here
were in a relatively good agreement with the Kobayashi
classification (Figure 7), with a marked exception of SbCl3,
which is a moderate acid by AN = 61.4, but was found very
active (A-1) by Kobayashi.
Although most 31P NMR spectra featured a single peak (Figure
6 SI b), in some cases - several signals were observed (Figure 6
Si a). A weighted average of chemical shifts (taking into
account chemical shifts and peak areas) was initially
considered as the most accurate method of reporting the AN
value. However, this methodology would not be suitable for
LCCs, where partial overlap with signals from P888O was
anticipated, preventing meaningful integration of signals. Thus,
AN was calculated based on the chemical shift of the signal
with the largest peak area, related to the Lewis acidic species
of highest concentration. 31P NMR chemical shifts of P222O and
metal chloride solutions in DCE are listed in Table 1 SI, with
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

PleaDsaeltdoonnTortaandsjaucsttimonasrgins
Page 10 of 15
ARTICLE
Journal Name
Lewis acidity by Gutmann acceptor number approach
View Article Online
DOI: 10.1039/C7DT02408H
Kobayashi(scale(
A>2( A>3( B>3( A>1(
Methodology. To measure acceptor numbers of LCCs, small
quantities (1 - 3 wt %) of P222O were dissolved in neat LCCs and
31P NMR spectra were recorded (Table 1 SI). This methodology,
developed and optimised for halometallate ionic liquids, was
shown to be sensitive to their speciation (with the exception of
chloroaluminate ionic liquids).21,63
A>1(
110"
100"
90"
80"
70"
60"
50"
40"
30"
20"
10"
0"
With regards to P888O-MClx LCCs studied in this work, a
potential difficulty might had been in the signal from the
ligand (P888O) obscuring the signal of the probe (P222O). To
establish the relative difference in 31P NMR chemical shifts
between signals from P888O and P222O complexes, an LCC
based on the probe itself, P222O-AlCl3, χAlCl3 = 0.50, was
synthesised, and its 31P NMR spectrum was compared to
P888O-AlCl3, χAlCl3 = 0.50. The 31P NMR signals for P222O-AlCl3,
χAlCl3 = 0.50 (80.37 and 82.28 ppm, recorded at 55 °C) were ca.
5 ppm downfield with respect to the analogous peaks from
P888O-AlCl3 χAlCl3 = 0.50 (75.20 and 77.38 ppm - see Table 2).
This difference was expected to be sufficient for the detection
of a probe signal in P888O-MClx LCCs. Indeed, low intensity
peaks originating from the probe were easily distinguished in
all P888O-MClx LCCs, upfield from the high intensity signals from
the ligand. These signals were easily identified, as they were
not present in 31P NMR spectra of neat LCCs, and their
intensity was increasing with the increasing probe
concentration (Figure 7 SI).
"
"
"
"
"
3
l
"
"
"
"
r
3"
3
l
aC
l
C
C
ate
TiCl4
In
Al
SbCl5
SnCl4
SnCl2 ZnCl2 SbCl3
w
G
Figure 7. Acceptor Numbers of metal halides, measured as 1,2-dichloroethane
solutions, and water, compared to classification by Kobayashi.58 ANs of SbCl5 and water
cited after Gutmann and co-workers.19
)
AN values of metal chlorides measured in DCE (this work) and
in Et2O (Beckett et al.)59 are compared in Table 4. Acceptor
number for AlCl3 was independent of solvent (AN = ca. 86), in
contrast to AN values for TiCl4 and SnCl4, which were
significantly lower when measured in Et2O compared to DCE
(Table 4).
Reactions of LCCs with a base/spectroscopic probe. Acceptor
numbers for LCCs and their corresponding metal chlorides are
listed in Table 5. To facilitate the discussion, data were
additionally plotted on two separate bar chats (Figure 8).
In general, although LCCs are adducts of Lewis acidic metal
chlorides with either a hard base (P888O) or a soft base (P888),
these adducts have the same or higher AN values as neat
metal chlorides. An in-depth analysis is required to analyse the
meaning of the AN number, considering a wide variety of
Lewis acidic centres and species compared here. Bearing in
mind a very large excess of LCC compared to the probe, three
modes of interaction between LCCs and P222O may be
envisaged, as detailed below.
