The effect of structural modifications on the thermal stability, melting points and ion interactions for a series of tetraaryl-phosphonium-based mesothermal ionic liquids

Article abstract of DOI:10.1039/c7cp06278h

A family of mesothermal ionic liquids comprised of tetraarylphosphonium cations and the bis(trifluoromethanesulfonyl)amidate anion are shown to be materials of exceptional thermal stability, enduring (without decomposition) heating in air at 300 °C for three months. It is further established that three specific structural elements-phenoxy, phenacyl, and phenyl sulfonyl-can be present in the cation structures without compromising their thermal stability, and that their incorporation has specific impacts on the melting points of the salts. Most importantly, it is shown that the ability of such a structural component to lower a salt melting point is tied to its ability to lower cation-cation repulsions in the material.

Full text of DOI:10.1039/c7cp06278h

View Article Online
View Journal
PCCP
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. H. Davis, Jr, B.
Rabideau, K. N. West, E. A. Salter, A. C. Stenson, A. Wierzbicki, M. Soltani, C. G. Cassity, B. Siu, J. L.
McGeehee, K. J. Strickland and M. Vo, Phys. Chem. Chem. Phys., 2017, DOI: 10.1039/C7CP06278H.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
1 7 January 2016 Pages 1–636
Royal Society of Chemistry peer review process and has been
accepted for publication.
PCCP
Accepted Manuscripts are published online shortly after
Physical Chemistry Chemical Physics
www.rsc.org/pccp
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 1463-9076
PERSPECTIVE
Darya Radziuk and Helmuth Möhwald
Ultrasonically treated liquid interfaces for progress in cleaning and
separation processes
rsc.li/pccp

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 1 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
0.
The effect of structural modifications on the thermal stability,
melting points and ion interactions for a series of tetraaryl-
phosphonium-based mesothermal† ionic liquids
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
www.rsc.org/
Cody A. Cassity,a Benjamin Siu,b Mohammad Soltani,a Jimmy L. McGeehee,a Katie J. Strickland,a
Matt Vo,a E. Alan Salter,a Alexandra C. Stenson,a Andrzej Wierzbicki,a Kevin N. West,b Brooks D.
Rabideau,*b and James H. Davis, Jr.*a
A
family of mesothermal ionic liquids comprised of tetraarylphosphonium cations and the
bis(trifluoromethanesulfonyl)amidate anion are shown to be materials of exceptional thermal stability, enduring (without
decomposition) heating in air at 300oC for three months. It is further established that three specific structural elements –
phenoxy, phenacyl, and phenyl sulfonyl – can be present in the cation structures without compromising their thermal
stability, and that their incorporation has specific impacts on the melting points of the salts. Most importantly, it is shown
that the ability of such a structural component to lower a salt melting point is tied to its ability to lower cation-cation
repulsions in the material.
+
PPh4 (tetraphenylphosphonium; TPP) cation and a handful of
variants thereof were shown by us to be thermally stable
organo-ion salts, amidst a growing realization in the chemical
Introduction
During the 1960s, researchers from the United States Air Force
Research Laboratory, Shell Oil Company, Monsanto, and
others conducted extensive studies on the thermal stability of
small organic molecules.1-5 In broad terms, those studies
established the superior thermal stability of aromatic species
over those that are aliphatic in character, as well as the types
of heteroatom-containing groups which tend to be thermally
robust versus those which are not. Those studies were
contemporaneous with successful efforts by C. S. Marvel to
develop thermally robust polymers, the structural and
compositional elements of which closely align with those
found to be conducive to thermal stability in small molecules.6-
10 We note that several polymers invented during that period –
PEEK, PES, etc. (Figure 1) – have since become commercially
available, and demand for them is forecast to grow.11
community that typical ILs are not as stable at high
temperatures as is often asserted.13 Indeed, at elevated
temperatures common IL cations – including imidazoliums,
quaternary ammoniums, and tetraalkylphosphoniums14,15
–
decompose by anion-involved retro-SN2 or E2 reactions,13 to
which perarylated species would not be subject. Subsequent
research further validated the high thermal stability of the
parent tetraphenylphosphonium cation16 (first investigated in
an IL context by Wilkes, et al.),17,18 and even more recent work
has shown that triarylsulfonium salts can likewise exhibit
degrees of thermal and oxidative stability superior to common
ILs, as well as polymers such as PEEK and PES.19 Finally, and
most recently, we have established a hierarchy of the relative
thermal stabilities of a number of anions which are or which
might be used in creating thermally stable ILs.20
Polymers such as those stood alone for decades as the most
thermally stable organic materials known. However, in 2013 it
was shown that a small number of molten salts/ionic liquids
could be added to this list.12 Specifically, ILs based upon the
Figure 1. Repeat unit structure of thermally stable polymers
PEEK and PES.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 2 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
Our initial decision to explore perarylated onium cations as a derivatives to further explore the connection between cation
promising basis for thermally stable ILs was prompted by the core-structure functionalization, thermal stability, and salt Tm.
realization that these entities should be resistant to the two Our efforts have focussed on further elaboration of cation
most common mechanistic pathways by which IL cations structures using phenoxy appendages, but also on diarylketone
(heretofore regarded as the thermal weak point of ILs) are and sulfone groups (as access to requisite starting materials
known to undergo decomposition: de-quaternization by the permit). The latter two appendage types are likewise known
SN2 or E2 cleavage of an organic appendage from the through high-temperature polymer research to be thermally
heteroatom centre of positive charge of the onium ion in robust. Once prepared, each new salt was to be subjected to
question.13 But, frustratingly, these ions are also large and thermal stressing for a 90-day period, in air, at 300oC, and
rigid, factors that doubtlessly contribute to the relatively high assessed for mass loss and changes in chemical composition or
melting points (usually > 100 oC) their salts typically exhibit. structure (our now-standard methodology for assessing
Consequently, we made early efforts to depress Tm by grafting thermal stability).12,19,20 Here we report the results of this
alkyl substituents (in particular) onto one or more of the arene study.
rings of the parent peraryl onium cation frameworks. It was
our hope that doing so would induce Tm depression by Experimental
decreasing the cation symmetry and perhaps by increasing Materials and representative synthesis
their number of energetically accessible rotational/vibrational Commercial reagents were obtained from Aldrich Chemical,
modes. However, it quickly (and, in retrospect, unsurprisingly) Frontier Scientific, and TCI and used without further
1
became apparent that the thermal stability of these salts was purification. H, 19F, 31P, and 13C NMR were recorded on a 500
subject to the same ‘rules’ found by Marvel and others to MHz JEOL spectrometer using CDCl3 as a solvent at room
govern the thermal stability of polymers and organic small temperature. All chemical shifts for 1H and 13C NMR were
molecules; in short, aliphatic substituents were thermally reported downfield using tetramethylsilane (TMS, at
δ = 0.00
unstable at the high temperatures of particular interest to us ppm). Mass spectra were obtained using a Thermo Scientific
(>200oC). This was also true for a large number of other ion trap mass spectrometer set to positive mode ESI-MS with a
possible organic functional group types.6-10, 12
mobile phase of 100% LC-MS grade methanol. Elemental
analyses were obtained as a commercial technical service from
Atlantic Microlabs, Inc., Norcross, Georgia, USA.
