Nanomechanical Properties of Standard and Strained

0 downloads 0 Views 179KB Size Report
bonded to a handle wafer via an insulating layer of 145 nm SiO2. SOI wafers are fabricated using wafer bonding and the hydrogen implantation ion cut layer ...
ECS Transactions, 64 (5) 27-32 (2014) 10.1149/06405.0027ecst ©The Electrochemical Society

Nanomechanical Properties of Standard and Strained SOI Films Fabricated by Wafer Bonding and Layer Splitting M.A. Mamun1,2, K. Zhang2,3, H. Baumgart2,3, and A.A. Elmustafa1,2 Department of Mechanical and Aerospace Engineering, Old Dominion University, Norfolk, Virginia 23529, USA 2 Department of Electrical and Computer Engineering, Old Dominion University, Norfolk, Virginia 23529, USA 3 Applied Research Center, Thomas Jefferson Lab, Newport News, Virginia 23606, USA 1

Biaxial tensile strained Si layers grown epitaxially on SiGe have attracted renewed attention for high performance CMOS devices. Strain engineering for mobility enhancement is an attractive option to improve CMOS device performance. Epitaxial Si1-xGex and most recently Ge-free bi-axially strained sSOI are among the serious candidates for higher mobility channel material. We demonstrate how the fabrication of thin single crystal Si devices by wafer bonding and hydrogen ion cut film exfoliation technique is affecting their nanomechanical properties compared to bulk single crystal Si. Although both the handle wafer and the bonded thin device layer are single crystal Si, it turns out that their nanomechanical properties differ considerably. Using nanoindentation, we measured the properties of SOI and bi-axially tensile strained sSOI and benchmarked the results against single crystal bulk Si. The experimental results indicate that the hardness of the strained thin Si films varies significantly. Introduction

High demands for improved performance, better power consumption, and miniaturized devices prompted the migration to SOI technology. The SOI fabrication process helps in achieving greater performance and offers less power consumption compared to the bulk Si Process. In SOI fabrication technology, transistors are built on a single crystal Si layer bonded to a handle wafer via an insulating layer of 145 nm SiO2. SOI wafers are fabricated using wafer bonding and the hydrogen implantation ion cut layer splitting technology. Strained Si-on-Insulator (sSOI) technology has been developed for high mobility channel devices [1,2,3]. Biaxial tensile strained Si layers grown epitaxially on SiGe substrates have attracted considerable attention for high performance CMOS devices. Strain engineering for mobility enhancement is an attractive option to improve CMOS device performance by introducing lattice strain into the Si channel. In this paper, we have analyzed how the fabrication of very thin single crystal Si device films by wafer bonding and ion cut film exfoliation technique through hydrogen implantation is affecting their nanomechanical properties in comparison to bulk single crystal Si. Although both the handle wafer and the bonded thin device layer are single crystal Si, it turns out that their nanomechanical properties differ considerably. Using nanoindentation, we measured the properties of regular relaxed non-strained SOI films of

27

Downloaded on 2015-10-20 to IP 54.152.109.166 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

ECS Transactions, 64 (5) 27-32 (2014)

88nmthickness and bi-axially tensile strained Silicon-on-Insulator layers (sSOI 15nm, 70nm, and 100 nm) and benchmarked the results against bulk Si.

Sample Fabrication

A schematic process flow of the ion cut fabrication process is depicted in Figure 1. Biaxially tensile strained SOI (sSOI) films were obtained by epitaxially growing Si films on a relaxed Si1-xGex donor wafer. The sSOI films for this study were grown epitaxially on a Si0.8Ge0.2 substrate. However, the highly strained SOI (sSOI) sample of 15 nm thickness was grown epitaxially on a Si0.6Ge0.4 substrate. Subsequently the donor wafer was split off with the hydrogen implantation ion cut exfoliation technique. Finally the cleaved surface was finished with etching processes to completely remove all traces of the straining Si1-xGex layer. In our case the bi-axially tensile strained sSOI films were obtained with a fabrication sequence of epitaxially growing 150Å to 1000Å strained Si films on a relaxed 20% Ge containing Si1- xGex buffer layer on a donor wafer. During epitaxy the Si lattice stretches to match the larger Si1-xGex lattice. The larger lattice constant of Ge produces a 4.1% lattice mismatch with the Si crystal. Following successful bonding of both wafers, the donor wafer was split off with the ion cut exfoliation technique. The surface is then finished with an etching process to completely remove all traces of the Si1-xGex film, resulting in a Ge-free bi-axially strained Si film on amorphous SiO2 insulator [4, 5]. The insulating buried oxide beneath the SOI film is on the order of 145 nm thick.

