Ionic liquids: Differential scanning calorimetry as a new indirect method for determination of vaporization enthalpies

Article abstract of DOI:10.1021/jp300834g

Differential scanning calorimetry (DSC) has been used to measure enthalpies of synthesis reactions of the 1-alkyl-3-methylimidazolium bromide [C _nmim][Br] ionic liquids from 1-methylimidazole and n-alkyl bromides (with n = 4, 5, 6, 7, and 8). The optimal experimental conditions have been elaborated. Enthalpies of formation of these ionic liquids in the liquid state have been determined using the DSC results according to the Hess Law. The ideal-gas enthalpies of formation of [C_nmim][Br] were calculated using the methods of quantum chemistry. They were used together with the DSC results to derive indirectly the enthalpies of vaporization of the ionic liquids under study. In order to validate the indirect determination, the experimental vaporization enthalpy of [C₄mim][Br] was measured by using a quartz crystal microbalance (QCM). The combination of reaction enthalpy measurements by DSC with modern high-level first-principles calculations opens valuable indirect thermochemical options to obtain values of vaporization enthalpies of ionic liquids. ? 2012 American Chemical Society.

Full text of DOI:10.1021/jp300834g

Article
pubs.acs.org/JPCB
Ionic Liquids: Differential Scanning Calorimetry as a New Indirect
Method for Determination of Vaporization Enthalpies
Sergey P. Verevkin,*,† Vladimir N. Emel’yanenko, Dzmitry H. Zaitsau, and Ricardas V. Ralys
Department of Physical Chemistry, University of Rostock, Dr-Lorenz-Weg. 1, 18059 Rostock, Germany
Christoph Schick†
Department of Physics, University of Rostock, Wissmarsche Str. 43-45, 18057 Rostock, Germany
S
* Supporting Information
ABSTRACT: Differential scanning calorimetry (DSC) has been used to
measure enthalpies of synthesis reactions of the 1-alkyl-3-methylimidazo-
lium bromide [Cnmim][Br] ionic liquids from 1-methylimidazole and
n-alkyl bromides (with n = 4, 5, 6, 7, and 8). The optimal experimental
conditions have been elaborated. Enthalpies of formation of these ionic
liquids in the liquid state have been determined using the DSC results
according to the Hess Law. The ideal-gas enthalpies of formation of
[Cnmim][Br] were calculated using the methods of quantum chemistry.
They were used together with the DSC results to derive indirectly the
enthalpies of vaporization of the ionic liquids under study. In order to
validate the indirect determination, the experimental vaporization enthalpy
of [C4mim][Br] was measured by using a quartz crystal microbalance
(QCM). The combination of reaction enthalpy measurements by DSC
with modern high-level first-principles calculations opens valuable indirect
thermochemical options to obtain values of vaporization enthalpies of ionic liquids.
management for the reactor design and operation is a crucial
point leading to high quality of the resulting product and pre-
venting of a thermal runaway. Thus, a systematic accumula-
tion of reaction enthalpy data of ILs synthesis is apparently
required and calorimetry is a suitable experimental tool for this
purpose.
Differential scanning calorimetry (DSC) is widely used for
studies of heat capacities, enthalpies of phase transitions
(fusion, mesomorphic, vaporization), and chemical reactions,
etc.4 This technique offers important advantages, such as the
small quantity of sample necessary for the experiment (a few
milligrams), handiness of manipulation, rapidity of measure-
ment, versatility, and precision of temperature control. DSC is
also very useful in the study of physical-chemical processes,
such as the determination of the kinetics and thermodynamics of
polymerization,5 due to the fact that these types of reactions are
exothermic and so they can be easily followed isothermally and
dynamically by DSC. This study reports an elaboration and
development of a DSC procedure for reliable measurements of
reaction enthalpies of ILs synthesis.
1. INTRODUCTION
Ionic liquids (ILs) exhibit unique physical-chemical properties
such as negligibly low vapor pressure at ambient temperatures
and thermal stability at elevated temperatures. ILs might be
useful as environmentally benign solvents that could replace
volatile organic compounds. ILs have also widespread potential
for practical applications such as heat transfer and storage
medium in solar thermal energy systems as well as for many
other areas such as fuel cells and rechargeable batteries. By
varying the length of the alkane chains of the cationic core and
the type or size of the anion, the solvent properties of ILs can
be tailored to meet the requirements of specific applications to
create an almost infinite set of “designer solvents”. Systematic
studies of thermodynamic properties within ionic liquid homo-
logical series (e.g., imidazolium-, pyridinium-, ammonium-, or
pyrrolidinium-based ILs) are of interest, because they are
required for the validation and development of the molecular
modeling1 and first-principles methods2 toward this new class
of solvents.
Currently, the most widely used starting materials for
imidazolium-based ionic liquids are imidazolium halides (mostly
bromides and chlorides). The latter are commonly synthesized by
a quaternization reaction of a substituted imidazole with alkyl
halides. In most cases the IL synthesis reactions are highly
exothermic and the kinetics was shown to be fast.3 The heat
Received: January 25, 2012
Revised: March 12, 2012
Published: March 21, 2012
© 2012 American Chemical Society
4276
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
Figure 1. A cycle of the thermochemical measurements and calculation leading to the vaporization enthalpies of ILs.
In this work, the enthalpies of reaction of 1-methylimidazole
with the linear butyl-, pentyl-, hexyl-, heptyl-, and octyl bromides:
corresponding to this exothermic heat effect was then inte-
grated over the time range of the effect and a total enthalpy of
reaction was determined from the area of the DSC peak. The
heat released during the IL synthesis reaction 1 was related to
the molar amount of a stoichiometrically deficient reactant in
the starting formulation.
1‐Me‐imidazole + CH − (CH )
2 n−1
− Br = [C mim][Br]
3
n
(1)
were determined in the liquid phase with a commercially availa-
ble DSC. Using these results, the molar enthalpies of formation,
ΔfH°m(liq), of the [Cnmim][Br] family were derived according
to the Hess law. Experimental values of ΔfH°m(liq) together
with the gaseous enthalpies of formation ΔfH°m(g) of these ILs
calculated by first-principles composite CBS-QB3 method were
used to derive enthalpies of vaporization of [Cnmim][Br]
indirectly. In order to validate this indirect procedure, the
enthalpy of vaporization of [C4mim][Br] was also measured
using a quartz crystal microbalance (QCM). A cycle of the
thermochemical measurements and calculation leading to the
vaporization enthalpies of ILs is presented in Figure 1.
In the present work, the enthalpy of reaction was measured
using a Mettler-Toledo 822 heat flux DSC with a MultiSTAR
FRS5 sensor and a mechanical refrigerating system. Before mea-
surements on the ILs, two standard materials, indium (99.999%
pure) and zinc (99.999% pure), were used to calibrate the tem-
perature and heat flow scales of the DSC.
ILs are most commonly synthesized by a quaternization
reaction of a substituted amine (or phosphine, N-heterocycles,
etc.) with alkyl halides, alkyl sulfates, alkylmethanesulfonates,
etc. As a rule, such N-alkylation reactions are highly exothermal
and very fast even at room temperature. For example, the
reaction of MeIm with trifluoromethanesulfonate7 at room
temperature is extremely fast, and also the reaction of MeIm
with diethyl sulfate is fast,8 but the reaction of MeIm with butyl
chloride is very slow.9 As a consequence, two different pro-
cedures have to be developed to prepare samples for DSC
studies, depending on the reaction rate at room temperature.
The first one was applied for the slow reactions, and the second
procedure requiring additional precautions in order to preclude
any reaction before equilibration of the DSC measuring cell.
