A comparative analysis of oxidation rates for thin films of SiGe versus Si

Phys. Status Solidi A, 1–6 (2012) / DOI 10.1002/pssa.201200092 p s sa

statu

s

soli

di

www.pss-a.comph

ysi

ca

ical Section onPhotovoltaics

Part of TopAdvanced Silicon Materials for Electronics and

applications and materials science

A comparative analysis of oxidationrates for thin films of SiGe versus Si

Ethan Long*, Augustinas Galeckas, and Andrej Yu. Kuznetsov

Department of Physics/Centre for Materials Science and Nanotechnology, University of Oslo,

P.O. Box 1048 Blindern, 0316 Oslo, Norway

Received 29 April 2012, revised 8 May 2012, accepted 8 May 2012

Published online 14 August 2012

Keywords nanostructuring, silicon germanium, thermal oxidation, thin films

* Corresponding author: e-mail [email protected], Phone: þ47 22 85 28 40, Fax: þ47 22 85 28 60

In order to evaluate the role of Ge as a catalyst or inhibitor for

the oxidation process in SiGe, oxidation rates for sub-100-nm

films of SiGe are examined and compared to previous reports

and established models for Si oxidation. Values for the Ge

concentration in the pile-up layer at the oxidation interface are

considered as well as the more traditional approach of

considering the Ge content in the as-grown SiGe film. The

experimental results presented here indicate that oxidation rates

for SiGe closely match those of Si and provide evidence that the

presence of Ge in very thin films of SiGe does not lead to

enhanced or retarded oxidation rates as compared to Si. This

comparative analysis is performed with a focus on oxidation of

epitaxial thin films of Si1�xGex in dry O2 at 1 atm at 800, 850,

900, 950, and 1,000 8C.

� 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1 Introduction There has been substantial interest inthe oxidation of silicon germanium (SiGe) as it pertains totransistor and nanowire design and fabrication. Commercia-lization of FINFETs and incorporation of Ge into Si CMOSprocessing has been motivated by SiGe’s utility in creatinghigh mobility channels [1], reduction of parasitic effects likedrain induced barrier lowering in deep sub-micron devices[2], and creation of other novel gate architectures andnanowire transistors [3, 4]. The potential of SiGe core–shellconstructions in photovoltaics is also a new and promisingapplication [5]. All of this adds to the importance of researchinto the dynamics of oxidation of SiGe.

Although the oxidation process of Si is well understoodand described by the Deal–Grove [6] and Massoud et al. [7]models, the oxidation of SiGe is not as well characterized. Avariety of oxidation conditions have been investigated,including dry [8], wet [9], ozone [10], fluorine [11] ambients,oxygen plasma [12], UV assisted oxidation [13], a variety ofpressures [14], initial germanium concentrations, and oxidethicknesses, as well as SiGe-on-insulator [15]. Several workshave used analyses of the flux of Si into the oxide, thediffusion of O in SiO2, and the diffusion of Si in SiGe and Geto model the oxidation process in SiGe [11, 16–19]. Despiteconsiderable effort to study oxidation of SiGe, a general lackof consensus remains regarding oxidation rates for SiGe ascompared to that of Si.

This paper addresses the oxidation rate of SiGe underdry thermal oxidation by examining new data obtainedin the present work and that from literature. Morespecifically, oxidation rates of sub-100-nm films of SiGewith varying Ge concentrations in the initial SiGe layer, aswell as in the pile-up layer, are presented and analyzedalongside results from previous studies as well as establishedmodels for oxidation of Si. An emphasis is placed on thinoxides due to their relevance to modern technologicalapplications.

2 Experimental Epitaxial layers of Si1�xGex on a100 nm Si buffer layer were grown on (100) oriented Sisubstrates by molecular beam epitaxy. Samples were takenfrom two wafers which had Si1�xGex layer thicknesses of80 nm and 70 nm for x¼ 0.15 and x¼ 0.20, respectively.Control samples were taken from a third Si (100) wafer. Thethermal oxidation was performed in a tube furnace atambient pressure and flushed with dry O2. For a givenoxidation time and temperature, three samples, Si,Si0.85Ge0.15, and Si0.80Ge0.20, were oxidized simultaneouslyin order to ensure identical oxidation conditions for the threestoichiometries. Oxidation was carried out at 800, 850, 900,950, and 1000 8C, with the oxidation times adjusted toprovide a series of oxide thicknesses between 0 and 70 nmfor each temperature.