Table 4. Comparison of ANs measured for selected metal chlorides in 1,2-
dichloroethane solution, ANDCE (this work) or diethyl ether solution, ANEt2O (Beckett et
al.)57
MClx
TiCl4
AlCl3
SnCl4
ANDCE
108.8
85.6
ANEt2O
70
87
59
79.2
This can be justified through differences in coordination
chemistry of the three metal chlorides. Diethyl ether is known
to coordinate weakly to Lewis acidic metal centres, forming
hexacoordinate [MCl4(Et2O)2] adducts with Ti or SnIV
chlorides,60,61 and a tetracoordinate [AlCl3(Et2O)] adduct with
aluminium chloride.62 Since equimolar amounts of the
phosphine oxide probe were used, it is possible that Lewis
acidity of TiCl4 and SnCl4 was weakened and softened by the
remaining Et2O ligand (Eq 18). In contrast, in the [AlCl3(Et2O)]
adduct, ether was likely to be entirely substituted by P222O (Eq
19).
[MCl4(Et2O)2] + P222O ⇄ [MCl4(Et2O)(P222O)] + Et2O
[AlCl3(Et2O)] + P222O ⇄ [AlCl3(P222O)] + Et2O
Eq 18
Eq 19
This may suggest that using DCE as a non-coordinating solvent
may give results that are more true to ‘neat’ metal chlorides.
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 11 of 15
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
Table 5. Gutmann AN values for metal chlorides, and for LCCs based on P888 or P888
ligands, combined with a range of metal halides at different χMClx values.
O
View Article Online
DOI: 10.1039/C7DT02408H
140"
130"
120"
110"
100"
90"
180"
MClx"="1.00"
!
P888O4MClx"
085"
!MClx"="0.50""
AN
MClx
χMClx
086"
!MClx"="0.60""
neat MClx
P888
O
P888
AlCl3
0.50
0.60
1.00
85.4
94.7
83.3
101.9
80"
70"
85.6
60"
50"
40"
30"
20"
10"
0"
GaCl3
0.50
0.60
0.67
0.75
1.00
87.4
96.3
105.0
107.3
87.8
95.6
102.3
104.9
75.9
74.3
InCl3
0.50
1.00
75.3
67.5
"
3
"
"
"
"
"
l
3"
l
l
aC
C
C
G
TiCl4
In
Al
SbCl3
SnCl2
SnCl4
ZnCl2
140"
130"
120"
110"
100"
90"
80"
70"
60"
50"
SbCl3
0.50
0.60
1.00
69.3
76.2
58.3
-
P8883MClx"
61.4
75.2
SnCl2
SnCl4
0.50
1.00
74.9
77.6
0.50
0.60
1.00
79.6
77.0
-
-
79.2
108.8
66
TiCl4
0.50
0.60
1.00
109.80
-
122.1
135.9
40"
30"
20"
10"
ZnCl2
0.50
0.60
1.00
63.3
65.8
64.3
-
0"
"
"
3
"
"
"
"
"
3"
3
l
aC
l
l
C
C
TiCl4
In
Al
SbCl3
SnCl2
SnCl4
ZnCl2
G
Figure 8. ANs of metal halides (MClx, ꢀMClx = 1.00), measured as 1,2-dichloroethane
solutions, and LCCs of a general formula L-MClx, ꢀMClx = 0.50 and 0.60, measured neat.
L = trioctylphosphine oxide, P888O (top) or trioctylphosphine, P888 (bottom).
In the simplest mode, coordinationally unsaturated Lewis acids
undergo a simple acid-base addition, exemplified in Eq 20.
[InCl3L]
+
P222O → [InCl3L(P222O)]
Eq 20
Another mode of reaction between an LCC and the probe is
the substitution of a ligand in the coordinationally saturated
complex. In halometallate ionic liquids, such case reported was
only in chloroaluminate(III) system, where P222O replaced the
chloride in [AlCl4]-, owing to high oxophilicity of AlIII (Eq 21).21,63
This reactivity was anticipated for P888O-MClx LCCs (M = InIII,
SnIV or TiIV), and for P888-MClx LCCs (M = InIII or SnII), which exist
as coordinationally unsaturated monomeric complexes - see
Table 3. All bar one LCCs from this group had AN values the
same (within error bars) as the corresponding metal chloride
(Figure 8, top). The exception was [InCl3(P888)], with its AN
value lower than that of InCl3 (67.5 vs. 74.3 - see Figure 8,
bottom). It can be attributed to a very good orbital overlap
between InIII and the phosphorus, which leads to lowering
electrophilicity of this LCC. Noteworthy, AN value for another
tetracoordinate indium species, [InCl4]-, was reported at a
much lower value of AN = 57.1,21 which may be attributed to
the negative charge on the tetrachloroindate(III) anion.