2.2.1. Slow Reactions. Since the reaction rates at room
temperature were considered as negligible, the reactants can be
mixed without any precautions. Typically, 6−15 mg of starting
materials was placed in a standard 40 μL aluminum DSC pan
which was weighed with an electronic balance with the resolu-
tion of 0.000 005 g. Air buoyancy corrections were made,
using known densities of materials. Charging of the pan with
the starting materials was convenient to perform with one-way
syringes. The pan was hermetically sealed with a cover to
eliminate any evaporation of the reagents. The pan was
weighed before and after each experiment to detect any mass
loss during the measurements. Experiments with an occasional
mass loss were rejected. Measurements were performed in the
2. EXPERIMENTAL SECTION
2.1. Materials and Chemicals. Samples of 1-methylim-
idazole (MeIm) and n-alkyl bromides (R-Br) were of
commercial origin, and they were distilled under reduced
pressure prior to use and stored under dry nitrogen. The degree
of purity was determined using a gas chromatograph (GC)
equipped with a flame ionization detector. No impurities
(greater than 0.05 mass %) could be detected in the samples
used for the measurements. The purity of the reactants (es-
pecially water traces) was decisive for the quality of the
measurements by any method, but it was of crucial importance
for the DSC method because of the very tiny amounts of
samples used. The filling station for the DSC pans was in a
glovebox under protecting nitrogen atmosphere.
2.2. Differential Scanning Calorimetry: Measurements
of Reaction Enthalpy. The general principles for the heat-flux
DSC measuring the enthalpies of chemical reactions are well
documented.4,6 In short, in order to obtain a reaction heat ef-
fect in a dynamic temperature scan mode, the temperature
difference between a sample pan and an empty reference pan
was measured and converted into a heat flux value. The peak
4277
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
range 273−500 K with scanning rates between 10 and
50 K·min−1.
DSC experiment independent of the molar ratio of the
reactants.
2.2.4. Solvent or Solventless? Commonly, the quaterniza-
tion reactions are carried out in batch reactors. Organic solvents
(e.g., toluene, ethanol) have often been used in order to
improve the heat removal, to increase safety of the process, and
to obtain high quality ILs. Also, many ILs or their intermediates
can often be quite viscous. By using a solvent, the product can
be kept in a relatively low viscosity liquid solution that can be
more easily processed.11 However, the solvent use leads to in-
creased reaction times due to the dilution and requires quanti-
tative removal in an additional process step.
2.2.2. Fast Reactions. Due to the fact that for fast reactions
part of the heat could be released already by charging of the pan
at room temperature, the reactants have to be thoroughly
separated inside the pan before the onset of the measurement.
Any separation membrane or any other mechanical facility is
hardly possible to install inside of the tiny (40 μL) aluminum
pan. In preliminary experiments, a separation of reactants in the
pan with a wax baffle has heavily aggravated the acquisition of
the reaction peak because of an additional contribution from
fusion of the wax. This idea was rejected. Instead, silicone oil
has turned out to be a very practical thing for the reactants'
separation. Indeed, the silicone oil is nonvolatile, chemically
inert, and thermally stable, and, as a rule, it has negligible mis-
cibility with chemicals common for the ionic liquids synthesis.
Moreover, different types of commercially available silicone oils
have broad variations of densities. Hence, it is easy to find a
specific silicone oil with the density just between the values of
the two starting reactants. Thus, the thorough separation of the
reactants in the pan was achieved according to the “sandwich”
layers having the denser compound on the bottom of the pan,
the silicone oil layer in between, and the second reactant as the
upper layer. In the course of the experiment, heating of the pan
inside the DSC led to an intensive mixing of the ingredients
due to thermal convection. In such a way, a controlled reaction
start and a measurement of the total heat release was achieved
even for very fast reactions. Preliminary test runs with the pan
filled only with a portion of one of the reactants together with
the silicone oil have confirmed the absence of any detectable
mixing effects, providing easy acquisition of the reaction peak
without the necessity of taking into account additional heat
effects. Thus, the reaction enthalpy (ΔrH°m) was determined
by integration of the experimentally obtained heat flux signal,
assuming heat effects due to mixing was negligible.
For our DSC studies, use of the solvent has been of crucial
importance since the highly volatile alkyl halides were used as
alkylating agents (see reaction 1). DSC experiments at neat
conditions were hardly possible due to the risk of possible pan
explosion. Common solvents mentioned above are as a rule
also too volatile. Another apparent problem is the miscibility of
the reactants; e.g., for the synthesis of the ionic liquid 1-butyl-3-
methylimidazolium chloride from (MeIm) and 1-chlorobutane
already for a MeIm conversion of 8%, the neat synthesis lead to
two phases either rich in IL or 1-chlorobutane, whereby MeIm
is soluble in both phases. Only an addition of 20 vol % of
ethanol led to a single-phase synthesis.12 Such a complex phase
behavior during the DSC study could impact the acquisition of
the reaction peak. To overcome these two problems, a very low
volatile solvent, which is miscible with the reactants and the
resulting ionic liquid, is required. It has turned out that the
most suitable solvent for the imidazolium-based ILs studied in
this work was another ionic liquid, 1-ethyl-3-methylimidazo-
lium bis(trifluoromethylsulfono)imide ([C2mim][NTf2])
(99.9%, Iolitec). This solvent was ideal to the decrease the
total pressure in the reacting system. It was miscible with the
precursors and the resulting ILs. In addition, the [C2mim]-
[NTf2] as the solvent was very advantageous for a high
conversion and selectivity of the MeIm alkylation with alkyl
halides. We tested the influence of the amount of [C2mim]-
[NTf2] (between 10 and 70 wt %) on the reproducibility of the
measured enthalpy of reaction and selected to keep dilution of
the reacting mixture on the level of >50−70 wt % where the
measured reaction enthalpies were most reproducible.
For some reactions which could slowly start already at room
temperature, it was practical to dilute the reactants with the
final product (ionic liquid) in order to slow down the reaction
rate and to avoid an uncontrolled conversion within the baseline
stabilization period in the DSC.
2.2.3. Molar Ratio of Reagents. Syntheses of ILs are usually
performed in an excess of the highly volatile alkylating reagent
in order to reach high degrees of conversion and to easily purify
the prepared IL by gentle heating in vacuum. However,
sometimes reactions have to be performed at the excess of
MeIm to provide homogeneity of the reaction mixture at the
end of the experiment.10 For our DSC studies of the energetics
of the IL synthesis reactions, an excess of the volatile com-
ponent seems to be impractical. Indeed, in the course of the
temperature scan it elevates the total pressure in the pan and
the pan usually explodes on heating. In contrast, at the excess of
the substituted amine (MeIm in this case), the pressure inside
of the pan remains moderate and it keeps tight during the
experiment. From a chemical point of view, the molar reaction
enthalpy should be independent of the excess of reactants,
provided that complete conversion of the stoichiometrically
deficient reactant was achieved. We tested the experimental
conditions using as an example the molar ratios MeIm:C4H9Br
at the three levels 0.3:1, 0.9:1, and 1.3:1. We have not found
any differences in the measured reaction enthalpies within the
boundaries of the experimental uncertainties of 0.5 kJ·mol−1,
proving the completeness of the IL synthesis reaction in the
2.2.5. Temperature Range: Reactions in Isothermal or in
Scanning Mode? There are at least four main prerequisites for
reliable and reproducible measurements of reaction enthalpies
by DSC: (1) high selectivity of the chemical reaction; (2) com-
pleteness of the chemical reaction; (3) well-defined and sharp
reaction peak; and (4) stable baseline with well-defined on-set
and end-set of the reaction peak.
Fortunately, the two first requirements, selectivity and high
conversion degree, could be generally fulfilled for IL synthesis
reactions13 provided that the reaction is performed at a suitable
temperature (or temperature range). Our preliminary experi-
ments (MeIm + C4H9Br) were carried out under isothermal
conditions at temperatures commonly recommended for syn-
thesis of imidazolium-based ILs. However, the main problem
with isothermal measurements results from the ill-defined
behavior of the signal following the initial approach to steady
state of the instrument with the sample. In isothermal con-
ditions, we also obtained very broad reaction peaks with an
ambiguous conversion of reactants and unclear end set. Thus,
our further experiments were performed under nonisothermal
conditions with dynamic temperature ramps which provide the
most reproducible DSC curves. We have studied different
4278
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
temperature profiles with scanning rates from 5 to 50 K·min−1.