2 E. Long et al.: Oxidation rates contrasted: SiGe versus Sip

hys

ica ssp st

atu

s

solid

i a

Oxide thicknesses were measured by ellipsometry at632.8 nm and 830.0 nm wavelengths and 708 angle ofincidence with each sample being measured five times ateach wavelength in order to account for random error in themeasurements. The native oxides were not removed prior tooxidation but their thicknesses have been taken into accountwhen calculating oxidation rates.

The measured delta and psi data from ellipsometry wasanalyzed by calculating the minimum mean squared error(MSE) as compared to simulated data using a brute forcemethod. The analysis used a three layer model and treated theoxide, pile-up, and SiGe layer thicknesses as well as theindex of refraction for the pile-up layer as unknowns.

The indices of refraction for thermal oxide and the Sisubstrates used in the analysis were determined by using asingle layer model to find the minimum MSE for thecollection of measurements for all of the Si only samples. Inthis way, the indices of refraction for the oxide and substratewere determined from a collection of 270 sets of delta andpsi values.

A total of 83 samples were analyzed with a subset (16samples) being used to check the ellipsometry results againstthose found using Rutherford backscattering spectrometry(RBS). The RBS measurements were done with 2 MeV 4Heþ

ions backscattered into the detector at 1008 relative to theincident beam direction. This glancing-angle detectorgeometry was used to provide enhanced depth resolution.The oxide thicknesses were determined by fitting simulatedcurves to the experimentally determined RBS spectra. Withthe assumption that the RBS values represent the true values,the sample standard deviation in the ellipsometry measure-ments was calculated to be 3.8 nm.

The Ge concentrations of the SiGe layers were measuredusing X-ray diffraction (XRD). A Bruker AXS X-raydiffractometer in double axis configuration and CuKa1radiation was used to conduct 2u–v scans of the (004) peakfor each sample. The incident beam was conditioned toremove CuKa2 radiation using a Gobel mirror and Gemonochromator. A point source detector in ‘‘1D’’ mode wasused on the secondary side to maximize the count rates. Theprofiles from the 2u–v scans were fit using a three layermodel and the Leptos simulation software with data forlattice constants from Dismukes et al. [20] Reciprocal spacemaps focusing on the �(113) and �(224) peaks from alimited number of samples were used to confirm that theSiGe layers were fully strained before and after oxidation.

3 Results and discussion The lack of consensusabout the oxidation rate of SiGe versus that of Si under drythermal oxidation is clear, given reports of it being the sameas [8, 21, 22], faster than [23–27], and slower than [15, 28,29] that of Si. Discussions of either enhanced or retardedoxidation rates for SiGe, relative to that of Si, commonlyfocus on two plausible explanations: firstly, interstitialversus vacancy mechanisms for diffusion of Si in SiGe andthe implications of the Ge concentration dependent latticeparameter of SiGe, and secondly, the difference in bonding


energy of Si to Ge, as opposed to that of Si to Si [8, 18, 22, 23,25, 26].

The models proposed for the oxidation processes in SiGe[11, 16–19] consider the rate at which Si is incorporated intothe oxide relative to the diffusivity of Si in SiGe and Ge.These models rely on accounting for the disposition of Si andGe in the various layers formed by the oxidation process,including consumption of Si and Ge by the oxide. Thesimplifying assumption that the oxide is composed exclu-sively of SiO2 is supported by empiric evidence [14, 19, 21–24, 27–32]. The exceptions to this assumption, wherein bothSiO2 and GeO2 are found, are native oxides [19, 27, 33, 34],those formed under high oxidant pressures [14, 27], andthose formed at low temperatures [11, 32]. Gibbs freeenergies are commonly used to explain the preferentialformation of SiO2 over that of GeO2 [14, 19, 32]. The presentwork focuses on oxidation conditions that lead to theexclusive formation of SiO2, ensuring that the statedsimplifying assumption remains valid for the analysis herein.

