RCA Wet Bench

1. Theory

• Definition: The RCA wet bench removes organics, particles, native oxide and metals from silicon with two heated peroxide chemistries, SC-1 (ammonia) and SC-2 (hydrochloric acid), usually with an HF dip between them; it was developed by Kern around 1965 and published in 1970 (Semiconductor History Museum; Oestling).
• SC-1: Hydrogen peroxide oxidizes the surface while ammonia slowly etches it, so particles are lifted off, and at high pH particles and surface are both negatively charged and repel (MIT; Kim et al.).
• SC-2: The acidic peroxide solution dissolves alkali ions and metals that SC-1 leaves, such as Na, K, Ca, Al, Fe and Mg (Oestling; MIT).
• Why it matters: Particles cause most yield loss, mobile ions shift thresholds and deep-level metals cut carrier lifetime (Oestling; MIT).

2. Working Principle

• Pre-clean: Piranha or Nanostrip removes resist and heavy organics (Oestling; Rochester).
• SC-1: About 1:1:5 ammonia:peroxide:water at 70 to 80 ${}^{\circ }\mathrm{C}$ for about 10 min (Oestling; Georgia Tech; Rochester).
• HF dip: Dilute HF about 1:50 or BOE strips the chemical oxide (Rochester; Georgia Tech).
• SC-2: About 1:1:6 hydrochloric acid:peroxide:water at 70 to 80 ${}^{\circ }\mathrm{C}$ (Oestling; Georgia Tech).
• Rinse and dry: Quick-dump rinses in water above 18 $\mathrm{M\Omega }\mathrm{cm}$ between baths, then spin rinse dry (Rochester; Oestling).

3. Applications

• Pre-furnace clean: Before oxidation and diffusion, where metals would diffuse deep into the wafer (MIT).
• Post-CMP and post-strip clean: Lowering Ca and particles after CMP (MIT; Defect Sources and Classification page).
• Gate-oxide preparation: Limiting mobile ions that shift threshold voltage (Oestling).

Bath Chemistry

Name	Formula	Related Formulae	Variables/Constants
SC-1 (APM)	$\mathrm{N}{\mathrm{H}}_4\mathrm{OH}:{\mathrm{H}}_2{\mathrm{O}}_2:{\mathrm{H}}_2\mathrm{O}=1:1:5\text{to}1:2:7$	$T=70\text{to}80{}^{\circ }C,t\approx 10\min$	$\begin{array}{r}Oestling:ratios,temperature,organics,\\ metalcomplexingandparticledislodging,\\ highN{H}_{4}OHetchingandrougheningSi;\\ GeorgiaTech:1:1:5at75to80{}^{\circ }Cfor10\\ to20min;Rochester:5:1:1at75{}^{\circ }Cfor\\ 10min;MIT:oxidizestoachemicaloxide\\ whileetching\end{array}$ $\begin{array}{r}(myreading:theperoxideoxidizeswhile\\ theammoniaslowlyetchestheoxide,so\\ thesurfaceisrenewedandparticlesare\\ undercut)\end{array}$
Oxide strip	$\mathrm{HF}:{\mathrm{H}}_2\mathrm{O}\approx 1:50,t=1\min$	$r_{BOE}\approx 20\mathring{\mathrm{A}}/s(6:1)$	$\begin{array}{r}Oestling:diluteHFabout1:50fornative\\ oxide;Rochester:50:1for1min,Teflon\\ orpolypropylenetank;GeorgiaTech:6:1\\ BOEfor30satabout20\mathring{\mathrm{A}}/s;MIT:HFdip\\ removestheSC-1chemicaloxide\end{array}$ $\begin{array}{r}(assumed:1\mathrm{nm}chemicaloxide;computed:\\ 0.5sin6:1BOE,soa30sdipisalarge\\ over-etch)\end{array}$
SC-2 (HPM)	$\mathrm{HCl}:{\mathrm{H}}_2{\mathrm{O}}_2:{\mathrm{H}}_2\mathrm{O}=1:1:6\text{to}1:2:8$	$\text{removes Na, K, Ca, Al, Fe, Mg}$	$\begin{array}{r}Oestling:ratiosat70to80{}^{\circ }C,alkali\\ ionsandA{l}^{3+},F{e}^{3+},M{g}^{2+};MIT:tracemetals\\ SC-1missesandNa,K,Ca;GeorgiaTech:\\ 1:1:6,plastictweezers\end{array}$ $\begin{array}{r}(myreading:thelowpHkeepsoxidized\\ metalsdissolvedasionsinsteadof\\ redepositing)\end{array}$