It has turned out that the most reproducible results could be
obtained at scanning rates above 30 K·min−1. At increasing
scanning rates, thermal lag increases too and may result in
serious smearing of the curves. Thus, in this work the reaction
enthalpy measurements were performed at 50 K·min−1 heating
rate in the temperature range from room temperature up to 523 K.
In general, the selection of the upper temperature of the
DSC investigation depends crucially on the thermal stability of
the ILs under study. For the reactions of the imidazolium-based
halides (reaction 1), we have deliberately set the end
temperature to 523 K in order to obtain complete conversion.
But it should be mentioned that this end-set temperature was
well below the decomposition temperatures (around 560 K) of
[Cnmim][Br].14 It is quite obvious that, in studies of ILs where
decomposition is expected, the upper limit of the DSC study
should be adjusted to lower values which have to be at least
20−30 K below the expected decomposition temperature.
In general, the selection of the lowest start temperature of
the DSC investigations depends on the reaction rate. As a rule,
after the beginning of the experiment the calorimeter needs
certain time to reach steady-state conditions at the start tem-
perature and to stay in this state at least for 2 min. It is
important that the onset of the reaction peak is outside of this
equilibration period. If the reaction under study is too fast, the
lowest temperature of the DSC runs could be also selected
below room temperature. In this case, the mixing of the
reactants during room temperature handling of the sample pan
had to be prevented; see section 2.2.2.
2.2.6. Proof of Completeness of the IL Synthesis Reaction.
In each experiment, we have optimized the temperature range
and the scanning rate in order to achieve the maximal possible
conversion of the precursor which was in stoichiometric
deficiency. In order to check the completeness of the reaction,
the aluminum pan with the sample was taken from the DSC
and cooled down to room temperature. After weighing of the
pan, its top cover was punctured and the pan was heated at
373−393 K in a special vacuum oven for several hours until a
constant mass was reached. In such conditions, the IL and the
silicone oil could not be evaporated from the pan. Thus, the
mass loss from the pan was ascribed to the loss of the MeIm
which was usually taken in excess. Having the record of masses
before and after the experiment, the degree of conversion was
calculated. As a rule, the level of conversion in all experiments
was about 99% (with the estimated uncertainty of 0.5%). Small
corrections for the incomplete conversion have been taken into
account for each reaction under study.
2.3. QCM: Measurements of Vaporization Enthalpy.
Enthalpies of vaporization of [C4mim][Br] was measured using
the QCM technique.15 A sample of the IL was placed in an
open cavity of a thermostated metal block and it was exposed to
a vacuum system at 10−5 Pa. The quartz crystal microbalance
was placed directly over the measuring cavity containing the IL.
The change in the vibrational frequency of the crystal Δf was a
measure of the amount of IL deposited on the cold QCM. The
value of Δf was measured as function of time at different
temperatures, T. The change of the vibrational frequency Δf
was directly related to the mass deposition Δm on the crystal
surface. Using the frequency change rate df/dt measured by the
QCM, the molar enthalpy of vaporization, ΔlgHm(T0), was
obtained15 by
g
g
°
pm 0
l
⎛
⎜
⎝
⎞
⎟
⎠
Δ H° (T ) − Δ C
T
⎛
⎜
⎝
⎞
⎟
⎠
df
dt
m
0
1
T
1
l
ln
T
= A′ −
−
R
T
0
g
°
⎛
⎞
Δ C
pm
T
T
0
l
+
ln
⎜
⎟
⎠
R
⎝
(2)
with a constant A′, which is essentially unknown, including all
empirical parameters which are specific for the apparatus and
the substance under study. T0 appearing in eq 2 is an arbitrarily
chosen reference temperature (we used 298 K in this work).
The value ΔlgCp°m = Cp°m(g) − Cp°m(l) is the difference of the
molar heat capacities of the gaseous Cp°m(g) and the liquid
phase Cp°m(l), respectively. The frequency change rate df/dt was
measured in a few consequent series with increasing and
decreasing temperature steps. In order to detect a possible
decomposition of IL under the experimental conditions, the
residual IL in the cavity and the IL deposit on the QCM were
analyzed by ATR-IR spectroscopy. No changes in the spectra
have been detected with ILs under study in this work. The
primary experimental data are listed in Table S1 (Supporting
Information).
2.4. Computations. Standard ab initio molecular orbital
calculations were performed using the Gaussian 03 Revision 04
program package.16 Conformations of molecular and ionic
species were optimized on the B3LYP/6-31+G(d,p) level of the
theory. After that, the lowest energy conformations were
calculated on the CBS-QB3 level. The CBS-QB3 theory uses
geometries from B3LYP/6-311G(2d,d,p) calculation, scaled
zero-point energies from B3LYP/6-311G(2d,d,p) calculation
followed by a series of single-point energy calculations at the
MP2/6-311G(3df,2df,2p), MP4(SDQ)/6-31G(d(f),p), and
CCSD(T)/6-31G levels of the theory.17 The calculated values
of enthalpy and Gibbs energy of ions and ion pairs are based on
the electronic energy calculations obtained by the CBS-QB3
method using standard procedures of statistical thermody-
namics.18 The CBS-QB3 calculations were carried out using
Linux cluster (AMD Opteron processor) in the University of
Rostock (Germany). During calculations, eight processors with
2.5 GB RAM for each processor were used. The calculating
procedure took from 1 day to 1 week depending on the size of
molecule.
2.2.7. Summary of the Experimental Conditions. Accord-
ing the results of the screening of the experimental parameters
that has been performed in this work, the following optimal
conditions for DSC studies of the IL synthesis reactions of
starting with precursors MeIm and alkyl bromides could be
recommended: (1) ratio of precursors (MeIm)(R-Br) = 0.3
mol:1.0 mol; (2) dilution with a suitable low-volatile solvent
about 70 wt %; (3) temperature range 373−523 K; and (4)
heating rate 50 K·min−1.
In these conditions, a sharp reaction peak with well-defined
onset, end set and, baseline could be achieved providing the
completeness of the chemical reaction and an easy to draw
baseline for peak integration. These optimal conditions have
been also tested for pyridinium-, pyrrolidinium-, and
piperidinium-based ionic liquids synthesis reactions with alkyl
halides (chlorine, bromine, and iodine) successfully.
3. RESULTS AND DISCUSSION
Ionic liquids are often considered as the new class of environ-
mentally friendly solvents due to their molecularly “tunable”
properties and lack of volatility. New applications are being
developed at a rapid pace. However, the lack of kinetic and
thermodynamic data of their synthesis with emphasis on reaction
4279
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
engineering and process intensification is acknowledged in the
literature.11 Only a few experimental kinetic and thermochem-
ical studies of the imidazolium halide synthesis reaction 1 were
published until December 2011.7,9,11,12,19−24 Schleicher and
Scurto11 performed extended studies on kinetics and solvent
effects in the synthesis of [C6mim][Br] in 10 different solvents:
acetone, acetonitrile, 2-butanone, chlorobenzene, dichloro-
methane, dimethyl sulfoxide, ethyl formate, ethyl lactate,
methanol, and cyclopentanone. The first thermochemical
study of the reaction 1 was published by Waterkamp et al.19
Formation of [C4mim][Br] by reaction of MeIm and 1-bromo-
butane was performed with a reaction enthalpy of −96.0 kJ·mol−1
at 338−358 K and 1−2 bar, but unfortunately, no experimental
details are available in this work.19 They also estimated the
adiabatic temperature for a runaway reaction for this system
and suggested a way to intensify the synthesis of [C4mim][Br]
by using a continuously operating microreactor system. Further
and very promising developments of heat pipe controlled
syntheses of ionic liquids in microchannel reactors were re-
provided very small corrections not larger than 0.3−0.4
kJ·mol−1 and these corrections are well within the boundaries
of the DSC experimental uncertainties of 1−2 kJ·mol−1 (see
Table 1). For this reason, it was reasonable to ascribe in this
work the enthalpies of reactions measured by DSC to the
reference temperature without any temperature correction.
