Kinetic Analysis of Crystallization Processes in In60Se40 Thin Films for Phase Change Memory (Pram) Applications
Irene W. Muchira1, Walter K. Njoroge2, Patrick M. Karimi3
1Department of Electrical and Electronics Engineering, Kirinyaga University College, Kerugoya, Kenya
2Department of Physics, Kenyatta University, Nairobi, Kenya
3Institute of Energy Studies and Research, Nairobi, Kenya
To cite this article:
Irene W. Muchira, Walter K. Njoroge, Patrick M. Karimi. Kinetic Analysis of Crystallization Processes in In60Se40 Thin Films for Phase Change Memory (Pram) Applications. Advances in Materials. Vol. 5, No. 4, 2016, pp. 18-22. doi: 10.11648/j.am.20160504.11
Received: June 23, 2016; Accepted: July 7, 2016; Published: July 28, 2016
Abstract: In the present work, a systematic investigation of crystallization kinetics of In60Se40 alloy has been made. Thin films of In60Se40 alloy were prepared by thermal evaporation using Edward Auto 306 evaporation system. Electrical measurements at room temperature and upon annealing at different heating rates were done by four point probe method using Keithley 2400 source meter interfaced with computer using Lab View software. The dependence of sheet resistance on temperature showed a sudden drop in resistance at a specific temperature corresponding to the transition temperature at which the alloy change from amorphous to crystalline. The transition temperature was also found to increase with the heating rates. From the heating rate dependence of peak crystallization temperature (Tp) the activation energy for crystallization was determined using the Kissinger analysis. The films were found to have an electrical contrast of about six orders of magnitude between the as-deposited and the annealed states, a good quality for PRAM applications. The activation energies were determined to be 0.354 ± 0.018 eV.
Keywords: Chalcogenide Materials, Phase Change Memory, Crystallization Temperature, Transition Temperature
Chalcogenide glasses containing sulfur (s), selenium (se) or tellurium (Te) constitute a rich family of vitreous semiconductors. There has been intense research activity based on these glasses (Lathrop and Eckert 1989; Rao and Mohan 1980; Asokan et al., 1889; Balasubramanian and Rao 1994; Kolobov et al., 2005 and Kumar et al., 2008) in the view of basic physics as well as device technology. The freedom allowed in the preparation of glasses in varied composition brings about changes in their short-range order and thus results in variation of their physical properties. It is therefore easy to tailor their various properties as desired for technological applications.
These materials have a wide range of applications such as making optical fibres, memory devices, reversible phase change optical recording (Suri et al., 2006). Besides the wide commercial device applications like switching, memory and xerography etc of Se, it also exhibits a unique property of reversible transformation (Aggarwal and Sanghera, 2002; Tanaka, 1989) a property that makes it very useful in memory devices. Se based glasses can be heated above glass transition temperature and then quenched to amorphous or cooled slowly to crystalline with each of these two reversible states having distinct optical and electrical properties which is the basis for a digital phase change memory device. It is known that addition of a metal [indium (In), bismuth (Bi), antimony (Sb), silver (Ag), tin (Sn), arsenic (As), lead (Pb), zinc (Zn)] or chalcogen (S, Te) to Se can bring remarkable alterations in properties and reduce ageing. The effects of impurities on the electronic properties of amorphous chalcogenide glasses have been the subjects of serious debate ever since their discovery.
2. Materials and Methods
2.1. Thin Film Deposition
Indium and Selenium were prepared by Thermal evaporation technique. Microscope slides having dimensions 2.54cm x 7.62cm were used as substrates. The substrates were thoroughly cleaned before placing them on substrate holder directly above the evaporation boat. About 0.1g of the fine powder of a mixture of Indium and Selenium was placed on the molybdenum boat and the material heated. The vapour deposited on the substrates forming thin films of In60Se40.
2.2. Sheet Resistance Measurement
Electrical resistivity measurements of thin films were done using the four point probe arrangement adopting the van der pauw method. With a symmetrical square geometry adopted, the probe leads were connected to the Keithly Source Meter for voltage and current measurements. Figure 1 shows a schematic diagram of four-point probe resistivity measurement. Current of 1.0 x 10-10 A was applied through the contacts A and B and potential drop across D and C measured as shown in figure 1 (a). The switching of the probe tips on the sample was done and the same amount of current applied through the contacts A and D. Potential drop was again measured across the contacts B and C as illustrated in figure 1 (b). The measured values of current, voltage drops and film thickness were used to compute sheet resistivity. The sheet resistance was also measured upon annealing. Thermal annealing was done by placing In60Se40 thin films in a quartz tube of the electric furnace where argon was ambient. The furnace was programmed to heat the sample at a particular heating rate. The heating rates ranged from 25°C to 250°C.
Measurement of voltage and current was done using of Keithly SourceMeter which was configured to source current and measure voltage simultaneously on the thin film sample. A four wire remote sensing technique was adopted to minimize errors due to potential drops in the test leads when sourcing or reading voltages. Four wire remote sensing ensured that the programmed voltage was delivered to the thin film under test.
