Low Stress In-situ Boron doped Poly SiGe layers for

ECS Transactions, 35 (30) 45-52 (2011) 10.1149/1.3653922 © The Electrochemical Society

Low Stress In-situ Boron doped Poly SiGe layers for MEMS Modular Integration with CMOS S. N. R Kazmi, A. A. I. Aarnink, A.Y. Kovalgin, C. Salm and J. Schmitz MESA+ Institute for Nanotechnology, University of Twente, P.O . Box 217, 7500 AE Enschede, The Netherlands. Email: [email protected]

In-situ boron doped LPCVD polycrystalline silicon-germanium (poly Si30Ge70) layers are deposited from silane (SiH4) and germane o (GeH4) with fixed GeH4 to SiH4 partial pressure ratio at 430 C and 0.2 mbar. The layers exhibit resistivities less than 1 mΩ-cm with a uniform boron distribution over the film thickness. The effect of the diborane (B2H6) partial pressure on the properties of the SiGe alloy is investigated. The layers deposited at low partial pressures of B2H6 exhibit very low stress with a trend from tensile to compressive with increasing B2H6 partial pressure, accompanied by a phase transition from polycrystalline to amorphous, allowing to tune for minimal stress.

Introduction Polycrystalline silicon-germanium (poly SiGe) is emerging as an attractive structural material for post processing microelectromechanical systems (MEMS) on top of foundry fabricated CMOS [1-4] apart from its potential use as metal-oxide semiconductor field effect transistor (MOSFET) gate material [5-6]. The post processing approach of MEMS with CMOS requires a low thermal budget < 450 °C to avoid CMOS degradation [7, 8]. A standard poly-Si based MEMS process typically requires deposition temperatures above 600 °C and annealing temperatures above or at 900 °C to meet the requirements of low-resistivity, low-tensile-stress for MEMS structural films [9]. The lower deposition temperatures of doped poly SiGe allow its use in CMOS post processing [10, 11]. In this paper we present a study of in-situ doping of SiGe using Ar-diluted B2H6. All experiments are carried out at 430 °C and 0.2 mbar with fixed SiH4 and GeH4 partial pressures of 7.1x10-2 mbar and 3.5x10-2 mbar respectively to ensure 70% Ge contents in the deposited layers. The electrical and mechanical properties of the deposited layers are reported.

Experimental In-situ boron doped poly SiGe layers were deposited on 100 mm single side polished oriented Si wafers (381±15 µm, n-type/phosphorus doped, 1-10 Ω-cm) with 100 nm of thermally grown oxide. The wafers were directly loaded into a custom built hot-wall horizontal Low Pressure Chemical Vapor Deposition (LPCVD) reactor

45 Downloaded 08 Dec 2011 to 130.89.194.102. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp

ECS Transactions, 35 (30) 45-52 (2011)

-3

(Fig. 1), maintained at a base pressure of 10 mbar, after cleaning in 99% HNO3 for 5 minutes followed by DI water rinse and N2 drying. The pressure inside the furnace was raised to 10 mbar with 150 sccm of N2 flow to uniformly heat up the wafers to 430 oC for 30 minutes. A thin (few nm) amorphous silicon layer, acting as nucleation layer [11], was deposited at 0.5 mbar and 430 oC for 10 min with 88 sccm of SiH4 flow. The in-situ boron doped poly Si30Ge70 was then deposited without vacuum break from pure SiH4 and pure GeH4 gasses and with varied B2H6 partial pressure. All the gasses were introduced from the front side of the LPCVD tube. A total of 13 wafers were LPCVD deposited in each run including 4 dummy wafers, 2 in front and 2 at the back of wafer boat. The gas depletion effect was minimized using root blowers. The aim was to find optimized deposition condition that leads to the desired electrical and mechanical properties of deposited layer suitable for post processing compatible MEM devices on top of foundry fabricated CMOS.

Fig 1: Simplified view of the employed LPCVD system. The thickness of deposited layers was measured using a Dektak 8.0 surface profilometer after a masked etch of SiGe in SF6 and O2 plasma. The stress in the deposited layers was calculated using Stoney’s formula [12] by measuring the wafer curvature in two orthogonal directions with Dektak 8.0. The resistivity was measured by the four probe measurement method, averaged over nine points across the wafer. Cross sectional high resolution secondary electron microscopy (HRSEM) images were taken to observe the morphology of in-situ doped SiGe layers. X-ray diffraction (XRD) analysis was carried out with Philips XRD model expert system II using Cu K-α line of wavelength 1.54 Å to obtain information about the crystallinity of the deposited samples. The depth profile of Ge/Si contents were determined by X-ray Photoelectron Spectroscopy (XPS) across the entire thickness using 5-keV argon sputtering. Secondary Ion Mass Spectroscopy (SIMS) was also performed to determine the Si, Ge and boron concentrations in the deposited layer. SIMS analysis was performed using 3 keV O2+ primary ions bombardment with positive mode. The first ~10 nm of the profiles are unreliable due to transient instrumental effects and also the profile region close to oxide, within the oxide and after oxide is less reliable due to charging and matrix effects.