It is apparent from Table 1 that the enthalpies of the ILs
synthesis reactions were independent of the chain length of the
alkyl bromide (within the boundaries of the experimental
uncertainties). Our result for the [C4mim][Br] is in good agree-
ment with the value ΔrHm(298 K) = −(87.7
1.6) kJ·mol−1
determined in an isoperibol calorimeter.22 It should be men-
tioned that the study of reaction 1 in the isoperibol calorimeter
is generally more complex in comparison to the DSC applied in
this work. In ref 22 the ΔrHm was determined in a series of five
experiments at 334 K and then adjusted the value to the
reference temperature 298 K. The reaction was performed at
the 1.50−1.65 excess of MeIm to provide homogeneity of the
reaction mixture at the end of the experiment. However, the
degree of 1-bromobutane conversion at 334 K was only about
95%. In order to obtain enthalpy of reaction according to eq 1,
the authors22 were also challenged to take into account the
following: the enthalpy of mixing of 1-bromobutane with
MeIm; the enthalpy of dissolution of [C4mim][Br] in IL +
MeIm solution; the incomplete conversion of 1-bromobutane;
and the enthalpy of fusion of [C4mim][Br]. The most
important steps 1 and 2 were taken into account with a series
of calorimetric measurements on model mixtures.22 The
enthalpy of fusion of [C4mim][Br] was also obtained from
adiabatic calorimetry.23 In spite of the complexity of the
isoperibol calorimetric study, their result for the quaternization
reaction according to eq 1 was in agreement with our DSC
result (see Table 1) within the combined experimental errors.
In contrast, reaction enthalpy for the [C4mim][Br] synthesis
reaction measured by Waterkamp et al.19 was about 10 kJ·mol−1
less negative in comparison with those from this work, but
absence of experimental details in ref 19 did not allow to
explain the difference. Summarizing, values of ΔrHm listed in
Table 1 are consistent and precise and they are important for
technical as well as for theoretical purposes.
ported by Love et al.7,20,21 In this work, we have focused our
̈
experimental efforts on the precise quantitative determina-
tion of the heat effects, as presented in eq 1, for ILs synthesis
reactions.
3.1. DSC Results for ILs Synthesis Reactions. The
reaction enthalpies according to eq 1 were determined in a
series of 3−5 DSC runs applying the optimized conditions
described in section 2.2.7. The experimental results are col-
lected in Table 1.
Table 1. Experimental Results for Measurements of Reaction
Enthalpies, ΔrHm(liq), Using DSC
a
system (1) +
(2)
molar
ratio
solvent (wt
%)
ΔrHm(liq)
Tmax (K)
(kJ·mol−1
)
MeIm + BuBr
MeIm + PenBr
MeIm + HexBr
0.40
0.44
0.32
0.36
70
70
70
75
448
433
433
433
5
5
2
2
−89.0 0.8
−91.6 1.8
−89.4 2.2
−87.5 1.8
MeIm +
HepBr
MeIm + OctBr
0.30
75
458
5
−89.2 2.2
a
Uncertainties are twice the standard deviations of the mean value.
3.2. Molar Enthalpies of Formation of ILs from DSC
Measurements. Differential scanning calorimetry (in compar-
ison to the isoperibol reaction calorimetry) is a simple and
reliable experimental technique which allows examining thermal
properties of different substances, among them are thermal
effects of different phase transitions. Obviously the idea to
apply the DSC for the investigation of thermochemistry of ILs
appears to be very attractive. The ΔrHm values (see Table 1)
are practically important for the heat management, assessment
of thermal runaways and the quality of the products. However,
at the same time the ΔrHm results are also of theoretical
importance to derive molar enthalpies of formation of ILs,
ΔfH°m from DSC measurements. Enthalpies of formation of
ionic liquids in combination with the high-level ab initio
calculations have opened a new way to obtain enthalpies of
vaporization of ILs.2 Traditionally, the combustion calorimetry
is used as the well-established method to provide enthalpies of
formation.2,27−29 However, this method is sophisticated and
needs extended experience to obtain reliable results. Also, the
method has several shortcomings and restrictions. The most
crucial prerequisite for the combustion calorimetry is the
quality and amount of sample required for the study. In order
to complete the experiment successfully, about 10−20 g of the
As a matter of fact, the enthalpy of reaction, ΔrHm, obtained
by integration of the DSC peak is referred to the temperature
between the onset and end-set of the peak. In most cases we
measured symmetric reaction peaks and we could assign the
ΔrHm to the temperature of the peak maximum, Tmax. To
calculate the enthalpy of reaction ΔrHm at T = 298 K for the
reactions of MeIm with alkyl bromides, the difference, ΔrCpm,
between heat capacities of the ionic liquid Cp1(IL) and the
precursors Cp1(MeIm) and Cp1(R-Br) is required:
−1
Δ H (298 K)/J·mol = Δ H (T
)
r
m
r
m
max
− Δ C (T
− 298 K)
r p,m max
(3)
For example, with the values Cp1 = 162.1 0.6 J·K−1·mol−1
for 1-butyl bromide,25 Cp1 = 147.2 0.6 J·K−1·mol−1 for
MeIm,26 and Cp1 = 308.7 1.0 J·K−1·mol−1 for [C4mim][Br],23
the value ΔrCp,m = −(0.6
1.3) J·K−1·mol−1 was estimated.
Similar results were also obtained for other alkyl bromides: all
ΔrCp,m values were negative and of the order of 1 J·K−1·mol−1.
Application of these ΔrCp,m values for adjustment of the mea-
sured reaction enthalpies to the reference temperature 298 K
4280
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
Table 2. Thermodynamic Data for IL Synthesis Reaction: Total Enthalpies H298 and Reaction Enthalpies, ΔrH°m(g), and
Enthalpies of Formation, ΔfH°m(g), Calculated Using the CBS-QB3 Composite Method and Properties Derived from This
Calculation: Enthalpies of Formation, and Enthalpies of Formation, ΔfH°m(liq), and Enthalpies of Vaporization, ΔglH°m, of
[Cnmim][Br] (in kJ·mol−1)
ILs
H298 (hartrees)
ΔrH°m(g)
ΔfH°m(g)
ΔrH°m(liq) (DSC)
ΔfH°m(liq), eq 4
ΔlgHm, eq 6
ΔlgHm, eq 9
a
a
a
a
[C1mim][Br]
[C2mim][Br]
[C3mim][Br]
[C4mim][Br]
[C5mim][Br]
[C6mim][Br]
[C7mim][Br]
[C8mim][Br]
−2.877.746896
−2.916.976825
−2.956.202413
−2.995.427190
−3.034.652454
−3.073.876527
−3.113.100986
−3.152.325444
−15.9
−24.5
−27.5
−28.5
−30.3
−29.4
−29.3
−29.3
68.0
33.6
(−81.0)
(−87.0)
(−89.0)
(−69.7)
(−106.4)
(−140.1)
(137.7)
(143.3)
a
a
a
a
(140.0)
(145.8)
a
a
a
a
10.6
(150.7)
(148.4)
−10.5
−32.5
−51.5
−71.5
−91.6
−89.0
−91.6
−89.4
−87.5
−89.2
−162.1
−191.1
−212.9
−235.2
−263.6
151.6
158.6
161.4
163.7
172.0
152.2
157.4
161.1
164.0
170.8
a
Assessed values (see text). Values in bold are recommended for further thermochemical calculations.
highly pure (99.9%) sample is required. Such amount of ILs is
very expensive and also methods for purity attestations of the
sample are still not sufficiently developed. Another issuein
order to get an acceptable accuracy of the results (with an
uncertainty of 1−3 kJ·mol−1 in ΔfH°m), the combustion energies
have to be measured with an error of 0.01−0.02%. Such result
is already achieved for C-, H-, O-, and N-containing com-
pounds using the static bomb calorimeter.27−29 For the ILs
additionally containing Cl, Br, I, and S atoms, a special
procedure utilizing a rotating bomb calorimeter is required.
