Thermal phase behavior of 1-butyl-3-methylimidazolium hexafluorophosphate: Simultaneous measurements of the melting of two polymorphic crystals by Raman spectroscopy and calorimetry

Article abstract of DOI:10.1016/j.cplett.2013.08.052

The thermal phase behavior of 1-butyl-3-methylimidazolium hexafluorophosphate, which forms three polymorphic crystals (α, β, and γ), has been re-investigated by the simultaneous measurements of Raman spectroscopy and calorimetry. The peak assigned to the phase change from β to γ phase is found to be exothermic and, in striking contrast with a previous report, the peak for this transition is observed near 255 K by recooling and subsequent reheating of the β phase. This finding enabled separate measurements of the melting points (285.8 and 285.3 K), fusion enthalpies (13.1 and 22.6 kJ mol^-1), and entropies of the β and γ phases, respectively.

Full text of DOI:10.1016/j.cplett.2013.08.052

Chemical Physics Letters 584 (2013) 79–82
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
Thermal phase behavior of 1-butyl-3-methylimidazolium
hexafluorophosphate: Simultaneous measurements of the melting
of two polymorphic crystals by Raman spectroscopy and calorimetry
a
b,
⇑
Takatsugu Endo , Keiko Nishikawa
a
Department of Chemical Engineering and Materials Science, The University of California, One Shield Avenue, Davis, CA 95616, USA
Graduate School of Advanced Integration Science, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The thermal phase behavior of 1-butyl-3-methylimidazolium hexafluorophosphate, which forms three
Received 11 June 2013
In final form 14 August 2013
Available online 22 August 2013
polymorphic crystals (
spectroscopy and calorimetry. The peak assigned to the phase change from b to
exothermic and, in striking contrast with a previous report, the peak for this transition is observed near
55 K by recooling and subsequent reheating of the b phase. This finding enabled separate measurements
a
, b, and
c
), has been re-investigated by the simultaneous measurements of Raman
c
phase is found to be
2
ꢀ1
of the melting points (285.8 and 285.3 K), fusion enthalpies (13.1 and 22.6 kJ mol ), and entropies of the
b and phases, respectively.
c
Ó 2013 Elsevier B.V. All rights reserved.
1
. Introduction
Room temperature ionic liquids (RTILs) are a new class of liq-
of these phases based on Raman spectroscopy. This implies that
Raman spectroscopy can easily distinguish the [C mim]PF crystal-
4 6
line states of interest. Thus the thermal phase behavior linking
with the conformational change of the cation and the crystal struc-
uids. Though they are composed solely of ions, their melting points
fall within the range of ambient temperature, typically below
tures of [C
cies still remained between the calorimetric and NMR data
[12,14,15]. Calorimetric measurements of the phase indicated
4 6
mim]PF seemed to be well understood, but discrepan-
3
73 K. Since RTILs have outstanding properties as liquids, such as
negligible vapor pressures, high thermal/chemical/electrochemical
stabilities, and characteristic solubilities, they are regarded as hav-
ing potential utility as electrolytes, chemical reaction media, and
lubricants [1–4].
c
that it was generated via an endothermic phase change (gaining
high thermal energy) from the b phase [12]. However, the results
of NMR experiments suggested that the
c phase was the most sta-
1
-Butyl-3-methylimidazolium
hexafluorophosphate
([C4-
ble phase and that the b phase was a metastable phase and was to
mim]PF , Figure 1) is a representative RTIL. This material consists
6
be eventually relaxed to the
ature [14,15].
c phase over a wide range of temper-
of relatively simple ions, but its thermal phase behavior, which is
one of the fundamental physical properties of RTILs, is complicated
and confusing [5–7]. The thermal behavior of this RTIL significantly
depends on its thermal history and has been known to have at least
two different crystalline phases [5–10], but only one crystal struc-
ture had been studied until recently [5,11].
The present study has intended to re-investigate the thermal
phase behavior of [C mim]PF to interpret the contradiction be-
4
6
tween the previous calorimetric and NMR data. The behavior has
been studied using calorimetry with high temperature stability
and by a simultaneous analysis of Raman spectra. The sample
was recooled and reheated repeatedly in the crystalline state so
as to make reliable observations of solid–solid phase changes.