Figure 1. Schematic depiction of wafer bonding technology and ion cut cleaving process through Hydrogen implantation ( source: O. Moutannabir priv.communication).

28

Downloaded on 2015-10-20 to IP 54.152.109.166 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

ECS Transactions, 64 (5) 27-32 (2014)

Experimental Modern Nanoindentation technique facilitated experimental testing of the mechanical properties of microscopic structures, thin wires, cell walls in plants, red blood cells, and field emission in accelerator physics, which formerly was impossible. In general, instrumented indentation is used to measure hardness and elastic modulus, but it can also be used to investigate rate-sensitive plasticity or creep of thin films and bulk materials. Advances in fabrication and processing of nanostructured materials, nanoelectronics, and nanorods resulted in the rapid development of experimental methods for characterizing their properties. Nanoindentation testing technique is used in this research to investigate the mechanical properties of Silicon-on-Insulator (SOI) and strained Silicon-on-Insulator (sSOI) thin films. A nanoindenter XP equipped with a three sided Berkovich diamond tip was used in conjunction with the continuous stiffness method in depth control mode to measure the hardness and modulus of the SOI and sSOI thin films. The CSM method allows continuously evaluating the mechanical properties of materials as a function of the contact depth as detailed elsewhere [6]. Results and Discussion Four samples in addition to bulk Si were tested, which comprised one regular nonstrained SOI sample of 88 nm thickness and three strained sSOI samples of 100nm, 70nm, and 15 nm thicknesses. The FE-SEM micrographs of Figure 2 depicts a crosssectional cleavage view of a 100 nm thick sSOI sample. A total of 15 indents with maximum indentation depth of 500 nm were performed on each sample. The allowable drift rate and the strain rate for loading were specified as 0.05 nm/s and 0.05 s-1 respectively. During the loading of the indenter the material undergoes elastic and plastic deformation. Among the nanomechanical properties of interest in this study of these SOI and sSOI films are the hardness and the modulus. The nanoindentation hardness is defined as the indention load divided by the projected contact area of the indenter tip and the modulus is calculated from the contact stiffness and the contact area. The hardness and modulus results versus contact depth of indentation are shown in Figures 3 & 4. The scatter in the nanomechanical measurements from different indents on each sample is illustrated as 3σ error bar, where σ is the standard deviation. The hardness of bulk Si was measured as 12 GPa, which is in agreement with literature published hardness values [7] and remains flat to a contact depth of less than 15 nm. For a contact depth of indentation between 30nm and 100 nm, the three different thin SOI film samples (88 nm thick non-strained SOI, 70nm and 100 nm strained sSOI) suffered a softening effect because the indenter is approaching the softer buried Silicon dioxide SiO2 (BOX) insulator. The hardness of these films continued to increase until a dominant bulk Si substrate effect is observed at a contact depth of indentation of 150 nm in the underlying handle wafer. The 15 nm sSOI sample is too thin and immediately exhibited a hardness value of the buried oxide SiO2 insulator to which the 15nm sSOI film is firmly attached by the bonding interface. The hardness of the 15 nm sSOI continued to increase and reached the hardness of the bulk Si substrate at a contact depth of 130 nm. The measured modulus of bulk Si is in the range of 170 GPa, which is also in agreement with modulus values published in the literature [7]. For the three film samples (88 nm SOI, 70 and 100 nm sSOI), the measured moduli depicted composite moduli between a modulus of bulk Si and a SiO2

29

Downloaded on 2015-10-20 to IP 54.152.109.166 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

ECS Transactions, 64 (5) 27-32 (2014)

layer (130 GPa) for a contact depth of up to 50 nm. For a contact depth of indentation greater 50 nm the moduli continued to increase to the bulk Si substrate modulus. Since the 15 nm sSOI film is too thin, the modulus immediately depicted the modulus of the SiO2 layer and continued to increase to the bulk Si modulus for a depth of indentation larger than 50 nm. For the case of the 15nm thick sSOI single crystal film the substrate dominance effect is inevitable especially for this hard ultra-thin crystalline Si film on a softer buried SiO2 substrate as predicted by the literature [8]. For the nanoindentation hardness measurements displayed in Figure 3 primarily the 15 nm sSOI film stands out exhibiting the substrate dominance effect with a greatly reduced hardness. For the experimentally determined modulus values shown in Figure 4 it is evident that not only the thin 15 nm sSOI film but instead all sSOI and SOI films of various thicknesses behave differently from bulk Si. But again the 15 nm sSOI film differs with the modulus values from the thicker measured sSOI films.

