Si structures

Applied Surface Science 253 (2007) 7957–7963 www.elsevier.com/locate/apsusc

Analysis of excimer laser annealing of amorphous SiGe on La2O3//Si structures L. Fornarini a,*, J.C. Conde b,1, S.Chiussi b, P. Gonzalez b, B. Leon b, S. Martelli c b

a Enea-Frascati, Via Enrico Fermi 45, I-00044 Frascati (Roma), Italy Dpto. Fı´sica Aplicada, E.T.S.I.I. University of Vigo, Lagoas Marcosende 9, E-36200 Vigo, Spain c Centro Sviluppo Materiali,Via di Castel Romano 100, I-00128 Roma, Italy

Available online 24 February 2007

Abstract The reduction of complementary metal oxide semiconductor dimensions through transistor scaling is in part limited by the SiO2 dielectric layer thickness. Among the materials evaluated as alternative gate dielectrics one of the leading candidate is La2O3 due to its high permittivity and thermodynamic stability. However, during device processing, thermal annealing can promote deleterious interactions between the silicon substrate and the high-k dielectric degrading the desired oxide insulating properties. The possibility to grow poly-SiGe on top of La2O3//Si by laser assisted techniques therefore seems to be very attractive. Low thermal budget techniques such as pulsed laser deposition and crystallization can be a good choice to reduce possible interface modifications due to their localized and limited thermal effect. In this work the laser annealing by ArF excimer laser irradiation of amorphous SiGe grown on La2O3//Si has been analysed theoretically by a numerical model based on the heat conduction differential equation with the aim to control possible modifications at the La2O3//Si interface. Simulations have been carried out using different laser energy densities (0.26–0.58 J/cm2), different La2O3 film thickness (5–20 nm) and a 50 nm, 30 nm thick amorphous SiGe layer. The temperature distributions have been studied in both the two films and substrate, the melting depth and interfaces temperature have been evaluated. The fluences ranges for which the interfaces start to melt have been calculated for the different configurations. Thermal profiles and interfaces melting point have shown to be sensitive to the thickness of the La2O3 film, the thicker the film the lower the temperature at Si interface. Good agreement between theoretical and preliminary experimental data has been found. According to our results the oxide degradation is not expected during the laser crystallization of amorphous Si0.7Ge0.3 for the examined ranges of film thickness and fluences. # 2007 Elsevier B.V. All rights reserved. Keywords: Poly SiGe; La2O3; Excimer laser annealing; Crystallization

1. Introduction During the last few years a big effort has been produced worldwide for making excimer laser annealing (ELA) a reliable process to be used in complementary metal oxide semiconductor (CMOS) technology. In fact the technique offers many advantages compared to conventional thermal anneal such as junction depth control and high dopant activation [1,2].

* Corresponding author. Tel.: +39 06 94005196; fax: +39 06 94005312. E-mail address: [email protected] (L. Fornarini). 1 Enea guest. 0169-4332/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2007.02.063

The technique involves very complex physical mechanisms and it is based on the laser irradiation of amorphous films followed by the solidification process [3–5]. The application of the correct amount of energy density cause the melting of a film area with a steep thermal gradient and a sharp transition from the liquid to the crystalline phase [6–12]. It is known the dielectric properties and thermodynamic stability of La2O3 can be modified during conventional thermal device processing, interdiffusion and interfacial phases can take place so degrading the oxide properties. In the case of ELA such problems are much reduced and possibly avoided. The sample is heated just for few tens of nanoseconds and due to the short times involved any

7958

L. Fornarini et al. / Applied Surface Science 253 (2007) 7957–7963

interdiffusion process can take place only if the material reaches the melting point (MP) [13]. In fact only in the melted material diffusion rates are high enough to cause species interdiffusion and therefore interface degradation. In order to improve the excimer laser annealing it is necessary to accurately predict the thermal profile of the process in order to know for which, fluence and film thickness the materials at the interface could reach the melting point. In this study we have considered as initial hypothesis for the La2O3//Si interface stability that the melting temperature must not be reached in this region. By means of accurate numerical analysis we explored the threshold laser energy densities for which, the melting point is reached at the interfaces for the structures: La2O3//Si, La2O3// SiO2//Si and amorphous SiGe (a-SiGe)//La2O3//Si, and different a-SiGe and La2O3 film thickness. 2. Excimer laser annealing: numerical analysis When an absorbing solid is irradiated by a laser pulse, energy is initially absorbed by the electronic system and then transferred to the lattice. The absorbed energy is nearly instantaneously converted in local heat, which can diffuse in the sample by thermal conduction. If laser beam size is much larger than the absorption length, it is homogeneous in energy and the lateral heat diffusion is neglected we can describe the heating process by the onedimensional heat diffusion equation [14]: rðTÞc p ðTÞ

