Non-PFC Plasma Chemistries for Patterning Complex Materials/Structures (Task Number: )

Transcription

1 PIs: Non-PFC Plasma Chemistries for Patterning Complex Materials/Structures (Task Number: ) Jane P. Chang, Chemical and Biomolecular Engineering, UCLA Graduate Students: Jack Chen, PhD student, Chemical and Biomolecular Engineering, UCLA Nathan Marchack, PhD candidate, Chemical and Biomolecular Engineering, UCLA Undergraduate Students: Michael Paine, UG student, Chemical and Biomolecular Engineering, UCLA Other Researchers: (TBD) 1

2 Objectives Assess the thermodynamic feasibility of patterning etchresistant materials (complex materials and structures) Identify the non-pfc alternative for through silicon via etch Validate the theoretical assessment by performing etching experiments of these materials by industrial sponsors Features Goal Ref. 1 Etch rate 50 um/min I. Sakai, J. Vac. Sci. Technol. A, Aspect ratio (Anisotropy) I. Sakai, IEEE, Side wall angle 90 o ±10 o M. Hooda, J. Vac. Sci. Technol. A, 2010

3 ESH Metrics and Impact 1. Reduction in the use of PFC gases by focusing on non- PFC chemistries 2. Reduction in emission of PFC gases to environment 3. Reduction in the use of chemicals by tailoring the chemistries to the specific materials to be removed 3

4 TSV (Through Silicon Via) TSV for 3D-LSI technology road map [1] From 2-D to 3-D packaging [2] 2-D 3-D According to Moore s Law, the density of IC devices doubles every two years, the 3-D packaging system is required. [1] M. Motoyoshi, IEEE, 2009; [2] P. Garrou, Samsung website, 2010 Advantages: Smaller footprint and form-factor Less weight and power consumption Potentially lower cost

5 Challenges in TSV High etch rate (~1 μm/min) High aspect ratio (>10) [1, 2] Greenhouse gas usage [3] Silicon size (inch) Thickness (um) CO 2 increases gradually Requirement to increase etch rate, increase aspect ratio and reduce the usage of global warming gases [1] M. Motoyoshi, IEEE, 2009; [2] P. Garrou, Samsung website, 2010; [3] NOAA Earth System Research Laboratory, 2012

6 TSV Process Options DRIE (Deep Reactive Ion Etching) [4] Laser drill [5] TSV Processes Bosch DRIE [6] Cryogenic DRIE [7] Laser drill [5] Width <5mm <5mm <15mm Aspect ratio >20:1 >10:1 >7:1 Sidewalls Vertical (90 o ) Vertical &smooth 80 o -90 o Temp. affection Negligible -110 o C ~ -130 o C Yes Process ICP ICP Laser Potential applications High-end High-end Low-cost Example Microprocessors Microprocessors DRAM/Flash As the critical dimension continues to decrease, DRIE becomes the most feasible technology for TSV [4] Applied Materials Inc., website, 2012; [5]C. Tang, J Micromech Microeng,2012; [6] R. Landgraf, IEEE, 2008; [7] M. Walker, Proc. SPIE, 2001

7 Systematic Approach - Thermodynamics Thermodynamic approach can be systematic - If such data is available - NIST-JANAF Thermo-chemical tables - HSC Chemistry for windows, chemical reaction and equilibrium software with extensive thermo-chemical database - FACT, Facility for Analysis of Chemical Thermodynamics - Barin and Knacke tables (thermo-chemical data for pure substances and inorganic substances) - Determination of dominant surface/gas-phase species - Assessment of possible reactions Graphical Representation of thermodynamic analysis - Richardson Ellingham diagram - Pourbaix diagram - Volatility diagram 12

8 A Case Study Cu-Cl of Volatility Diagram [8] 100 o C 150 o C 200 o C Target the volatile species Control pressure, temperature and gas composition [8] N.S. Kulkarni, J. Electrochem. Soc., 2002

