Introduction To Deep Reactive Ion Etching

Introduction to Deep Reactive Ion Etching After Hibert (2002)

Felix Lu AQT/Duke April 22, 2008

http://www.oxfordplasma.de/i mages/scetch/icp_r_ww.gif

After Gale

http://www.micromagazine.com/archive /05/12/0512MI35d.jpg

Topics • Motivation and applications for Deep RIE • Requirements using DRIE • Deficiencies of standard wet/dry etch processes • Optimization of etch rate, smoothness and selectivity • The Bosch and Cryo DRIE processes • Summary Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Motivation & Applications for DRIE Trench capacitors

DRAM micrograph at left shows cross section of ~60:1-deep trench capacitor. SEM images at right show Al2O3 thicknesses proving 100% step coverage. [http://www.micromagazine.com/ar chive/02/06/lead.html] http://www.clarycon.com/Resources/Slide2s.jpg

http://www.semiconductortechnology.com/contractor_images/sts/3_ga as-via.jpg

After Walker

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

SOIMUMPS backside etching – Backside etching of SOI MUMPS die for releasing and metallization of mirror surfaces – Need to etch ~500-700 µm through Si substrate. Evaporated Au

BACK substrate

SOI (mirror) Au pads

SOI (mirror)

SOI (mirror)

Au pads

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Au pads

Desirable characteristics for etching high aspect ratio features •

Relatively high etch rate – standard RIE ≈1 µm/min; DRIE
~2 to >20 µm/min [1] • Cost effectiveness

•

– higher density of reactants Anisotropic etch independent of crystal orientation – Vertical sidewalls/ability to control taper (≈90 deg vertical sidewalls, [2]) – Control lateral etch rate • ion motion normal to surface & protected sidewalls

•

•

High mask etching selectivity (120-200:1 for SiO2 [1]) – Thin mask more convenient – Ion bombardment not dominant, balancing of chemical sputtering and ion bombardment Relatively high smoothness on sidewalls and bottom – Depends on application requirements – Control of diffusion profiles for ions and radicals – May have a tradeoff with etch rate

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Wet and dry etching features

After Kovacs et al. (1998) Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Extending the RIE process RIE
dry etch, anisotropic, independent of crystal orientation Deep vertical etching achievable [3] – SLOW (~0.5-1 µm/min) http://www.ee.byu.edu/c leanroom/everything_w afers.parts/v_groove

User controllable

Increase etch rate: After Bruce Gale, U of Utah.

Increase reactants: Increase gas flow (pressure)

increase ion energies Increase RF power More radicals

More collisions (less directional)

Higher energy ions Decrease in anisotropy

< 20:1

Less mask selectivity

(Ayon,/ 1999) Felix Lu / Applied Quantum Technologies / Duke University April 2008

High density plasma requirements for faster etching • Typical RIE : Capacitively Coupled Plasma –ion energy and density of radicals COUPLED.

High Density Plasma ICP reactor

• Inductively Coupled Plasma (ICP) : control for plasma confinement • Substrate bias: control for ion bombardment • Radicals
chemical reaction (higher selectivity) • ICP desirable because: – High density of radicals (~10×) [6] without high density of high energy ions. [7] – Ion bombardment at low levels [7] for

http://www.ece.neu.edu/edsnu/hopwood/icp-labpage.html

Ion Assisted Chemical Etching

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Ion assisted chemical etching Enhanced etch rate not explained by summing XeF2 etch rate and Ar+ etch rate. XeF2 flux decay

~5-6× sum of individual etch rates

Spontaneous etch rate of XeF2 at 50 mTorr on Tungsten ~sum of XeF2 and Ar+ etch rates

-123 °C

Artifact from measurement

After Coburn and Winters (1979)

Ion enhanced etch rate at low temperature.

Ion assisted chemical etching

Bensaoula (1986) Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Ion enhanced chemical etching models Ions & electrons Damage to Si surface and/or SF6

SF6

F- ion F F F S FF

F

Enhanced dissociation and/or adsorption

Volatile product F F

Damage may also enhance removal

“Damage Enhancement model”

Chemisorbed F

Implanted ions provide the energy to chemically sputter the substrate material.