Representative synthesis: TPP-p-POP∙Tf2N, IL 4
A large, thick-walled, high-pressure glass tube with a PTFE
screw-cap was charged with a magnetic stir bar and 5.00g
(0.0191 mol) triphenylphosphine, 5.00g (0.0200 mol) 4-
bromodiphenyl ether, 0.13g (0.0006 mol) anhydrous NiBr2, and
6.5 mL of ethylene glycol. The combination was agitated for
approximately 5 min using a mechanical vortex mixer to
improve the homogeneity of the reactant mixture. The tube
was then immersed in a thermostatically controlled oil bath
heated to 180oC. The degree of immersion was such that all of
the tube contents were below the level of the heated oil in the
bath.
Note that per Marcoux and Charette, careful
temperature control is important for an optimal outcome.21
The tube and its contents were heated at 180oC for 4 h after
which time the tube was carefully removed
(Caution:
dangerously hot!) from the oil bath and allowed to cool to
ambient temperature. After cooling, the tube was charged
with ca. 10 mL of saturated aqueous NaBr and ca. 15 mL of
CH2Cl2 and thoroughly agitated. After allowing the aqueous
and organic phases to separate, the aqueous phase was
removed using a separatory funnel and the organic phase
recovered and washed twice more with aqueous NaBr. After
the final separation, the organic phase was dried over
anhydrous Na2SO4, the latter removed by filtration, and the
organic phase evaporated to dryness, leaving a white solid
presumed to be the bromide salt of the desired cation (8.28g,
0.0162 mol, 84.8% yield (unoptimized).
Figure 2. Structures of cations in mesothermal ionic liquids 1-9. The counter-ion
of each of these species was Tf2N- [bis(trifluoromethanesulfonyl)amidate].
Nevertheless, by happenstance we had also prepared IL
4
(Figure 2) in our initial structure-property screening
activities.12 Despite bearing an appended phenoxyether group,
it, like the parent PPh4Tf2N salt (
thermally stable. Better still, inclusion of the phenoxyether
substituent imbued with a lower Tm. Bearing in mind
Marvel’s work, we note that the phenoxyether functional
1), proved to be quite
4
group in 4 was one observed by him to be thermally robust
when used in polymers.6-10 Considering the foregoing, it
Without further purification, the white solid from the previous
step was suspended in 50 mL of near-boiling water. While
+
seemed useful to undertake the synthesis of additional PPh4
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 2
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 3 of 13
Journal Name
View Article Online
DOI: 10.1039/C7CP06278H
ARTICLE
stirring the suspension, small portions of methanol were
added until the solid was completely dissolved. Then, to the
stirred, hot solution was added, in one portion, 50 mL of an
aqueous solution of KTf2N (5.80g, 0.0182 mol, 20% molar
excess). A white suspension formed instantly. Upon cooling
(but with continued stirring) the suspension coalesced into a
dense, viscous ivory oil. The aqueous supernatant was
decanted, the oil taken up in CH2Cl2 and dried over Na2SO4.
After removal of the latter by filtration, the CH2Cl2 was
removed using a rotary evaporator to produce a clear, viscous
ivory oil. Over the course of about a week at ambient
temperature, the oil solidified into a polycrystalline white mass
(10.84g, 0.0152 mol, 94.0%). Tm (DSC): 102.1oC. Anal. calcd.
for C32H24F6NO5PS2: C, 54.01; H, 3.40; N, 1.97. Found: C, 54.27;
Figure 3. Positive mass spectra of IL
300oC for 90 days.
7
before (A) and after (B) heating in air at
The samples were analyzed using a Thermo Scientific ion trap
mass spectrometer set to positive mode ESI-MS with a mobile
phase of 100% LC-MS grade methanol. Mass spectra of
samples before and after long term heat exposure were
compared to confirm the presence of the parent cation and
detect any compositional changes possibly caused by thermal
degradation. An example spectrum is shown in Figure 3. Any
peaks that could be degradation products were compared to a
list of precursor materials and potential molecular fragments
that could have been produced during ionization.
1
H, 3.42; N, 1.96. NMR: H (CDCl3, 500 MHz): δ 7.82-7.88 (m,
3H), (m, 2H), 7.69-7.77 (m, 6H), 7.55-7.63 (m, 6H), 7.45-7.53
(m, 2H), 7.38-7.44 (m, 2H), 7.16-7.26 (m, 3H), and 7.10-7.14
(m, 2H) ppm; 13C (CDCl3, 125 MHz): δ 164.14, 153.60,
136.45(d), 135.40(d), 133.99(d), 130.45(d), 130.24, 125.66,
120.70, 119.33(q; C-F), 118.49, 117.69(d), 108.72(d) ppm; 31P
Differential scanning calorimetry
Melting points, glass transitions and enthalpies of fusion were
measured using Differential Scanning Calorimetry (DSC) on a
TA Instruments Q2000. Entropies of fusion were calculated as
the enthalpy of fusion divided by the absolute temperature. In
this work, the melting points are reported as the transition
from a solid state (crystalline or amorphous) to the isotropic
liquid state. 1-20 mg samples were loaded into open DSC pans
and heated to 110°C for 10 minutes to remove any water
absorbed from the environment or volatile contaminants. The
samples were then cooled to -80°C and heated at a ramp rate
of 10°C/minute to 300°C. Their melting points are reported as
the melting onset temperature as calculated by the analysis
software. All measurements were carried out under a nitrogen
atmosphere. The reported values are the average of at least 3
measurements and the standard uncertainty is calculated as
the standard deviation of the mean.
(CDCl3, 202 MHz):
δ
23.22 ppm. 19F (CDCl3, 470 MHz): δ -78.61
ppm. All other new salts were prepared in an analogous
manner. Complete characterization data for them is provided
as Supporting Information.
Mass spectroscopy
Approximately 2 mg samples of each salt collected before and
after heating were added to 1 ml of LC-MS grade methanol in
microcentrifuge tubes. The samples were then agitated in an
ultrasonic bath for 15 minutes before being centrifuged for 2
minutes. Any sample that did not fully dissolve was set aside to
reach saturated concentration. All samples were diluted by a
factor of 104 by serially diluting 10
ꢀL of solution to 1 ml using
LC-MS grade methanol. After each dilution the ultrasonic
agitation and centrifugation steps were repeated.
Thermogravimetric analysis
TGA curves were measured on a Netzsch TG 209 F1 thermal
gravitational analyzer. Samples of 1-20 mg were loaded into
open ceramic TGA pans and heated at 10°C per minute from
ambient to 500°C under air. The reported decomposition
temperatures are calculated as the onset temperature (Tonset
after any solvent losses.
)
Evaluation of long-term thermal stability
Per the procedure utilized in our prior studies – one modelled
on that of Fox and co-workers22 – on longer-term IL thermal
stability, samples (0.20g – 0.50g) of each salt were charged
into new 15 mL porcelain crucibles (which were left
uncovered), then heated in a muffle furnace at 300oC in air for
90 days. At the end of the thermal stressing regimen the
samples were removed from the oven, cooled to ambient
temperature, and weighed to determine what mass loss, if
any, had occurred. Further, the appearance of each sample
was noted, and a complete new panel of NMR spectra (1H-,
13C-, 31P-, and 19F-NMR) and ESI-MS was obtained on each. The
mass loss data for the compounds is given in Table 1.