The basis of modeling the oxidation process relies on thebalance between the rate at which Si is converted to SiO2 andthe flux of Si from the SiGe layer to the oxidation front. Forthe present discussion, the oxidation front is defined as theinterface between the oxide, SiO2, and the underlying SiGe.Under the oxidation conditions considered here, the rate atwhich Si diffuses through the initial SiGe layer towards theoxidation front is not sufficient to replace all of the Si that isincorporated into the oxide by conversion to SiO2. Thisimbalance will act to increase the Ge content at the oxidationfront and thus increase the diffusion of Si towards theoxidation front until the rate at which Si is consumed byoxidation is balanced with the diffusion of Si in the SiGelayer adjacent to the oxidation front [19]. The region ofenhanced Ge concentration adjacent to the SiO2 growth frontthat results from this process is alternately referred to as apile-up region, snow plow effect, or germanium rich layer.Although the formation of a pile-up region adjacent to theoxidation front is widely accepted and recognized, thosestudies that have investigated the influence of Ge concen-tration on the oxidation rate of SiGe have only considered theGe concentration in the initial SiGe layer and not theconcentration in the pile-up layer at the oxidation interface.If Ge acts as a catalyst or inhibitor for oxidation of SiGe, thenthe critical concentration will be that at the oxidation frontrather than that in the initial SiGe layer. Furthermore, one hasto consider that the diffusivity of Si in SiGe is a functionof the Ge concentration, with the diffusivity increasingsubstantially as Ge concentration increases [35]. If theoxidation rate of SiGe is also a function of Ge concentration,then there will occur a feedback loop between the tendencyof the variable diffusivity of Si in SiGe to change the flux ofSi arriving at the oxidation front by altering the Geconcentration in the pile-up layer and the tendency of theGe concentration at the oxidation front to enhance ordiminish the oxidation rate and thus alter the flux of Si intothe oxide [19]. Consideration of such a feedback loop and therecognition of the existence of a pile-up region with less than

www.pss-a.com

Phys. Status Solidi A (2012) 3

Original

Paper

100% Ge suggests that the oxidation rate of SiGe isindependent of the Ge content at the oxidation interface.

Rather than relying solely on theory and modeling, it isinstructive to consider the available data as a whole. In orderto evaluate the influence of Ge on the oxidation rate, onemust first have an estimate of Ge concentration at theoxidation front. It has been shown that the Ge concentrationin the pile-up layer is significantly higher than, and onlyweakly correlated to, the concentration in the initial SiGelayer [19, 36]. The Ge content in the pile-up is also largelyindependent of the oxide thickness [19, 36]. This leavestemperature as the primary determining factor for the Geconcentration at the oxidation front. Referring to the dataobtained in the present study, Fig. 1 shows the dependenceof the Ge content in the pile-up layer on the oxidationtemperature along with a linear estimation of the correlationbetween the two.

So, when comparing oxide thickness or oxidation rate tooxidation time, SiGe samples oxidized at a given tempera-ture may be treated as having a nearly constant Ge contentat the oxidation interface and may be compared to the Sidata points for the same temperature. Figure 2(a)–(e) showoxide thickness versus time for sub-50-nm oxides for fivetemperatures between 800 and 1000 8C. In each case, datafrom the present study is presented alongside availablevalues from the literature. Predictions for oxidation of Siusing both Deal–Grove and Massoud models using constantsfrom those publications are shown in order to provideperspective. The data presented in Fig. 2(a)–(e) is labeled inaccordance with the Ge content in the initial SiGe layer, asthis has been the approach adopted in most of the literature.

For 800 8C, where the pile-up region is about 64% Ge,Fig. 2(a) shows a significantly larger variation between datafrom different studies than between Si and SiGe. Theliterature data [8] for Si and SiGe with 15% Ge in the initiallayer show essentially the same oxidation rates, which arealso in close agreement with the Massoud and Deal–Grove

Figure 1 (online color at: www.pss-a.com) The Ge concentrationin the pile-up layer, Xpu, of SiGe samples plotted against oxidationtemperature. Data measured by XRD is labeled by the Ge contentin the initial SiGe layer and is presented alongside a linear fit to thedata.

www.pss-a.com

predictions. The oxides from the present study are consist-ently thicker than as predicted by simulations, but, there is nosystematic difference between the oxidation rate in samplesof Si and that in samples of SiGe.

This remains true at 850 8C, where the pile-up region isabout 57% Ge. The data from the present work, in Fig. 2(b),shows SiGe with the same oxidation rate as that of Si, though,the oxides are thicker than as predicted by simulation. Datafrom the literature is limited, but helps emphasize that thereis similar deviation between measured oxide thicknesses andthose predicted by simulation for both SiGe and Si.

At 900 8C, where the pile-up region is about 50% Ge, theliterature data for both SiGe and Si for very short oxidationtimes (less than about 500 s) is in good agreement with thesimulations for Si. At longer oxidation times, there is data forSiGe that agrees very well with the Massoud model as well asdata that shows somewhat higher oxidation rates. The datafrom the present work shows thicker oxides than predicted,but with oxides on Si being slightly thicker than those on20% SiGe and thinner than those on 15% SiGe, though, thisvariation is small as compared to the uncertainty in themeasurements.