Volatile product

“Reactive Spot model” [Tachi (1985)]

[Coburn and Winters (1978)]

DRIE processes take advantage of ion assisted chemical etching Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Overview of DRIE processes Sidewall protection because fluorine radicals spontaneously etch Si. SF6 / C4F8

SF6 / O2 plasma F-

Condensed n-CF2 polymer

Condensed SiOxFy

F-

Mask

Mask

Significantly reduced spontaneous etch rate

Si

Si

@ ~ -110°C

A.k.a. “Cryo process”

A.k.a. “Bosch”, “Pulsed” or “Time multiplexed” process

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

DRIE parameters • High plasma density at low pressure – – – – –

low pressure reduces ion scattering maintains ion trajectory as mostly vertical better control of etch profiles improves transport of species into deep trenches Low P
fast pumping or low flow rate • Low flow rate reduces etch rate

• SF6 used as isotropic etchant due to low toxicity compared to F2. • O2 typically used with SF6 to : – Combine with SFn and CFn so that F does not combine with them
keeps F concentration high. – Passivates surfaces where mask has eroded – Reacts with polymer film to keep it from getting too thick.

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Fluorine reactivity with Si and SiO2 as a function of Temperature -eE’a/kT 1/2 Etch Rate = C1nFT e Constant with weak T dependence

C1 2.86×10-22 6.14×10-23

Si SiO2

Density of F atoms (3×1021/m3)

Ea’(eV) 0.108 0.163

condensation of SF6

Substrate temperature (°C)

[After Roth (2001)]

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

350

300

250

200

150

100

50

0

-50

-100

-150

-200

Si SiO2

-250

At -110°C , >100× drop in Si etch rate by F radicals.

etch rate (µm/min)

Etch rate using fluorine radicals 1.E+01 1.E+00 1.E-01 1.E-02 1.E-03 1.E-04 1.E-05 1.E-06 1.E-07

Cryo process data SF6 DRIE

Si etch rate increases by > 2× with decreasing T.
presumably due to Ion assisted enhancement.

SiO2 etch rate decreases by ~5× with decreasing T.
Presumably due to F not efficiently reacting with SiO2 compared with Si

Sidewall etching (R) effectively goes to zero at T < 90°C. Decreasing T

After Tachi (1987)

Should not go below -130 °C as SF 6 will condense on wafer [7] Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Bosch process details – High mask selectivity over Si etching (at least 50:1 if not 100:1) possible • “soft” teflon like polymer ( low energy ion bombardment for removal) • Low energy bombardment does not significantly erode masking materials. • Harder (more polymerized teflon based polymers) polymers would require larger ion bombardment energies and the masking selectivity suffers. [2]

– Alternating of etch and passivation steps allows easier and dynamic optimization of process. – Using the two steps simultaneously causes extinction of the amount of radicals by chemical recombination. [2] – This alternating sequence allows of RIE lag • the duty cycle of the step varied to adjust for trench widths (which are proportional to the amount of passivation at the bottom of the trench).

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Bosch process artifacts Bosch DRIE 20 µm via

After Qu (2006)

Without polymer

With polymer

Bosch Process scalloping

After Hibert (2002)

After Lietaer

Scallop period is determined by duty cycle.

www.alcatelmicromachining.com/amms_en/download/docs/news/doc148.pdf Felix Lu / Applied Quantum Technologies / Duke University / April 2008

DRIE artifacts Aspect Ratio Dependent Etching (ARDE)

Via etching

Chambers et al., Surface Technology Systems, Advanced Packaging, 2005 http://ap.pennnet.com/Articles/Article_Display.cfm?Secti on=Articles&Subsection=Display&ARTICLE_ID=225422

After Walker (2001)

Smaller opening
fewer ions
lower etch rate. Felix Lu / Applied Quantum Technologies / Duke University / April 2008

DRIE factors and tradeoffs • Maximize smoothness? – Bosch – reduce duty cycle (thus etch rate) for smaller scallops. – Cryo – intrinsically smoother than Bosch structures

• Maximize mask selectivity? – Less ion bombardment, more chemical activity

• Maintain 90°walls? – Balance ion bombardment with chemical etching

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Comparison of Bosch and Cryo DRIE processes Cryo etch rate > 5µm/min [20]

Bias is higher for Bosch
consistent with lower mask selectivity. Bosch process alternates between etch and passivation steps – which allows tuning of duty cycle to accommodate deep features. After Walker (2001)

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

UNC Alcatel DRIE system ?

http://www.alcatelmicromachining.com

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

Summary • DRIE main advantages over other wet and dry processes are fast etching speed with freedom to tune selectivity, smoothness, and have vertical sidewalls. • A remote high density plasma is independently controlled along with a substrate bias to balance radicals and ion bombardment. • The combination and judicious tuning of chemical and physical etching produces an enhanced etching rate with “smooth” vertical sidewalls • Cryo process has smoother sidewalls, however Bosch process allows dynamic optimization to account for RIE lag.

