Supplement to Manuscript No.: A16.03.0272
Title:
Glass transition and stable glass formation of tetrachloromethane
Author(s):
Y.Z. Chua, M. Tylinski, S. Tatsumi, M.D. Ediger, C. Schick
Purity of the vapor-deposited glass
Glasses of tetrachloromethane (CCl4) produced by physical vapor deposition have been
characterized by in situ AC chip nanocalorimetry. In order to monitor the gas composition in
the chamber during deposition and during the temperature scans, a residual gas analyzer
(RGA) (Model QMS 200 Prisma, Pfeiffer Vacuum, USA) was added as an extension to the
measurement setup.
During the deposition, the RGA monitors the gases in the chamber. FIG. S1(a) shows a
typical data set during deposition of CCl4, at substrate temperature 76.4 K and deposition time
of about 400 s resulting in a ca. 400 nm film, while (b) shows an “empty” deposition
measurement at 10 K. The “empty” deposition measurement is an experiment with empty
sample reservoir but otherwise identical procedures (e.g. substrate temperature, deposition
time) as for the data of FIG. S1(a). This “empty” deposition measurement is to clarify the
source of the observed species, whether from vacuum chamber (background) or from CCl4
sample. Interestingly during this “empty deposition”, FIG. S1(b), no ion currents above 4 pA
are observed. The ion current measured is about three orders of magnitude smaller than that
shown in FIG. S1(a), indicating absence of serious leaks in the system. Therefore it can be
determined, that all the existing peaks in FIG. S1(a) originate from the CCl4 sample.
In FIG. S1(a), the measured mass spectrum (solid red line, left axis) is compared with the
mass spectrum of CCl4 from NIST chemistry webbook1 (solid blue line, right axis). Other
peaks, that are not consistent with mass spectrum of CCl4, have been verified by an ex situ gas
chromatography – mass spectroscopy measurement, shown in FIG. S2 and FIG. S3.
Due to the limitation of m/z = 80 for the RGA, this causes uncertainty on the result of RGA.
Therefore, CCl4 is additionally measured ex situ with gas chromatography – mass
spectroscopy (GC-MS). CCl4 is used without further purification.
An Agilent 5977E GC-MS was used for the analysis, with ionization achieved by electron
impact at 70 eV. The capillary column used was a J&W Low Bleed (HP5-ms,
30 m × 0.25 mm ID, 0.25 µm thick film). The operating conditions are as followed: injection
1
port temperature at 200 °C; 10 µL injection volume of sample vapor (at room temperature),
column: isothermal at 25 °C; helium carrier gas flow: 1 mL/min. The split/splitless injector
was operated in the 1:10 split mode. The chromatogram of CCl4 is shown in FIG. S2.
Tetrachloromethane
-8
10
100
(a)
Peaks as in CCl4 mass spectrum
CCl4
C3 H6 O
-9
80
C4H10O
CHCl3
60
-10
10
40
Rel. Intensity
Ion current in A
10
-11
10
20
-12
10
0
0
10
20
30
40
50
60
70
80
50
60
70
80
m/z
Empty measurement
-8
10
(b)
-9
Ion current in A
10
-10
10
-11
10
-12
10
0
10
20
30
40
m/z
FIG. S1. (a) (Left axis) Ion current measured by the RGA during the deposition of the CCl 4 as shown in
solid red line. The black indication arrows show the consistent peaks with mass spectrum of CCl4 from
NIST chemistry webbook, as confirmation that CCl4 is deposited. (Right axis) Mass spectrum of CCl4
from NIST chemistry webbook1 is shown in solid blue lines Mass spectra for acetone, C 3H6O (solid green
lines), tert-butanol, C4H10O (solid cyan lines) and chloroform, CHCl3 (solid magenta lines) are also
included. (b) Mass spectrum of “empty” deposition measurement at 10 K.
2
For identification purposes, the mass spectrum of each peak was recorded in the total ion
current mode of the mass spectrometer, within a m/z range of 10 to 300. Identification of
compounds was achieved using NIST mass spectral database and then by comparing the mass
spectra of the chromatographic peaks with CCl4 sample, FIG. S3.
