1. Exceptional Kinetic Stability of
Vapor Deposited As2S3 Films
Paarth Thapar, Dmytro Savytskii,
William Heffner, Himanshu Jain
This work has been supported by the National Science Foundation through the
International Materials Institute for New Functionality in Glass (DMR-0844014).
Department of Materials Science and Engineering
Lehigh University, Bethlehem, PA
2. Introduction
Kearns et al, J. Chem. Phys. 127, 154702 (2007)
Schematic representation of the effect
of cooling rate on enthalpy and volume.
General notion: Thin film are less
kinetically stable than the source
bulk glass because of the very high
cooling rate when the vapor deposits
on the surface
3. Motivation
University of Wisconsin-Madison group:
Swallen et al, Science 315, 353 (2007)
Kearns et al, J. Chem. Phys. 127, 154702 (2007)
Glass annealed at 328 K
after few days
Vapor deposited film into a DSC pan, rate 5 nm/s
Ordinary Glass,
40 K/min
TNB: 1,3-bis-(1-naphthyl)-
5-(2-naphthyl) benzene
Tonset of the glass transition region
Glass annealed
at 296 K after
6 months
4. Kinetic stability: response during heating
University of Wisconsin-Madison group:
Swallen et al, Science 315, 353 (2007)
Kearns et al, J. Chem. Phys. 127, 154702 (2007)
Tg onset vapor dep. film
Tg onset aged glass
Tg onset ordinary glass
5. Goal
• Just like some organic glasses, is it possible to
produce more kinetically stable glass films of
chalcogenide glass in comparison to bulk glass?
– No one seems to have investigated this question with
inorganic materials
• Determine the effect of substrate temperature on
the onset temperature of glass transition (Tonset)
• Practical issues:
– How can we perform DSC measurements on films?
6. Methods
Characterization:
Materials:
Arsenic sulfide, As2S3
on Aluminum foil
Thermal evaporation:
Edwards E306A vacuum
deposition system
Differential Scanning
Calorimetry (DSC)
150 200 250
-0.16
-0.14
-0.12
-0.10
-0.08
-0.06
-0.04
-0.02
0.00
0.02
HeatFlow(W/g)exo-->
Temperature (C)
Run 1
Run 2
Tonset
~10 mg of sample
20 punch holes – film/Al foil
Run 1: heat flow of the as
prepared glass / aged glass
/ film
Run 2: heat flow of the
“fresh” glass after
controlled 5K /min cooling
7. 34 36 38 40 42 44 46
196
198
200
202
204
206
208
210
212
214
216
218
Run 1
Run 2
OnsetTemperature(C)
% As
Heated Substrate Results!
Initial Impressions
Tsubstrate = Tg-50 = 164⁰C
Practically the same
chemical composition as
Evap 6
Bulk
Room temperature
deposited films on Al foil
Ts=1640C
Run1
Ts=1640C
Run2
8. 50 75 100 125 150 175 200 225 250
-0.12
-0.10
-0.08
-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
HeatFlow(W/g)
Temperature (o
C)
Ts = 160o
C
Ts = 120o
C
Ts = 80o
C
Ts = 54o
C
Ts = 23o
C
Tonset of Glass Transition with various Ts
As Ts ↑ Tonset ↑: Kinetic
stability criteria used by
Wisconsin-Madison group
T substrate (⁰C) Tonset (⁰C)
160 212
120 205
80 199
54 198
23 198
9. Residual strain with various Ts
As Ts ↑ , Exothermal peak
area ↓, onset of peak ↑,
Tonset of Tg ↑
Area taken from onset of exothermic peak to onset of Tg
T
substrate
(⁰C)
Tonset of
exothermic
(⁰C)
Tonset
of Tg
(⁰C)
Area
(W⁰C/g)
23 70 198 3.38
54 87 198 2.52
80 112 199 1.7
120 143 205 .79
160 188 212 .09
Exothermal peak attributed
to residual strain relaxation
Tonset of exothermic peak
Tonset of exothermic peak
50 75 100 125 150 175 200 225 250
-0.12
-0.10
-0.08
-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
HeatFlow(W/g)
Temperature (o
C)
Ts = 160o
C
Ts = 120o
C
Ts = 80o
C
Ts = 54o
C
Ts = 23o
C
10. 150 160 170 180 190 200 210 220 230 240 250
-0.12
-0.10
-0.08
-0.06
-0.04
-0.02
0.00
0.02
0.04
HeatFlow(W/g)
Temperature (o
C)
Deposition at RT & 15 min anneal
Deposition at 160o
C & no anneal
Compare RTS films to HTS films
Tonset= 212OC
Deposition at 1600C
can be seen as
deposition at RT and
annealing for 15 min
11. 180 190 200 210 220 230 240 250
-0.04
-0.02
0.00
0.02
HeatFlow(W/g)
Temperature (o
C)
15 min
30 min
1 hr
2 hr
Annealing of room temperature deposited films
Tonset1?
Tonset2? Some exothermal peak
in temperature range of
glass transition region?
exothermal peak
disappears as tanneal ↑
Exothermal
peak
12. Conclusions
• For As2S3 films deposited on high temperature
substrates Tonset > that of the films prepared at RT, even
higher than the Tonset of the bulk glass. Thus, glassy films
prepared on heated substrates can be kinetically more
stable as reported previously for organic glasses First
observation of higher kinetic stability of vapor deposited
inorganic glass films than the bulk glass!
• However, the quantification of kinetic stability of
inorganic glass films is complicated by exothermic
relaxation of internal strains in glass structure.