9 A Case Study Cu-Cl-H of Volatility Diagram [8] (Cu-Cl-H) Proposed two-step reaction (2002): Log P H2 = -20 Log P H2 = -15 Log P H = -30 Log P H = -25 Log P H2 = -10 Log P H = -20 Log P H2 = -5 1)Cu (c) + 2Cl CuCl 2(c) at low temp 2)3 CuCl 2(c) + 3H Cu 3 Cl 3 + 3HCl Experimentally verified by Hess in 2007/2011 Addition of reactive chemistry (H 2 /H) changed the reaction kinetics and resulted in higher etch rates [8] N.S. Kulkarni, J. Electrochem. Soc., 2002

10 Important Chemistries for Silicon Etch [9] Chemistries Radical Volatile Product Inhibitor H 2 H H Si x F y CH 4 CH 3 CH 2 SiH 4 Si x C y H z F 2 F Si x F y F, NF 2 Si x N y F z F, SiF 3 Si x F y CF 4 F, CF 3 Si x C y F z SF 6 F, SF 5 Si x S y F z S 2 F 2 F, S 2 F Si x S y F z Cl 2 Cl SiCl 4 Cl Br 2 Br SiBr 4 Br CBr 4 Br, CBr 3 SiBr 4 Si x C y Br z, Br Fluorine-based gases remains the most effective chemistry Chemistries Radical Volatile Product Inhibitor CHF 3 CF 2 HF, Si x C y H z CFH, CH 2 F 2 HF Si x C y F z H a C CH2 CH 3 F HF, H CFH 2 Si x C y F z H a COF 2, O 2 F, CF 4 /O 2 F, O, CF 3 OF, F 2, Si x O y F z CHF, HF, CF 4 /H 2 F, H, CF 3 Si x C y F z SF 6 /O 2 SF 5, O, F SOF 4, Si x O y F z SF 6 /H 2 SF 5, H, F HF, Si x S y F z SF 6 /N 2 SF 5, N 2, F Si x S y F z SF 6 /CHF 3 SF 5, F, CF 2 HF, Si x C y F z CBrF 3 F, Br SiBr 4 Si x C y F z Br, Si x C y Br z [9]H. Jansen, J. Mrcromech. Microeng CCl 4 CCl 3, Cl, C SiCl 4 Cl

11 Can SF 6 be replaced for Si etching Chemistries Atmospheric conc. in 2005 (ppt) Con. since 1994* & 1998 (ppt) Annual emission in late 1990s (Gg) Rafactive efficiency (W/m 2 ppbv) Lifetime (year) Global warming potential CO 2 278x x10 6 * - - variable 1 [10] CH 4 7x x10 3 * [10] N 2 O 275x x10 3 * [10] CHClF 2-105x10 3 * [10] CF ~ , [11] CCl 2 F 2-503x10 3 * [10] C 2 F ~ , [11] CHF ~ [11] SF ~ , [11] <0.1 - ~ [11] Ref. GWPs are one type of simplified index based upon radiative properties which estimate the potential future impacts of gases is promising but it is a Greenhouse gas. [10] United Nations Environment Program(UNEP), 2010; [11] W. Tsai, J Hazard Mater, 2008

12 The potential use of in TSV and its reaction products are toxic in Toxic Substances Control Act (TSCA) +2H 2 O 3HF + HNO 2 ; + 3H 2 O 6HF + NO + NO 2 ; 2 + 3H 2 N 2 + 6HF Title Authors Year Journal Citation,the greenhouse gas missing from Kyoto Environmental and health risk analysis of, a toxic and potent greenhouse gas Abatement of Greenhouse Gases [12] : BOE Edwards Thermal Processing System Prather, M.J. and J. Hsu 2010 Tsai, W. T Catalytic Decomposition Systems Geophysical Research Letters Journal of Hazardous Materials Plasma Abatement Systems Not effective in decomposing CF 4 The catalyst lifetime only 18 months For semiconductor industry in plasma SRC/SEMATECH Engineering Research Center for Environmentally Benign etching Semiconductor process and Manufacturing chamber cleaning [12] Vasilis Fthenakis, U.S. DOE, 2001