Quantum-based calculations
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 4 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
The isolated cations were optimized in the gas phase using the
B3LYP density functional method23-26 and the 6-31G* basis set
as implemented in Spartan ’08 (Wavefunction, Inc., Irvine, CA);
isodensity electrostatic potential energy maps (elstats) were
phenoxy-modified materials were ones in which two phenoxy
substituents were appended to each cation core, one of which
was p-phenoxy in all cases (using commercially sourced by
specially-made diphenyl(p-phenoxyphenylphosphine),§ while
the position of the second phenoxy group was varied from
then generated. All structures except 5-7 were confirmed as
ortho to meta to para (5, 6, and 7, respectively). It was our
stable by computing analytic vibrational frequencies. Dipole
moments were computed using Gaussian16 (Gaussian, Inc.,
Wallingford, USA), with the origin placed at the center of nuclear
charge (as is default).
expectation that the incorporation of two potentially ‘floppy’
appendages might lead to still lower values of Tm than the
inclusion of just one. Characterization data for all final
products (1H-, 13C-, 31P-, and 19F-NMR, ESI-MS, DSC, TGA, and
CHN analysis) are provided as Supporting Information.
Molecular Dynamics (MD) Methodology
Table 1. Mass loss from open containers over 90 days at 300oC versus Tonset
Liquid-phase MD simulations of IL systems were carried out
under the general AMBER force field (GAFF)27 applied to both
anions and cations. Charges were derived using the Restrained
Electrostatic Potential (RESP)27 procedure, utilizing the
Hartree-Fock model and 6-31G* basis set in Gaussian09 (v.09,
revision E.01, Gaussian, Inc., Wallingford, USA).28 Additionally,
all charges were scaled by 80%, since it was recently shown
that GAFF can accurately determine thermodynamic and
transport properties of many ILs when this is done.29 For each
system, 105 IL pairs were placed inside a cubic simulation box
with a side length of 55 Å using PACKMOL.30 Simulations were
performed in LAMMPS31 using periodic boundary conditions
and a 12 Å cut-off. Long-ranged electrostatic interactions were
calculated using the particle-particle particle-mesh (PPPM)32
technique with an accuracy of 10-4. The system was first
minimized using a conjugate gradient method until the relative
changes in the energy were less that 10-4. This was followed by
an equilibration run in the NPT ensemble at 500 K and 1 atm
for 1 ns using a time step of 1 fs. Corresponding damping
parameters of 100 fs and 1000 fs were used to maintain the
temperature and pressure, respectively. The production run
consisted of a slow cooling of the system from 500 K to 400 K
over the course of 20 ns. Thermodynamic data were collected
every 100 fs and structural data every 20 ps for later analysis.
Compound
Mass loss (%)
11.2
Tonset (oC)
416
434
440
443
a
1
2
3
4
5
6
7
8
9
25.0
47.6
9.5
22.0
21.6
a
22.2
a
10.3
447
0.0
436
a
TGA data were acquired at the request of
a
reviewer after our initial
, or 7
experiments were concluded. By that time, no sample of compound
5
,
6
remained. In addition, their synthesis requires a specialty starting material, all of
which was likewise exhausted by that point.
Thermogravimetric Analysis and Long-term Thermal Stressing
A recent study16 carried out by collaborators of ours in
Germany clearly established two key points pertinent to the
present class of ILs. First, the primary mode of thermal mass
loss for the parent salt PPh4 Tf2N-
+
(
1) is demonstrably
evaporation without decomposition. Second, when IL
1
is
thermally pushed to the point of decomposition (which occurs
at temperatures >400oC), the Tf2N- anion decomposes while
the cation remains intact. This conclusion is manifestly
apparent on the basis of the comparative decomposition
temperature of
1 versus CsTf2N, the cation of which clearly
cannot decompose. With both of these materials, a strong
mass loss onset is seen around 420oC, only ascribable under
the circumstances to the decomposition of Tf2N-. Quite
significantly, as shown in Table 1, the Tonset of the present new
ILs is in the same thermal regime. Accordingly, it seems
apparent that with them as well, anion decomposition is
responsible for mass losses occurring above ca. 400 oC.
Insofar as the 300oC thermal behaviour of bulk samples of the
materials is concerned, Table 1 indicates that the relative
thermal stabilities of the compounds varied widely, at least as
measured from the standpoint of mass loss. Out of ten total
Results and discussion
The structure of each IL cation is shown in Figure 2. All were
isolated and studied as Tf2N- salts, that anion having recently
been established in separate work to be stable at high
temperatures for long periods of time.14 Save for the parent
salt 1, one of three functional group types was tethered to the
+
various salt PPh4 cation cores: phenoxy, phenyl ketone, or
phenyl sulfone. Aided by the commercial availability of the
requisite building blocks, we were able to prepare a wider
assortment of phenoxy ether variants (
2-7) than those with
phenyl ketone (8) or phenyl sulfone (9) substituents.
materials, four had mass losses of ca. 10% or less (
1,
4,
8,
9),
The various phenoxy-appended cations, in turn, may be placed
into one of two subcategories. In the first, a single phenoxy
ether appendage was attached (respectively) in positions
ortho, meta, and para to the tetraphenylphosphonium core,
allowing us to determine what effects (if any) this type of
positional variation alone might have on the characteristics of
while others reached as high as 50% (3). However, in light of
the paper by Wasserscheid et al.16 and the TGA data for the
new ILs, it seems apparent that the open-container, bulk-
material mass losses experienced by ILs 2-9 are evaporative in
character, as opposed to being by thermal decomposition
followed by the volatilization of the ensuing decomposition
products. To further validate this possibility, samples of the ILs
the resulting salts (2, 3, and 4). The second subcategory of
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 4
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 5 of 13
Journal Name
View Article Online
DOI: 10.1039/C7CP06278H
ARTICLE
were heated to 300oC for 90 days, still under air, but in loosely
capped weighing bottles. The outcome was revealing. When
cooled and opened, there was essentially no mass loss (<2%
with any compound), as would be expected had gaseous by-
products of thermal decomposition formed and accumulated
in the container. Furthermore, NMR analysis of the materials
indicated that no changes in them occurred during the course
of the thermal stressing. We thus conclude that the mass loss
associated with these salts during the open-vessel heating was
from evaporation of ‘intact’ ions, and not by-products of their
thermal decomposition. The distinction is not trivial; such a
result would suggest that the long-term thermal stability of
these ILs in practical closed-system applications (cooling
systems, hydraulics, etc.) could be very high indeed.