Tetrachloromethane
85000
4.68
80000
2.84
Abundance
75000
3.00
70000
65000
3.90
60000
55000
50000
0
2
4
6
8
10
12
14
16
18
20
t r in s
FIG. S2. Chromatogram of tetrachloromethane vapor at 25 °C. The peaks with the printed retention
times, tr, correspond to acetone (tr = 2.84 s), tert-butanol (tr = 3.00 s), chloroform (tr = 3.90 s) and
tetrachloromethane (tr = 4.68 s).
Tetrachloromethane
100
(a)
43
t r = 2.84 s
80
Rel. Intensity
Acetone (C3H6O)
60
40
58
20
15
27
0
0
10 20 30 40 50 60 70 80 90 100 110 120 130
m/z
3
Tetrachloromethane
100
59
(b)
t r = 3.00 s
Rel. Intensity
80
tert-butanol (C4H10O)
60
40
18
20
31
41
0
0
10 20 30 40 50 60 70 80 90 100 110 120 130
m/z
Tetrachloromethane
100
(c)
t r = 3.90 s
80
Rel. Intensity
83
Chloroform (CHCl3)
60
28
40
18
20
47
35
0
0
10 20 30 40 50 60 70 80 90 100 110 120 130
m/z
Tetrachloromethane
100
117
(d)
t r = 4.68 s
Rel. Intensity
80
Tetrachloromethane (CCl4)
60
40
82
47
20
35
0
0
10 20 30 40 50 60 70 80 90 100 110 120 130
m/z
FIG. S3. Ion mass spectra obtained for (a) the first peak (acetone, tr = 2.84 s), (b) the second peak (tertbutanol, tr = 3.00 s), (c) the third peak (chloroform, tr = 3.90 s) and (d) the fourth peak (CCl4, tr = 4.68 s)
by GC-MS.
4
The analysis of ex situ GC-MS measurement shows presence of acetone (C3H6O), tert-butanol
(C4H10O) and chloroform (CHCl3) in the CCl4 sample. With this, the black indication arrows
in FIG. S1(a) show relatively good consistency with the expected mass spectra of CCl 4,
C3H6O, C4H10O and CHCl3, as measured from GC-MS. The CHCl3 might be present as
byproduct due to synthesis of CCl4. As a rough estimation from the area of the ion current
peaks, the impurities, as ratio to CCl4, are C3H6O about 0.05 %, C4H10O about 0.03 %, CHCl3
about 0.01 % and other traces about <0.01 %. This rough estimation is sufficient to confirm
the purity of CCl4 sample, where the amount of impurities is too small to influence the result
of vapor-deposited CCl4 measurements.
Next, the desorption of the different species on slow heating during AC calorimetric
measurements was studied with the RGA. FIG. S4(a) and (b) show the RGA scans after
depositing a sample at 76.4 K and for an empty sample reservoir measurement, respectively.
Again for the “empty” scan, the ion currents are about three orders of magnitude smaller than
after depositing a sample. For the empty scan at about 35 K, there is an increase in ion current
for several species, especially mass of 28(14) which might be nitrogen. Even though the
signals are very small, in order to avoid co-deposition of these species, the CCl4 was
deposited above this temperature.
The ion currents from the RGA on slow heating of a CCl4 sample, which is deposited at
76.4 K and cooled to 60 K, are shown in FIG. S4(a) together with the calorimetric signal. The
heat capacity of the as-deposited CCl4 up to 85 K (black dashed line, left axis) and of an
ordinary glass (red dashed line, left axis) up to high temperature. A glass transition is
observed from glass to supercooled liquid at about 80 K for the as-deposited sample and for
the ordinary glass as detailed in the main text. At about 100 K, the heat capacity decreases
that shows possible crystallization. However at about 120 K, the sample desorbs as denoted
by the rapid drop of the heat capacity. In parallel, the RGA (solid lines, right axis) shows
peaks for species with m/z = 12, 16, 28, 35, 44 and 47, again indicating desorption of CCl4 as
these masses are peaks of CCl4. This confirms that the result from the heat capacity
measurement showed the glass transition of CCl4.
5
Tetrachloromethane
Tg,dyn = 82 K
1.5x10
Ion current in A
Heat capacity in a.u.