Felix Lu / Applied Quantum Technologies / Duke University / April 2008

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Gregory T. A. Kovacs, Nadim I. Maluf, Kurt E. Petersen, “Bulk Micromachining of Silicon”, Proceedings of the IEEE, Vol 86, No. 8, August 1998, p. 1536 Franz Laermer and Andrea Urban, Robert Bosch Gmbh, “Milestones in Deep Reactive Ion Etching”, Transducers’05, 13th international conference on solid state sensors, actuators, microsystems, Seoul, Korea, june 5-9, 2005, p. 1118 Roger Shile, MEMSTALK posting; [email protected] Tue Mar 6 20:19:47 2001 ] Bruce K. Gale, Dry etching. (presentation slides), Fundamentals of Micromachining, BIOENG 6421, The University of Utah A. A. Ayon, R. Braff, C. C. Lin, H. H. Sawin, and M. A. Schmidt, Characteriation of a time multiplexed inductively coupled plasma etcher, Journal of the Electrochemical Society, 146, (1) 339-349 (1999) Scott Smith, Ph.D. Thesis, “inductively coupled plasma etching of III-N semiconductors”, 1999, NCSU Martin J. Walker, “Comparison of Boasch and cryogenic processes for patterning high aspect ratio features in silicon”, © 2001 by the Society of Photo-opical Instrumentation Engineers, P. O. Box 10, Bellingham, Washington 98227 J. W. Coburn and Harold F. Winters, “Ion- and electron-assisted gas surface chemistry – An important effect in plasma etching”, J. Appl. Phys. 50 (5) May 1979, p. 3189 A. Bensaoula, A. Ignatiev, J. Strozier, and J. C. Wolfe, “Low Temperature ion beam enhanced etching of tungsten films with Xenon Difluoride”, Appl. Phys. Lett. 49 (24) 15 Dec 1986, p. 1663 Shin’ichi Tachi, Kazunori Tsujimoto, and Sadayuki Okudaira, “Low temperature reactive ion etching and microwave plasma etching of silicon”, Appl. Phys. Lett. 52 (8) 22 Feb 1988 J. Reece Roth, Industrial Plasma Engineering, CRC Press 2001 Hongwei Qu, Ph.D. Thesis, “DEVELOPMENT OF DRIE CMOS-MEMS PROCESS AND INTEGRATED ACCELEROMETERS”, U. of Florida, 2006 Cyrille Hibert, “State of the Art DRIE processing”, CMI Annual Review, 18 May 2004 Cyrille Hibert, “Dry Etching in MEMS fabrication”, CMI Comlab Review, 4 June 2002 Sami Franssila, “Introduction to Microfabrication”, John Wiley 2004 Shin’ichi Tachio and Sadayuki Okudaira, “Chemical Sputtering of silicon by F+, Cl+, and Br+ ions: Reactive spot model for reactive ion etching”, J. Vac. Sci. Tenol. B 4 (2) mar/Apr 1986, p. 459 S. A. McAuley, H. Ashraf, L. Atabo, A. Chambers, S. Hall, J. Hopkins, and G. Nicholls, “Silicon micromachining using a high ensity plasma source”, J. Phys. D: Appl. Phys: 45 (2001) 2769-2774 Ranganathan Nagarajan, Krishnamachar Prasad, Lioa Ebin, Balsubramaniam Narayanan, “Development of dual etch via tapering process for through-silicon interconnection”, Sensors and Actuators A 139 (2007) 323-329 Daniel L. Flamm, Mechanisms of silicon etching in fluorine and chlorine containing plasmas”, Pure & Appl. Chem. Vol. 62, No. 9, pp. 1709-1720, 1990 L. Sainiemi, and S. Franssila, “Mask Material effects in cryogenic deep reactive ion etching”, J. Vac., Sci. Technol. B 25 (3) May/ Jun 2007, p. 801 Felix Lu / Applied Quantum Technologies / Duke University / April 2008