@T ¼ r  ðKðTÞrTÞ þ Sðx; tÞ; @t

(1)

where K(T) is the thermal conductivity, T the temperature, r(T) the density, cp(T) the specific heat capacity. S(x, t) = (1  R(T))P(t) ea(T)x is the heat generation function at position x and at the time t, which depends on the optical absorption coefficient a(T), the surface reflectivity R(T) and the time evolution of the laser beam power P(t). If vaporization and radiative losses are neglected the following boundary conditions hold @T ¼ 0; Tðx; tÞjx;t¼0 ¼ T in ; Tðx; tÞjx ! 1;t ¼ T in ; @x x¼0;t where Tin is the initial temperature.

Due to the dependence of the material properties from temperature and the presence of phase transitions during the process the equation cannot be analytically solved and the temperature distribution can be determined only by means of numerical methods. To calculate the temperature profiles we made use of the finite elements method [15–19] using the ANSYS1 (10.0) program [20], the temperature field in the isotropic material is obtained in the nodes of a previously defined mesh (2000 rectangular elements and 3003 nodes; vertical distance between nodes of 5 nm and the horizontal length of 40 nm with a central node). Thermal and optical properties of a-SiGe and Si are well known [21–28]. Focusing our attention on La2O3, the value of its thermal conductivity is only reported for a compositional range of 0–30 mol% of LaO1.5 of the thorium–lanthanum mixed oxide (ThO2–LaO1.5) [29–31]. In order to approximate the lanthanum oxide thermal conductivity, we have extrapolated these values from 30 to 100 mol% LaO1.5. The lanthana optical properties are obtained from literature [32–34]. The thermo-physical properties of materials used in this paper are shown in Table 1. In Table 2 the optical properties are reported for the wavelength (l) 193 nm and at room temperature. 2.1. La2O3//Si theoretical analysis and preliminary experimental results The numerical analysis was carried out for different La2O3 film thickness (5, 10, 15 and 20 nm) deposited on a crystalline Si substrate (5 mm) and with a laser beam (wavelength l = 193 nm, pulse length FMHW = 20 ns, energy density (F) 0.080 < F (J/cm2) < 0.32) impinging at normal incidence with respect to the film surface. Fig. 1(a) and (b) shows the results of temperature versus time for the energy density of 0.26 J/cm2 and La2O3 film thickness of 5 and 10 nm respectively at different depths. Up to the energy density of 0.26 J/cm2 and the examined range of La2O3 film thickness neither the oxide film nor the silicon surface is melted. At the laser energy density of 0.26 J/ cm2 and La2O3 film thickness of 5 nm the Si begins to melt at the interface with La2O3. See Fig. 1(a). At the same energy density but for La2O3 film thickness of 10 nm the La2O3 free surface starts melting while the Si interface is still completely solid. See Fig. 1(b).

Table 1 Thermal and physical properties of materials used in this paper Material (reference)

Density, r (kg/m3)

Thermal conductivity, K (W/m K)

Specific heat, Cp (J/(kg K)

Melting temperature, Tm (K)

La2O3 [35,36]

6510

1.08  1.906  104T + 2.283  108T2

2578

231.14  103

SiO2 [35,37,38]

2200

1940

140.0  103

a-Si0.7Ge0.3 [21,22]

3240

1.15 + 1.343  103 (T  300) [300–1000 K] 2.09 [>1000 K] 10.503  1.4332  102T + 7.7068  106T2

Si [28]

2320 [Tm]

370 + 39.5  103T  42  105/T3 [300–2578 K] 931 + 0.256  T  24.0  106 T2 [300–2000 K] 406 + 0.587 T  0.562 103T2 + 2.2 107T3 699.55 Exp [2.375  104T) [T > 300 K] 1.05 [>Tm]

1.525  105 T1.226 [T < 1200 K] 901 T0 502 [T > 1200 K] 5.3 + 2.93  102 (T  Tm) [>Tm]

Latent heat of fusion (J/Kg)