This procedure has now succeeded for the S-containing ILs30
and further development for the Cl-, Br-, and I-containing ILs is
now in progress in our laboratory. In this context, we put focus
of our efforts on the development of the new and concurring
DSC-based thermochemical method leading to values of enthalpies
of formation and further, indirectly, to vaporization enthalpies.
As a representative example, we consider once more the reaction
according eq 1 to show the way how the enthalpy of reaction of
synthesis of [C4mim][Br] is used to obtain its enthalpy of forma-
tion. According to Hess’s Law the enthalpy of reaction is calculated
as follows:
on MD techniques, as well as for anchoring parameters for pVT
equations of state for neat ILs and their mixtures.33 Unfortunately,
the available data on vaporization enthalpies of ILs are rather a
minefield than a consistent collection of data.33 For example, the
experimental results from different techniques reported for the most
frequently studied homologues series [Cnmim][NTf2] are in
disagreement15,34 of 5−7 kJ·mol−1, but the data from the direct
drop-calorimetric study35 differ even by 30−40 kJ·mol−1. This fact
shows clearly the necessity of a more conclusive investigation of the
vaporization enthalpies of ILs using reliable experimental techniques.
At the same time, any indirect option for obtaining enthalpies of
vaporization from empirical, half-empirical, or theoretical methods is
of high value to test the consistency of the data.
3.3.1. Vaporization Enthalpy from Enthalpy of Formation,
ΔfH°m(liq). One of the first indirect options to determine
enthalpy of vaporization is the combination of combustion
calorimetry with the high-level quantum chemical calculations.
This method has turned out to be a valuable thermochemical
option to derive vaporization enthalpies.2,27−31 The procedure
is based on experimental results from combustion calorimetry and
first-principles calculations according to the general equation
Δ H° = Δ H° ([C mim][Br], liq)
g
l
r
m
f
m
4
Δ H° (g) = Δ H° (liq) − Δ H
f
m
f
m
m
(5)
− Δ H° (MeIm, liq) − Δ H° (BuBr, liq)
f
m
f
m
where ΔfH°m(liq) is measured by combustion calorimetry and
ΔfH°m(g) is calculated by high-level first-principles calculations
(e.g., CBS-QB composite method). Then the enthalpy of
vaporization, ΔlgHm, of the IL can be estimated from eq 5
(4)
where ΔrH°m(liq)exp = −(89.0 0.8) kJ·mol−1 was measured in
this work using the DSC in the liquid phase, ΔfH°m(MeIm,liq) =
(70.7
1.1) kJ·mol−1 was measured using the combustion
calorimetry recently,26 and ΔfH°m(BuBr,liq) = −(143.8
1.3)
g
l
Δ H° = Δ H° (g) − Δ H° (liq)
kJ·mol−1 was known from the literature.31,32 From eq 4 the
m
f
m
f
m
(6)
enthalpy of formation of the IL was calculated:
Comprehensive studies of the thermochemical properties of
[Cn mim][N(CN)2 ],2, 2 9 [Cn mim][NO3 ],27 , 28 and
[C1,4Pyrrolidinium][N(CN)2]36 have been performed success-
fully for validation of this procedure.
Δ H° ([C mim][Br], liq) = (−89.0)
f
m
4
+ (70.7) + (−143.8)
= −(162.1 1.9) kJ·mol
In this work, we have extended this procedure for combina-
tion of the high-level quantum chemical calculations with the
results from DSC as a valuable thermochemical option to
derive vaporization enthalpies. Instead of the combustion
calorimetry used earlier to derive ΔfH°m(IL,liq) this concurring
procedure is based on the DSC experimental determination of
the reaction enthalpies (see section 3.1) and ΔfH°m(IL,liq) (see
section 3.2). As an example, for the ionic liquid [C4mim][Br]
using the following data: ΔfH°m(liq) = −162.1 1.9 kJ·mol−1
derived from DSC calorimetry (see chapter 3.2); and ΔfH°m(g) =
−10.5 4.0 kJ·mol−1 calculated by first-principles calculations
−1
In the same way, the enthalpies of formation of other ILs
[Cnmim][Br] were estimated (see Table 2). These values can
be used (together with results from ab initio calculations) to
obtain the enthalpies of vaporization for these ILs indirectly.
3.3. Indirect Determination of Molar Enthalpies of
Vaporization of [Cnmim][Br] from DSC Results. Enthalpy
of vaporization data should be intensively measured, as they are
of paramount importance for the proper calibration and valida-
tion of new force fields used in computational simulations based
4281
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
(CBS-QB3 method, see Table 2), the enthalpy of vaporization
of the IL could be estimated from eq 6
Having established the values of ΔrH°m(liq)exp and ΔrH°m(g),
we have derived the vaporization enthalpy (e.g., for [C4mim][Br])
provided that the following balance holds:
g
l
Δ H = Δ H° (g) − Δ H° (liq)
m
f
m
f
m
g
l
Δ H° (liq) = Δ H° (g) − Δ H
f
m
f
m
m
(7)
−1
= 151.6 4.4 kJ·mol
The enthalpy of reaction was calculated using the Hess law:
This indirect value is in a good agreement with the result
ΔlgHm([C4mim][Br]) = 152.7 3.0 kJ·mol−1 measured with
QCM in this work. This good agreement demonstrates in a
convincing way the successful application of the DSC method
developed in this work to obtain reliable vaporization enthalpies of
ILs in the proposed indirect way.
g
Δ H° (liq) = [Δ H° (g)(IL) − Δ H (IL)]
r
m
exp
f
m
m
1
g
− [Δ H° (g)MeIm) − Δ H (MeIm)]
f
m
m
1
g
− Δ H° (g)(BuBr) − Δ H (BuBr)]
f
m
m
1
(8)
3.3.2. Vaporization Enthalpy from the Enthalpy of
Reaction, ΔrH°m(liq), Measured by DSC. The indirect option
to determine enthalpy of vaporization using eq 6 suggested in
section 3.3.1 has one handicap: it is based on the value of the
gaseous enthalpy of formation, ΔfH°m(g), which is calculated
using ab initio calculations. There are some ambiguities in
respect to procedures of converting results of first-principles
calculations (total energies E0 at T = 0 K and enthalpies H298 at
T = 298 K) into the desired value of ΔfH°m(g). It is well
established that in standard Gaussian-n theories, theoretical
standard enthalpies of formation, ΔfH°m(g), are calculated
through atomization reactions, and isodesmic and bond
separation reactions.37 However, it has turned out37,38 that
gaseous enthalpies of formation calculated from the standard
atomization procedure deviate very often systematically from
the available experimental results.39 The deviation can be posi-
tive or negative, depending on the homologous series under
study. It has been also often discussed that the enthalpies of
formation obtained from the first-principles calculations are
very sensitive for the choice of the bond separation or isodesmic
reactions used for this purpose.38−40 Possible reasons for these
deviations are usually given by different qualities and accuracies
of the experimental data involved in the estimations, as well as a
imbalance of electronic energies calculated for the reaction
participants. That is why we are somewhat hesitant to use
atomization reactions and isodesmic and bond separation
reactions to obtain ΔfH°m(g) of ionic liquids without careful
testing of these reactions. At the same time, the logic of the
DSC experiment lends itself to an elegant solution of the
aforementioned problem. Indeed, using the DSC we have
measured the experimental reaction enthalpy in the liquid
phase, e.g., for [C4mim][Br] according to eq 1. It is apparent
that the same reaction enthalpy, but in the gaseous phase, could
be obtained from the first-principles calculations. In this work,
we have calculated H298 at T = 298.15 K for all reaction
participants in eq 1 using the CBS-Q3B composite method.