Mechanisms and Contamination Effects

Name	Formula	Related Formulae	Variables/Constants
SC-1 etch and particle removal	$r_{etch}\uparrow \text{with}[\mathrm{N}{\mathrm{H}}_4\mathrm{OH}],T$	$\text{high pH}\Rightarrow \text{negative particle and surface charge}$	$\begin{array}{r}Kimetal.:afewangstromofSietched,\\ etchrisingwithN{H}_{4}OHandfallingwith{H}_2{O}_2,\\ removaldependingonetchandpH,\\ negativelychargedparticlesrepelled;\\ ChristensonandSmith:etchandroughening\\ risingwithTandN{H}_{4}OH,unaffectedby{H}_2{O}_2\\ (seetheQ/A)\end{array}$ $\begin{array}{r}(myreading:undercutliftstheparticle\\ andlikechargeskeepitfromreturning)\end{array}$
Metal contamination and threshold shift	$\Delta V_{th}=\frac{q{Q}_M}{C_{ox}}$	$N_{metal}\sim 10^{10}\mathrm{c}{\mathrm{m}}^{-2}$	$Q_M:mobileionarealdensity$ $\begin{array}{r}{C}_{ox}:oxidecapacitanceperarea;Oestling:\\ Q_M=6.5\times 10^{11}\mathrm{c}{\mathrm{m}}^{-2}in10\mathrm{nm}oxidegivingabout\\ 0.3V,metallevelsabout1\times 10^{10}\mathrm{c}{\mathrm{m}}^{-2};MIT:\\ Ca1.7\times 10^{11}tobelow2\times 10^{10}\mathrm{c}{\mathrm{m}}^{-2}afterRCA,\\ deep-levelmetalsloweringlifetime;\\ thresholdrelation(computedcheck)\end{array}$ $\begin{array}{r}(computed:C_{ox}=3.45\times 10^{-7}F/c{m}^2,so\Delta V_{th}=\\ 0.30V,matchingOestling)\end{array}$
Metal mobility	$L=\sqrt{Dt}$	$D_{Cu}\left(900{}^{\circ }C\right)\approx 10^{-4}\mathrm{c}{\mathrm{m}}^2/s$	$\begin{array}{r}MIT:Cudiffusivityandabout780\mu \mathrm{m}inone\\ minute,morethanwaferthickness;\\ getteringbymobilizing,diffusingand\\ trapping\end{array}$ $(computed:\sqrt{1\times 10^{-4}x60}=0.077cm,775\mu \mathrm{m})$

Particles, Rinse and Economics

Name	Formula	Related Formulae	Variables/Constants
Killer particle size	$d_{crit}=\frac{1}{2}{F}_{min}$	$\sim 75\%\text{of yield loss from particles}$	$\begin{array}{r}{F}_{min}:minimumfeaturesize;Oestling:\\ criticalparticlesizehalftheminimum\\ feature,about75percentofyieldloss\\ fromparticledefects;MIT:poorRCA\\ raisedlight-pointdefectsfrom21to\\ about1012\end{array}$ $\begin{array}{r}(computed:21to1012isa48-fold\\ increase)\end{array}$
Rinse water	${\rho }_{DI}>18\mathrm{M\Omega }\mathrm{cm}$	$\text{QDR}5\min \text{between baths}$	$\begin{array}{r}Oestling:DIwaterabove18\mathrm{M\Omega }\mathrm{cm};\\ GeorgiaTech:runDIto18MOhmbefore\\ rinsing;Rochester:5minquick-dumprinse\\ aftereachbath\end{array}$ $\begin{array}{r}(myreading:rinsesstopeachreactionand\\ avoidcarryingchemicalsintothenext\\ bath)\end{array}$
Yield economics	$\Delta P=N_w{N}_{chip}52p\Delta Y$	$N_w=1000/\text{week},N_{chip}=100,p=50\text{USD}$	$N_w:wafersperweek,N_{c}hip:chipsperwafer$ $p:priceperchip$ $\begin{array}{r}\Delta Y:yieldgain;Oestling:about260\\ milliondollarsofoutputandabout10\\ millionfroma3.8percentgain\end{array}$ $\begin{array}{r}(computed:1000x100x52x50=2.6\times 10^8,\\ and3.8percentofitis9.9\times 10^6)\end{array}$

Important Questions

Ques: What does each step of the RCA clean remove?