1543.4

1413.7  103

1683

1801.0  103

L. Fornarini et al. / Applied Surface Science 253 (2007) 7957–7963

7959

Table 2 Optical properties of materials used in this paper, l = 193 nm,T = 300 K Material

Index of refraction, h

Extinction coefficient, k

Absorption coefficient, a (cm1)

Optical penetration depth, dO (106 cm)

La2O3 Si SiO2 a-Si0.7Ge0.3

2.35 0.883 1.00 1.01

0.19 2.09 2.77 2.46

1.23  10 5 1.36  10 6 1.80  10 6 1.6  10 6

8.08 0.73 0.55 0.63

Increasing the film thickness up to 15 nm the La2O3 film is totally melted while the temperature at the interface is further decreased. Fig. 1(c)–(f) shows the evolution of the melting depth versus time for F = 0.3 J/cm2 and 0.32 J/cm2 and La2O3 film thickness of 5, 10, 15 and 20 nm, respectively.

As it is shown in these figures, when the La2O3 film thickness is kept constant and the laser energy is increased from 0.30 to 0.32 J/cm2 the melting time of the La2O3 surface does not considerably change while the melting time at the Si interface is significantly increased. Accordingly, the melting depth remains unchanged in

Fig. 1. La2O3//Si system, 5, 10, 15, 20 nm thick La2O3 film: (a and b) temperature vs. time at 0.26 J/cm2 for 5 and 10 nm La2O3 thick, respectively; (c–f) melting front for 5, 10, 15, 20 nm thick La2O3 film and 0.26 < F (J/cm2) < 0.32.

7960

L. Fornarini et al. / Applied Surface Science 253 (2007) 7957–7963

Fig. 2. TOF-SIMS (ION-TOF) analysis of 10 nm La2O3 film deposited on a Si substrate and irradiated with a ArF-Excimer laser (193 nm, FMHW = 20 ns) with different energy densities.

the La2O3 film but it considerably increases in the Si substrate. On the other hand, if the La2O3 film thickness is increased while keeping the same laser energy density, the behaviour discussed above is reversed: with the growth of the La2O3 film both the melting time and melting depth of the La2O3 film increase; while at the Si interface neither the melting time nor the melting depth appreciably change. Results indicate two main factors contribute to the temperature distributions in the two layers: La2O3 film thickness and the laser energy density.

Such findings have been compared to preliminary experimental data. In order to confirm theoretical data, a 10 nm La2O3 was deposited on a Si wafer and irradiated by a Lambda Physik 220i ArF-Excimer laser (193 nm and pulse duration of 20 ns) with different energy densities. Detailed information on the experimental set-up used for processing the samples can be found in Ref. [39]. Interface modifications have been analysed by means of TOF-SIMS measurements. They have been performed with a TOF-SIMS IV (ION-TOF) using a Ga gun for analysing in a 100mm  100 mm square of a 300 mm  300 mm crater achieved with Cs gun sputtering. Results of the analysis are shown in Fig. 2 for different laser energy densities. They show up to 0.25 J/cm2 little modifications can be noticed. At the fluence of 0.25 J/cm2 Si start to diffuse into the La2O3 layer, while at 0.35 J/cm2 a complete intermixing of Si and La2O3 is detected. According to theoretical data (Fig. 1(d)) the La2O3 film start melting at the fluence of 0.30 J/cm2 and at 0.35 J/cm2 it is completed melted. At this fluence a complete intermixing of the two materials is similarly expected as it appears from the experimental data reported in Fig. 2. From this figure we can also assume that as soon as the silicon substrate start melting at the interface Si is able to propagate into the oxide film. 2.2. La2O3//SiO2//Si theoretical analysis As a thin SiO2 layer is often deposited between the Si and the lanthana film in order to prevent the intermixing of the two

Fig. 3. La2O3//SiO2//Si system, for 10 and 15 nm thick La2O3 film and 5 nm thick SiO2: 3a, 3b: temperature vs. time at 0.26 J/cm2, 10 and 15 nm thick La2O3 film, respectively; 3c, 3d: melting front for 10 and 15 nm thick La2O3 respectively, 0.26 < F (J/cm2) < 0.32.

L. Fornarini et al. / Applied Surface Science 253 (2007) 7957–7963

materials during thermal processing, a thin layer of silicon dioxide (5 nm) has then been introduced between the La2O3 film and the substrate in our calculations to check for the differences in the thermal process produced by the addition of this layer. The numerical analysis was carried out at the same laser conditions as those used in the previous case. Fig. 3(a)–(d) shows the temperature versus time and the evolution of the melting depth versus time for a La2O3 film thickness of 10 and 15 nm obtained when the SiO2 layer is interposed between the lanthanum oxide film and the Si substrate.