13 Plasma Torch Abatement of & SF 6 [13] Cbefore Cafter Destruction and removal efficiency (DRE) = 100 C % before 71.6% is a Greenhouse gas but it can be destroyed nearly 100% [13] Yong C. Hong, IEEE, 2005

14 Etch Rate of Si /poly-si * in F-based chemistries Etchant Etch rate Power Pressure (nm/min) (W) (mtorr) Ref SF [g] SF 6 /O 2 (25%) [d] SF 6 /O 2 (25%) [d] SF 6 /O 2 /CHF [c] SF 6 (40sccm)/O 2 (14sccm)/CHF 3 (17sccm) [j] SF 6 (80%)/C 2 Cl 3 F 3 (20%) [i] SF 6 (10%)/CBrF 3 (90%) [h] SF 6 (6sccm)/HBr(10sccm)/Cl 2 (70sccm) * [e] SF 6 (10%)/CBrF 3 (80%)/Ar(10%) [h] CF 4 30 [k] CF 4 * 20 [f] CF 4 /O [k] CF 4 /O [k] CF 4 (90%)/O 2 (12%)/N 2 (8%) * 260 [e] CBrF [h] CBrF [g] F/F [k] /O 2 44 [k]

15 Etch Rate of Si/SiO 2 /Si 3 N 4 /SiC by Etchant Etch rate_si (nm/min) Etch rate_sio 2 (nm/min) Etch rate_si 3 N 4 (nm/min) Etch rate_sic (nm/min) Power Pressure (mtorr) Ref (100%) W/cm [l] (25%)/N W/cm [l] (25%)/Ar W/cm [l] (25%)/He W/cm [l] (25%)/O W/cm [l] (25%)/N 2 O W/cm [l] W 12 [m] /O W 2 [n] (60%)/CH 4 (40%) W 6 [m] (7sccm) W 100 [o] (300sccm) W 1000 [p] (500sccm)/O 2 (50%) W 1000 [p] (200sccm) W 1 [q] (10%)/O 2 (90%) W 1 [q] The addition of O 2 seems to increase Si, SiO 2 and Si 3 N 4 etch rate

16 Silicon Etching by Fluorine Main reactions of silicon etching by F 2 Reactions G(eV) log(k) 1 Si(c) Si Si(c) F 2 SiF Si(c) + F SiF SiF + F SiF SiF 2 + F SiF SiF 3 + F Electron impact dissociate energy of F 2 [14] Reactions G(eV) F 2 + e 2F + e 1.6 Fluorine is the most effective etching chemistry for silicon, especially atomic fluorine, as produced by plasma [14] Parker J., J. Quant. Electron., 1973

17 Production of Fluorine from SF 6 and Thermal Dissociate Rxns G(eV) Log(K) S1 SF 6 SF 5 + F S2 SF 5 SF 4 + F S3 SF 4 SF 3 + F S4 SF 3 SF 2 + F S5 SF 2 SF + F S6 SF S + F Thermal Dissociate Rxns G(eV) Log(K) N1 NF 2 + F N2 NF 2 NF + F N3 NF N + F [CRC handbook, 2010] [CRC handbook, 2010*; Y. Tanaka, IEEE, 1997] Electron Impact Dissociate Rxns G(eV) S1 SF 6 + e SF 5 + F + e 4.00 S2 SF 5 + e SF 4 + F + e 2.27 S3 SF 4 + e SF 3 + F + e 3.47 S4 SF 3 + e SF 2 + F + e 2.92 S5 SF 2 + e SF + F + e 4.01 S6 SF + e S + F + e 3.52 For SF 6, electron impact dissociate energy is comparable and slightly higher than that of equilibrium data (~0.4eV) It is possible that the available equilibrium data on can be used as a guide to the production of fluorine in a plasma