Organics and particles, oxide, then metals. Oestling lists SC-1, ammonia, peroxide and water at 1:1:5 to 1:2:7 and 70 to 80 ${}^{\circ }\mathrm{C}$, as removing organics, complexing some metals and dislodging particles, and SC-2, hydrochloric acid, peroxide and water at 1:1:6 to 1:2:8, as removing alkali ions and Al, Fe and Mg; dilute HF near 1:50 removes native oxide. The MIT lecture explains that SC-1 grows a thin chemical oxide while etching, so particles lift off, and that an HF dip between baths exposes fresh silicon. Georgia Tech and Rochester run SC-1 at 75 to 80 ${}^{\circ }\mathrm{C}$ for 10 to 20 min, 50:1 HF or 6:1 BOE, then SC-2, with rinses between. MIT reports calcium falling from $1.7\times 10^{11}$ to below $2\times 10^{10}$ $\mathrm{c}{\mathrm{m}}^{-2}$ after RCA, a post-CMP case where such residues are a known defect source (Defect Sources and Classification page). Reference: M. Oestling, ‘Wafer Clean and Wet Processing’, IH2655 Lecture 3, KTH (Spring 2012): contaminants as particles, organics, heavy metals, alkali ions and native oxide, killer particle size half the minimum feature, piranha H2SO4:H2O2 2:1 to 4:1, SC-1 NH4OH:H2O2:H2O 1:1:5 to 1:2:7 at 70 to 80 C for 10 min removing organics, complexing metals and dislodging particles, high NH4OH etching Si and roughening, SC-2 HCl:H2O2:H2O 1:1:6 to 1:2:8 at 70 to 80 C removing alkali ions and Al, Fe, Mg, dilute HF about 1:50 for native oxide, Ohmi sequence with O3/H2O and HF/H2O2, DI water above 18 MOhm cm, 20 to 50 kHz or about 1 MHz agitation, metal contamination about 1e10 atoms/cm2, about 75 per cent of yield loss from particles, yield economics example, Q_M = 6.5e11 cm-2 in 10 nm oxide giving about 0.3 V shift (fetch-tool summary); MIT OpenCourseWare 6.774 Physics of Microfabrication: Front End Processing (Fall 2004), cleaning and gettering lecture transcript: hot sulfuric-peroxide resist strip, SC-1 at about 5:1:1 water:H2O2:NH4OH and 80 C for 10 min oxidizing the surface to a chemical oxide while etching to lift particles, HF dip between steps, SC-2 at 6:1:1 for trace metals and Na, K, Ca, deep-level Fe, Cu, Au, Zn lowering lifetime, Ca from 1.7e11 to below 2e10 atoms/cm2 after RCA, a poor RCA raising light-point defects from 21 to about 1012, Cu diffusivity about 1e-4 cm2/s at 900 C, gettering steps and intrinsic gettering temperatures (fetch-tool summary); Georgia Tech IEN Micro/Nano Fabrication Facility, ‘RCA clean’ procedure: SC-1 1:1:5 NH4OH:H2O2:H2O at 75 to 80 C for 10 to 20 min for organics and particles, 6:1 BOE for 30 s at about 20 A/s for the thin oxide, SC-2 1:1:6 HCl:H2O2:H2O at 75 to 80 C for metals, DI rinse at 18 MOhm, plastic tweezers for BOE and SC-2, about 45 min per clean (fetch-tool summary); University of Rochester URnano, ‘SOP RCA clean bench’: Nanostrip 5 min, QDR 5 min between steps, SC1 5:1:1 at 75 C for 10 min, 50:1 HF for 1 min, SC2 6:1:1 at 75 C for 10 min, spin rinse dry, HF tank in Teflon or polypropylene only (fetch-tool summary); Defect Sources and Classification page of this vault.

Ques: Does hydrogen peroxide concentration change the SC-1 etch rate?