1.0
3
5
0.5
0.0
6
-0.5
-100
-50
0
50
Exo Up
Universal V4.5A TA
Temperature (°C)
As might be the case with any IL, a larger liquidus range is
likely to make for a more versatile material. Consequently, we Figure 4. DSC curves for compounds 3, 5 and 6. The lines have
are eager to ascertain if we can identify structure-property
relationships that will allow us to reach lower pour point
values with future IL designs without compromising high
thermal stability. Accordingly, as was the case in a recent
study focussed on high-T-stable triarylsulfonium ILs, the
present compounds provided us another opportunity to
conduct an exhaustive thermodynamic analysis of their the
entropic and enthalpic contributions to these compounds’
melting points and to couple this analysis with insight from
quantum and molecular dynamics calculations.
been offset but not rescaled. Note that the y-axis scale is
significantly smaller than for Figures 5 and 7, due to the weak
nature of the thermal events.
7
8
9
Thermodynamics of Melting
Figure 3 shows the DSC curve for
contain the parent ion and the o-POP and p-POP substitutes
cations, respectively. In Figure 4, the DSC curves
demonstrating weak glass transitions for compounds
comprised of the m-POP ( ), o,p-di-POP ( ) and m,p-di-POP (6)
1, 2 and 4, compounds which
3
5
substituted cation are shown. Lastly, the melting phenomena
observed on the DSC curves for bistriflimide salts of the p,p-di-
POP
(7), the benzoyl (8) and the p-phenylsulfonyl (9)
Figure 5. DSC curves for compounds
been offset but not rescaled.
7, 8 and 9. The lines have
substituted cation are shown in Figure 5.
Table 2 shows the melting points and enthalpies and entropies
of fusion for compounds and ; the properties of
compound
, (Ph)4P+Tf2N-, are included as reference so that
1
1,
2,
4,
7,
8
9
1
the impact of the replacement of one of the phenyl groups
with a phenoxyphenyl [POP] moiety in the parent cation can
be assessed. With replacement of one phenyl group by o-
2
4
phenoxyoenyl [o-POP], the new salt (
~30°C higher than that of However, replacing one phenyl
group by p-POP ( ), the melting point decreases by ~33°C. We
note that compounds
replacement of phenyls by p-POP and o-POP) and
2) has a melting point
1.
4
3
(phenyl replaced by m-POP),
5
(dual
(dual
6
replacement of phenyls by p-POP and m-POP) were not
observed to exhibit a strong transition from an ordered solid
to an isotropic liquid, although weak, low-temperature glass
transitions were observed for them at 10.1°C, 20.4°C, and
7.9°C, respectively, and these varied as much as 10°C
depending on the thermal history of the sample. This
behaviour is likely due, in part, to the asymmetry of the cation
Figure 3. DSC curves for compounds
been offset but not rescaled.
1, 2 and 4. The lines have
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 6 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
frustrating solid-phase packing, resulting in the formation of a Finally, for the salt containing the p-benzoyl group (
8) the
matrix that ossifies into an amorphous solid at low melting point decreases by ~61°C relative to 1, while for the
temperatures. Interestingly, and a bit surprisingly,
p-POP groups on the central phosphorous, has a melting point increases
very close to that of the , which has only one p-POP group.
7
, with two salt with the p-phenylsulfonyl unit (
9
), the melting point
by
8°C.
4
Table 2. Melting points, enthalpies and entropies of fusion for salts 1, 2, 4, 7, 8 and 9, along with standard uncertainties as calculated by
the standard deviation of the mean.
Compound
Cation
Tm
Tm
± uc
(K)
ΔHfus
± uc
ΔSfus
± uc
(°C)
(K)
(kJ/mol)
(kJ/mol)
(J/mol K)
(J/mol K)
+
1
2
4
7
8
9
PPh4
135.0
164.7
102.1
94.8
408.2
437.9
375.3
368.0
346.8
416.2
±0.1
±0.1
±0.2
±0.8
±0.4
±1.2
32.3
50.5
33.8
34.9
19.6
22.1
±0.1
±0.4
±0.3
±0.5
±0.4
±0.7
79.2
115.4
90.1
94.9
56.5
53.1
±0.3
±0.9
±0.8
±1.5
±1.1
±1.7
PPh3-o-POP +
PPh3-p-POP+
PPh2-(p-POP)2
PPh3-Ph-p-CO-Ph+
PPh3-Ph-p-SO2-Ph+
+
73.6
143.0
As the foregoing behaviours are more complex than as to be Although most of the comparisons show significant changes in
explained by manifestly obvious structure/property both ΔHfus and ΔSfus, the changes are (paradoxically) in the
relationships, we decided to employ
a combination of same direction and nearly offset one another in driving change
thermodynamic analysis, quantum calculations and molecular in Tm.
dynamics simulations to help clarify how the structural The first entry in Table 3 shows the relative change in ΔHfus and
variations between these cations result in the observed ΔSfus when one of the four phenyl groups in the cation of
1
is
outcomes.
replaced by an o-POP group (1ꢀ2). Both increase significantly
For solid/liquid equilibrium at the melting point, the Gibbs free (56% and 46%, respectively). The increase in ΔSfus for the o-
energy of fusion is 0, and the relationship between the melting phenoxy substituted cation, even relative to the p-phenoxy
point and enthalpy and entropy of fusion is given by:
fus
substituted cation (discussed below) may be attributed to
additional asymmetry introduced because the ortho position
imposes a “kink” in the cation structure, similar to the way in
which cis-double bonds or thioether moieties impose a kink in
long chain alkanes, phenomena that our group has used in the
past to lower the melting points of lipidic ionic liquids.32,33
However, in the present materials this increase in ΔSfus is
overwhelmed by an increase in enthalpy, resulting in a melting
point increase that is, in the final analysis, enthalpically driven.
ꢁH
Tm
=
ꢁS fus
Thus, relative difference in ΔHfus and ΔSfus for a series of similar
compounds can help to reveal the influences of structural
changes on Tm. To facilitate comparisons, relative changes in
melting points are shown in Table 3 along with the ratios of
the ΔHfus and ΔSfus where in salts
2,
4
and
7 are compared to
salt 1, and 8 and 9 are compared to
4. The latter are all para
substituted, and only the link between the phenyl rings varies:
, an oxygen atom; , a carbonyl group; , a sulfonyl group.
4
8
9
Table 3. Comparison of the effects of structural variations on the thermal and thermodynamic properties of fusion for salts 2, 4, and 7
relative to 1, and 8 and 9 relative to 4, and analysis of the dominant driving force for the effects. Although a driving force can be
ascertained, each of the comparisons has enthalpy and entropy changes of the same magnitude which are nearly offsetting.
Comparison
i vs. ref
ΔTref
(K)
Tm,i/Tm,ref x 100% ΔHfus,i/ ΔHfus,ref
100%
x
ΔSfus,i/ ΔSfus,ref
100%
x
Driving Force
i
ꢀ
2 vs. 1
4 vs. 1
7 vs. 1
+29.7
-32.9
-40.2
107%
92%
90%
156%
105%
108%
146%
114%
120%
Enthalpically driven Tm increase
Entropically driven Tm decrease
Entropically driven Tm decrease
8 vs. 4
9 vs. 4
-28.5
+40.9
92%
58%
65%
63%
59%
Enthalpically driven Tm decrease
Entropically driven Tm increase
111%
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 6
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 7 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
Additional insights into the origin of the increases in ΔSfus and
ΔHfus are provided by quantum calculations and molecular
dynamics simulations (vide supra).