According to the Hess law, the enthalpy of this reaction was
calculated as follows:
From eq 6, the enthalpy of vaporization of the IL has been
obtained in eq 9:
g
Δ H (IL) = −Δ H° (liq) + Δ H° (g)
m
r
g
m
exp
r
g
m
1
+ Δ H (MeIm) + Δ H (BuBr)]
m
m
1
1
= −(−89.0) + (−28.5) + (55.0) + (36.7)
−1
= 152.2 4.1 kJ·mol
(9)
with values ΔlgHm(MeIm) = 55.0 0.3 kJ·mol−1 measured in
our lab recently,26 and ΔlgHm(BuBr)] = 36.7
0.1 kJ·mol−1
available from the literature.40
Thus, using the value ΔrH°m(g) calculated with CBS-QB3 in
the gas phase and enthalpies of vaporization of MeIm and
alkyl bromides available from the literature (see Table S2,
Supporting Information), and the enthalpy of reaction
ΔrH°m(liq)exp measured using the DSC in the liquid phase,
the enthalpies of vaporization of the ILs under study have been
derived (see Table 2).
In addition, we compared vaporization enthalpies derived
using two DSC-based procedures: from eq 6 and from eq 9
(see Table 2). Both methods provide the values in very close
agreement; however, we should recommend eq 9 because the
ΔrH°m(g) values for this procedure have been calculated
directly from the H298 avoiding any recalculating procedure.
3.4. Consistency of Enthalpies of IL Synthesis
Reactions. Structure−property relations within a homologous
series are always a valuable tool to get insight into the consistency
of experimental and calculated results. The systematic study of
the [Cnmim][Br] family in the current work has allowed to
correlate the enthalpies of IL synthesis reactions, ΔrH°m, in the
liquid and in the gaseous phase (see Table 2) with the number
of C atoms, Nc, in the alkyl chain of the cation, (see Figure 2).
It is apparent that gaseous enthalpies, ΔrH°m(g), calculated by
using the CBS-QB3 composite method for reactions according
to eq 1 show a quite linear dependence for Nc ≥ 3. However,
reaction enthalpies with methyl and ethyl bromides are
noticeably less negative (see Figure 2). Enthalpies ΔrH°m(liq)
measured by DSC for Nc = 4−8 are obviously indistinguishable
within the boundaries of their experimental uncertainties. But
what trend for reaction enthalpies could be expected for Nc < 4
in the liquid phase? Unfortunately, methyl, ethyl, and propyl
bromides are too volatile in order to perform the DSC studies
properly. That is why we studied in this work only reactions
with Nc ≥ 4 (see Figure 2). However, it is quite interesting to
assess at least the level of thermodynamic properties for
[C1mim][Br], [C2mim][Br], and [C3mim][Br], where the
DSC experiments were not feasible. To do this, we assumed
the similarity of tendencies for ΔrH°m(g) and ΔrH°m(liq) for
Δ H° (g) = H ([C mim][Br], g) − H (MeIm, g)
r
m
298
4
298
−1
− H (BuBr, g) = −28.5 kJ·mol
298
It is important to realize and to emphasize that the ΔrH°m(g)
was obtained using first principles directly from the calculated
H
298 bypassing the conventional calculation of the enthalpies of
formation with the help of atomization, isodesmic, or other
procedures. As a consequence of this trick, the resulting value
of ΔrH°m(g) was not affected at all by any choice and quality of
the auxiliary quantities commonly required for the atomization,
bond separation, etc., reactions.
4282
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
ization enthalpy for Nc ≥ 4 is expressed by the following
equation:
g
−1
Δ H (298 K)/kJ·mol = 135.3 + 4.3N
m
c
1
(r = 0.989)
from which enthalpies of vaporization of other representatives
of the [Cnmim][Br] series with Nc ≥ 4 can be calculated.
3.5. Quantum Chemical Calculations. Gas-Phase
Enthalpies of Reaction and Formation. Thermodynamic
properties of the reaction 1 participants were calculated using
high-level first-principles calculations. The CBS-QB3 composite
method is one of the most powerful methods for calculation of
large molecules of ILs in a reasonable time. A careful con-
formational analysis of the [Cnmim ] cations has been perform-
ed at the level RHF/3-21G* at 0 K. Molecular structures and
relative energies of all conformations for the cations formed by
rotation of alkyl groups around N−C and C−C bonds of the
butyl group (CCCN and CCNC dihedrals) by 360° have been
studied with 10° steps starting from the coplanar conformation.
The conformers with the lowest energies were chosen for
further CBS-QB3 calculations. The halogen anion was placed
near the atoms with the highest positive charge. In this way,
from 10 to 20 conformers of ionic liquids were generated,
which were optimized on B3LYP/6-31+G(d,p) level of theory.
After that, the conformers with the lowest energy were
calculated on the CBS-QB3 level.
3.5.1. Enthalpy of Formation ΔfH°m(g,298 K): Improve-
ment of the Atomization Procedure. We calculated the total
energies E0 at T = 0 K and enthalpies H298 at T = 298 K using
the CBS-QB3 (Table 2). Atomization reactions are the simplest
way to obtain theoretical standard enthalpies of formation,
ΔfH°m(g), from the ab initio results.37,46 A more sophisticated
option is to use a set of isodesmic reactions, bond separation
reactions, or homodesmic reactions to derive theoretical
enthalpies of formation.37 Unfortunately, for ionic liquids the
choice of species with reliable thermodynamic data which could
be involved in these reactions is very restricted. In this work, we
applied the atomization procedure (AT), e.g., for [C2mim][Br]:
Figure 2. Enthalpies of reaction, ΔrH°m, of [Cnmim][Br] synthesis
△
○
reaction: ( ) ΔrH°m(g) for Nc = 1−8; ( ) ΔrH°m(liq) for Nc = 4−8;
●
( ) ΔrH°m(liq) for Nc < 3.
Nc < 4. The assessed values of ΔrH°m(liq) are given in paren-
theses in Table 2. The latter values have allowed to estimate
enthalpies of vaporization for [C1mim][Br], [C2mim][Br], and
[C3mim][Br] using eqs 6 and 8 (Table 2). The chain length
dependence of vaporization enthalpy for the [Cnmim][Br]
series is presented in Figure 3. As can be seen in Figure 3,
[C mim][Br] = 6C + 11H + 2N + 1Br
(10)
2
In contrast to the bond separation and isodesmic reactions,
the atomization reaction suggests the constant quality of the
data on the right side (the constituted atoms with the well
established thermodynamic values) the same for all compounds
under study.37 However, the AT procedure for estimation of
enthalpies of formation, ΔfH°m(g,298 K), of ILs have to be
tested for validity with the reliable experimental data. In Table S3
(in the Supporting Information), we collected a set of reliable
enthalpies of formation for the molecular Br-containing
compounds. Results of the CBS-QB3 calculations (Table S3)
revealed that enthalpies of formation calculated from the
atomization procedure are systematically less positive than those
from the experimental data. But, a remarkable linear correlation
was found between experimental and calculated by AT
enthalpies of formation (see Table S3, and Figure 3):
Figure 3. Enthalpies of vaporization ΔlgH°m(298 K), of [Cnmim]-
○
●
[Br]: ( ) [Cnmim][Br] with Nc = 1 and 2; ( ) [Cnmim][Br] with
Nc ≥ 3.