Editor's Notes
Here we report on the stability of As2S3 films, vapor deposited at various Ts.
Schematic representation of the effect of cooling rate on enthalpy and volume. Also shown are the results for conventional, unstable !filled circle", and stable !open circle" vapor-deposited glasses. The dashed and dotted lines represent the extrapolated liquid and glass behavior, respectively. The intersection of these lines defines the fictive temperature of a glass created by cooling a liquid.
The general notion about glass is that vapor deposited films are kinetically and thermodynamically less stable than the bulk form. However, a glass group in the University of Wisconsin-Madison observed that organic glass films vapor deposited onto a substrate at temperature Ts=Tg-50 were more stable than the bulk form…ordinary glass.
It has been observed that thermal deposition of organic TNB [1,3-bis-(1-naphthyl)-5-(2-naphthyl)benzene] at a substrate temperature of Tg-50 leads to films with a substantially higher onset temperature (Tonset) of glass transition, which is the criterion that the University of Wisconsin-Madison group used for kinetic stability. Unexpected result because usually, expect films to have a lower Tg than bulk/aged glass
... The proposed explanation for this kinetic stability is that the increased temperature of the substrate allows for movement of atoms on the nanoscale, which allows the glass to relax and therefore give this higher Tonset.
The goal of our work was to create and characterize films of inorganic, chalcogenide glasses, which have the kinetic stability proposed above. Arsenic sulfide chalcogenide glasses have been widely studied for their optical properties and have been used as materials for waveguides, holographic storage, and photo induced effects. Although much research has been done on films of arsenic sulfide, no one has looked at the stability of the arsenic sulfide films in the same way as that of the University of Wisconsin-Madison group.
An Edwards E306A vacuum deposition system was used to conduct vapor deposition. Glass of particle size 500-700μm was used as the source material during vapor deposition.
The Differential Scanning Calorimeter (DSC) requires about 10mg of sample to provide a meaningful measurement. Due to this restriction, the substrate was moved 6.5 cm away from the source to produce a thick film. Depositions were performed with the aluminum foil as the substrate. The thickness of the film after each deposition was between 6 and 9 μ.
An steel hole punch (5mm diameter) was used to punch holes in the glass film. 18 circles of aluminum foil and film were stacked inside the DSC pan.
DSC analysis (TA Instruments Q2000) of each sample consisted of two scans on a heat-cool-heat-cool measurement program. The first scan (“run 1”) heated the as deposited samples from 10⁰C to 275⁰C at a heating rate of 10⁰C/min to allow the vapor deposited films to completely turn to the superliquid state. The samples were then cooled at a rate of 5⁰C/min to 10⁰C. The second scan (“run 2”) was performed using the same parameters. Run 2 showed the heat flow of a freshly made glass for comparison.
The heat flow signal was used to determine the onset temperature of glass transition Tonset, which is the criterion that the University of Wisconsin-Madison group used for kinetic stability. Slide also shows the method which was used by the TA Analysis software to determine the Tonset.
EDS measurements of the samples were conducted by using the Hitachi 4300 scanning electron microscope. EDS spectra were measured for 4-5 different spots on the sample. The compositions were then averaged for each film to give an average composition per film. As a standard, an EDS spectrum of the bulk glass was measured. The software calculated a chemical composition of 65%S, 35%As, whereas the known composition of this glass is 60%S, 40%As.
By using DSC and EDS spectra Tonset and chemical composition was determined for the bulk sample and several depositions. Films from evaporation 6, 7, and 10 have an increasing As content (compared to the bulk standard). This shows that the composition of the prepared films is arsenic rich. It is important to note that the value of the Tonset of the bulk sample (upon the first run) is significantly higher than the second run, whereas the first and second run of the room temperature substrate films are within a single error range. By comparing our data with Skripachev et al, we see that there is a trend that By deviating from the stoichiometric composition of arsenic sulfide, the Tonset of glass transition decreases.
DSC curve: Run 1 is significantly different for the HT substrate film, than for the LT substrate film Tonset is significantly higher than observed before for 1st run.
Another thing to note is that the second run (fresh glass) for both the RTS and HTS are almost equivalent.
The Tonset of the endothermic event for the as-prepared HTS film is measured to be 216⁰C, while the Tonset of the fresh glass is 198⁰C
We saw that the as-deposited HTS film had a significantly higher Tonset than the film prepared at a room temperature substrate, and even the bulk sample. This higher Tonset falls within the category of kinetically stable, as defined by the University of Wisconsin-Madison group. They had proposed that this increase in the onset temperature is due to increase movement on the nanoscale during the deposition itself (due to the higher substrate temperature). This may be a valid explanation because the temperature of Tg-50 is also the temperature at which glasses are annealed.
An increasing of overshoot in the glass transition endothermic peak occurs for films deposited at higher Ts compared to the glass transition endothermic peak for films deposited at lower Ts. This overshoot shows that the films behave as aged glass on the DSC curve. The higher Tonset and the increase in the Tg overshoot suggest the films prepared at high Ts are kinetically more stable than films deposited at RT, as they require a higher temperature to cause rearrangement of molecules.
Strain energy: at lower Ts, molecules are frozen as they hit the substrate. Ts ↑, molecules can move around during deposition less residual strain in film
Annealing glasses allows the glass itself to relax and remove strains within its structure.
Deposition at 1600C can be seen as deposition at RT and annealing for 15 min.
As tanneal ↑ we get rid of the exothermic reaction
We received our first data point at Tsubstrate = 164⁰C (Tg-50). Film deposited at high substrate temperature fits the criteria of kinetically stable