We reported in 2010 by the simultaneous measurements of Ra-
man spectroscopy and calorimetry that [C mim]PF had three
4 6
polymorphic crystals with different cation conformations [12].
The melting-curve traces for the b and
equal melting points (T ), have been distinguished, and the T
sion enthalpies ( fusH), and entropies ( fusS) of the b and phases
have been determined.
c
phases, which have nearly
0
Namely, the gauche–trans (GT), trans–trans (TT), and gauche –trans
m
m
, fu-
0
(
G T) conformations were assigned to the
a
, b, and
c
phases,
D
D
c
0
respectively. The
c
phase with G T conformers of the cation was as-
signed to the reported crystal structure.
Saouane et al. [13] recently analyzed the X-ray single-crystal
2
. Experimental
mim]PF was prepared by metathesis of 1-butyl-3-methyl-
structures of the
a and b phases and confirmed our assignment
[
C
4
6
⇑
Corresponding author.
E-mail address: k.nishikawa@faculty.chiba-u.jp (K. Nishikawa).
imidazolium chloride with sodium hexafluorophosphate. The ionic
liquid was purified by washing several times with distilled water
0
009-2614/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cplett.2013.08.052

8
0
T. Endo, K. Nishikawa / Chemical Physics Letters 584 (2013) 79–82
ꢁ
226.5 K during heating (so-called cold crystallization). After crys-
tallization, it passes through two endothermic peaks that have
been attributed to the phase changes from
a to b (250.3 K), b to
4 6
Figure 1. Chemical structure of [C mim]PF .
c
(276.1 K), and finally melts at ꢁ285.3 K [12]. The crystalline
a,
b and
c phases show characteristic Raman spectra (Figure 2b).
1
These spectra are distinguished in the regions from 580 to 640
and using activated charcoal. It was characterized by H NMR (JEOL
ECX400) and elemental analysis (Perkin-Elmer 2400). 1H-NMR
ꢀ1
and 300 to 350 cm . The cation conformers in these Raman spec-
0
tra were assigned to GT, TT, and G T for the
a, b and c phases,
(
(
(
(
DMSO-d
s, 1H), 4.10 (t, 2H), 3.79 (s, 3H), 1.71 (m, 2H), 1.21 (m, 2H), 0.84
t, 3H). Elemental analysis for C P, calcd. (found): C 33.81
33.77), H 5.32 (4.96), N 9.86 (9.81). Note that HF was generated
6
, d/ppm relative to TMS): 9.02 (s, 1H), 7.68 (s, 1H), 7.61
respectively, by calculations using the density functional theory
[
12]. Recent X-ray analyses of single crystals have shown that
8
15 6 2
H F N
the Raman spectral assignments are basically correct, but there
still remained other minor coexisting conformers, particularly in
the b phase [13].
It was necessary to provide reproducible Raman spectra of the
crystalline phases that exist in the temperature range above the
in the elemental analysis measurement by this instrument, which
caused underestimation of the existence of H. No chlorine ions
were detected upon the addition of an aqueous AgNO solution
3
ꢀ3
to the RTIL. The sample was dried under vacuum (ꢁ10 Pa) at
13 K for one day before use and handled in a glove-box under
atmosphere to avoid atmospheric moisture. After preparation,
the water content in RTIL was ꢁ140 ppm, as measured by Karl
Fischer titration using a coulometer (Mettler–Toledo model DL39).
A self-made apparatus [16] combining a Raman spectrometer
and a laboratory-made calorimeter was used. This spectrometer
was equipped with an optical fiber (HoloLab 5000, Kaiser Optical
Systems) and a GaAlAs diode laser (wavelength: 785 nm). A spec-
tiny endothermic peak, 276.1 K, which were assigned to the
c
3
phase [12]. This temperature range is narrower than 10 K and over-
laps with the premelting region, where the sample starts partial
melting below its melting point so that liquid-like Raman spectra
are often observed. In addition, the endothermic peak is so small
that fluctuations in the laser energy used for the Raman measure-
ments sometimes interfere with the observation of the peak. Fur-
thermore, the melting points, which are often a good indicator
for distinguishing different polymorphic crystals, need special cau-
N
2
ꢀ1
trum ranging 100–3450 cm
was simultaneously measurable
ꢀ1
tion in this case, because the melting points of the b and c crystals
with resolution of 4 cm by adopting a multiplex grating backed
up by a CCD camera. A calorimeter with a temperature controller
with high stability, which was applicable for simultaneous mea-
surements of vibrational energy and calorimetric data, was used;
see Ref. [17] for the original design. Thermo modules were used
in these instruments as heat flow sensors and heat pumps, en-
abling calorimetric measurements with proper sensitivity and
temperature stability. The apparatus constant for quantitative
measurements of enthalpy was determined using the method
developed by Tozaki and Sou [18]. The temperature was calibrated
using the onset temperature of the melting of distilled water. The
stabilities of the baseline and the temperature of the present calo-
are so close that their correct assignment needs a special caution;
see later.
