Ionic liquids of superior thermal stability

Article abstract of DOI:10.1039/c3cc44118k

In a head-to-head study, ILs based upon tetraarylphosphonium cations exhibit superior long-term thermal stability compared to familiar salts based upon imidazolium, quaternary ammonium, or tetraalkylphosphonium cations. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3cc44118k

View Article Online
View Journal
ChemComm
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Davis, A. Mirjafari, C. G. Cassity, N.
Mobarrez and R. A. O'Brien, Chem. Commun., 2013, DOI: 10.1039/C3CC44118K.
This is an Accepted Manuscript, which has been through the RSC Publishing peer
review process and has been accepted for publication.
Accepted Manuscripts are published online shortly after acceptance, which is prior
Chemical Communications
www.rsc.org/chemcomm
Volume 46 | Number 32 | 28 August 2010 | Pages 5813–5976
to technical editing, formatting and proof reading. This free service from RSC
Publishing allows authors to make their results available to the community, in
citable form, before publication of the edited article. This Accepted Manuscript will
be replaced by the edited and formatted Advance Article as soon as this is available.
To cite this manuscript please use its permanent Digital Object Identifier (DOI®),
which is identical for all formats of publication.
More information about Accepted Manuscripts can be found in the
Information for Authors.
Please note that technical editing may introduce minor changes to the text and/or
graphics contained in the manuscript submitted by the author(s) which may alter
content, and that the standard Terms & Conditions and the ethical guidelines
that apply to the journal are still applicable. In no event shall the RSC be held
responsible for any errors or omissions in these Accepted Manuscript manuscripts or
any consequences arising from the use of any information contained in them.
ISSN 1359-7345
COMMUNICATION
FEATURE ARTICLE
J. Fraser Stoddart et al.
Directed self-assembly of a
ring-in-ring complex
Wenbin Lin et al.
Hybrid nanomaterials for biomedical
applications
1359-7345(2010)46:32;1-H
www.rsc.org/chemcomm
Registered Charity Number 207890

ChemComm
Page 1 of 3Journal Name
Dynamic Article Links ►
View Article Online
DOI: 10.1039/C3CC44118K
Cite this: DOI: 10.1039/c0xx00000x
www.rsc.org/xxxxxx
ARTICLE TYPE
Ionic liquids of superior thermal stability
Cody G. Cassity,a Arsalan Mirjafari,b Niloufar Mobarrez,a Katie J. Strickland,a Richard A. O’Brien,a and
James H. Davis, Jr.*a
Received (in XXX, XXX) Xth XXXXXXXXX 20XX, Accepted Xth XXXXXXXXX 20XX
5
DOI: 10.1039/b000000x
accompanied by decomposition.6
In
a head-to-head study, ILs based upon tetraaryl-
50 Believing these observations to signal an exceptional level of
robustness for TAP salts, we prepared an initial fourꢀmembered
set (1 ꢀ 4) of ILs.7 Three of these manifested (gravimetrically)
superior inꢀair stability at high temperatures when benchmarked
in sideꢀbyꢀside experiments against representative imidazolium,
55 tetraalkylammonium, and tetraalkylphosphonium ILs.
phosphonium cations exhibit superior long-term thermal
stability compared to familiar salts based upon imidazolium,
quaternary ammonium, or tetraalkylphosphonium cations.