Chloroindate(III) ionic liquids have been used as water stable,
mild Lewis acids.64,65
[AlCl4]-
+
P222O → [AlCl3(P222O)] + Cl-
Eq 21
The [AlCl3(P222O)] complex was formed, resulting in high
acceptor number (AN = 96), despite the fact that ionic liquids
with [AlCl4]- anion are typically catalytically inactive - in
contrast to very active ionic liquids with the [Al2Cl7]- anion.1 In
this context, high AN reported for [AlCl4]- was considered a
‘fake reading’,21,63 where interaction with the probe was not
reflected in interaction with most substrates in catalytic
reactions. In L-AlCl3 ꢀAlCl3 = 0.50 LCCs, P222O may replace either
a chloride in [AlCl4]- or L in [AlCl3L], both resulting in the
formation of [AlCl3(P222O)], and in consequence - in AN value
the same as AlCl3 (Figure 8). It is not immediately obvious
whether this value corresponds to actual catalytic activity.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins

PleaDsaeltdoonnTortaandsjaucsttimonasrgins
Page 12 of 15
ARTICLE
Journal Name
Ionic liquids based on [GaCl4]- gave low acceptor number (AN =
45),21 indicating that P222O is neither displacing the chloride,
nor does it form a pentacoordinate adduct in an easy manner.
However, ANs for L-GaCl3 ꢀGaCl3 = 0.50 LCCs are even higher
than that recorded for GaCl3 (Figure 8). It is unclear whether
the ligand (L) be replaced by P222O to form [GaCl3(P222O)], or a
pentacoordinate [GaCl3L(P222O)] is formed (Eq 20) - in either
case the AN values are unexpectedly high, especially in the
light of earlier-reported lack of catalytic activity of L-GaCl3
ꢀGaCl3 = 0.50 LCCs in carbocationic reactions.{Matuszek:2015ir,
Hogg:2015co}
Finally, the third case of P222O interaction with LCCs is the
reaction with bi- and oligonuclear complexes with chloride
bridges. They are nominally coordinationally saturated, but
bridging chloride bonds are easily broken by any base (viz.
‘masked’ Lewis acids66). This type of Lewis acidity is prevalent
in chlorometallate ionic liquids with oligomeric anions, such as
[Al2Cl7]- or [Zn2Cl6]2-.2
In LCCs, oligomeric species were postulated for LCCs based on
AlCl3, GaCl3, SbCl3 and ZnCl2 (Table 3). Furthermore, the
existence of bridging structures was suggested for P888O-SnCl2
and P888-TiCl4 LCCs.
It could be envisaged that P222O breaks the chloride bridge to
form a simple adduct, as exemplified in Eq 22. In this case, the
Lewis acidity measured by AN should be the same as for the
corresponding metal chloride, which is the case for L-SbCl3 and
L-ZnCl2 LCCs.
View Article Online
DOI: 10.1039/C7DT02408H
Scheme 1. The model Diels-Alder reaction of cyclopentadiene (CPD) and ethyl acrylate
(EA) used in this study.
Both the reaction rate and the endo:exo selectivity are
sensitive to Lewis acid strength, therefore Diels-Alder reaction
is amongst the more popular test reactions for grading
catalytic activity of Lewis acids.67 Amongst the studied LCCs, a
wide spectrum of catalytic performance was observed, from
lack of reactivity and selectivity, through to moderate activity,
to full conversion and high exo:endo selectivity (Figure 9). It
was concluded that the type of the reaction, and reaction
conditions, are indeed suitable to distinguish between
different Lewis acid strengths.
100
80
60
40
20
0
90
80
70
60
50
40
30
20
10
0
P888O$MClx"""!MClx$=$0.50$$
[Ga2Cl7]-
+
P222O → [GaCl3(P222O)] + [GaCl4]-
Eq 22
However, in other cases AN values increase from metal
chloride, to χMClx = 0.50, and further to χMClx = 60 (Figure 8).