Conversely, when the phenoxy group is introduced para
(1ꢀ
4), there is only a modest 5% increase in the ΔHfus, likely
due to weaker cation-anion interactions interactions arising
from the presence of the p-phenoxy group. Likewise, the ΔSfus
increases by a modest 14%, thus driving the melting point
decrease (entropically). Our interpretation is that asymmetry
introduced by the addition of the p-phenoxy group likely
frustrates packing in the solid phase, a phenomenon often
associated with melting point decreases.
Notably, addition of the phenoxy group in the para position
leads to an enthalpy increase that is 10 times smaller than
substitution in the ortho position. Thus, it seems apparent
that the increase in ΔHfus for
2 relative to 1 cannot simply be
due to increased intermolecular forces brought about by
additional van der Waals interactions involving the structurally
and compositionally identical added groups.
In fact,
Figure 6. Electrostatic potential energy plots of cations
computational results (below) indicate that the dominant
interaction driving this change concerns cation-cation
corresponding to salts
1
–
9
. Common colour scale for
1
-7: 96
(red) to 387 (blue) kJ/mol; common scale for
to 387 (blue) kJ/mol.
8 and 9: -4 (red)
interactions, which are significantly more repulsive in
they are amongst the other cations. Interestingly, compound
, which is structurally similar to , save that it has two p-POP
groups, has nearly identical Tm, ΔHfus and ΔSfus values
compared to those of , indicating that the introduction of an
2 than
7
4
decrease was larger (entropically driven), the melting point
increased (ΔT ~41°C).
4
Cation geometries and electrostatic potentials
Optimized structures and electrostatic potential energy maps
additional p-POP group has little thermodynamic effect.
The second set of comparisons provided in Table 3 relate 8 and
(elstats) of cations 1-9 as determined by the B3LYP/6-31G*
9
to
type on the properties, rather than the ring position of the
linking group. Both and have a more polar linking group
than ; however, the effects of the structural alteration are
quite different and illustrate the complexity of analysing even
subtle structural changes for these species. For the
4, and serve to elucidate the effect of the linking group
model are shown in Figure 6; elstat ranges are summarized in
Table 4. Dipole moments are presented in Table 5.
+
PPh4 , the cation of the parent salt
8
9
1, is the simplest and most
4
symmetric (S4 point group), and its properties are presented as
a reference for the series. The elstat energies on the surface
4ꢀ8
of
1 range from 216 to 372 kJ/mol, with maxima near the
comparison, the replacement of the oxygen linkage with a
central phosphorous; the least-positive values are on the faces
of phenyl groups, with typical values ranging from 220 to 250
kJ/mol. The maximum elstat values are found near the
phosphorus atom in all cations of the series.
carbonyl group results in a significant melting point decrease
of ~29°C, with decreases in both ΔHfus and ΔSfus. The ΔHfus of
is 58% of that of , indicating
while the ΔSfus is 63% that of
8
4
4
that both were significantly altered. However, the enthalpy
was decreased more, resulting in a melting point decrease that
The ortho-substituted cation of
2 adopts a “tucked”
is enthalpically driven. Similarly, both the ΔHfus and ΔSfus of
decreased relative to
respectively, of those of
9
conformation (Figure 6), which hides the lone pairs of the
phenoxy oxygen, limiting their external exposure and potential
4
, with values that are 65% and 59%,
for interaction with other cations. For
2, the elstat values span
4
. Conversely, since the entropy
from 145 to 350 kJ/mol – a lower range than that of
1, but
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 8 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
with a substantially higher floor as compared to
least-positive values are located on the π faces of the distal phenyl group are analogous to that of
phenoxy moiety, where values range typically from 155 to 170 The elstat values of the para-substituted cation of
3 or 4. The phenyl groups and on the faces of the substituted central
2.
4
span a
kJ/mol for either the syn or anti face (syn/anti - relative range of 103 to 359 kJ/mol, with least-positive sites at the lone
position of the π face to the lone pair region of the oxygen). pair region of the oxygen (~105 kJ/mol) and on the syn π face
Values on the π faces of the three unsubstituted phenyl groups of the phenoxy moiety (115-135 kJ/mol). Another less-positive
range roughly from 210 to 230 kJ/mol, while the π faces of the region of the cation is located on the anti π face of the distal
substituted phenyl group on the central phosphorous fall in a phenoxy (145-155 kJ/mol). Values on the other π faces are
similar range.
analogous to that of
The dual-substituted cations
energies (ca. 96 to 343 kJ/mol) that fall below the ranges of
and The least-positive sites are at the para phenoxy
oxygens (~100 kJ/mol), followed by the meta phenoxy oxygen
of at about 110 kJ/mol. As in , the lone pair region of the
2 and 3.
5-7 have similar surface elstat
Cation
Surface Elstat Range
(kJ/mol)
Difference from 2
2
(kJ/mol)
from Salt:
3
.
Lowest
+216
+145
+111
+103
+96
Highest
+372
+350
+361
+359
+337
+347
+345
+373
+387
Lowest
+71
Highest
+22
1
2
3
4
5
6
7
8
9
6
2
-
-
ortho phenyoxy oxygen of
6
faces the interior of the cation
Typical elstat values on the
-34
-42
+11
+9
and is largely hidden.
unsubstituted phenyl groups of
than that found on cations
The ketone and sulfone moieties in respective cations
greatly impact the low end of the elstat range versus that of
. In the case of , the elstat value is minimal at the carbonyl
oxygen (16 kJ/mol); negative elstat values are found at the
sulfone oxygens of (-4 kJ/mol), suggesting that a favourable
5-7 are ~10-15 kJ/mol lower
2-4.
+97
8
and 9
+96
+16
1
8
-4
9
Table 4. Comparison of the most and least positive elstat
values on the surfaces of the four cations, along with how the
electrostatic interaction with another cation is possible
through the sulfone. The syn π faces are also less-positive
surface charges in 1, 3 and 4 compared to that of 2.
regions, with typical values of 130-140 kJ/mol on that of
120-130 kJ/mol on that of . Interestingly, the elstat values on
the π faces of the unsubstituted phenyl groups of
closely resemble those of than of the phenoxy series
Molecular Dynamics Results
9 and,
8
8
and
9
more
1
2-4.
Table 5. Dipole moments of cations.1
The structure of the different ILs was assessed using radial
distribution functions (rdf), which are displayed in Figure 7.
These functions were averaged over the range of 400-500 K, in
which range the ILs remained in the liquid state. Furthermore,
no significant variations in the values could be detected over
this range. Figure 7a shows the rdf between the phosphorous
atom of the cation and the nitrogen atom of the anion for the
different ILs. Most of the variation between the different ILs
occurs in the far shoulder of the first peak at around 9 Å, which
corresponds to the tip of the functionalized phenyl group. The
Cation from Salt:
Dipole Moment
(Debye)
0.00
1
2
3
4
5
6
7
8
9
2.25
4.79
5.09
4.47
3.91
4.79
height of this shoulder follows the following order:
2 > 1,3 > 4
7.94
> 8,9 > 7. Figure 7b shows the rdf between the phosphorous
11.76
atom of the cation and the connecting atom of the
functionalized phenyl group on the cation. This rdf can be
viewed as a form of orientational alignment of the IL cations.