18 Can SF 6 be Replaced by? 400 Si SF 6 4 SiS 2 (c) ΔG = -7.0eV Reactions G(eV) log(k) 1 Si(c) Si Si(c) F 2 + SiF Si(c) SF S 2 Si(c) Si(c) SF S 2 Si(c) SiF Si Si 3 N 4 (c) Δ G = -8.6eV Reactions G(eV) log(k) 1 Si(c) Si Si(c) F 2 SiF Si(c) N 4 Si 3 (c) Si(c) N 4 Si 3 (c) SiF Log P SiF4, SiF 2 (atm) Log P SiF4, SiF 2 (atm) ' 1 Si (c) S 2 Si (c) SiF 2 S 2 Si (c) SiF Si ' Log P SF6 (atm) Si (c) Si 3 N 4 (c) Si 3 N 4 (c) SiF 2 SiF Si Log P NF3 (atm) is more capable of removing silicon via the formation of However, another significant reaction product from reaction with is Si 3 N 4, which has to be removed and competes for fluorine

19 Can SF 6 be Replaced by? 400 Si SF 6 4 SiS 2 (c) + 4F SF 6 ΔG = -7.0eV Δ G = - 1.9eV Reactions G(eV) log(k) 1 Si(c) Si Si(c) F 2 + SiF Si(c) SF S 2 Si(c) Si(c) SF S 2 Si(c) SiF Log P SiF4, SiF 2 (atm) ' 1 Si (c) S 2 Si (c) SiF 2 S 2 Si (c) SiF Si Log P SF6 (atm) Si Si 3 N 4 (c) + 4F Δ G = -8.6eV Δ G = -0.2eV Reactions G(eV) log(k) 1 Si(c) Si Si(c) F 2 SiF Si(c) N 4 Si 3 (c) Si(c) N 4 Si 3 (c) SiF Log P SiF4, SiF 2 (atm) ' Si (c) Si 3 N 4 (c) Si 3 N 4 (c) SiF 2 SiF Si Log P NF3 (atm) is more capable of removing silicon via the formation of However, another significant reaction product from reaction with is Si 3 N 4, which has to be removed and competes for fluorine

20 A Strategy to Remove Si 3 N 4 by O 2 addition Formation of Si 3 N 4 (c) ΔG(eV) Si(c) Si 3 N 4 (c) -8.6 Removal of Si 3 N 4 (c) by F ΔG(eV) 1 6 Si 3 N 4 (c) + 4F Removal of Si 3 N 4 (c) Removal of SiO 2 (c) Total reaction Removal of Si 3 N 4 (c) & SiO 2 (c) by & O 2 ΔG (ev) 1 6 Si 3 N 4 (c)+ 7 6 O SiO 2 (c)+ 2 3 NO SiO 2 (c) NO 2 F+ 1 2 NOF N Si 3 N 4 (c)+ 7 6 O NO NO 2 F+ 1 2 NOF N From the analysis of ΔG, the non-volatile byproducts, Si 3 N 4 could be removed by - O 2 (as a mixture)

21 Surface of Silicon ( /O 2 ) O N 2 N O 2 F 2 F N 2 /N O 2 /O F 2 /F Si Si 3 N 4 Si subtrate SiO 2 Si SiO 2 Si 3 N 4 Gas phase The interaction of N/O/F need to be determined Surface Since Si(c), SiO 2 (c) and Si 3 N 4 (c) are non-volatile compounds, they all need to be removed by the plasma

22 ΔG(eV) [15] NO 2 F NOF Gas phase - O 2 / From the FTIR of /O 2 silicon etching[17], NO 2 F, NOF and NO can be found when the with O 2 addition. The following reactions have been proposed, O + NF x NOF + (x-1)f 2O + NF x NO 2 F + (x-1)f with Oxygen-300K G(eV) log(k) A O + NOF + 2F B O + NF 2 NOF + F C O + NF NOF D 2O + NO 2 F + 2F E 2O + NF 2 NO 2 F + F F 2O + NF NO 2 F K G(eV) log(k) G NF 2 + F H NF 2 NF + F I NF N + F With O 2 addition, [F] is increased and nitrogen forms NOF and NO 2 F which are stable volatile species. [15] NIST-JANAF Thermo-chemical tables, 2012; [17] D. Linaschke, Proc. 23rd Eur. Photovoltaic Solar Energy Conf. Exhib Log P NF3, NF 2, NF, NOF (atm) Significant increase in [F] NF 2 B NF 2 E F NOF NO 2 F NF x D G-H-I Log P O (atm) A