For 950 8C, where the pile-up region is about 43% Ge,Fig. 2(d) shows only data from the present work as no datawas located in the literature for these oxidation conditionsand oxide thicknesses. The trends of oxides being thicker forSiGe and Si than as predicted by simulation, as well as therebeing no discernible difference in the oxidation rates for Siand SiGe, remain true for 950 8C.

There is somewhat more literature data available for1000 8C, where the pile-up region is about 36% Ge. The datafrom the present study, shown in Fig. 2(e), is in line with datafrom the literature. The data appears to agree very closelywith simulation for oxide thicknesses below 20 nm, while thedifference between measured and simulated oxide thick-nesses increases with longer oxidation times.

Note that, for all of the temperatures discussed here,the deviation between data from different studies is morepronounced than the deviation between SiGe and Si orbetween SiGe and simulations for Si. These observations arebetter illustrated by considering a wider range of oxidethicknesses. Figure 3 shows data from the present work withvalues collected from the literature as well as predictions foroxidation of Si using both Deal–Grove and Massoud models.The oxide thickness scale for Fig. 3 is increased to just over300 nm in order to explicitly change the perspective on therelation between oxidation rates of Si compared to those ofSiGe. It is immediately clear that there is a large variation inthe values for a variety of initial Ge concentrations, but thatthe data is generally in line with predictions for oxidation ofSi. In this broader view, the correlation between the modelsand the measured data for the thin oxides (<100 nm) that arerelevant to modern technological applications is qualitat-ively quite good. It is critical to point out that values forSi or low initial Ge concentrations (<4% Ge) are scatteredamongst those for medium initial Ge concentrations(5–15% Ge.) Values for high initial Ge concentrations



hys

ica ssp st

atu

s

solid

i a

Figure 2 (online color at: www.pss-a.com) Oxide thickness versus time for SiGe with various initial layer Ge contents and for Si. Thepanels are separated by oxidation temperature as follows: (a) 800 8C, (b) 850 8C, (c) 900 8C, (d) 950 8C, and (e) 1000 8C. Data from thisstudy is presented alongside published data and the Deal–Grove and Massoud models for Si. The data is grouped by Ge content in theinitial SiGe layer. Values from various published works are grouped together to ensure readability in the graphs at the expense of moreprecise correlation between data points and references.

(>20% Ge) are less prevalent but consistent with low initialGe concentrations. There is some data that is indicative ofself-limiting oxidation in oxides thicker than 100 nm, but thisis offset by additional data that shows continued oxidation.These observations should highlight the importance ofconsidering the Ge concentration in the pile-up region rather


than that in the initial SiGe layer when evaluating theinfluence of Ge concentrations on the oxidation rate. Giventhat the Ge concentration in the pile-up layer from anoxidation at 1,000 8C can be estimated as being about 36%,comparison of the data in Fig. 3 to the simulations for Siwould suggest that there is little or no difference in the

www.pss-a.com

Phys. Status Solidi A (2012) 5

Original

Paper

Figure 3 (online color at: www.pss-a.com) Oxide thickness versustime for Si1�xGex (0< x< 0.5) at 1000 8C. The labeling is similarto that in Fig. 2. Oxide thicknesses up to 300 nm are presentedfor comparison to predictions by the Deal–Grove and Massoudmodels for Si.

oxidation rates of Si and SiGe. Or, more precisely, that thevariation in the measurements between studies is signifi-cantly larger than any difference between Si and SiGe.

Another way to view the data is to compare the oxidationrate to the oxide thickness for multiple temperatures.Massoud used this sort of analysis effectively to demonstrateand model high oxidation rates for very thin oxides [7].Oxidation rate data from this study and literature, foroxidation temperatures of 800, 900, and 1000 8C, is shown inFig. 4 alongside predictions for Si using the Massoud modeland assuming the presence of a native oxide layer. As hasalready been stated, there is some deviation between themeasured data and the model predictions. However, the trend

Figure 4 SiGe oxidation rates plotted as a function of oxide thick-ness for oxidation in dry O2 at 1 atm at 1000, 900, and 800 8C. Datafrom the present work is presented alongside data collected fromthe literature as well as model [7] predictions for Si. The Ge contentin the initial SiGe layers is between 8% and 50%.

www.pss-a.com

of increasing oxidation rates with increasing temperature,as well as the tendency for the oxidation rate to fall asoxides thicken, is consistent between the measured data andsimulations. Neither growth rate reduction nor growth rateenhancement is conclusively present for the oxidation rate ofSiGe versus that of Si. It is reasonable to conclude that theMassoud model approximates the oxidation rate of SiGe aswell as it does that of Si for the temperature range between800 and 1000 8C.