1B3LYP/6-31G* model. Origin located at centre of nuclear charge.
The meta-substituted cation of
3 adopts an elongated
Salt
noticeable alignment of cations with other cations. This
ordering increases in going to and . Interestingly,
2, which does not show a first peak at 6 Å, shows no
geometry with both π faces and phenoxy oxygen lone pairs
accessible for interaction. The elstat energies range from 111
to 361 kJ/mol. The least-positive sites are located on the lone
pair region of the oxygen (~118 kJ/mol) and on the syn π face
of the phenoxy moiety, where typical values range from 115 to
130 kJ/mol. Another less-positive region is located on the anti
π face of phenoxy group, with representative values ranging
from 135 to 150 kJ/mol. Values on π faces of the three central
3,
7,
4,
8
9
this increase in cation-cation ordering is also strongly
correlated with the reduction in the Tm and reinforces the idea
that “electrostatic exposure” has a significant effect on the IL
properties. Salt
9 is the only IL that does not show this strong
correlation between cation ordering and a lowering of Tm.
However, in the following section, we show that the addition
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 8
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 9 of 13
Journal Name
View Article Online
DOI: 10.1039/C7CP06278H
ARTICLE
of the sulfonyl linker leads to a very strong cation-cation
ordering that makes it fundamentally different from the other
ILs.
(a)
(b)
Figure 7. Radial distribution functions between (a) the phospohorous center of the cation and the anion’s nitrogen and (b) the cation’s
phosphorous and the linking atom of the Ph-(R)-Ph group.
To assess the IL structures in further detail, spatial distribution significant loss of anion localization at the apex of the
functions (sdf) were constructed (see Figure 8) showing the functionalized phenyl group. This coincides with the reduction
preferred three-dimensional locations of the anion nitrogen in the rdf shoulder height in Figure 7a and further supports the
(blue) and the cation phosphorous (orange) relative to the idea that articulation of the phenoxy group might frustrate
cation. The top row shows locations that are 3.5 times the bulk order in this direction, driving down Tm. Salts
probability and the bottom row 2 times. Notice in Figure 8 that additionally show significant ordering of secondary cations
has a near perfect tetragonal structure with anions strongly (orange in Figure 8) near the functionalized phenyl group. Note
preferring to reside in the four crevices of the tetragonal that does this in a way that does not disrupt cation-anion
cation (top), and secondarily at the tips of the phenyl groups structure at the top tip of the cation, while this location is
(bottom). These two locations correspond to the first peak and completely blocked in , again suggesting a fundamentally
the far shoulder of the rdf in Figure 7a. Both and show a different liquid structure for the latter.
8 and 9
2
8
9
3
4
Figure 8. Spatial distribution functions showing the preferred locations of nitrogen atoms of the anion (blue) and the
phosphorous atoms of other cations (orange) relative to the cation. The top row show surfaces that are 3.5 times the bulk
density and the bottom row 2 times the density.
The enthalpy of the liquid state as a function of temperature Additionally, salts
for each of the ILs is given in Figure 9. Salt has the lowest
enthalpy of the cations, followed by 4, 7 and then
8
and
9
have the highest enthalpy of the ILs.
3
1,
2.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 10 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
Values in parenthesis are the van der Waals (vdw) and
electrostatic (elec) contribution to the overall pairwise energy.
Table 7 shows a breakdown of the changes in the average
pairwise interaction energies between the different ion types
relative to
1. Going from 1 to 2 is shown to be slightly
unfavourable (+3.5 kJ/mol), which agrees with the increase in
∆Hfus that is seen from experiment (+18.2 kJ/mol). The
breakdown shows that this is largely caused by an increase in
cation-cation energy (+14.0 kJ/mol) and particularly the
electrostatic contribution to this interaction (+20.4 kJ/mol). No
significant change is seen for the cation-anion interaction (-0.4
kJ/mol). The anion-anion contribution, while more favourable
(-10.1 kJ/mol – likely due to the increased lattice spacing due
to the larger cation of
2), is unable to overcome the large
increase in cation-cation energy.
Figure 9. Liquid phase enthalpy as a function of temperature
for the different salts.
Going from
there is no source of experimental confirmation for this, since
was never observed to crystallize. Regardless, this is
1 to 3 is predicted to be largely favourable, though
3
Table 6 shows the liquid phase entropy of salts 1, 2 and 4 at
predicted to be the case, driven by a strong decrease in the
electrostatic component of the cation-cation interactions.
Since the electrostatic interactions between the cations are
entirely repulsive, this can be interpreted as a strong easing of
this repulsion. In changing the phenoxy group from the ortho-
to the para- position, the electronegative oxygen atom is
pushed farther from the positive phosphorous centre,
resulting in a larger electrostatic moment. Hence, through
selected temperatures using the two-phase thermodynamic
method (see Supporting Information for details).36-38 As one
might expect, these values are larger at the higher
temperatures, but there is little difference between the three
ILs. The bulk contribution to the entropy comes from
vibrational components (>65%) and, in comparing
, appears proportional to the number of vibrational degrees
of freedom.
1 vs. 2 and
4
proper alignment, cations in
dipole moment to significantly reduce the electrostatic
repulsions, while cations in are unable to do so effectively.
4 can take advantage of this
Salt
Temp
Strans
Srot
Svib
Stotal
(K)
(J/mol K) (J/mol K) (J/mol K) (J/mol K)
2
1
2
400
300
400
500
300
400
500
168
144
169
193
141
168
193
163
138
167
187
137
164
188
629
551
761
937
557
761
946
960
832
This view agrees with the increase in the rdf peak in Figure 7b
and could explain why there is a significantly weaker cation-
1097
1318
835
cation electrostatic repulsion in
snapshots of these cation alignments are shown in Figure 10.
The MD results show that going from to is slightly
4 relative to 2. Representative
1
4
4
1093
1327
favourable (-10.7 kJ/mol), and is driven by increased
favourability of anion-anion interactions (-8.4 kJ/mol). The
experiments show only a very slight increase in ∆Hfus between
the two ILs (+1.5 kJ/mol). Again, it is reasoned that this
lessening of electrostatic repulsions experienced by the anions
is caused by increased lattice spacing resulting from the larger
cation size.
Table 6. Liquid phase entropy as a function of temperature as
computed by the two-phase thermodynamic method.