23 Si 3 N 4 -O-300K ΔG (ev) log(k) 1 Si(c) + O SiO SiO 2 (c) SiO + O Si 2 N 2 O(c) Si(c) N O Si 2 N 2 O(c) O SiO 2 (c) N Si 2 N 2 O(c) O SiO N Log P SiO2, Si, SiN, Si 2 N (atm) ΔG(eV) [15] SiO Si 3 N Si(c) Si 2 N 2 O(c) 5 SiO Log P O (atm) SiO 2 (c) Surface of Si 3 N 4 -O & SiO 2 -N SiO 2 -N-300K ΔG (ev) log(k) Si 3 N 4 (c) SiN N SiO 2 (c) + N SiN + O Si 2 N 2 O(c) N 1 3 Si 3 N 4 (c) O Si 2 N 2 O(c) SiN O SiO 2 (c) + N 1 2 Si 2 N 2 O(c) O Log P SiN (atm) SiO 2 (c) SiO 2 (c) is the most stable species in both systems [15] NIST-JANAF Thermo-chemical tables, 2012; [16] Huheey, The Strengths of Chemical Bonds, 1958 Bond Energy(eV) [16] Si-O 4.69 Si-N Si 2 N 2 O(c) 9 SiN Log P N (atm) Si 3 N 4 (c) 6

24 ΔG(eV) [15] SiOF Surface of Si 3 N 4 -F & SiO 2 -F Bond Energy(eV) [16] Si-F 5.86 Si 3 N 4 -F-300K G(eV) log(k) Si 3 N 4 (c) + 4F N Si 3 N 4 (c) + 2F SiF N SiO 2 -F-300K G(eV) log(k) 13 SiO 2 (c) + 4F + O SiO 2 (c) + 2F SiF 2 O O Log P Si, SiN, Si2 N (atm) SiF 2 Si 3 N 4 (c) Log P F (atm) Log P Si, SiOF2, (atm) SiF 2 O SiO 2 (c) 13 Log P F (atm) The pressure of volatile products, SiF 2 and SiF 2 O, are comparable in both systems [15] NIST-JANAF Thermo-chemical tables, 2012; [16] Huheey, The Strengths of Chemical Bonds, 1958

25 Si 3 N 4 -O 2 -F-300K G (ev) log(k) 1 Si(c) O 2 SiO SiO 2 (c) SiO O Si 2 N 2 O(c) Si(c) N O Si 2 N 2 O(c) O 2 SiO 2 (c) N Si 2 N 2 O(c) O 2 SiO N SiO 2 (c) + 4F + O SiO 2 (c) + 2F SiF 2 + O Si 3 N 4 (c) O 2 SiO N Si 3 N 4 (c) O Si 2 N 2 O(c) N Log P SiO2, Si, SiN, Si 2 N (atm) Si(c) Surface of Si 3 N 4 -O 2 -F & SiO 2 -N 2 -F Si 2 N 2 O(c) Si 3 N 4 (c) SiO 2 (c) SiO SiF 2 P F =0.76 torr P F =7.6 mtorr P F =7.6 x mtorr Log P O2 (atm) SiO 2 -N 2 -F-300K G (ev) log(k) Si 3 N 4 (c) SiN N SiO 2 (c) N 2 SiN + O Si 2 N 2 O(c) N Si 3 N 4 (c) O Si 2 N 2 O(c) SiN O SiO 2 (c) N Si 2 N 2 O(c) O Si 2 N 2 O(c) + 4F N Si 2 N 2 O(c) + 2F SiF N Si 3 N 4 (c) + 4F N Si 3 N 4 (c) + 2F SiF N Si 3 N 4 (c) could be oxidized by O 2 and then be removed by F. Log P SiN (atm) P F =0.76 torr P F =7.6 mtorr SiO 2 (c) P F =7.6 x mtorr Log P N2 (atm) 7 Si 2 N 2 O(c) Si 3 N 4 (c) SiF SiN SiF 2 6