4 Conclusions The Ge concentration at the oxidationfront is substantially higher than that in the initial SiGe layerfor thin films of SiGe subjected to dry thermal oxidation. Assuch, the Ge concentration in the pile-up layer is integral todetermination of whether Ge acts as a catalyst or an inhibitorin the oxidation process of SiGe.

With this consideration, once the measurement resultsfrom the present work and from the literature are consideredas a whole, there is no obvious correlation between Geconcentration in the initial SiGe layer or the pile-up layer andoxidation rates of thin films of SiGe. There is substantiallymore variation within the data presented here and betweenstudies than any Ge content induced oxidation rateenhancement or retardation. Use of the Massoud model withmodel parameters for Si give very reasonable and practicalestimations for the oxidation of thin films of SiGe.

Acknowledgements The authors gratefully acknowledge:A. Azarov and F. Kløw at the University of Oslo for providing RBSdata, the NSE-IM2NP team and NanoTecMat platform forproviding SiGe samples, and the Norwegian Research Councilfor financial support through the FRINAT program.

References

[1] K. L. Wang, S. G. Thomas, and M. O. Tanner, J. Mater. Sci.:Mater. Electron. 6, 311 (1995).

[2] S. E. Thompson, M. Armstrong, C. Auth, M. Alavi, M.Buehler, R. Chau, S. Cea, T. Ghani, G. Glass, T. Hoffman,C.-H. Jan, C. Kenyon, J. Klaus, K. Kuhn, Z. Ma, B. Mcintyre,K. Mistry, A. Murthy, B. Obradovic, R. Nagisetty, P. Nguyen,S. Sivakumar, R. Shaheed, L. Shifren, B. Tufts, S. Tyagi, M.Bohr, and Y. El-Mansy, IEEE Trans. Electron Devices 51,1790 (2004).

[3] S. Balakumar, K. D. Buddharaju, B. Tan, S. C. Rustagi, N.Singh, R. Kumar, G. Q. Lo, S. Tripathy, and D. L. Kwong, J.Electron. Mater. 38, 443 (2009).

[4] N. Singh, K. D. Buddharaju, S. K. Manhas, A. Agarwal, S. C.Rustagi, G. Q. Lo, N. Balasubramanian, and D.-L. Kwong,IEEE Trans. Electron Devices 55, 3107 (2008).

[5] A. I. Hochbaum and P. Yang, Chem. Rev. 110, 527 (2010).[6] B. E. Deal and A. S. Grove, J. Appl. Phys. 36, 3770 (1965).[7] H. Z. Massoud, J. D. Plummer, and E. A. Irene, J. Electro-

chem. Soc. 132, 2685 (1985).[8] F. K. LeGoues, R. Rosenberg, and B. S. Meyerson, Appl.

Phys. Lett. 54, 644 (1989).[9] F. K. LeGoues, R. Rosenberg, T. Nguyen, F. Himpsel, and B.

S. Meyerson, J. Appl. Phys. 65, 1724 (1989).[10] J. M. Madsen, Z. Cui, and C. G. Takoudis, J. Appl. Phys. 87,

2046 (2000).



hys

ica ssp st

atu

s

solid

i a

[11] S. J. Kilpatrick, R. J. Jaccodine, and P. E. Thompson, J. Appl.Phys. 81, 8018 (1997).

[12] P. C. Chen, J. Y. Lin, Y. J. Hsu, and H. L. Hwang, Mater. Res.Soc. Symp. Proc. 281, 485 (1993).

[13] L.-P. Chen, Y.-C. Chan, S.-J. Chang, G.-W. Huang, and C.-Y.Chang, Jpn. J. Appl. Phys. 37, L122 (1998).

[14] D. C. Paine, C. Caragianis, and A. F. Schwartzman, J. Appl.Phys. 70, 5076 (1991).

[15] T. Shimura, M. Shimizu, S. Horiuchi, H. Watanabe, K.Yasutake, and M. Umeno, Appl. Phys. Lett. 89, 111923(2006).

[16] M. A. Rabie, Y. M. Haddara, and J. Carette, NSTI-Nanotech.3, 21 (2005).