Figure 2. FE-SEM micrograph of cleaved cross-section of 100 nm thick sSOI single crystal Si film on buried SiO2 bonded to a handle wafer. 14

Hardness (GPa)

12

10

14

12

10

8

8 sSOI 100 nm sSOI 70 nm sSOI 15 nm SOI 88 nm Si

6

4

6

2

0

4

0

50

10

100

20

150

30

200

40

250

50

300

350

Contact Depth (nm)

30

Downloaded on 2015-10-20 to IP 54.152.109.166 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

ECS Transactions, 64 (5) 27-32 (2014)

Figure 3. Plot of nanoindentation hardness versus contact depth for SOI and sSOI single crystal Si films on buried oxide.

200

Modulus (GPa)

150

200

100 150 100

50

sSOI 100 nm sSOI 70 nm sSOI 15 nm SOI 88 nm Si

50 0

0

0

50

100

150

200

50

40

30

20

10

250

300

350

Contact Depth (nm)

Figure 4. Experimental data plotting modulus versus contact depth for SOI and sSOI single crystal Si films on buried oxide. Conclusion Using nanoindentation, we measured the properties of SOI and bi-axially tensile strained sSOI and benchmarked the results against single crystal bulk Si. The experimental results indicate that the hardness of the strained thin Si films varies significantly. The ultra-thin strained sSOI film of 15nm behaves differently and exhibits a pronounced substrate dominance effect from the softer buried SiO2 substrate, to which the single crystal 15nm Si film is attached by a strong hydrophilic wafer bonding interface. Acknowledgments The authors would like to acknowledge the college of William and Mary for the FE-SEM microscopy images. References 1.

T. Akatsu, J.M. Hartmann, A. Abbadie, C. Aulnette, Y.M. LeVaillant, D. Rouchon, Y. Bogumilovicz, L. Portigliatti, C. Colnat, N. Boudou, F. Lallement, F. Triolet, Ch. Figuet, M. Martinez, P. Nguyen, C. Delattre, K. Tsyganenko, B. Berne, F. Allibert Ch. Deguet, M. Kennard, E. Guiot, F. Metral and I. Cayrefourcq ; ECS Transactions, Vol. 3, No 6, 107 – 117 (2006)

31

Downloaded on 2015-10-20 to IP 54.152.109.166 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

ECS Transactions, 64 (5) 27-32 (2014)

2. Mazure and A.J. Auberton-Herve, Proc. of 35th European Solid-State Device Research Conf. ESSDERC, p. 29 (2005) 3. T. A. Langdo, M. T. Currie, Z.-Y. Cheng, J.G. Fiorenza, M. Erdtmann, G. Braithwaite, C.W. Leitz, C.J. Vineis, J. A. Carlin, A. Lochtefeld, M. T. Bulsara, I. Lauer, D.A. Antoniadis, M. Somerville, Solid-State Electronics, 48 , 1357 (2004) 4. I. Cayrefourcq, A. Boussagol and G. Celler, in SiGe and Ge: Materials, Processing, and Devices, ECS Trans. 3, (7), 399 (2006) 5. N. Miller, K. Tapily, H. Baumgart, G. K. Celler, F. Brunier, and A. A. Elmustafa, Mater. Res. Soc. Symp. Proc. Vol. 1021, 1021-HH05-24 (2007) 6. W. C. Oliver and G. M. Pharr, Journal of Materials Research 19, 3 (2004). 7. P. Mishra, S. R. Bhattacharyya and D. Ghose, Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms 266 1629 (2008) 8. I. Manika, and J. Maniks, Journal of Physics D: Applied Physics, 41 (7) 074010 (1-6) (2008)

32

Downloaded on 2015-10-20 to IP 54.152.109.166 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).