7961

This system presents a thermal behaviour similar to the one previously examined. The energy density of 0.26 J/cm2 is the threshold energy for the melting of a La2O3 film 10 nm thick. The increase of the thickness of this film causes an increasing in the melting time at the free surface and melting depth inside the La2O3 film. It can be observed in Fig. 3(a) and (b) that for the fluences in the range from 0.26 to 0.32 J/cm2 the temperature at the La2O3// SiO2 interface is always lower than the SiO2 MP. The energy threshold for the Si MP at the SiO2//Si interface, is reached for F > 0.26 J/cm2.

Fig. 4. a-Si0.7Ge0.3//La2O3//Si system, 30 nm thick a-Si0.7Ge0.3 on 5 nm thick La2O3 and 50 nm thick a-Si0.7Ge0.3 on 20 nm thick La2O3: (a) 30 nm thick a-Si0.7Ge0.3: temperature vs. time at 0.36 J/cm2 and at different depths; (b): melting front for 30 nm thick a-Si0.7Ge0.3, 0.36 < F (J/cm2) < 0.56; (c and d) 50 nm thick a-Si0.7Ge0.3: temperature versus time at 0.28 and 0.47 J/cm2 respectively and at different depths; (e and f) melting front for 50 nm thick a-Si0.7Ge0.3, 0.26 < F (J/cm2) < 0.50.

7962

L. Fornarini et al. / Applied Surface Science 253 (2007) 7957–7963

If the La2O3 thickness is kept constant, an increase in the energy density, from 0.30 to 0.32 J/cm2, does not change significantly neither the melting time at the La2O3 surface nor the melting depth in the La2O3 film. An effective increment of the melting time at the Si interface and an increase in the melting depth is on the other hand observed in the Si substrate, see Fig. 3(c) and (d). As in the previous point, if the energy density is kept constant, the melting time of the La2O3 film and the melting depth are increased with the growth of the La2O3 layer, but at the SiO2//Si interface the melting time and the melting depth of Si are practically the same. If the previous La2O3//Si system is compared to the La2O3// SiO2//Si system, it is possible to observe the introduction of the SiO2 layer has only slightly increased the temperatures at the La2O3 surface (see Figs. 1(b) and 3(a)). Comparing Figs. 1(e) and (f) and 3(c) and (d) (same total oxide thickness) it is also possible to observe the Si melting front at their respective interfaces present a similar shape and the melting time is comparable in the two cases. On the other hand the temperature at the La2O3//Si interface for a film thickness of 15 or 20 nm (Fig. 1(e) and (f)) is close to the temperature of the La2O3 + SiO2//Si interface at 10 nm + 5 nm or 15 nm + 5 nm, respectively (Fig. 3(c) and (d)). These results could be due to the fact the thermal conductivity of SiO2 is lower than the silicon one while the thermal conductivity and enthalpy of fusion of the La2O3 layer are close to those of the SiO2 film (Table 1). Fig. 3(c) and (d) shows at the La2O3//SiO2 the two oxides do not reach the MP in this case indicating that the thin layer of SiO2 can actually improve the quality of the oxide interface. 2.3. a-SiGe//La2O3//Si theoretical analysis The temperature distribution and the melting front are studied for the a-Si(1x)Ge(x)//La2O3//Si system. The aSi(1x)Ge(x) melting point (1543.4 K) was calculated through the linear approximation: MPSi(1x)Ge(x) = MPSi  (1  x) + MPGe  (x) (where x = 0.3, is the fractional composition of germanium). The laser energy density, 0.26 < F (J/cm2) < 0.56, is applied to different thickness of a-Si0.7 Ge0.3: 30 and 50 nm and on a 20 and 5 nm thick La2O3 film, respectively (Si bulk: 5 mm). Fig. 4(a)–(f) shows the results obtained after one pulse laser irradiation. For a a-Si0.7Ge0.3 film 30 nm thick, the threshold energy necessary to reach the a-Si0.7Ge0.3 surface MP is about 0.36 J/ cm2 as it is shown in Fig. 4(a) and (b). Fig. 4(b) present the evolution of the melting depth versus time for energy densities in the range 0.36 < F (J/cm2) < 0.56. The a-Si0.7Ge0.3 film completely melts at the laser fluence of about 0.42 J/cm2. The La2O3 and Si MP at their respective interfaces are not reached in this energy range. Increasing the a-Si0.7Ge0.3 and La2O3film thickness to 50 and 20 nm, respectively, we observe that a-Si0.7Ge0.3 MP is reached at its surface with energies lower than 0.26 J/cm2 while it completely melts at 0.28 J/cm2, Fig. 4(c).