The sources disagree. Kim et al. report that the SC-1 etch rate rises with ammonia and falls with peroxide, and that more peroxide lowers copper precipitation but also lowers particle removal. Christenson and Smith found that etch rate and roughening rose with temperature and ammonia but were unaffected by peroxide concentration. Both agree that ammonia and temperature drive the etch, which Oestling also links to roughening. The difference may come from range: a peroxide effect could appear when peroxide is low enough that oxidation, not oxide dissolution, limits the etch (my reading). Within normal ratios such as 1:1:5, peroxide is therefore more likely a weak factor, with ammonia and temperature the main controls, and peroxide mainly protecting against metal deposition (my reading). Reference: H. T. Kim, G. H. Park, B. J. Cho, J. H. Lee, M. S. Kim, J. Y. Kim and J. G. Park (Hanyang University), ‘Prevention of Metal Contamination in Sub 50 nm SC1 Cleaning Process’, 228th ECS Meeting (2015), abstract: SC1 etching a few angstrom of Si, etch rate rising with NH4OH and falling with H2O2, particle removal depending on etch rate and pH with negatively charged particles repelled at high pH, Cu precipitates falling but particle removal also falling at high H2O2 (fetch-tool summary); K. K. Christenson and S. M. Smith (FSI International), ‘Effects of SC-1 Dilution and Temperature Variations on Etch Rate and Surface Haze’, MRS Proc. 386, 135 (1995), abstract: slight Si etch and roughening a concern, roughening and etch rising with temperature and NH4OH, unaffected by H2O2 concentration (fetch-tool summary); M. Oestling, ‘Wafer Clean and Wet Processing’, IH2655 Lecture 3, KTH (Spring 2012): contaminants as particles, organics, heavy metals, alkali ions and native oxide, killer particle size half the minimum feature, piranha H2SO4:H2O2 2:1 to 4:1, SC-1 NH4OH:H2O2:H2O 1:1:5 to 1:2:7 at 70 to 80 C for 10 min removing organics, complexing metals and dislodging particles, high NH4OH etching Si and roughening, SC-2 HCl:H2O2:H2O 1:1:6 to 1:2:8 at 70 to 80 C removing alkali ions and Al, Fe, Mg, dilute HF about 1:50 for native oxide, Ohmi sequence with O3/H2O and HF/H2O2, DI water above 18 MOhm cm, 20 to 50 kHz or about 1 MHz agitation, metal contamination about 1e10 atoms/cm2, about 75 per cent of yield loss from particles, yield economics example, Q_M = 6.5e11 cm-2 in 10 nm oxide giving about 0.3 V shift (fetch-tool summary).

Ques: Why does SC-1 remove particles so well?

By etching under them and repelling them. The MIT lecture describes SC-1 as oxidizing the silicon surface to a chemical oxide while slowly etching it, so particles are released as the surface recedes. Kim et al. state that particle removal depends on both etch rate and pH, and that at high pH particles carry a strongly negative charge that repels them from the similarly charged surface. The etch is only a few angstrom, but Christenson and Smith warn that even this slight etch roughens silicon, more so at higher temperature and ammonia concentration. Oestling adds that agitation at 20 to 50 $\mathrm{kHz}$ or about 1 $\mathrm{MHz}$ improves removal. The particle that matters is about half the minimum feature (Oestling). Reference: MIT OpenCourseWare 6.774 Physics of Microfabrication: Front End Processing (Fall 2004), cleaning and gettering lecture transcript: hot sulfuric-peroxide resist strip, SC-1 at about 5:1:1 water:H2O2:NH4OH and 80 C for 10 min oxidizing the surface to a chemical oxide while etching to lift particles, HF dip between steps, SC-2 at 6:1:1 for trace metals and Na, K, Ca, deep-level Fe, Cu, Au, Zn lowering lifetime, Ca from 1.7e11 to below 2e10 atoms/cm2 after RCA, a poor RCA raising light-point defects from 21 to about 1012, Cu diffusivity about 1e-4 cm2/s at 900 C, gettering steps and intrinsic gettering temperatures (fetch-tool summary); H. T. Kim, G. H. Park, B. J. Cho, J. H. Lee, M. S. Kim, J. Y. Kim and J. G. Park (Hanyang University), ‘Prevention of Metal Contamination in Sub 50 nm SC1 Cleaning Process’, 228th ECS Meeting (2015), abstract: SC1 etching a few angstrom of Si, etch rate rising with NH4OH and falling with H2O2, particle removal depending on etch rate and pH with negatively charged particles repelled at high pH, Cu precipitates falling but particle removal also falling at high H2O2 (fetch-tool summary); K. K. Christenson and S. M. Smith (FSI International), ‘Effects of SC-1 Dilution and Temperature Variations on Etch Rate and Surface Haze’, MRS Proc. 386, 135 (1995), abstract: slight Si etch and roughening a concern, roughening and etch rising with temperature and NH4OH, unaffected by H2O2 concentration (fetch-tool summary); M. Oestling, ‘Wafer Clean and Wet Processing’, IH2655 Lecture 3, KTH (Spring 2012): contaminants as particles, organics, heavy metals, alkali ions and native oxide, killer particle size half the minimum feature, piranha H2SO4:H2O2 2:1 to 4:1, SC-1 NH4OH:H2O2:H2O 1:1:5 to 1:2:7 at 70 to 80 C for 10 min removing organics, complexing metals and dislodging particles, high NH4OH etching Si and roughening, SC-2 HCl:H2O2:H2O 1:1:6 to 1:2:8 at 70 to 80 C removing alkali ions and Al, Fe, Mg, dilute HF about 1:50 for native oxide, Ohmi sequence with O3/H2O and HF/H2O2, DI water above 18 MOhm cm, 20 to 50 kHz or about 1 MHz agitation, metal contamination about 1e10 atoms/cm2, about 75 per cent of yield loss from particles, yield economics example, Q_M = 6.5e11 cm-2 in 10 nm oxide giving about 0.3 V shift (fetch-tool summary).