For example, there is a marked increase in AN values for the
P888-TiCl4 LCCs, with dinuclear complexes present, but AN for
P888O-TiCl4 LCC, with mononuclear complexes, remains the
same as TiCl4. Similar behaviour was reported for
chlorometallate ionic liquids, where AN was increasing with
TiCl4 SnCl4 SnCl2 InCl3 ZnCl2 AlCl3 GaCl3 SbCl3
Figure 9. Catalytic performance of LCCs of a general formula P888O-MClx χMClx = 0.50 in
Diels-Alder reaction, in terms of conversion and stereoselectivity. Reaction conditions:
cyclopentadiene (1.586 g, 24 mmol), ethyl acrylate (1.602 g, 16 mmol), catalyst 10
mol% per EA, ice bath, 1500 rpm, 5 min, conversion and selectivity determined by GC.
ꢀ
MClx in a step-wise fashion, correlated with the appearance of
different Lewis acidic species.21,39,53 Evidently, the probe must
have the ability to detect the oligomeric species, for example
following reaction as per Eq 23.
Conversion of dienophile in the Diels-Alder reaction catalysed
with LCCs based on different metals was plotted against AN
values measured for these LCCs (Figure 10). Despite limited
number of data points, most results fit well to Boltzmann
sigmoidal curve. The same type of AN-conversion correlation
has been reported for Diels-Alder reaction catalysed with
Lewis acidic borenium ionic liquids.68 There are, however, two
outliers: P888O-AlCl3 ꢀAlCl3 = 0.50 and P888O-GaCl3 ꢀGaCl3 = 0.50,
which had no catalytic activity despite high AN values. This is
also in agreement with earlier studies, where L-AlCl3 and L-
GaCl3 LCCs at ꢀMCl3 = 0.50 were catalytically inactive, in
contrast to ꢀMCl3 > 0.50 compositions.16,17,69
2[Ga2Cl7]-
+
P222O → [GaCl3(P222O)GaCl3] + 2[GaCl4]-
Eq 23
In conclusion, numerical AN values are underpinned by the
probe partaking in several different reactions, and being in
different coordination environment. With such variety of Lewis
acids, it was interesting to investigate whether there would be
any correlation of AN with catalytic activity of the LCCs.
LCCs as catalysts in a model Diels-Alder reaction
LCCs of the general formula P888O-MClx ꢀMClx = 0.50 were
selected for the catalytic study, being the most numerous
group, with relatively easy to understand speciation. The
cycloadditon of cyclopentadiene (CPD) to ethyl acrylate (EA)
was used as the model Diels-Alder reaction (Scheme 2). It was
carried out solventless, in ice-bath, with diene in excess. The
LCC catalyst (10 mol% per EA) was dissolved in neat reactants.
12 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 13 of 15
Dalton Transactions
Please do not adjust margins
Journal Name
ARTICLE
The relationship between the speciation and LewiVsieawcAidrtiictlye O(AnliNne)
in LCCs was analysed by considering DreOaI:c1t0io.1n039o/fC7tDhTe02P420282OH
probe with species present in each LCC. To form an adduct
with a Lewis acid, the probe must undergo one of three
reactions: (1) addition to coordinationally unsaturated Lewis
acid, (2) substitution of a ligand in a coordinationally saturated
Lewis acid and (3) addition to a Lewis acidic centre by breaking
a weak bridging chloride bond. It was observed that AN values
obtained following reaction (1) or (3) were in good correlation
with the dienophile conversion in Diels-Alder reaction (used as
an alternative measure of Lewis acidity), whereas AN values
obtained through reaction (2) were overestimated.
In this scoping study, several potentially interesting catalytic
LCCs were identified. For example, the P888O-InCl3 system was
more acidic than chloroindate(III) ionic liquids (AN = 75.3 vs.
57),21 and could replace them as mild and water-stable Lewis
acidic catalyst. The P888O-TiCl4 system (albeit still moisture-
sensitive) was of much higher stability compared to neat TiCl4,
whilst retaining very high Lewis acidity, and could be used as a
safer and less hazardous alternative. In summary, this
contribution opens up new avenues of research into LCCs,
both on speciation and applications in Lewis acid catalysis.
Adj. R-Square = 0,99772
R2#(adj)#=#0.99772#
100
80
60
40
20
0
TiIV#
SnIV#
SnII#
InIII#
ZnII#
AlIII#
SbIII#
GaIII#
60
70
80
90
100
110
AN
AN#
Figure 10. Dienophile conversion in Diels-Alder reaction catalysed with LCCs of a
general formula P888O-MClx χMClx = 0.50, plotted against AN values of these LCCs.