the first representative, methyl bromide, is slightly out of the
linear correlation. However, such an anomaly was also observed
for vaporization enthalpies of the [Cnmim][NTf2] series of
ionic liquids,41 as well as for the molecular liquids such as
aliphatic esters42,43 and aliphatic nitriles44 and this fact might be
caused by the high dipole moments of these species. The
vaporization enthalpy of [C2mim][Br] already fits very well the
linear behavior. The chain length dependence of the vapor-
−1
Δ H° (g)(exp)/kJ·mol = 1.002 × Δ H° (g)(AT)
f
m
f
m
+ 28.2
(r = 0.9981)
(11)
Using this correlation, the “corrected” enthalpies of forma-
tion of the [Cnmim][Br] have been calculated (see Tables 2
and 3) and these values are now in good agreement with the
4283
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
in the ideal gaseous state has been calculated (Table 4). These
values are in accord with Kp = 1.4 × 10−54 (at 298 K) for
[C4mim][N(CN)2] derived in our previous work.2 Such
extremely low values allow us to conclude with high reliability
that the degree of dissociation of the ion pair is zero for the
ionic liquids under study; i.e., [Cnmim][Br] exist exclusively as
ion pairs in the gaseous phase. This theoretical finding was
proved by QCM measurements, where [C4mim][Br] was
successfully measured without thermal decomposition (tested
by ATR IR) and agreement between the direct measurement by
QCM and indirect DSC result was the validation of all
assumptions applied in this work.
Table 3. Results of Calculation of the Enthalpy of Formation
(in kJ·mol−1) of Br-Containing ILs in the Gaseous Phase at
298 K, CBS-QB3
ILs
ΔfH°m from AT
ΔfH°m from ATcorr
[C1mim][Br]
[C2mim][Br]
[C3mim][Br]
[C4mim][Br]
[C5mim][Br]
[C6mim][Br]
[C7mim][Br]
[C8mim][Br]
39.7
5.4
68.0
33.6
−17.6
−38.6
−60.6
−79.6
−99.6
−119.6
10.6
−10.5
−32.5
−51.5
−71.5
−91.9
3.6. Summary of Benefits for Using DSC for Indirect
Evaluation of ΔlgH°m. There are several very important
advantages of using the DSC technique for the indirect evaluation
of the vaporization enthalpies of ionic liquids:
experimental results. Having established this “modified atom-
ization procedure”, we could use eqs 11 and 6 for estimation of
the enthalpies of formation of the Br-containing compounds
from the CBS-QB3 energies without creating any “suitable”
bond separation or isodesmic reactions. Thus, the CBS-QB3
method combined with the modified atomization procedure
has been applied in this work for reliable calculations of
ΔfH°m(g, 298 K) of the ILs (see Tables 2 and 3).
3.5.2. Quantum Chemical Calculations for Degree of
Dissociation of the Ion Pair. Enthalpy of vaporization of
[C4mim][Br] was measured in this work in the temperature
range 427−467 K using the QCM. At such elevated
temperatures, thermal dissociation of the ILs according to
reaction R1
(1) Experimental ΔrH°m(liq)exp measurements using DSC
are less demanding compared to the combustion
calorimetry.
(2) Purity requirements for the chemicals are less rigorous
compared to the combustion calorimetry.
(3) ΔrH°m(g) is obtained using first principles directly from
the calculated H298 (avoiding the usual calculation of the
enthalpies of formation with help of atomization,
isodesmic, etc., procedures.
(4) ΔlgHm of the starting chemicals are well-known or easy to
be measured.
(5) As a rule, DSC calorimeter is available as basic
+
−
[C mim][Br] ⇌ [C mim] + [Br]
(R1)
equipment in the laboratories working with ILs.
2
2
could be expected. However, in our recent studies2,27,29,36 we
have shown that the aprotic ILs such as [C4mim][N(CN)2],
[Pyrr1,4][N(CN)2], and [Cnmim][NO3] exist in the gaseous
phase as ion pairs and not as separated ions. In this work, cal-
culations for the free cation [C2mim]+, the free anion [Br],¯ and
the ion pair [C2mim]+[Br]¯ have been performed (Table 4).
4. CONCLUSIONS
Standard DSC equipment has been applied for the express
determination of reaction enthalpies of ionic liquids synthesis
reactions in the systems quaternary amine + R-Hal (Hal = Cl,
Br, I). Experimental parameters important for the precise
determination of the heat effects were carefully studied and the
optimal conditions have been recommended. The combination
of the DSC method with the high-level first-principles
calculations provided a consistent set of molar enthalpies of
vaporization and enthalpies of formation for the homologous
series of the imidazolium-based ILs. The indirect procedure to
derive vaporization enthalpies from DSC measurements was
successfully validated with the experimental QCM method.
Combination of differential scanning calorimetry with the first-
principles calculations allows now a rapid accumulation of the
needed thermodynamic data. This procedure provides indis-
pensable data material for testing first-principles procedures
and molecular dynamic simulations techniques in order to
understand thermodynamic properties of ionic liquids on a
molecular basis.
Table 4. CBS-QB3 Calculations of the Thermodynamic
Properties of the Ions [Br]−, [C2mim]+, and the Ionic Pair
[C2mim][Br] Used for Prediction of the Equilibrium
Constant Kp of Dissociation of the Ionic Liquid According to
Reaction 1
R1
properties
298 K
779.8
447 K
617.9
ΔrG°m, kJ·mol−1
ΔrH°m, kJ·mol−1
ΔrS°m, J·mol−1·K−1
Kp
703.2
759.5
−257.0
316.7
6.7 × 10−73
2.4 × 10−137
Results of the total electronic energy at 0 K, the molecular
structures in the lowest energetic state, and all vibrational
frequencies of each species have been obtained (Table 4). On
the basis of these quantum mechanical results, the molar Gibbs
energy, the molar enthalpy, and the molar entropy at 298 K
have been calculated using standard procedures of statistical
thermodynamics. The purpose of this procedure was to obtain
values of the molar standard Gibbs energy of reaction ΔrG°m,
the standard molar entropy of reaction ΔrS°m, and the molar
reaction enthalpy ΔrH°m for the process of dissociation of the
ion pair, e.g., [C2mim]+[Br],¯ into ions in the gaseous phase
according to reaction R1. The chemical equilibrium constant
Kp = 2.4 × 10−137 (at 298 K) and Kp = 6.7 × 10−73 (at 447 K)
ASSOCIATED CONTENT
■
S
* Supporting Information
Table S1, the results of the temperature-dependent studies of
df/dt using the QCM for [C4mim][Br]. Table S2, the results
for alkyl halides derived from CBS-QB3 and experimental
values for ΔlgH°m alkyl halides. Table S3, results of calculation
of the enthalpy of formation bromo-containing compounds in
the gaseous phase at 298.15 K. Figure S1, the IR spectra for
[C4mim][Br] during enthalpy of vaporization investigation
with QCM method, and Figure S2, correlation between
calculated and experimental enthalpies of formation of
4284
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285

The Journal of Physical Chemistry B
Article
(25) Paulechka, Y. U.; Kabo, G. J.; Blokhin, A. V.; Firaha, Dz. S.
J. Chem. Eng. Data 2011, 56, 4891−4899.
bromo-containing compounds in the gaseous phase at 298.15
K. This material is available free of charge via the Internet at
http://pubs.acs.org.
(26) Verevkin, S. P.; Zaitsau, Dz. H.; Emel’yanenko, V. N.; Paulechka,
Y. U.; Blokhin, A. V.; Bazyleva, A. B.; Kabo, G. J. J. Phys. Chem. B 2011,
115, 4404−4411.
AUTHOR INFORMATION
(27) Emel’yanenko, V. N.; Verevkin, S. P.; Heintz, A.; Schick, C.
J. Phys. Chem. B 2008, 112, 8095−8098.
■
Corresponding Author
(28) Emel’yanenko, V. N.; Verevkin, S. P.; Heintz, A.; Voß, K.;
Schulz, A. J. Phys. Chem. B 2009, 113, 9871−9876.