Thus, the calorimetric measurement of this small endothermic
peak was repeated several times, and different Raman spectra were
obtained in the higher temperature region than the peak, including
the spectra of the b,
c and liquid-like states (or mixtures of b and c,
see Figure S1 for details). The results indicate that the endothermic
peak observed just below the melting point does not represent the
phase change from b to
change from b to is exothermic and appears at nearly identical
temperature with that from to b (255 K). Figure 3 shows calori-
c. In fact, the true peak for the phase
c
a
metric traces with their thermal histories, which enables the
observation of the exothermic peak corresponding to the phase
change from b to c.
rimeter were estimated to be ꢁ5
l
W and ꢁ0.001 K, respectively.
ꢀ1
The temperature was controlled at 20 mK s
.
First, the b phase was obtained according to the same proce-
dure, as shown in Figure 2a. After the first endothermic peak at
3
. Results
ꢁ
255 K, which indicates the phase change from a to b, the sample
was cooled down to 194 K. It was then heated again, and a rela-
tively large and broad exothermic peak was observed at 255 K. Fur-
ther heating enabled the observation of the melting at 285.3 K,
which is basically equal to that displayed in Figure 2a, but higher
Figure 2a shows typical calorimetric curves for [C
4
mim]PF
6
ꢀ1
measured at a scan rate of 20 mK s . The results are nearly the
same as those obtained at a slower scan rate (5 mK s ) [12]. While
ꢀ1
[
C
4
mim]PF
6
does not crystallize during cooling, it crystallizes at
at a scan rate of 20 mK s 1. The trace was initiated for the liquid state near room temperature. The inset is an enlargement of
ꢀ
Figure 2. (a) Calorimetric curves for [C
4
mim]PF
6
the tiny endothermic peak just below the melting point. (b) Raman spectra of [C
4 6
mim]PF . Colors correspond to the following phases: black: liquid (302 K), red: crystal a
ꢀ
1
(
229 K), blue: b (254 K), and green: (280 K). Two peaks are observed in the Raman spectra of the a phase ranging 300–350 cm . The larger peak is assigned to the GT
c
conformer, while the smaller peak is assigned to other minor conformers. The GT conformer becomes dominant with decreasing temperature [12,13]. (For interpretation of
the references to colour in this figure legend, the reader is referred to the web version of this article.)

T. Endo, K. Nishikawa / Chemical Physics Letters 584 (2013) 79–82
81
The estimated T
listed in Table 1. The
larger than that for the b crystal. This estimate agrees well with
m
,
D
D
fusH, and
D
fusS values for these phases are
ꢀ1
fusH value for the c crystal is 9.5 kJ mol
ꢀ1
the
consistency thus confirms that the exothermic peak represents
the phase change from b to . The fusH values reported for [C4-
mim]PF ranged widely from 9.21 to 19.91 kJ mol [7–10,12,19].
DH value of the phase change from b to c, 9.7 kJ mol . This
c
D
ꢀ
1
6
These differences must have originated partially from variations
in the purity of the samples as well as those in the experimental
methods and procedures of data analyses, but inadvertent observa-
tion of other unidentified phase changes is probably a major
contributor.
4
. Discussion
Figure 3. Calorimetric curves with different thermal histories from that shown in
Figure 2a. The trace was initiated from the glassy state at 194 K. The inset
corresponds to that in Figure 2a. The small fluctuations in the heat flow observed in
the inset may be ascribed to the melting of a little amount of water originating from
the atmospheric moisture.