The threshold to reach the La2O3 MP at the a-Si0.7Ge0.3// La2O3 interface is about 0.47 J/cm2, Fig. 4(d). As it is shown in Fig. 4(e) and (f), it is possible to notice that increasing the laser energy the temperature both at the surface and at the two interfaces increases as well as the melting time and melting depth for each material. Comparing these results between them and with those of the previous sections they evidence the a-Si0.7Ge0.3 film has a lower thermal conductivity with respect to the lanthanum oxide. In fact the light at this wavelength (193 nm) is completely absorbed in the first 10 nm of a-Si0.7Ge0.3 and the difference in temperature is mainly due to the different thermal properties between the two films. With the here examined film thickness, the simulations show the Si MP is not reached at the La2O3//Si interface in the range 0.26 < F (J/cm2) < 0.50. 3. Conclusions The above results allow us to obtain the energy threshold to reach the Si MP at the different interfaces and knowing if interface degrading can be possible in the different systems. La2O3//Si laser annealing shows that for energies lower than 0.26 J/cm2 interface modification is very difficult while interface damage would be possible in the 0.30 < F (J/cm2) < 0.32 energy range, but it is unlikely because the La2O3 MP is not reached at this interface. These results are in good agreement with preliminary experimental data, as it appears comparing results reported in Fig. 1(d) with those shown in Fig. 2. La2O3//SiO2//Si laser annealing shows in the range 0.26 < F (J/cm2) < 0.30 the interface modification is rather unlikely at the La2O3//SiO2 interface for La2O3 thickness of 10 and 15 nm. As in the previous case, the La2O3 MP is not reached at this interface. While the Si MP is reached in whole this energy range, the SiO2 remains always in the solid phase. The introduction of a thin SiO2 film seems to prevent the degradation of the La2O3 interface. a-Si0.7Ge0.3//La2O3//Si laser annealing: the threshold energy densities necessary to cause melting of the a-Si0.7Ge0.3 at the aSi0.7Ge0.3//La2O3 interface (i.e. total a-SiGe film melting) is about 0.44 J/cm2 for a lanthana thickness of 30 nm. Increasing the a-Si0.7Ge0.3 thickness to 50 nm, the threshold energy to reach a-Si0.7Ge0.3 MP at the a-Si0.7Ge0.3//La2O3 interface decreases and it is about 0.28 J/cm2. The La2O3 and Si MP are not reached in both cases. The energy density for which the La2O3 or Si start melting are much higher than the respective energies necessary to totally melt the a-Si0.7Ge0.3 layer therefore the oxide degradation is not expected in the energy densities ranges of interest for a-SiGe crystallization. More experiments are in progress to validate the theoretical results. References [1] S. Sedky, A. Witvrouw, K. Baert, Sens. Actuators A: Phys. 97/98 (2002) 503–511.