10 A number of prospective applications of ionic liquids – as heat
transfer fluids, for example1 – require or might benefit from their
use at high temperatures. Consequently, it is little wonder that an
oftꢀstressed attribute of these salts is their stability under such
conditions. And, while it is true that numerous ILs have been
15 reported to be stable in their liquid form to temperatures of 400
oC or higher – these claims must be critically assessed with
respect to several factors.2 Because this parameter is routinely
established by thermogravimetric analysis (TGA), it is vital to
note that the sample purity, rate of heating, and the nature of the
20 atmosphere over the sample (air or inert gas) can all markedly
affect the experimental outcome.2,3 For example, as recently
noted by Maton, De Vos, and Stevens, the decomposition
temperature of the common IL [EMIM]Tf2N has been reported as
o
o
o
o
o
410 C, 419 C, 439 C, 440 C, and 455 C.2 However, using
25 isothermal gravimetric analysis, MacFarlane et al. showed that, in
air, there is already 1% mass loss of [EMIM]Tf2N in only 1 h at
295 oC, and in 10 h at 234 oC.3a
Figure 1. Structures of the cations in ILs 1 through 5. The anions of all
new and reference ILs is Tf2Nꢀ
In this light, it seems that many ILs – especially those for which
TGA data was reported some time ago – may be far less
30 thermally stable than believed, as discussed in the Maton, et al.,
review.2 Overall, it seems clear that in the final analysis,
establishing the practical thermal stability of an IL comes down
to observing its response to conditions of real duress: holding
macroscopic samples of it at high temperatures, for prolonged
35 times, and in air, all conditions to which it might be exposed in a
realꢀworld application.
Building upon our longstanding experience in designing ions for
ILs with specific properties,4 we take up here the question of
whether it might be possible to design or identify molecular ions
40 durable enough for extended highꢀtemperature use. Taking into
account what is known about the mechanisms by which ILs
decompose, we investigated tetraarylphosphonium cations in the
present study. Our choice was guided by the longstanding use of
the parent tetraphenylphosphonium (TPP) ion to form stable salts
45 with even highly reactive organometallic anions, suggesting a
utilitarian degree of chemical inertness on its part.5 Further, we
noted that the high melting points of known TPP and related
tetraarylphosphonium (TAP) salts are rarely denoted as being
As noted, four TPP/TAP salts were initially prepared. The first,
60 TPP[Tf2N], was described earlier but the only thermophysical
o
data published for it are dissonant values of Tm (160 C8a,b and
137 oC8c).8d That we could find, only one other ILꢀfocused paper
made use of a TAPꢀtype salt, in that case not as an IL proper but
as a transition state analog of a reactive intermediate.9
65 The basis for the inclusion of 1 in the present study is selfꢀ
apparent, the cation being the simplest embodiment of the
tetraarylphosphonium genre. The second salt, 2, was selected to
probe what thermal impact (if any) there might be from having an
alkyl group appended to the parent cation core. In a similar vein,
70 3 was studied to determine the effect of having a heteroatom
substituent tethered to the basic cation core. Finally, in order to
determine the influence of having a heteroatom in one of the
aromatic rings, we prepared 4.
The syntheses of the salts were straightforward. [TPP]Tf2N (1)
75 was prepared from commercially available [TPP]Br by
metathesis using KTf2N in water. On mixing hot aqueous
solutions of these, the product instantly precipitates as a white
solid which is isolated by filtration, then recrystallized from
acetone. The related TAP salts (2ꢀ4) were all prepared first as
This journal is © The Royal Society of Chemistry [year]
[journal], [year], [vol], 00–00 | 1

ChemComm
Page 2 of 3
View Article Online
DOI: 10.1039/C3CC44118K
bromides by the method of Marcoux and Charette (Eqn. 1), then 35 425 oC over a period of 2 h (3.5 oC/min). Based upon the
subjected to anion exchange in the same fashion.10 They too
calibration study, at this Tstage the crucible contents (ILs) would
be at 350 oC. Samples were maintained at that temperature for 48
hours then cooled, weighed, and returned to the heat stage for
another 48 hours (96 hours of thermal stressing in toto); Note that
40 most IL thermal studies described as “long term” are of 6ꢀ20 h
duration. Samples were reꢀpositioned on the stage every 12 h to
offset any chance of uneven heating due to localized hot spots or
unusual air flow dynamics around the vessels. Within the molten
salts, Marangoni flow was often clearly apparent.
form roomꢀtemperature crystalline solids upon recrystallization.
Characterization of each was straightforward, and consisted of
1
5
DSC, Hꢀ, 13Cꢀ, and 31PꢀNMR. All were consistent with those
expected for the proposed structures and compositions.