The exotherm during heating indicates that the phase change is
not a reversible phase transition but an irreversible phase change.
The former occurs at a temperature where the Gibbs free energies
of two phases are equal, whereas the latter is a phase change from
one phase with a higher Gibbs free energy to that with a lower one.
In addition, a phase transition is always accompanied by heat gen-
eration during cooling and heat absorption during heating. The
4 6
irreversible phase change of [C mim]PF , which can be observed
over a wide temperature range, was already pointed out by our
NMR spectroscopic measurements [14,15]. The spin–lattice relaxa-
1
3
31
tion times of the C nuclei in the cation and the P nuclei in the
anion in each crystalline state of the RTIL were measured, and
the phase change from the b to
20–280 K. This finding explains the reason for the reported incor-
rect assignment of the very small endothermic peak [12]. Namely,
we had accidentally observed the phase just above the endother-
mic peak, because the phase change from b to can, but not nec-
c was observed in the range of
2
c
c
essarily, occur at any temperature in that range. As mentioned
above, this temperature range falls within the premelting region,
which could act as a trigger for the phase change.
Figure 4. (a) Calorimetric curve for the b to
below and above the exothermic peak shown in (a). The blue curve represents the b
crystal while the green curve does the one. (For interpretation of the references to
c phase change. (b) Raman spectra
There is a significant difference in the
DfusS values for the b and
c
c phases, which can be explained by considering their crystal
colour in this figure legend, the reader is referred to the web version of this article.)
structures [5,13]. The analysis of the crystal structures of the two
phases revealed disordered structures for both the cations and an-
ions in the b phase. Highly disordered structures lead to an in-
crease in entropy. As a result, the difference in the entropy,
DfusS,
in intensity. Near the exothermic peak, the Raman spectrum repro-
between the crystal and liquid states decreases at the melting
point. One should note that the disordered structures in the b
phase imply that the ions have high mobility in this phase, being
consistent with the NMR spectroscopic data [14,15]. The decrease
ducibly changes from the b to
Namely, the 624 cm peak, which is characteristic of the b phase,
c phase, as shown in Figure 4.
ꢀ1
vanishes through the exothermic peak, followed by the appearance
ꢀ1
of the 600 cm peak, which indicates the transformation to the
phase [12].
c
in the values for both
D
fusH and
DfusS values in the b phase in com-
parison with those for the
c
phase coincidentally result in the same
It is thus demonstrated that the melting behavior with the ther-
mal history displayed in Figure 2a is attributable to the b crystal,
melting temperature for the two crystals, which confuses the iden-
tification of these two polymorphic phases.
while that shown in Figure 3 is assigned to the
c crystal. Note that
the former was previously assigned to the phase [12]. The melt-
c
ing temperatures for these different polymorphic crystals are
nearly identical: 285.8 ± 0.7 K for the b phase and 285.3 ± 0.7 K
for the
c phase, as estimated from the peak tops [12]; nevertheless,
their intensities are significantly different.
Table 1
m
Melting points (T ), fusion enthalpies (DfusH), and entropies (DfusS) of the b and c
crystals. Values are taken from the peak top, and those in parentheses are from onsets
of the peak. Experiments were performed three times with different weights of the
sample to estimate standard deviations.
T
m
D
fus
H
D S
(J K mol )
fus
ꢀ1
(kJ molꢀ1)
ꢀ1
(K)
b
c
285.8 ± 0.7 (284.1 ± 1.0) 13.1 ± 0.7 (13.1 ± 0.7) 45.9 ± 2.6 (46.1 ± 2.6)
285.3 ± 0.7 (283.4 ± 0.9) 22.6 ± 1.6 (22.6 ± 1.6) 79.1 ± 5.6 (79.5 ± 5.4)
4 6
Figure 5. Schematic phase change diagram for [C mim]PF .