(29) Verevkin, S. P.; Emel’yanenko, V. N.; Zaitsau, Dz. H.; Heintz,
A.; Muzny, C. D.; Frenkel, M. L. Phys. Chem. Chem. Phys. 2010, 12,
14994−15000.
*E-mail: sergey.verevkin@uni-rostock.de. Fax: +49 381 498
6524. Tel: +49 381 498 6508.
Notes
The authors declare no competing financial interest.
†Department of Life, Light and Matter, Faculty of Interdiscipli-
nary Research, University of Rostock.
(30) Zaitsau, Dz. H.; Emel’yanenko, V. N.; Verevkin, S. P.; Heintz, A.
J. Chem. Eng. Data 2010, 55, 5896−5899.
(31) Bjellerup, L. Acta Chem. Scand. 1961, 15, 231−241.
(32) Thermochemistry of Organic and Organometallic Compounds;
Cox, J. D., Pilcher, G., Eds.; Academic Press: New York, 1970.
(33) Esperanca, J. M. S. S.; Santos, L. M. N. B. F.; Rebelo, L. P. N.
J. Chem. Eng. Data 2010, 55, 3−12.
ACKNOWLEDGMENTS
■
This work has been supported by the German Science
Foundation (DFG) in frame of the priority program SPP
1191 “Ionic Liquids” and by Research Training Group 1213
“New Methods for Sustainability in Catalysis and Technique”
(DFG).
̀
(34) Zaitsau, Dz. H.; Verevkin, S. P.; Emelyanenko, V. N.; Heintz, A.
Chem. Phys. Chem. 2011, 12, 3609−3613.
(35) Santos, L. M. N. B. F.; Lopes, J. N. C.; Coutinho, J. A. P.;
Esperanca, J. M. S. S.; Gomes, L. R.; Marrucho, I. M.; Rebelo, L. P. N.
J. Am. Chem. Soc. 2007, 129, 284−285.
(36) Emel’yanenko, V. N.; Verevkin, S. P.; Heintz, A.; Corfield, J.-A.;
Deyko, A.; Lovelock, K. R. J.; Licence, P.; Jones, R. G. J. Phys. Chem. B
2008, 112, 11734−11742.
REFERENCES
■
(1) Maginn, E. J.; Elliott, J. R. Ind. Eng. Chem. Res. 2010, 49, 3059−
3078.
(37) Ohlinger, W. S.; Klunzinger, P. E.; Deppmeier, B. J.; Hehre,
W. J. J. Phys. Chem. A 2009, 113, 2165−2175.
(2) Emel’yanenko, V. N.; Verevkin, S. P.; Heintz, A. J. Am. Chem. Soc.
2007, 129, 3930−3937.
(38) Verevkin, S. P.; Emel’yanenko, V. N.; Pimerzin, A. A.;
Vishnevskaya, E. E. J. Phys. Chem. A 2011, 115, 1992−2004.
(39) Dorofeeva, O. V.; Kolesnikova, I. N.; Marochkin, I. I.; Ryzhova,
O. N. Struct. Chem. 2011, 22, 1303−1314.
(40) Wadso, I. Acta Chem. Scand. 1966, 20, 544−552.
(41) Verevkin, S. P.; Zaitsau, Dz. H.; Emel’yanenko, V. N.; Schick,
C.; Liu, H.; Maginn, E.; Bulut, S.; Krossing, I.; Kalb, R. J. Am. Chem.
Soc. In press.
(42) Krasnykh, E. L.; Verevkin, S. P.; Koutek, B.; Doubsky, J. Vapor
Pressures and Enthalpies of Vaporization of a Series of the n-Alkyl
Acetates. J. Chem. Thermodyn. 2006, 38, 717−723.
(43) Mansson, M.; Sellers, P.; Stridh, G.; Sunner, S. J. Chem.
Thermodyn. 1977, 9, 91−97.
(3) Renken, A. Chem. Eng. Process. 2007, 46, 840−845.
(4) Differential Scanning Calorimetry; Hohne, G., Hemminger, W.,
̈
Flammersheim, H.-J., Eds.; Springer: Berlin, 2003; pp 298.
(5) Maffezzoli, A. J. Therm. Anal. 1996, 47, 35−49.
(6) Blackmond, D. G.; Rosner, T.; Pfaltz, A. Org. Process Res. Dev.
1999, 3, 275−280.
(7) Lowe, H.; Axinte, R. D.; Breuch, D.; Hofmann, C. Chem. Eng. J.
̈
2009, 155, 548−550.
(8) Große-Bowing, A.; Jess, A. Chem. Eng. Sci. 2007, 62, 1760−1769.
̈
(9) Große-Bowing, A.; Jess, A.; Wasserscheid, P. Chem. Eng. Technol.
̈
2005, 77, 1430−1439.
(10) Wasserscheid, P.; Keim, W. Angew. Chem., Int. Ed. 2000, 39,
3772−3789.
(44) Emel’yanenko, V. N.; Verevkin, S. P.; Koutek, B.; Doubsky, J.
Vapor Pressures and Enthalpies of Vaporization of a Series of the
Linear Aliphatic Nitriles. J. Chem. Thermodyn. 2005, 37, 73−81.
(11) Schleicher, J. C.; Scurto, A. M. Green Chem. 2009, 11, 694−703.
(12) Große-Bowing, A.; Jess, A. Green Chem. 2005, 7, 230−235.
̈
(13) Wasserscheid, P.; Keim, W. Angew. Chem., Int. Ed. 2000, 39,
3772−3789.
(45) Notario, R.; Castano, O.; Abboud, J.-L. M.; Gomperts, R.;
̃
Frutos, L. M.; Palmeiro, R. J. Org. Chem. 1999, 64, 9011−9014.
(46) Raghavachari, K.; Stephanov, B. B.; Curtiss, L. J. Chem. Phys.
1997, 106, 6764−6767.
(14) Ngo, H. L.; LeCompte, K.; Hargens, L.; McEwen, A. B.
Thermochim. Acta 2000, 357−358, 97−102.
̀
(15) Verevkin, S. P.; Zaitsau, Dz. H.; Emelyanenko, V. N.; Heintz, A.
J. Phys. Chem. B 2011, 115, 12889−12895.
(16) Frisch, et al. Gaussian 03, Revision B.04; Gaussian, Inc.:
Pittsburgh, PA, 2003.
(17) Montgomery, J. A.; Frisch, M. J. Jr.; Ochterski, J. W.; Petersson,
G. A. J. Chem. Phys. 2000, 112, 6532−6542.
(18) Statistical Mechanics; McQuarrie, D. A., Ed.; Harper & Row:
New York, 1976.
(19) Waterkamp, D. A.; Heiland, M.; Schluter, M.; Sauvageau, J. C.;
̈
Beyersdorff, T.; Thoming, J. Green Chem. 2007, 9, 1084−1090.
̈
(20) Hu, S.; Wang, A.; Lowe, H.; Li, X.; Wang, Y.; Li, C.; Yang, D.
̈
Chem. Eng. J. 2010, 162, 350−354.
(21) Lowe, H.; Axinte, R. D.; Breuch, D.; Hofmann, C.; Petersen,
̈
J. H.; Pommersheim, R.; Wang, A. Chem. Eng. J. 2010, 163, 429−437.
(22) Paulechka, Y. U.; Kabo, A. G.; Blokhin, A. V. J. Phys. Chem. B
2009, 113, 14742−14746.
(23) Paulechka, Y. U.; Kabo, G. J.; Blokhin, A. V.; Shaplov, A. S.;
Lozinskaya, E. I.; Vygodskii, Ya. S. J. Chem. Thermodyn. 2007, 39,
158−166.
(24) Kabo, G. J.; Paulechka, Y. U.; Kabo, A. G.; Blokhin, A. V.
J. Chem. Thermodyn. 2010, 42, 1292−1297.
4285
dx.doi.org/10.1021/jp300834g | J. Phys. Chem. B 2012, 116, 4276−4285