Tc  92 K
-9
2.5x10
-11
1.0x10
-12
5.0x10
-9
2.0x10
0.0
Tsubstrate = 76.4 K
60
70
80
90
100
110
120
T in K
AD
OG
2
12
14
16
28
32
35
44
47
-9
1.5x10
-9
1.0x10
Ion current in A
(a)
-9
3.0x10
-11
-10
5.0x10
0.0
60
80
100
120
140
160
180
200
T in K
Empty measurement
-11
2.0x10
(b)
-11
1.8x10
12
14
16
18
28
32
35
44
47
-11
1.6x10
-11
Ion current in A
1.4x10
-11
1.2x10
-11
1.0x10
-12
8.0x10
-12
6.0x10
-12
4.0x10
-12
2.0x10
0.0
0
20
40
60
80
100 120 140 160 180 200
T in K
FIG. S4. (a) (left and bottom axes) Heat capacity of as-deposited (black dashed line) and ordinary glass
(red dashed line) of CCl4 measured by AC chip calorimetry. (right and upper axes) RGA measurement of
CCl4. Inset shows the RGA measurement from 60 K to 120 K. (b) RGA measurement of empty sample
reservoir from 10 K to high temperature, with otherwise identical procedures to (a).
From the measurement of RGA and GC-MS, the amount of impurities in CCl4 sample was too
small to significantly influent the measurement results, while confirming the presence of
CCl4. The heat capacity measurement in FIG. S4(a) shows glass transition of CCl4, as verified
by the desorption of CCl4 at about 120 K.
6
Amorphous glass or plastic crystal?
CCl4 has different phases, especially a plastic crystal phase (Ib). This leads to the question
whether the as-deposited glass of CCl4 is an amorphous glass or a glass of plastic crystal.
Unfortunately, it is not possible unambiguously to create a plastic crystal phase from the
vapor deposited CCl4. Therefore the glass transition of both phases of CCl4 was not available
for comparison. We utilized ethanol, as it was possible to investigate the glass transitions of
the plastic crystal as well as of the supercooled liquid.2, 3 Ethanol was vapor-deposited at
substrate temperature of 87 K, otherwise the deposition conditions and scanning rate are kept
consistent as for CCl4.
Ethanol
Tg,dyn = 105 K Tg, PC = 112 K
(a)
Tsubstrate
= 87 K
Heat capacity in a.u.
Frequency = 20 Hz
Glass/Supercooled liquid
Plastic crystal (PC)
Crystal
80
85
90
95
100
105
110
115
120
T in K
Ethanol
Tg,dyn = 105 K Tg, PC = 112 K
Phase angle in °
(b)
5°
Tsubstrate
= 87 K
Frequency = 20 Hz
Glass/Supercooled liquid
Plastic crystal (PC)
Crystal
80
85
90
95
100
105
110
115
120
T in K
FIG. S5. (a) Heat capacity and (b) phase angle of glass/supercooled liquid (solid black line), plastic crystal
(PC) (solid red line) and crystal (solid green line) of ethanol. The dynamic glass transition temperatures of
supercooled liquid and plastic crystal at frequency of 20 Hz are 105 K and 112 K respectively.
7
FIG. S5 shows the heat capacity and phase angle of glass/supercooled liquid of ethanol,
indicated by the solid black line. Upon heating to slightly higher temperature at 110 K, the
supercooled liquid crystallized to plastic crystal (solid red line). The sample finally fully
crystallized to a crystal (solid green line) when heated above 113 K. The plastic crystal phase
is differentiated from glass/supercooled liquid by the broader step in heat capacity (or broader
phase minimum) at the glass transition, while fully crystallized phase does not show transition
step (or phase minimum).
The glass transition for the supercooled liquid has steep step and sharp phase minimum of
about 3.6 K in width, while the glass transition of the plastic crystal has a width of about
6.7 K. The curves for the supercooled liquid are consistent with the heat capacity
measurement of CCl4, shown in FIG 5 in the main manuscript, which also shows steep step in
heat capacity and a sharp phase minimum of comparable width of 3.4 K. We take this as a
strong indication that the as-deposited glass of CCl4 is an amorphous glass, instead of plastic
crystal.
References
1
N. M. S. D. Center and S. E. S. (director), Mass Spectra. (National Institute of Standards and
Technology, Gaithersburg MD, 2015).
2
O. Haida, H. Suga and S. Seki, J. Chem. Thermod. 9, 1133-1148 (1977).
3
M. A. Ramos, I. M. Shmytko, E. A. Arnautova, R. J. Jimenez-Rioboo, V. Rodrýguez-Mora,
S. Vieira and M. J. Capitan, J. Non-Crystall. Solids 352 (42-49), 4769-4775 (2006).
8