8
2
T. Endo, K. Nishikawa / Chemical Physics Letters 584 (2013) 79–82
A schematic phase change diagram for [C
in Figure 5. The phase change from to b always occurs near
50 K, being barely dependent on the scan rate of calorimetry. This
phase change would thus be a phase transition, as displayed by the
intersection of the two lines in the figure. The Gibbs free energy of
4
mim]PF
6
is depicted
are well interpreted by use of reported NMR and crystal structural
data. In addition, the information acquired in the present study
suggests that this peak is related to an increase in the disorder of
the ions, although the origin of the tiny endothermic peak ob-
served in the b phase still remains unsettled.
a
2
the b phase is always higher and steeper than that of the
due to the higher enthalpy and entropy, respectively. The phase
change from b to
and even with a different scan rate (10 mK s , data not shown).
Taking into account the NMR data that revealed a faster phase
change at 257 K than at 225 K [14], the temperature region around
c
phase
Now it would be possible to state that the complicated thermal
phase behavior of [C mim]PF , as one of the representative RTILs,
4 6
c
was found to occur at ꢁ 255 K reproducibly
has been clearly understood. The present study has been focused
on the behavior under atmospheric pressure, but the findings will
also be useful for the understanding under high pressure because
this RTIL has recently been one of the highlighted subjects of phase
behavior studies under high pressure [13,26–30].
ꢀ1
2
55 K seems to be favorable for the phase change. This is probably
related to nucleation and growth, because a phase change requires
nucleation and subsequent growth. The favorable temperature at
which each phenomenon occurs is sometimes different. In such a
case, a phase can maintain its state even though it is thermody-
Acknowledgements
The present study has been supported by the Ministry of Educa-
tion, Culture, Sports, Science and Technology of Japan, No.
25248003, Grant-in-Aid for Scientific Research (A) and JSPS Post-
doctoral Fellowships for Research Abroad.
namically metastable. For [C
appropriate temperature at which both nucleation and subsequent
growth are induced and lead to the phase change from b to . The
Gibbs free energy curves for both the b and crystals cross that of
4 6
mim]PF , 255 K appears to be an
c
c
the liquid at similar temperatures. Because the premelting phe-
nomenon exists in this RTIL, as it frequently occurs in other RTILs
Appendix A. Supplementary data
[
20,21], a phase state located in the temperature region just below
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.cplett.2013.08.052.
the melting point of the b phase would be rather unstable. In such a
region, different phases, or sometimes mixtures of phases, can be
observed via Raman spectroscopy when different measurements
are taken, as displayed in Figure S1.
References
The origin of the tiny endothermic peak, temporally assigned to
the phase change from b to
c
, is still unclear. Although it is hard to
[1] P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis, VCH-Wiley, Weinheim,
Germany, 2003.
obtain reproducible data near this temperature region, as men-
tioned above, but it can at least be concluded that this peak is
not assignable to any phase change that would affect the Raman
spectrum. For example, Figure S2 shows the results of the simulta-
neous measurement of the Raman spectrum and the calorimetry
for this peak. The Raman spectrum is that of the continuous phase
change of b to a phase that is partially mixed with a liquid compo-
nent as the temperature increases. There is no other distinct
change in the Raman spectrum throughout the peak. Therefore,
the endothermic peak does not accompany a clear conformational
change of the cation or a displacement of the relative position of
the ions, both of which should significantly affect the Raman spec-
tra of RTILs [22–25]. Nor is any endothermic peak observed clearly
[
2] B. Kirchner (Ed.), Topics in Current Chemistry, first edn., Springer, Berlin,
Germany, 2009.
[3] I. Minami, Molecules 14 (2009) 2286.
[
[
4] A. Kokorin (Ed.), Ionic Liquids: Applications and Perspectives, InTech, 2011.
5] A.R. Choudhury, N. Winterton, A. Steiner, A.I. Cooper, K.A. Johnson, J. Am.
Chem. Soc. 127 (2005) 16792.
[6] A. Triolo et al., J. Phys. Chem. B 110 (2006) 21357.
[
[
[
7] J. Troncoso, C.A. Cerdeiriña, Y.A. Sanmamed, L. Romaní, L.P.N. Rebelo, J. Chem.
Eng. Data 51 (2006) 1856.
8] G.J. Kabo, A.V. Blokhin, Y.U. Paulechka, A.G. Kabo, M.P. Shymanovich, J.W.
Magee, J. Chem. Eng. Data 49 (2004) 453.
9] H. Jin et al., J. Phys. Chem. B 112 (2008) 81.