Type
1→2
1→3
1→4
(vdw,elec)
(vdw,elec)
(vdw,elec)
+0.1 (-7.9,+8.0)
+14.0 (-6.4,+20.4)
-48.0 (-7.4,-40.6)
cation-
cation
cation-
anion
anion-
anion
total
Discussion of Computational Results
For the ortho-substituted cation
calculations and molecular dynamics simulations indicate that
the phenoxy ring exists primarily in bent or tucked
(2), both quantum
-0.4 (-3.1,+2.7)
-10.1 (+1.7,-11.8)
+3.5 (-7.8,+11.3)
-0.5 (-4.8,+4.3)
-8.7 (+1.7,-10.4)
-57.2 (-10.5,-46.7)
-2.4 (-6.3,+3.9)
-8.4 (+2.0,-10.3)
-10.7 (-12.3,+1.6)
a
configuration. The tucked configuration shields the lone pair of
electrons on the phenoxy oxygen; consequently, the presented
surface area of the ortho-substituted cation is more positive
than that of either the p- or the m-substituted cations as
shown in Table 4. Hence, the cation-cation interaction is more
Table 7. Breakdown of the change in pairwise interaction
energies, given in kJ/mol-pair, between the different ILs.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 10
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 11 of 13
Journal Name
View Article Online
DOI: 10.1039/C7CP06278H
ARTICLE
+
repulsive in
more efficiently packed in the crystalline lattice than is
resulting in the greater enthalpy of fusion. For both the ortho electrostatic and structural effects that alter the enthalpy and
) and para-substituted ( ) cations, molecular dynamics entropy of fusion, and thus the melting point, in dramatically
simulations indicate liquid-phase entropies higher than that of different ways, depending on the position of the
, in agreement with the higher entropy of fusion for and has it’s
substitution.§§ While the ortho-substituted species
2
and the more compact
2
is also presumably Relative to the parent PPh4 cation in
1
the introduction of a
4,
single phenoxy group on the cation can bring about
(
2
4
1
2
4
2
observed experimentally, as to be expected for a species with phenoxy group in a tucked conformation that inhibits the
additional atoms, vibrations and rotations. Furthermore, the electrostatic exposure of the lone electron pairs on the
extended orientation of the phenoxy groups in salts
(again observed in both quantum and molecular dynamics extended phenoxy groups. This allows for significantly
calculations) support the concept that asymmetry imparted by decreased cation-cation repulsion in and slightly more
an extended phenoxy group plays a role in the increase in the cation-anion interaction in . Quantum chemical calculations
entropy of fusion, likely due to frustrated packing in the solid of the isolated cations show that has a more positive
phase in addition to increased freedom in the liquid phase. exposed surface than does or . Similarly, the results from
molecular dynamics simulations of the liquid phase show that
Taken as a whole, the results of the thermodynamic analysis, the additional phenyl ring in does, in fact, lower the Van der
3
and
4
oxygen, the meta- (3) and para-substituted (4) species have
3
4
2
3
4
2
quantum calculations and molecular dynamics simulations Waals interaction energy, as it allows for more π face
present a complex yet complimentary picture of this family of interactions, however this effect is overwhelmed by the
thermally stable molten salts. Typically, with organic ions in increase in the electrostatic cation-cation interaction, making
common ionic liquids (i.e., imidazolium or pyridinium cations), the overall liquid phase cation-cation interaction less
the addition of pendant moieties on a parent ion would be favourable (more repulsive) than in
1. This, along with its
expected to marginally increase enthalpies of fusion (due to larger cation size, serves to increase cation-cation interaction
additional van der Waals interactions) and significantly energy, which increases the liquid phase enthalpy and
increase the entropy of fusion by frustrating packing, resulting subsequently the observed enthalpy of fusion. As these
in melting point decreases. However, in this series of salts, the effects overwhelm the entropy increases arising from
addition of pendant groups to the parent cation results in asymmetry and additional atoms and bonds, the enthalpy
properties that appear to be largely governed by differences in increase becomes the driving force for the increase in melting
cation-cation interactions.
point from
Conversely, when the phenoxy group is placed in the para
position ( ), molecular dynamics simulations show that the
1 to 2.
4
increase in electrostatic energy is much smaller, likely due to
the electrostatic exposure of the lone electron pairs on the
phenoxy oxygen. This increase in energy agrees with the
quantum calculations of the isolated cation, and is virtually
offset by the favourable Van der Waals interactions as seen in
both the molecular dynamics simulations and in the
experimental enthalpy of fusion. Here, entropy dominates a
melting point decrease. The lack of an observable first-order
(a)
solid/liquid transition for the meta-substituted cation (
consistent with the overall liquid phase energy (from MD) and
isolated cation charge (from quantum calculations) of being
significantly lower than or This decrease in energy
3) is
3
(b)
1,
2
4.
relative to the other species is primarily driven by smaller
cation-cation repulsion, also indicated by both quantum and
MD calculations. This would lead to a lower enthalpy of fusion
and, coupled with the asymmetry of the cation of 3, could
result in the formation of a gel or glass rather than an ordered
solid.
Figure 10. Representative snapshots showing (a) the failure of
To summarize, the relationship between Tm and substituents
on a core phenyl ring of the parent PPh4+ cation is less obvious
than one of simple substitution position; rather, it is tied more
closely to how the substitution affects the electronics and
structure of the cation, and in turn how these affect the liquid
phase properties. In general, with cation-anion interactions
two close cations of salt
2 to form an energetic pair and (b)
three cations of salt 4 forming an oriented trimer. Blue lines
show favourable alignments between oxygen atoms and
phosphorous atoms of separate cations.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 12 of 13
View Article Online
DOI: 10.1039/C7CP06278H
Journal Name
ARTICLE
being similar among the phenoxy-substituted species (
2
,
3
and
also thanks the donors of the Petroleum Research Fund,
administered by the American Chemical Society, for support of
this research. Finally, this work was also made possible in part by a
grant of high performance computing resources and technical
support from the Alabama Supercomputer Authority.
4) the melting point decreases as the cation-cation repulsive
interactions are decreased. Species 7, which contains the bis-
para-phenoxy substituted cation has a melting point and
thermodynamic properties very similar to , with slightly
4
Notes and references
higher enthalpy and enthalpy, as would be expected from
adding another bulky phenoxy group. The nexus between the
strength of the cation-cation interaction and melting point is
§
Frontier Scientific, Inc., Logan, UT, USA.
continued with 8, where substitution is at the para position,
§§ We note that in a related work by our group (accepted for
publication pending revision), the position of substitution on the
quinoline ring in a family of (quinoline)PPh3+Tf2N- ILs plays a
critical role in the thermal stability of the individual isomers.
but the linking group is a carbonyl group rather than an oxygen
atom. Radial distribution functions between cation centres
and the linker atoms of different cations suggest that this
reduction in cation-cation repulsions is
alignment of electrostatic moments between the neighbouring
cations. But, we note that compound , which has the highest
a result of the
References
9
alignment, is fundamentally different from the other salts in its
liquid phase structure, because of an unusually strong cation-
cation orientation that prevents formation of the tetragonal
structure that is common to the other cations.
1.
G. Baum and F. R. Short, Thermal stability of organic
compounds determined by the isoteniscope method,
Technical Report No. AFML-TR-65-347, Air Force Materials
Laboratory,Wright-Patterson Air Force Base, 1966.
E. S. Blake, W. C. Hammann, J. W. Edwards, T. E. Reichard
and M. R. Ort, J. Chem. Eng. Data, 1961, 6, 87.
I. B. Johns, E. A. McElhill and J. O. Smith, Ind. Eng. Chem.
Prod. Res. Dev., 1962, 1, 2.