45 The results of the extended time, inꢀair, highꢀT treatment are
summarized in Table 1. Except for methoxyꢀsubstituted 3, the
new TPP/TAP based ILs were remarkably stable as gauged by
mass loss, decreasing in each case by 4.3% or less during the 96 h
experiment. In contrast, the reference ILs incorporating [EMIM],
50 [BMIM], [P4444], and [Pyrr14] cations were all fully decomposed,
with > 90% mass losses in the same period. The extensive
decomposition of the reference ILs was visually apparent well
(Figure 2). Indeed, by the time 60 h had elapsed, the three Nꢀ
based standards consisted of only charred solids, while [P4444] had
55 been reduced to solids in only 12 h. In contrast, 1 and 4 returned
to their crystalline solid state (albeit somewhat discolored) upon
cooling. After 48 h, 2 had become orangeꢀbrown, and thereafter
exhibited no further change in appearance. On cooling, it became
a mechanically tough glass.
Thermogravimetric analyzers are generally designed to evaluate
milligramꢀsized samples, one sample at a time. However, the
10 present analyses were carried out open to the atmosphere on a
digital heat stage with integrated platinum RTD sensing, the latter
reliably providing stage temperature control to ±1% of the set
point. We embraced the digital heat stage for the present study
because it (unlike commercial TGA devices) allowed multiple,
15 macroscopic (1+ g) samples of ILs to be evaluated sideꢀbyꢀside
under identical conditions of heating and atmosphere/humidity/air
flow, such as might occur in a realꢀworld setting. Consequently,
the relative thermal stabilities of the ILs would be established
with a high degree of confidence, regardless of any fluctuations
20 in stage T or air flow around the apparatus that might occur.
Nevertheless, cognizant that the indicated stage temperature
might not reflect the temperature of the crucible contents due to
heat transfer limitations, calibration experiments were performed
that allowed us to establish that the contents reach thermal
25 equilibrium 75 oC lower than the stage set temperature.11
Table 1. Mass loss from ILs7 under longꢀterm thermal stress
ILa,b
Tm (oC)
Mass
loss
(%)c
48 h
1.8
Mass
loss
(%)c
96 h
2.3
PreꢀHeating
Appearance
(25 oC)
Postꢀ
Heating
Appearance
(25 oC)
Ivory
crystals
Ivory
crystals
Orangeꢀ
brown glass
Brown
glass
Brown
1
135.07
135.07
135.33
Colourless
crystals
Colourless
crystals
Colourless
crystals
Colourless
crystals
60 Figure 2. A picture worth a thousand words. One gram samples (at room
temperature) of three ILs after heating in new porcelain crucibles for 96 h
in air at 350 oC. l to r : [EMIM]Tf2N and [Pyrr14]Tf2N – nothing but black
solids remain of what began as colourless liquids; [TPP]Tf2N – only
slightly changed from its original white colour.
1c
0.5 d
2.1
0.8d
3.1
2
3
4
≈178
(dec.)
107.38
32.9
0.7
58.4
3.9
65 Significantly, the NMR spectra of 1 and 4 after the 96 h period
were unchanged, further indicating their stability under the
grueling conditions to which they were subjected. However, the
spectra of 2 – despite its very low mass loss – show differences
suggestive of a structural change to a portion of the cations that is
70 consistent with ionomer formation, an outcome worth separate
study. Finally, 3 was alone among the new ILs in exhibiting
extensive mass loss under the experimental conditions. Indeed, it
Colourless
crystals
crystals
5
7.72
(Tg)
657b
2.7
4.3
Ivory resin
Ivory resin
[P4444
]
86.4
81.6
76.9
35.1
94.5
92.4
97.8
93.7
Colourless
crystals
Colourless
liquid
Colourless
liquid
Colourless
liquid
Black char
Black char
Black char
Black char
[Pyrr14]
[EMIM]e
[BMIM]
ꢀ18.112
ꢀ33.113
ꢀ1.114
o
begins to show evidence of decomposition as low as T = 178 C.