[
[
10] U. Doma n´ ska, A. Marciniak, J. Chem. Eng. Data 48 (2003) 451.
11] S.M. Dibrov, J.K. Kochi, Acta Crystallogr. C C62 (2006) o19.
[12] T. Endo, T. Kato, K. Tozaki, K. Nishikawa, J. Phys. Chem. B 114 (2010) 407.
13] S. Saouane, S.E. Norman, C. Hardacre, F.P.A. Fabbiani, Chem. Sci. 4 (2013) 1270.
14] T. Endo, H. Murata, M. Imanari, N. Mizushima, H. Seki, K. Nishikawa, J. Phys.
Chem. B 116 (2012) 3780.
[
[
in the
c phase, as shown in the inset of Figure 3. In addition, taking
the fact into account that both the cation and anion tend to be dis-
ordered in the b crystal [13], this peak might possibly be related to
further displacements of the ions.
[15] T. Endo, H. Murata, M. Imanari, N. Mizushima, H. Seki, S. Sen, K. Nishikawa, J.
Phys. Chem. B 117 (2013) 326.
[
[
[
[
16] T. Endo, K. Tozaki, T. Masaki, K. Nishikawa, Jpn. J. Appl. Phys. 47 (2008) 1775.
17] S. Wang, K. Tozaki, H. Hayashi, H. Inaba, J. Therm. Anal. Calorim. 79 (2005) 605.
18] K. Tozaki, K. Sou, Jpn. J. Appl. Phys. 45 (2006) 5272.
19] D.M. Fox, W.H. Awad, J.W. Gilman, P.H. Maupin, H.C. De Long, P.C. Trulove,
Green Chem. 5 (2003) 724.
5
. Concluding remarks
[
[
[
20] K. Nishikawa, S. Wang, H. Katayanagi, S. Hayashi, H. Hamaguchi, Y. Koga, K.
Tozaki, J. Phys. Chem. B 111 (2007) 4894.
21] Y.U. Paulechka, A.V. Blokhin, G.J. Kabo, A.A. Strechan, J. Chem. Thermodyn. 39
4 6
The thermal phase behavior of [C mim]PF , which had been ad-
dressed by many groups, has been investigated to resolve the con-
tradictions in the reported findings. Simultaneous measurements
of Raman spectroscopy and calorimetry were performed repeat-
edly with recooling and reheating of the sample in the crystalline
state. Contrary to the existing assignment of the tiny endothermic
peak observed just below the melting point, ꢁ276 K, to the phase
(
2007) 866.
22] T. Endo, T. Kato, K. Nishikawa, J. Phys. Chem. B 114 (2010) 9201.
[23] T. Endo, T. Morita, K. Nishikawa, Chem. Phys. Lett. 517 (2011) 162.
[24] Y. Umebayashi, T. Fujimori, T. Sukizaki, M. Asada, K. Fujii, R. Kanzaki, S.
Ishiguro, J. Phys. Chem. A 109 (2005) 8976.
[
[
25] S. Hayashi, R. Ozawa, H. Hamaguchi, Chem. Lett. 32 (2003) 498.
26] O. Russina, B. Fazio, C. Schmidt, A. Triolo, Phys. Chem. Chem. Phys. 13 (2011)
12067.
27] L. Su et al., J. Chem. Phys. 130 (2009) 184503.
28] T. Takekiyo, N. Hatano, Y. Imai, H. Abe, Y. Yoshimura, High Pressure Res. 31
change from b to
surements has confirmed that the b to
c
, the present analysis by the simultaneous mea-
phase change actually oc-
[
[
c
curs exothermically during heating near 255 K. These observations
(
2011) 35.
have enabled us to distinguish the melting curves for each crystal
[29] L. Su et al., J. Phys. Chem. B 114 (2010) 5061.
[
30] R.G. De Azevedo, J.M.S.S. Esperança, V. Najdanovic-Visak, Z.P. Visak, H.J.R.
Guedes, M.N.D. Ponte, L.P.N. Rebelo, J. Chem. Eng. Data 50 (2005) 997.
and to derive their thermodynamic parameters. While the T
for each phase is quite similar, the fusH and fusS values of the
phase significantly exceed those of the b phase. These differences
m
value
D
D
c