In conclusion, when the P222O probe reacts with LCCs by a
simple addition to coordinationally unsaturated Lewis acid (Eq
20), or by breaking of the bridging halide bond (Eq 22 and 23),
the AN values correspond well to the catalytic activity of LCCs.
In contrast, in exceptional cases where the probe substitutes
the ligand (Eq 21), the AN values reported are artificially high
and do not correspond to the actual catalytic activity. This is
backed by our studies on borenium ionic liquids (containing
coordinationally unsaturated Lewis acidic centres), where very
good correlation between AN values and catalytic activity was
reported, with no outliers.68
Acknowledgements
LB acknowledges EPSRC for PhD funding. KM thanks the COST
Action (CM1206 EXIL - Exchange on Ionic Liquids) for
supporting STSM in the Queen’s University Belfast, UK, and the
National Science Centre (no. UMO-2016/20/T/ST8/00388) for
the doctoral scholarship. AC and KM acknowledge the Ministry
of Science and Higher Education, National Science Centre
(grant no. UMO-2014/13/N/ST8/00127) for funding. Solvay
Cytec are acknowledged for supplying phosphine and
phosphine oxide materials free of charge.
Conclusions
The strategy of generating liquid coordination complexes by
using ligands with octyl alkyl chains was successful for all
tested metal chlorides, and is likely to be applicable for a wide
array of metal salts. Notably, several P888-MClx LCCs underwent
a redox reaction, therefore the P888O ligand appears to be a
more robust choice.
Multinuclear NMR spectroscopy was used to gain some insight
into the speciation. Spectra of most metals were not
informative due to strong quadrupole effect. However,
building on 31P NMR spectroscopy and on the literature data, it
was suggested that most LCCs contain charge-neutral
coordination complexes. Earlier-reported L-AlCl3 and P888O-
GaCl3 systems, comprising cationic, anionic and neutral
species, appear to be the exception rather than the rule.
Furthermore, many complexes are inferred to be di- or
oligonuclear (Table 3). Further multi-technique studies are
required to confirm these preliminary conclusions.
Notes and references
‡
1
2
As per CSD search May 2017
T. Welton, Chem. Rev., 1999, 99, 2071.
J. Estager, J. D. Holbrey and M. Swadźba-Kwaśny, Chem. Soc.
Rev., 2014, 43, 847.
3
4
5
6
7
8
9
H.-P. Steinrück and P. Wasserscheid, Catal. Lett., 2015, 145,
380.
H. Olivier-Bourbigou, L. Magna and D. Morvan, Appl. Catal. A,
2010, 373, 1.
Z. C. Liu, R. Zhang, C. M. Xu and R. G. Xia, Oil Gas J., 2006,
104, 52.
Z. Liu, X. Meng, R. Zhang, C. Xu, H. Dong and Y. Hu, AIChE J.,
2014, 60, 2244.
J. Simley, https://www.uop.com/?press_release=honeywell-
uop-introduces-ionic-liquids, accessed June 2017.
F. Coleman, G. Srinivasan and M. Swadźba-Kwaśny, Angew.
Chem. Int. Ed. Engl., 2013, 52, 12582.
S. Yous, J. H. Poupaert, I. Lesieur, P. Depreux and D. Lesieur,
J. Org. Chem., 1994, 59, 1574.
Lewis acidity of both metal chlorides and corresponding LCCs
was quantified using Gutmann acceptor number approach. A
scale established for metal chlorides followed the order: TiCl4 >
SbCl5 > AlCl3 > SnCl4 > GaCl3 ≈ InCl3 ≈ SnCl2 > ZnCl2 > SbCl3 was
in a reasonable agreement with the Kobayashi scale. LCCs
10
A. G. Suárez, Tetrahedron Lett., 1999, 40, 3523.
D. W. Meek and R. S. Drago, J. Am. Chem. Soc., 1961, 83,
4322.
were found to be equally or more acidic than the 11
corresponding metal chlorides.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 13
Please do not adjust margins

PleaDsaeltdoonnTortaandsjaucsttimonasrgins
Page 14 of 15
ARTICLE
Journal Name
12
13
14
15
16
17
18
19
20
21
M. Dalibart, J. Derouault, P. Granger and S. Chapelle, Inorg.
Chem., 1982, 21, 1040.