Even so, its total mass loss (58.4%) was still less than those of the
75 benchmark ILs.
It concerned us that the most thermally vulnerable of the new ILs
were those with exo substituents. Accordingly, we prepared an
additional IL, 5, to further probe whether having any exoꢀring
substituent off of a TPP cation core would lead to lower thermal
80 stability; if so, options for designing additional highꢀT stable TPP
variants might be quite limited. So, in 5 a phenyl ether
substituent was grafted onto a TPP core in place of the methyl
ether of 2. Our choice was based on the knowledge that diphenyl
a All as Tf2Nꢀ salts. b Each IL was halideꢀfree (negative silver ion test). c
At 425 oC stage temperature in new porcelain crucibles (contents at 350
oC). d Platinum crucible. e Commercial material (Iolitec).
30
Samples of ILs 1ꢀ4, as well as the three reference ILs (all ca. 1 g)
were charged into new porcelain crucibles and placed on the heat
stage at room temperature. The stage temperature was ramped to
2
| Journal Name, [year], [vol], 00–00
This journal is © The Royal Society of Chemistry [year]

Page 3 of 3
ChemComm
View Article Online
DOI: 10.1039/C3CC44118K
ethers are among the most thermally stable organics known. We
were pleased to observe that when 5 was subjected to the same 96
h / 350 oC stress, its mass loss was quite low like those of 1 and 4,
and its 1Hꢀ and 31PꢀNMR spectra were unchanged.
administered by the American Chemical Society, for support of
this research. He also thanks Prof. Andre Charette for useful
clarifications concerning the synthesis of TAP bromides.
5
Encouraged by the results of the prolonged timeꢀatꢀT
experiments, we made an effort to ‘push the envelope’ of ILs 1
and 5 to the thermal limit of our present equipment. Accordingly,
samples of these were placed directly on the surface of the stage,
Notes and references
45 a Department of Chemistry, University of South Alabama, Mobile,
Alabama 36688 USA. Fax: 251-460-7359; Tel: 251-460-7427; E-mail:
jdavis@southalabama.edu
b Department of Chemistry and Mathematics, Florida Gulf Coast
University, Fort Myers, Florida 33965 USA. E-mail:
50 amirjafari@fgcu.edu
o
heated to 450 C, and kept at that temperature for 24 h. Under
10 these conditions, both salts are freeꢀflowing liquids of low
apparent viscosity and with good wetting properties. They
discolored only slightly, to about the same degree as was
observed in the 96 h test. Further, on cooling we were able to
recover more than 90% of the mass of each IL from the test
15 surface, and we believe that most of the mass lost was due to the
mechanics of their recovery. Most significantly, the NMR
spectra of 1 and 5 were unchanged from those of the pure, preꢀ
stressed salts (Figure 3).
† Electronic Supplementary Information (ESI) available: [NMR data;
remarks parenthetical to the main text]. See DOI: 10.1039/b000000x/
1
2
3
E. B Fox, A. E. Visser, N. J. Bridges, J. W. Amoroso, Energy
& Fuels, 2013, 27, 3385.
C. Maton, N. De Vos, C. V. Stevens, Chem. Soc. Rev., 2013,
doi: 10.1039/c3cs60071h
(a) K. J. Baranyai, G. B. Deacon, D. R. MacFarlane, J. M.
Pringle, J. L. Scott, Aust. J. Chem., 2004, 57, 145; (b) M.
Kosmulski, J. Gustafsson, J. B. Rosenholm, Thermochim.
Acta, 2004, 412, 47; (c) D. M. Fox, W. H. Awad, J. W.
Gilman, P. H. Maupin, H. C. De Long and P. C.
Trulove, Green Chem., 2003, 5, 724.
55
60
4
5
See: Z. Fei, P. J. Dyson, Chem. Commun., 2013, 49, 2594.