Results and Discussions The results of the experiments show that the deposition rate of in-situ boron doped -3 Si30Ge70 increases with increasing B2H6 partial pressure till 2.4x10 mbar and then -3 decreases with further increase of B2H6 partial pressure to 4.7x10 mbar. This increase in the deposition rate is attributed to the boron enhanced desorption of hydrogen atoms from the deposition surface [13]. Whereas the decrease in deposition rate is caused due to gas phase reactions. XRD analysis indicates a polycrystalline layer structure, as seen from the diffraction peaks (111), (220) and (311) in Fig. 2. The peaks lie closer to the Ge peak positions than Si peak positions, a clear indication that the material contains more Ge than Si. An average of 73% Ge content is calculated for polycrystalline samples using Vegard’s law [14]. Increasing diborane partial pressure, results in a drop and broadening of all peaks (Fig. 2). The material appears to become amorphous, perhaps partly microcrystalline, under these deposition conditions.

Fig 2: XRD foot prints of SiGe layers with varied B2H6 partial pressure. At low partial pressures of diborane (1.24x10-4 mbar and 1.9x10-4 mbar) the phenomenon of preferential crystallographic growth is notably observed. The XRD pattern in Fig. 3 shows a preferential V-shaped and columnar grain growth associated with (111) and (220) dominant peaks [15] at 1.9x10-4 mbar and 1.24x10-4 mbar of B2H6 partial pressures respectively. Such columnar grains can be etched very steeply and selectively to SiO2 at cryogenic temperatures with SF6 and O2 plasma chemistry [16].



(a)

(b)

Fig 3: XRD of poly SiGe layers with (a) 1.24x10-4 mbar (b) 1.9x10-4 mbar partial pressure of B2H6. HRSEM images of the deposited SiGe layers at varied B2H6 partial pressures are shown in Fig. 4. The transition from polycrystalline to amorphous phase with increased B2H6 partial pressure is obvious from these HRSEM images. The grain size in the layers deposited at 1.9x10-4 mbar is comparatively bigger than that of the layer with 4.7x10-4 mbar, i.e., few tens of nanometer.

(a)

(b)

(c) Fig 4: HRSEM images a) 1.9x10-4 mbar (b) 4.7x10-4 mbar (c) 4.7x10-3 mbar of B2H6 partial pressures.



The dependence of resistivities of in-situ doped layers on diborane partial pressure is plotted in Fig. 5. The resistivities of the in-situ doped layers at low B2H6 partial pressures (1.24x10-4 mbar and 1.9x10-4 mbar) is almost three orders of magnitude lower compared to the undoped layers. Initially the resistivity of the deposited layers drops to 0.8 mΩ-cm and 0.6 mΩ-cm with B2H6 partial pressure rises to 1.24x10-4 mbar and 1.9x10-4 mbar respectively. The decrease in the resistivities of the doped-layers is due to an increase in the boron concentration from 6.9x1020 cm-3 to 1.2x1021 cm-3, as found from SIMS with B2H6 partial pressures of 1.24x10-4 mbar and 1.9x10-4 mbar, respectively. The resistivity of the SiGe layers then starts to increase with increasing B2H6 partial pressure beyond 1.9x10-4 mbar and reaches 600 mΩ-cm, slightly higher than the resistivity of undoped poly SiGe layer, at 4.7x10-3 mbar of B2H6 partial pressure. The observed increase in the resistivity of deposited layers is associated with the transition from polycrystalline to amorphous phase with increased B2H6 partial pressure. This is in complete agreement with the cross-sectional HRSEM images and XRD results of these deposited layers. Therefore, it can be concluded that the B2H6 partial pressure of 1.9x10-4 mbar corresponds to a minimum of resistivity for doped SiGe layers at these deposition conditions.

Fig 5: Resistivity versus B2H6 partial pressures. A low tensile stress of ~12 MPa is observed in the doped SiGe layers for 1.24 x10-4 mbar of B2H6 partial pressure compared to undoped layers, ~55 MPa. The stress in deposited layers changes to low compressive (~3 MPa) with an increase in B2H6 partial pressure to 1.9x10-4 mbar. The stress further changes to even more compressive (~57 MPa) with B2H6 partial pressure reaches to 4.7x10-3 mbar. Fig. 6 shows the stress transition from tensile to compressive with increased B2H6 partial pressure.



Fig 6: Stress in SiGe film versus diborane partial pressure, deduced from wafer bow using Stoney’s formula [12]. Using XPS, the atomic concentrations of Ge and Si were determined to be 71%±2% and 29%±2% respectively, irrespective of the diborane flow. Fig. 7 shows the XPS depth profile, indicating no vertical concentration gradient.