[17] M. A. Rabie, Y. M. Haddara, and J. Carette, J. Appl. Phys. 98,074904 (2005).

[18] O. W. Holland, C. W. White, and D. Fathy, Appl. Phys. Lett.51, 520 (1987).

[19] E. Long, A. Azarov, F. Kløw, A. Galeckas, A. Y. Kuznetsov,and S. Diplas, J. Appl. Phys. 111, 024308 (2012).

[20] J. P. Dismukes, L. Ekstrom, and R. J. Paff, J. Phys. Chem. 68,3021 (1964).

[21] D. K. Nayak, K. Kamjoo, J. S. Park, J. C. S. Woo, and K. L.Wang, IEEE Trans. Electron Devices 39, 56 (1992).

[22] D. Nayak, K. Kamjoo, J. C. S. Woo, J. S. Park, and K. L.Wang, Appl. Phys. Lett. 56, 66 (1990).

[23] M. Spadafora, G. Privitera, A. Terrasi, S. Scalese, C.Bongiorno, A. Carnera, M. Di Marino, and E. Napolitani,Appl. Phys. Lett. 83, 3713 (2003).

[24] M. Spadafora, A. Terrasi, S. Mirabella, A. Piro, M. G.Grimaldi, S. Scalese, E. Napolitani, M. Di Marino, D. DeSalvador, and A. Carnera, Mater. Sci. Semicond. Process. 8,219 (2005).

[25] T. Shimura, Y. Okamoto, D. Shimokawa, T. Inoue, T. Hosoi,and H. Watanabe, ECS Trans. 33, 893 (2010).


[26] T. Shimura, Y. Okamoto, T. Inoue, T. Hosoi, and H. Wata-nabe, Phys. Rev. B 81, 033308 (2010).

[27] S. J. Kilpatrick, R. J. Jaccodine, and P. E. Thompson, J. Appl.Phys. 93, 4896 (2003).

[28] J. P. Zhang, P. L. F. Hemment, S. M. Newstead, A. R. Powell,T. E. Whall, and E. H. C. Parker, Thin Solid Films 222, 141(1992).

[29] Y. Zhang, C. Li, K. Cai, Y. Chen, S. Chen, H. Lai, andJ. Kang, J. Appl. Phys. 106, 063508 (2009).

[30] Y. S. Lim, J. S. Jeong, J. Y. Lee, H. S. Kim, H. K. Shon, H. K.Kim, and D. W. Moon, Appl. Phys. Lett. 79, 3606 (2001).

[31] Y. S. Lim, J. S. Jeong, J. Y. Lee, H. S. Kim, H. K. Shon, H. K.Kim, and D. W. Moon, J. Electron. Mater. 31, 529 (2002).

[32] M. K. Bera, S. Chakraborty, R. Das, G. K. Dalapati, S.Chattopadhyay, S. K. Samanta, W. J. Yoo, A. K. Chakraborty,Y. Butenko, L. Siller, M. R. C. Hunt, S. Saha, and C. K. Maiti,J. Vac. Sci. Technol. A 24, 84 (2006).

[33] I.-M. Lee and C. G. Takoudis, J. Vac. Sci. Technol. A 15,3154 (1997).

[34] R. J. Jaccodine and S. J. Kilpatrick, Proc. Electrochem. Soc.2005–05, 542 (2005).

[35] R. Kube, H. Bracht, J. Lundsgaard Hansen, A. NylandstedLarsen, E. E. Haller, S. Paul, and W. Lerch, J. Appl. Phys.107, 073520 (2010).

[36] E. Long, A. Galeckas, and A. Yu Kuznetsov, J. Vac. Sci.Technol. B 30(4), 041212 (2012).

[37] N. Daval, E. Guiot, K. K. Bourdelle, M. Kennard, I. Cayrefourcq,T. Akatsu, C. Mazure, H. Cerva, and A. Rucki, CrystallineDefects and Contamination: Their Impact and Control in DeviceManufacturing IV DECON 2005, edited by B. O. Kolbesen(Electrochemical Society, Pennington, 2005), Vol. 2005–10,pp. 42–51.

[38] B. G. Min, J. H. Yoo, H. C. Sohn, D. H. Ko, M. H. Cho, K. B.Chung, and T. W. Lee, Thin Solid Films 518, 2065 (2010).

www.pss-a.com

A comparative analysis of oxidation rates for thin films of SiGe versus Si

Documents

Transcript of A comparative analysis of oxidation rates for thin films of SiGe versus Si