2.3. Determination of Activation Energy of Crystallization
For stoichiometric combination of In60Se40 under investigation different temperature dependence of sheet resistance curves were obtained. The heating rates were varied between 2.5K/min and 12.5K/min. For each run, fresh specimens of the film were used. The peak crystallization temperature for each specific heating rate was obtained from the minimum of the first derivative of sheet resistance. The Kissinger plots were used in determining the activation energies for the sample.
2.4. Measurement of Film Thickness for as Deposited Films
The thickness of the as deposited films was determined using Alpha-Step IQ Surface profiler with a vertical resolution of 0.012 Å and vertical range of 100 Å-0.4 mm. The profiler compares the thickness of a blank glass slide with that having a film. The thickness was found to be 213.0 ± 0.1 nm.
2.5. Sheet Resistance Dependence on Temperature
Sheet resistance for as deposited and annealed films was measured using the four point probe following a procedure proposed by Van Dar Pauw and compared to those obtained for GST (Friendrich et al., 2000). The sheet resistance of as deposited films was 1.696 x 108±0.0005 (Ω/Sq) which is higher compared to that of GST at 0.9 x 108±0.0005 (Ω/Sq) which is widely used for PCM applications.
The obtained values of sheet resistance for as deposited films are approximately twice the sheet resistance of the widely used GST. This is an advantage with respect to PCM applications as high crystalline resistance reduces reset operating power.
2.6. Variation of Sheet Resistance with Temperature for In60Se40
Figure 2 shows variation of sheet resistance upon annealing for In60Se40 thin film at a heating rate of 5k/min. Annealing was done at a temperature range between 25°C to 250°C. Initially the film had a resistance of 1.72 x 108 Ω/Sq which reduced to 1.71 x 106 Ω/Sq. after annealing. The drop in resistance shows that there was a change of phase from amorphous to crystalline, however the change was not abrupt. Transition took place at temperature range between 84°C to 122°C. Transition temperature was obtained by differentiating the results obtained. The differentiated curve is shown in figure 3.
From figure 3 transition temperature for In60Se40 was 105.38 ± 0.54°C. The transition temperatures for In60Se40 was 105.38±0.54 which imply that a PRAM from this alloy will be stable at room temperature since the value of transition temperature is above the room temperature.
2.7. Comparison of Sheet Resistance for as Deposited and as Annealed Films
The table 1 give summary of sheet resistance for as deposited and annealed films. The values have also been compared to those of GST according to Friedrich et al. (2000). These values were used to compute sheet resistivity using the measured values of thickness.
Sheet resistivity for In60Se40 as deposited was 3.61 x 103Ωcm and 6.37 x 10-3Ωcm upon annealing showing an electrical contrast of six orders of magnitude. According to chung et al. (2008), an electrical contrast of at least three orders of magnitude is required for a phase change material. Therefore based on electrical contrast these alloys are suitable for PCM applications.
In60Se40 thin films were annealed at heating rates of 2.5, 5, 7.5, 10 and 12.5K/min within the temperature range of 25°C to 250°C.
2.8. The Activation Energy of Crystallization for In60Se40
The result for crystallization temperature dependence on heating rates for In60Se40 shows a positive shift in transition as the heating rate is increased for the heating rates 2.5, 5, 7.5, 10 and 12.5k/min. Figure 4 shows the various shifts as the heating rates are increased.
Figure 4 show the variation of sheet resistance upon annealing at different heating rates (2.5, 5, 7.5, 10 and 12.5k/min). A positive shift in the transition temperature with increase in heating rates is also observed for this alloy. The crystallization parameters for this alloy are tabulated as shown in table 3.
Figure 5 shows the Kissinger plot for the alloy.
The activation energy for In60Se40 was found to be 0.354 ± 0.018eV.
The values of transition temperature of the alloys at all heating rates: 1, 2.5, 5, 7.5 and 10K/min were much higher than the room temperature. This is an advantage for these alloys because it is essential to prevent self-transition of recording materials between the two phases. Hence one can expect the PCM made from this alloy to remain stable in its amorphous and crystalline states at the room temperature.
Since the crystalline SET state is a stable low resistance state, it is the stability of the quenched high resistance RESET phase which dominates retention issues (Burr et al., 2010). The amorphous phase suffers from two independent resistance altering processes: resistance drift and spontaneous crystallization (Burr et al., 2010). Although resistance drift does not cause any data loss, thermally activated crystallization leads to significant reduction in the resistance of the active layer, causing eventual retention failures for the binary storage.
3. Conclusion and Recommendation
Crystallization kinetics of In60Se40 thin films has been successfully investigated. The value of crystallization temperatures was much higher than room temperature. Therefore a PRAM cell from this material is stable and therefore this alloy is suitable for PRAM applications. These temperatures prevent self-transition of glassy alloys which is required for the application as stable glasses (shukla et al., 2000). The films also showed high crystalline resistance which reduces reset operating power. An electrical contrast of six orders of magnitude was also exhibited. According to chung et al. (2008) an electrical contrast of at least three orders of magnitude is required for a phase change material. A high activation energy of is 0.354 ± 0.018 eV was realised. Activation energy for crystallization is proportional to crystallization temperature and high crystallization temperature leads to high data stability.