H. M. A. Abood, A. P. Abbott, A. D. Ballantyne and K. S. Ryder, 45
Chem. Commun., 2011, 47, 3523.
P. Hu, R. Zhang, X. Meng, H. Liu, C. Xu and Z. Liu, Inorg.
Chem., 2016, 55, 2374.
Burford and A. Decken, Chem. Commun., 2012,V4ie8w, A7r9ti2cl2e .Online
DOI: 10.1039/C7DT02408H
A. Vogler and H. Kunkely, Coord. Chem. Rev., 2002, 230, 243.
44
F. Calderazzo, S. A. Losi and B. P. Susz, Helv. Chim. Acta, 1971,
54, 1156.
F. Calderazzo, S. Losi and P. B. Susz, Inorg. Chim. Acta, 1969,
3, 329.
D. Gordon and M. Wallbridge, Inorg. Chim. Acta, 1986, 111,
77.
M. F. Li and Y. K. Shan, Z. Kristallogr. New Cryst. St., 2006,
46
47
48
49
50
51
52
53
54
A. P. Abbott, J. C. Barron, K. S. Ryder and D. Wilson, Chem.
Eur. J., 2007, 13, 6495.
J. Hogg, F. M. Coleman, A. F. Ugalde, M. P. Atkins and M.
Swadzba-Kwasny, Green Chem., 2015, 17, 1831.
K. Matuszek, A. Chrobok, J. M. Hogg, F. Coleman and M.
Swadźba-Kwaśny, Green Chem., 2015, 17, 4255.
P. Hu, Y. Wang, X. Meng, R. Zhang, H. Liu, C. Xu and Z. Liu,
Fuel, 2017, 189, 203.
U. Mayer, V. Gutmann and W. Gerger, Monatsh. Chem.,
1975, 106, 1235.
I. López-Martin, E. Burello, P. N. Davey, K. R. Seddon and G.
Rothenberg, Chemphyschem, 2007, 8, 690.
J. Estager, A. A. Oliferenko, K. R. Seddon and M. Swadźba-
Kwaśny, Dalton Trans., 2010, 39, 11375.
221, 41.
Z. Zhu, W. Zhang and C. Y. Cheng, Hydrometallurgy, 2011,
105, 304.
M. Cargnello, T. R. Gordon and C. B. Murray, Chem. Rev.,
2014, 114, 9319.
A. D. Finke, D. L. Gray and J. S. Moore, Acta Crystallogr. E,
2016, 72, 35.
M. B. Alves, V. C. D. Soares, P. A. Z. Suarez and J. C. Rubim, J.
Raman Spectrosc., 2008, 39, 1388.
J. Estager, P. Nockemann, K. R. Seddon, M. Swadźba-Kwaśny
and S. Tyrrell, Inorg. Chem., 2011, 50, 5258.
J. Burgess and R. H. Prince, Zinc: Inorganic & Coordination
Chemistry, vol. 7, John Wiley & Sons, Ltd, Chichester, UK, 31st
edn. 2006.
22
23
C. L. Hussey, Adv. Molten Salt Chem., 1983, 5, 185.
J. Mason, Ed., Multinuclear NMR, Pleanum Press, New York,
1987.
24
25
J. Burt, W. Levason and G. Reid, Coord. Chem. Rev., 2014,
260, 65.
J. Burt, W. Levason, M. E. Light and G. Reid, Dalton Trans.,
2014, 43, 14600.
A. R. Barron, J. Chem. Soc., Dalton Trans., 1988, 3047.
S. P. Wicelinski, R. J. Gale and J. S. Wilkes, J. Electrochem.
Soc., 1987, 134, 262.
55
56
S. Martínez and F. J. Alguacil, J. Chem. Res., 2000, 188.
X. Chen, X. Bao, J.-C. Zhao and S. G. Shore, J. Am. Chem. Soc.,
2011, 133, 14172.
M. A. Beckett, D. S. Brassington, S. J. Coles and M. B.
Hursthouse, Inorg. Chem. Commun., 2000, 3, 530.
S. Kobayashi, T. Busujima and S. Nagayama, Chem. Eur. J.,
2000, 6, 3491–3494.
M. A. Beckett, D. S. Brassington, S. J. Coles and M. B.
Hursthouse, Inorg. Chem. Commun., 2000, 3, 530.