I. B. Johns, E. A. McElhill and J. O. Smith, Journal of
Chemical & Engineering Data, 1962, 7, 277.
J. J. Madison and R. M. Roberts, Ind. Eng. Chem., 1958, 50,
237.
2.
3.
4.
5.
6.
7.
Conclusions
The present results make it clear that the Tf2N- salts of the
+
herein-described PPh4 - based cations are materials of truly
remarkable thermal stability, being unaltered by heating, in
the presence of air, at 300oC for three months; indeed, based
upon the post-heating analysis of these salts, we have little
doubt that they would be stable indefinitely at that
temperature or perhaps ones even higher. This is especially
significant in light of the fact that, under identical conditions,
commercial samples of the ‘high-performance’ polymers PEEK
and PES were – at least by inspection – visibly more degraded
than were the present ionic liquids. In all, the new salts
described here may be the most thermally stable organic
materials heretofore described, or certainly amongst them.
This augurs well for their suitability for use in a host of high-
temperature applications.
P. E. Cassidy, Thermally Stable Polymers: Synthesis and
Properties, Marcel Dekker, Inc., 1980.
C. S. Marvel. In Proc. Stabil. Plast., Soc. Plast. Eng. Reg.
Tech. Conf.; H. Lenheis, Ed: Washington, D.C., 1967.
C. S. Marvel, J. Macromol. Sci., Part A, 1967, 1, 7.
C. S. Marvel, Pure Appl. Chem., 1968, 16, 351.
C. S. Marvel, Appl. Polym. Symp., 1973, No. 22, 47.
A. H. Tullo, Chem. Eng. News, 2016, 94 (9), 23.
C. G. Cassity, A. Mirjafari, N. Mobarrez, K. J. Strickland, R.
A. O'Brien and J. H. Davis, Chem. Commun., 2013, 49,
7590.
8.
9.
10.
11.
12.
From a more fundamental standpoint, the present findings
lead us to conclude that for relatively rigid, structurally similar
cations (roughly the same number of substituents, effective
size, sterics, etc.), increasing the electric dipole moment of the
cation allows them to align when in the liquid state, leading to
a more stable liquid state and, consequently, a lower melting
point; However, if the dipole moment introduced by the
structural modification is too strong (as is the case with
sulfonyl per the present report) significant changes take place
in the liquid state (and perhaps the solid state as well) that
worsen the thermodynamics of melting. Clearly, then,
achieving a balance between these considerations is to be
sought in their ongoing development. To that end, we have
begun the synthesis of new salts intended to further test and
refine this hypothesis, and we expect to report those results in
due course.
13.
14.
15.
16.
C. Maton, N. De Vos and C. V. Stevens, Chem. Soc. Rev.,
2013, 42, 5963.
K. J. Fraser , D. R. MacFarlane, Aust. J. Chem., 2009, 62,
309 .
C. J. Bradaric, A. Downard, C. Kennedy, A. J. Robertson, Y.
Zhou, Green Chem., 2003, 5, 143.
M. Scheuermeyer, M. Kusche, F. Agel, P. Schreiber, F.
Maier, H.-P. Steinrueck, J. H. Davis, F. Heym, A. Jess and P.
Wasserscheid, New J. Chem., 2016, 40, 7157.
C. A. Corley, R. E. Del Sesto, J. S. Wilkes, Proc. -
Electrochem. Soc., 2004, 24, 326.
R. E. Del Sesto, C. Corley, A. Robertson, J. S. Wilkes, J.
Organomet. Chem., 2005, 690, 2536.
B. Siu, C. G. Cassity, A. Benchea, T. Hamby, J. Hendrich, K.
J. Strickland, A. Wierzbicki, R. E. Sykora, E. A. Salter, R. A.
O'Brien, K. N. West and J. H. Davis, Jr., RSC Adv., 2017, 7,
7623.
17.
18.
19.
20.
A. Benchea, B. Siu, M. Soltani, J. H. McCants, E. A. Salter,
A. Wierzbicki, K. N. West and J. H. Davis, Jr., New J. Chem.,
2017, 41, 7844.
Acknowledgements
This material is based upon work supported by the National
Science Foundation under grant number CHE-1464740. JHD
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 12
Please do not adjust margins

Please do not adjust margins
Physical Chemistry Chemical Physics
Page 13 of 13
Journal Name
View Article Online
DOI: 10.1039/C7CP06278H
ARTICLE
21.
22.
D. Marcoux and A. B. Charette, Adv. Synth. Catal., 2008,
350, 2967.
E. B. Fox, L. T. Smith, T. K. Williamson, S. E. Kendrick,
Energy & Fuels, 2013, 27, 6355.
23.
24.
A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens.
Matter, 1988, 37, 785.
25.
26.
27.
28.
P. J. Stephens, F. J. Devlin, C. F. Chabalowski and M. J.
Frisch, J. Phys. Chem., 1994, 98, 11623.
S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58,
1200.
J. Wang, R. M. Wolf, J. W. Caldwell, P. A. Kollman and D.
A. Case, J. Comput. Chem., 2004, 25, 1157.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L.
Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, T. Vreven, J. A. Montgomery, J. E. Peralta, F.
Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,
V. N. Staroverov, R. Kobayashi, J. Normand, K.
Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J.
Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E.
Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K.
Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.
Dannenberg, S. Dapprich, A. D. Daniels, Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian
09, Revision E.01, Gaussian, Inc., Wallingford CT, 2009
K. G. Sprenger, V. W. Jaeger and J. Pfaendtner, J. Phys.
Chem. B, 2015, 119, 5882.
29.
30.
L. Martinez, R. Andrade, E. G. Birgin and J. M. Martinez, J.
Comput. Chem., 2009, 30, 2157.
31.
32.
S. Plimpton, J. Comput. Phys., 1995, 117, 1.
R. W. Hockney and J. W. Eastwood, Computer Simulation
Using Particles, CRC Press - Taylor & Francis Group, New
York, NY, 1998.
33.
34.
35.
M. J. Earle, J. M. S. S. Esperanca, M. A. Gilea, J. N.
Canongia Lopes, L. P. N. Rebelo, J. W. Magee, K. R. Seddon
and J. A. Widegren, Nature, 2006, 439, 831.
R. A. O'Brien, A. Mirjafari, K. M. Mattson, S. M. Murray, N.
Mobarrez, E. A. Salter, A. Wierzbicki, J. H. Davis and K. N.
West, J. Phys. Chem. B, 2014, 118, 10232.
S. M. Murray, R. A. O'Brien, K. M. Mattson, C. Ceccarelli,
R. E. Sykora, K. N. West and J. H. Davis, Jr., Angew. Chem.,
Int. Ed., 2010, 49, 2755.
36.
37.
38.
S.-T. Lin, M. Blanco and W. A. Goddard, III, J. Chem. Phys.,
2003, 119, 11792.
S.-T. Lin, P. K. Maiti and W. A. Goddard, III, J. Phys. Chem.
B, 2010, 114, 8191.
T. A. Pascal, S.-T. Lin and W. A. Goddard, III, Phys. Chem.
Chem. Phys., 2011, 13, 169
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 13
Please do not adjust margins