Fig 7: XPS of SiGe samples across the entire layer thickness. -4

SIMS analysis on samples with diborane partial pressure of 1.24x10 mbar and -4 1.9x10 mbar confirms the uniform depth profile of Si and Ge, and the measured values are in good agreement with the XPS findings. Fig. 8 shows the depth profile of Ge and Si for two different wafer positions in the wafer boat (first and last process wafer). The silicon peak near the interface (Figure 8b) originates from the silicon-CVD nucleation step (see Section Experimental). The boron depth profile, as in Fig. 9, was also examined with SIMS and found to be uniform throughout the SiGe layers. The concentration of boron was found to be 6.9x1020



-4

cm-3 for layers with a B2H6 partial pressure of 1.24x10 mbar and increased to 1.2x1021 -4 cm-3 for a B2H6 partial pressure of 1.9x10 mbar. However not all incorporated boron was electrically active. A saturation limit for the active boron concentration of about 5.0x1020 cm-3 was reported for poly SiGe layers deposited at 550 °C independent of the Ge content in the SiGe [17].

Fig 8: SIMS profiles of poly SiGe samples across the entire thickness (a) Ge depth profile (b) Si depth profile.

Fig 9: SIMS Boron depth profiles in poly SiGe layers deposited at 430 °C.

Conclusions We have deposited in-situ boron doped poly Si30Ge70 layers from SiH4, GeH4 and B2H6. The layers deposited at low diborane partial pressures exhibit the lowest stress and lowest resistivities reported until now in the literature for 70% Ge contents and temperatures below 450 oC without any further treatments (e.g., annealing, laser enhanced crystallization etc). The role of dopant partial pressure on deposition rate was



examined and found to increase with increaseing dopant partial pressure. The boron concentration at low B2H6 partial pressures was found to increase with the partial pressure of B2H6 accompanied by a slight decrease in resistivity. Moreover, the transition from polycrystalline to amorphous phase was observed due to the excess boron incorporation. Additionally, Si and Ge content were found to be uniformly distribution throughout the entire thickness of the layers. The SIMS depth profiles revealed a homogeneous distribution of boron atoms throughout the deposited layers. The boron-doped poly SiGe deposited at low B2H6 partial pressure showed its potential to be used as structural material for MEMS components for low-thermal-budget IC processes.

Acknowledgements The authors would like to thank Mark Smithers, Ite-Jan Hoolsema, Gerard Kip, Jiwu Lu all from MESA+ Institute of Nanotechnology and Jurgen van Berkum from MiPlaza Eindhoven for their help to characterize SiGe samples. This research work is financially supported by the Dutch Technology Foundation STW through project grant no. 10048 under “CMOS Receiver enhancement using arrays with MEMS” (CREAM).

References 1. S. Sedky, A. Witvrouw, A. Saerens, P. Van Houtte, J. Poortmans and K. Baert, J. Materials Research, 16, p.2607 (2001). 2. T..-J. King, R. T. Howe, S. Sedky, G. Liu, B. C. Y. Lin, M. Wasilik, and C. Duenn, IEEE International Electron Devices Meeting, p. 199, (2002). 3. Srinivasan Kannan, Craig Taylor and David Allred, J. Non-Crys. Solids, 352, p.1272 (2006). 4. M. A. Eyoum and T.-J. King, J. Electrochem. Soc., 151, p. J21, (2004). 5. Y. V. Ponomarev, P. A. Stolk, C. Salm, J. Schmitz, and P. H. Woerlee, IEEE Trans. on Electron Devices, 47, p.848 (2000). 6. C. Salm, D. T. van Veen, D. J. Gravesteijn, J. Holleman, and P. H. Woerlee, J. Electrochem. Soc., 144, p.3665 (1997). 7. H. Takeuchi, A. Wung, X. Sun, R. T. Howe, and T. J. King, IEEE Trans. on Electron Devices, 52, p.2081, (2005). 8. J. Schmitz, Nucl. Instr. And Meth., A 576, p. 142–149, (2007). 9. M. Biebl, G. T. Mulhern and R. T. Howe, 8th International Conference on Solid-state Sensors and Actuators - Transducers 95/Eurosensors IX, 1, p. 198, (1995). 10. T.-J. King, J. P. McVittie, K. C. Saraswat, and J. R. Pfiester, IEEE Trans. on Electron Devices, 41, p.228, (1994). 11. A. Kovalgin and J. Holleman, J. Electrochem. Soc., 153, p.363 (2006). 12. G. Stoney, Proc. R. Soc. London A, 82, p.172, (1909). 13. H. C. Lin, C. Y. Chang, W. H. Chen, W. C. Tsai, T. C. Chang, T. G. Jung, and H. Y. Lin," J. Electrochem. Soc., 141, p. 2559, (1994). 14. R. W. Cahn, Physical Metallurgy, North-Holland, Amsterdam (1965). 15. J. H. Wang, S. Y. Lien, C. F. Chen, and W. T. Whang," IEEE Electron Device Letters, 31, p. 38, (2010). 16. M. J. de Boer, J.G.E. Gardeniers, , H. V. Jansen, M.-J. Gilde, G. Roelofs, J. N. Sasserath, and M Elwenspoek, J. Microelectromech. Sys., 11, p.385 (2002). 17. A. Moriya, M. Sakuraba, T. Matsuura and J. Murota, Thin Solid Films, p. 541 (1999).