L. Fornarini et al. / Applied Surface Science 253 (2007) 7957–7963 [2] G. Fortunato, V. Privitera, A. La Magna, L. Mariucci, Cuscuna`, B.G. Svensson, E. Monakhov, M. Camalleri, A. Magrı`, D. Salinas, F. Simon, Thin Solid Films 504 (1/2) (2006) 2–6. [3] L.D. Laude, Excimer Lasers, Kluwer Academic Publishers, The Netherlands, 1994. [4] D.B. Chrisey, G.K. Hubler (Eds.), Pulsed Laser Deposition of Thin Films, John Wiley & Sons, New York, 1994. [5] J.C. Miller, R.F. Haglund, Laser Ablation and Desorption, Academic Press, USA, 1998. [6] G.L. McVay, A.R. DuCharme, Phys. Rev. B 9 (2) (1974) 627–631. [7] F. Foulon, E. Fogarassy, A. Slaoui, C. Fuchs, S. Unamuno, P. Siffert, Appl. Phys. A 45 (1988) 361–364. [8] S. de Unamuno, E. Fogarassy, Appl. Surf. Sci. 36 (1989) 1–11. [9] S.C. Jain, J.R. Willis, R. Bullough, Adv. Phys. 39 (2) (1990) 127–190. [10] R. Serna, A. Blanco, T. Missana, J. Solis, C.N. Afonso, A. Rodrı´guez, T. Rodrı´guez, M.F. da Silva, Appl. Phys. Lett. 68 (13) (1996) 1–3. [11] P. Boher, J.L. Stehle, E. Fogarassy, Appl. Surf. Sci. 138/139 (1999) 199– 205. [12] H.J. Osten, Thin Solid Films 367 (2000) 101–111. [13] R.F. Woods, J.R. Kirrkpatrick, G.E. Giles, Phys. Rev. B 23 (1981) 5555. [14] H.S. Carslow, J.C. Jaeger, Conduction of Heat in Solids, Oxford University Press, 1959. [15] L. Calcagnile, M.G. Grimaldi, P. Baeri, J. Appl. Phys. 76 (3) (1994) 1833– 1839. [16] D. Klinger, J. Auleytner, D. Zymierska, Cryst. Res. Technol. 32 (7) (1997) 983–987. [17] I.V. Grozescu, W.M.M. Yunus, M.M. Moksin, J. Phys. D: Appl. Phys. 33 (2000) 677. [18] J.C. Conde, P. Gonza´lez, F. Lusquin˜os, S. Chiussi, J. Serra, B. Leo´n, Appl. Surf. Sci. 248 (2005) 455. [19] J.C. Conde, P. Gonza´lez, F. Lusquin˜os, S. Chiussi, J. Serra, B. Leo´n, Appl. Surf. Sci. 248 (2005) 461.

7963

[20] ANSYS1, Analysis Guides and others, First Edition, SAS IP, Inc. #, 2006. [21] B. Abeles, Phys. Rev. 131 (5) (1963) 1096–1911. [22] J.P. Dismukes, L. Ekstrom, E.F. Steigmeier, I. Kudman, D.S. Beers, J. Appl. Phys. 35 (10) (1964) 2899–2907. [23] A.B. Djurisic, E. Herbert Li, Semicond. Sci. Technol. 16 (2001) 59–65. [24] R. Ahuja, C. Persson, A.F. da Silva, J.S. de Almeida, C.M. Araujo, B. Johansson, J. Appl. Phys. 93 (7) (2003) 3832. [25] A. Cuadras, J. Sancho, S. Bosch, B. Garrido, J.R. Morate, L. Fonseca, K. Pressel, Microelectron. Eng. 72 (2004) 185–190. [26] W. Szyszko, Appl. Surf. Sci. 90 (1995) 325–331. [27] M.W. Chase, NIST-JANAF Thermo-Chemical Tables, 4th ed., American Chemical Society and the American Institute of Physics, USA, 1998, pp. 1881–1887. [28] E.D. Palik, Handbook of Optical Constants of Solids, Academic Press, Inc., USA, 1985, pp. 465–478. [29] S. Fukushima, T. Ohmichi, M. Handa, J. Less-Common Met. 121 (1986) 631. [30] A. Lokwood, C. Wood, J. Vandersande, A. Zoltan, L. Danielson, V. Raag, D. Wittenberger, J. Less-Common Met. 126 (1986) 113. [31] P. Srirama, C.K. Mathews, J. Phys. D: Appl. Phys. 24 (1991) 2202. [32] G. Hass, J.B. Ramsey, R. Thun, J. Opt. Soc. Am. 49 (1958) 116. [33] T. Marcinow, K. Truszkowsha, Appl. Opt. 20 (1981) (1755). [34] A.A. Dakhel, J. Alloy. Compd. (2003) 233. [35] G.V. Samsonov, The Oxide Handbook, IFI/Plenum, 1973. [36] Y.S. Touloukian, Thermophysical Properties of High Temperature Solid Materials, No. 4, Macmillan Company, New York, 1967, pp. 226–233. [37] R.C. Weast, D.R. Lide, M.J. Astle, W.H. Beyer, CRC Handbook of Chemistry and Physics, CRC Press, Inc., Boca Raton, FL, 1989/1990. [38] V.E. Borisenko, P.J. Hesketh, Rapid Thermal Processing of Semiconductors, Plenum Press, New York, 1997. [39] E. Lo´pez, S. Chiussi, J. Serra, P. Gonza´lez, B. Leo´n, Thin Solid Films 508 (2006) 48–52.