26 Summary Gas phase NF 2 NF O + 2F NOF + F NF 2 NF O + 2F NO 2 F + F Surface Si O 2 /O N 2 /N F 2 /F SiO 2 (c) Si 3 N 4 (c) N 2 /N N 2 /N SiOF 2 / SiO 2 (c) SiO 2 (c) F 2 /F F 2 /F Si 3 N 4 (c) SiO 2 (c) The addition of O 2 increase significantly the [F] The removal rate of SiO 2 (c) and Si 3 N 4 (c) by F are comparable could replace the SF 6 with O 2 addition.

27 Reference c. Jansen, H., et al., The Black Silicon Method - a Universal Method for Determining the Parameter Setting of a Fluorine-Based Reactive Ion Etcher in Deep Silicon Trench Etching with Profile Control. Journal of Micromechanics and Microengineering, (2): p d. Mansano, R.D., P. Verdonck, and H.S. Maciel, Deep trench etching in silicon with fluorine containing plasmas. Applied Surface Science, : p e. Yeom, G.Y., Y. Ono, and T. Yamaguchi, Polysilicon Etchback Plasma Process Using Hbr, Cl2, and Sf6 Gas- Mixtures for Deep-Trench Isolation. Journal of the Electrochemical Society, (2): p f. Matsuo, P.J., et al., Role of N-2 addition on CF4/O-2 remote plasma chemical dry etching of polycrystalline silicon. Journal of Vacuum Science & Technology a-vacuum Surfaces and Films, (4): p g. Syau, T., B.J. Baliga, and R.W. Hamaker, Reactive Ion Etching of Silicon Trenches Using Sf6/O-2 Gas- Mixtures. Journal of the Electrochemical Society, (10): p h. Demic, C.P., K.K. Chan, and J. Blum, Deep Trench Plasma-Etching of Single-Crystal Silicon Using Sf6/O2 Gas-Mixtures. Journal of Vacuum Science & Technology B, (3): p i. Yunkin, V.A., D. Fischer, and E. Voges, Highly Anisotropic Selective Reactive Ion Etching of Deep Trenches in Silicon. Microelectronic Engineering, (1-4): p j. Legtenberg, R., et al., Anisotropic Reactive Ion Etching of Silicon Using Sf6/O-2/Chf3 Gas Mixtures. Journal of the Electrochemical Society, (6): p k. Moss, S.J. and A. Ledwith, The chemistry of the semiconductor industry. 1987, Glasgow: Blackie. xiv, 426 p. l. Langan et al., Method for plasma etching or cleaning with diluted NF.sub.3, US Patent , m. Wang, J.J., et al., High rate etching of SiC and SiCN in NF3 inductively coupled plasmas. Solid-State Electronics, (5): p n. Kim, B., et al., Use of a neural network to model SiC etching in a NF3 inductively coupled plasma. Modelling and Simulation in Materials Science and Engineering, (8): p

28 Future Plans Next Year Plans Identify potential low impact gases in target applications Perform thermodynamic calculations to assess potential impact and projected effectiveness Implement target chemistries and carry out plasma etching assessment Long-Term Plans Formulate the models to predict emission from plasma processes Assess the effectiveness of the plasma chemistries compared to that of the PFC gases

29 Publications, Presentations, and Recognitions/Awards Presentation in Gordon Research Conference(GRC), July 2012 Invited talk to AVS International Symposium, October 2012

30 Industrial Interactions and Technology Transfer Video conference to Intel, January 31, 2012 (Karson Knutson, Doosik Kim) Conference call with Novellus, March 2012 (Ron Powell, Roey Shaviv, Juwen Gao) Student interview at IBM, February 29, 2012 Student interview with Intel, March 2012 SRC Industrial Liaison, Satyarth Suri at Intel, April 2012 ERC Webinar, June 2012 Student will have poster in Gordon Research Conference, July, 2012 Conference call with Intel, September 2012 (Satyarth Suri, Bob Turkot) After studying the TSV by detailed thermodynamic analysis, the following research is kinetic measurements.