A. Yatsenko, L. Aslanov, M. Burtsev and E. Kravchenko, Zh.
Neorg. Khim+, 1991, 36, 2031.
S. I. Troyanov, G. N. Mazo, V. B. Rybakov and K. V. Budkina,
Koordinats. Khim+, 1990, 16, 466.
L. Hubener, H.-W. Lerner and M. Bolte, Acta Crystallogr. E,
2003, 59, m929–m930.
T. A. Zawodzinski Jr and R. A. Osteryoung, Inorg. Chem.,
1989, 28, 1710.
H. Q. N. Gunaratne, T. J. Lotz and K. R. Seddon, New J. Chem.,
2010, 34, 1821.
I. Mohammadpoor-Baltork, M. Moghadam, S.
Tangestaninejad, V. Mirkhani, A. R. Khosropour and A.
Mirjafari, C. R. Chim., 2011, 14, 568.
57
58
59
60
61
62
63
64
65
26
27
28
29
30
31
32
33
34
S. P. Wicelinski, R. J. Gale, S. D. Williams and G. Mamantov,
Spectrochim. Acta A, 1989, 45, 759.
C. Hardacre, R. W. Murphy, K. R. Seddon, G. Srinivasan and
M. Swadźba-Kwaśny, Aust. J. Chem., 2010, 63, 845.
F. Cheng, H. L. Codgbrook, A. L. Hector, W. Levason, G. Reid,
M. Webster and W. Zhang, Polyhedron, 2007, 26, 4147.
F. Cheng, A. L. Hector, W. Levason, G. Reid, M. Webster and
W. Zhang, Inorg. Chem., 2007, 46, 7215.
A. El-Hellani, J. Monot, R. Guillot, C. Bour and V. Gandon,
Inorg. Chem., 2013, 52, 506.
J. Yang, P. Tian, L. He and W. Xu, Fluid Phase Equilibr., 2003,
204, 295.
D. C. Apperley, C. Hardacre, P. Licence, R. W. Murphy, N. V.
Plechkova, K. R. Seddon, G. Srinivasan, M. Swadźba-Kwaśny
and I. J. Villar-Garcia, Dalton Trans., 2010, 39, 8679.
W. T. Robinson, C. J. Wilkins and Z. Zeying, Phys. Chem.
Chem. Phys., 1990, 219.
35
36
37
38
39
66
67
68
M. J. Ingleson, Top. Organomet. Chem., 2015, 49, 39.
G. Hilt and A. Nödling, Eur. J. Org. Chem., 2011, 2011, 7071.
K. Matuszek, S. Coffie, A. Chrobok and M. Swadźba-Kwaśny,
Catal. Sci. Technol., 2017, 7, 1045.
M. Lijewski, J. M. Hogg, M. Swadzba-Kwasny, P. Wasserscheid
and M. Haumann, RSC Adv., 2017, 7, 27558.
S. S. Chitnis, N. Burford, R. McDonald and M. J. Ferguson,
Inorg. Chem., 2014, 53, 5359.
R. M. Denton, J. An, B. Adeniran, A. J. Blake, W. Lewis and A.
M. Poulton, J. Org. Chem., 2011, 76, 6749.
K. L. Bamford, A. P. M. Robertson, H. A. Jenkins, B. O. Patrick
and N. Burford, Can. J. Chem., 2015, 93, 375.
M. Currie, J. Estager, P. Licence, S. Men, P. Nockemann, K. R.
Seddon, M. Swadźba-Kwaśny and C. Terrade, Inorg. Chem.,
2013, 52, 1710.
69
40
41
H. Arp, C. Marschner, J. Baumgartner, P. Zark and T. Müller, J.
Am. Chem. Soc., 2013, 135, 7949.
C. Gurnani, A. L. Hector, E. Jager, W. Levason, D. Pugh and G.
Reid, Dalton Trans., 2013, 42, 8364.
42
43
S. E. Denmark and J. Fu, J. Am. Chem. Soc., 2003, 125, 2208.
E. MacDonald, L. Doyle, S. S. Chitnis, U. Werner-Zwanziger, N.
14 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 15 of 15
Dalton Transactions
View Article Online
DOI: 10.1039/C7DT02408H
TOC figure: Lewis acidity of metal chlorides and their liquid coordination complexes was measured, and
related to their speciation and catalytic activity.
352x176mm (72 x 72 DPI)