Ques: Why must metals be removed before high-temperature steps?

Because they diffuse fast and damage devices. The MIT lecture notes that copper diffuses at about $10^{-4}$ $\mathrm{c}{\mathrm{m}}^2/s$ at 900 ${}^{\circ }\mathrm{C}$, which gives $\sqrt{Dt}$ = 775 $\mu \mathrm{m}$ in one minute (computed), more than the wafer thickness, and that deep-level metals such as Fe, Cu, Au and Zn act as recombination centres and cut lifetime, while alkali ions degrade gate oxides. Oestling gives an example: $6.5\times 10^{11}$ $\mathrm{c}{\mathrm{m}}^{-2}$ of mobile charge in a 10 $\mathrm{nm}$ oxide shifts threshold by about 0.3 V, which $q{Q}_M/C_{ox}$ reproduces as 0.30 V (computed). SC-2 is the step that removes Na, K and Ca most effectively (MIT), and residual metals end near $10^{10}$ $\mathrm{c}{\mathrm{m}}^{-2}$ (Oestling). Reference: MIT OpenCourseWare 6.774 Physics of Microfabrication: Front End Processing (Fall 2004), cleaning and gettering lecture transcript: hot sulfuric-peroxide resist strip, SC-1 at about 5:1:1 water:H2O2:NH4OH and 80 C for 10 min oxidizing the surface to a chemical oxide while etching to lift particles, HF dip between steps, SC-2 at 6:1:1 for trace metals and Na, K, Ca, deep-level Fe, Cu, Au, Zn lowering lifetime, Ca from 1.7e11 to below 2e10 atoms/cm2 after RCA, a poor RCA raising light-point defects from 21 to about 1012, Cu diffusivity about 1e-4 cm2/s at 900 C, gettering steps and intrinsic gettering temperatures (fetch-tool summary); M. Oestling, ‘Wafer Clean and Wet Processing’, IH2655 Lecture 3, KTH (Spring 2012): contaminants as particles, organics, heavy metals, alkali ions and native oxide, killer particle size half the minimum feature, piranha H2SO4:H2O2 2:1 to 4:1, SC-1 NH4OH:H2O2:H2O 1:1:5 to 1:2:7 at 70 to 80 C for 10 min removing organics, complexing metals and dislodging particles, high NH4OH etching Si and roughening, SC-2 HCl:H2O2:H2O 1:1:6 to 1:2:8 at 70 to 80 C removing alkali ions and Al, Fe, Mg, dilute HF about 1:50 for native oxide, Ohmi sequence with O3/H2O and HF/H2O2, DI water above 18 MOhm cm, 20 to 50 kHz or about 1 MHz agitation, metal contamination about 1e10 atoms/cm2, about 75 per cent of yield loss from particles, yield economics example, Q_M = 6.5e11 cm-2 in 10 nm oxide giving about 0.3 V shift (fetch-tool summary).