The tetraphenylphosphonium ion was first reported in 1928,
and its first use as an adjunct for crystallization in 1940. See:
(a) J. Dodonov and H. Medox, Berichte, 1928, 61B, 907; (b)
H. V. Medoks, A. V. Krylov, Zhur. Priklad. Khim., 1940, 13,
1529. (c) M. Sefkow, N. Borsuk, M. O. Wolff, Chimica Oggi,
2001, 19, 19.
65
70
6
7
See for example: T. Migita, T. Nagai, K. Kiuchi, M. Kosugi,
Bull. Chem. Soc. Jpn., 1983, 56, 2869.
As shown in Table 1, the melting points of 1ꢀ4 are > 100 C.
o
Although it is common to describe ionic liquids as salts that
are molten at temperatures < 100 oC, this demarcation is
wholly arbitrary; While its use is ocassionally convenient, it is
not defining.
75
8
(a) C. A. Corley, R. E. Del Sesto, J. S. Wilkes, Proc.
Electrochem. Soc., 2004, 24, 326; (b) R. E. Del Sesto, C.
Corley, A. Robertson, J. S. Wilkes, J. Organomet. Chem.,
2005, 690, 2536; (c) Weber, C. C., Masters, A. F.,
Maschmeyer, T., J. Phys. Chem. B., 2012, 116, 1858.
C. C. Weber, A. F. Masters, T. Maschmeyer, J. Phys. Chem.
B, 2012, 116, 1858.
80
20
Figure 3. Preꢀ (inset) and postꢀstressing (24 h, 450 oC direct heat) 31Pꢀ
NMR spectra of 1. Note that there is no change in the cation peak, and
that no Pꢀcontaining impurities are in evidence.
9
85
10 (a) F. Stazi, D. Marcoux, J.ꢀC. Poupon, D. Latassa, and A. B.
Charette, Angew. Chem. Int. Edit., 2007, 46, 5011; (b) D.
Marcoux, and A. B. Charette, J. Org. Chem., 2008, 73, 590;
(c) D. Marcoux, and A. B. Charette, Adv. Syn. Catal., 2008,
350, 2697: (d) A. B. Charette, J. C. Poupon, A. Boezio (2005)
PCT WO 2005097812; A1 20051020.
11 With 1g samples, the IL depths are too shallow to probe with
thermocouples. Increased samples sizes help, but at the
expense of increased sampleꢀsurface heat loss as well as
increased heat conduction issues in the samples (see refs. 1
and 2). IR thermometry was evaluated as an alternate means
to gauge T, but the area sampled was too large to confidently
assign the readout temperatures to the small sample areas
inside the crucibles.
12 D. R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, J.
Phys. Chem. B., 1999, 103, 4164.
13 H. L. Ngo, K. LeCompte, L. Hargus, A. B. McEwen,
Thermochim. Acta, 2002, 357ꢀ358, 97.
14 D. R. MacFarlane, D. R. Meakin, P. Amini, M. Forsyth, J.
Phys. Condens. Matter, 2001, 13, 8257.
We believe the present results support characterizing the new
TAP and TPP salts as the most, or certainly among the most,
25 thermally stable ionic liquids reported thus far. Their superiority
in this regard over a selection of standard ILs has been
quantitatively demonstrated, and it is visually apparent. Even so,
a focused effort is in order to determine if the liquid range of this
IL class can be generally broadened by decreasing their melting
30 points, but without sacrificing their high upper limits of thermal
stability. Already, the unexpectedly low Tg of 5 (7.72 oC),
combined with its stability to at least 450 oC, indicates that this is
a plausible objective. Regardless, it bears noting that the thermal
stability of the new ILs already rivals or may even exceed that of
35 some of the most advanced commercial heat transfer fluids, such
as Dow Syltherm 800® (rated to 400 oC). As a consequence,
further development of TPP/TAPꢀbased ionic liquids, especially
for use in highꢀT applications, is clearly warranted.
90
95
100
105
Acknowledgements
40 JHD thanks the donors of the Petroleum Research Fund,
This journal is © The Royal Society of Chemistry [year]
Journal Name, [year], [vol], 00–00 | 3