Ques: Who developed the RCA clean, and are its compositions recorded consistently?

Werner Kern at RCA, with a 1970 publication. The Semiconductor History Museum of Japan dates the development to about 1965 and the RCA Review paper, with Puotinen, to 1970, describing it as the standard for over 40 years in fused-quartz vessels at 75 to 80 ${}^{\circ }\mathrm{C}$. It lists SC-1 as ammonia, hydrochloric acid and peroxide, whereas Oestling, MIT, Georgia Tech and Rochester all give SC-1 as ammonia, peroxide and water, with hydrochloric acid only in SC-2. The museum entry is therefore more likely a transcription error (my reading). Mixing order also matters: Georgia Tech adds water, then ammonia or acid, then peroxide, and uses plastic tweezers in SC-2 and BOE. Reference: Semiconductor History Museum of Japan, ‘RCA Cleaning’ (2022): developed by W. Kern around 1965 and published in RCA Review 1970 with D. Puotinen, standard for over 40 years, fused quartz vessels at 75 to 80 C, SC-1 for organics and particles and SC-2 for metal ions, SC-1 listed as NH4OH, HCl and H2O2 (fetch-tool summary); M. Oestling, ‘Wafer Clean and Wet Processing’, IH2655 Lecture 3, KTH (Spring 2012): contaminants as particles, organics, heavy metals, alkali ions and native oxide, killer particle size half the minimum feature, piranha H2SO4:H2O2 2:1 to 4:1, SC-1 NH4OH:H2O2:H2O 1:1:5 to 1:2:7 at 70 to 80 C for 10 min removing organics, complexing metals and dislodging particles, high NH4OH etching Si and roughening, SC-2 HCl:H2O2:H2O 1:1:6 to 1:2:8 at 70 to 80 C removing alkali ions and Al, Fe, Mg, dilute HF about 1:50 for native oxide, Ohmi sequence with O3/H2O and HF/H2O2, DI water above 18 MOhm cm, 20 to 50 kHz or about 1 MHz agitation, metal contamination about 1e10 atoms/cm2, about 75 per cent of yield loss from particles, yield economics example, Q_M = 6.5e11 cm-2 in 10 nm oxide giving about 0.3 V shift (fetch-tool summary); MIT OpenCourseWare 6.774 Physics of Microfabrication: Front End Processing (Fall 2004), cleaning and gettering lecture transcript: hot sulfuric-peroxide resist strip, SC-1 at about 5:1:1 water:H2O2:NH4OH and 80 C for 10 min oxidizing the surface to a chemical oxide while etching to lift particles, HF dip between steps, SC-2 at 6:1:1 for trace metals and Na, K, Ca, deep-level Fe, Cu, Au, Zn lowering lifetime, Ca from 1.7e11 to below 2e10 atoms/cm2 after RCA, a poor RCA raising light-point defects from 21 to about 1012, Cu diffusivity about 1e-4 cm2/s at 900 C, gettering steps and intrinsic gettering temperatures (fetch-tool summary); Georgia Tech IEN Micro/Nano Fabrication Facility, ‘RCA clean’ procedure: SC-1 1:1:5 NH4OH:H2O2:H2O at 75 to 80 C for 10 to 20 min for organics and particles, 6:1 BOE for 30 s at about 20 A/s for the thin oxide, SC-2 1:1:6 HCl:H2O2:H2O at 75 to 80 C for metals, DI rinse at 18 MOhm, plastic tweezers for BOE and SC-2, about 45 min per clean (fetch-tool summary); University of Rochester URnano, ‘SOP RCA clean bench’: Nanostrip 5 min, QDR 5 min between steps, SC1 5:1:1 at 75 C for 10 min, 50:1 HF for 1 min, SC2 6:1:1 at 75 C for 10 min, spin rinse dry, HF tank in Teflon or polypropylene only (fetch-tool summary).

References

1. M. Oestling, “Wafer clean and wet processing,” IH2655 Lecture 3, KTH Royal Institute of Technology, 2012. Available: https://www.kth.se/social/upload/4f1d17ccf27654791f000000/Lecture%203.pdf
2. MIT OpenCourseWare, 6.774 Physics of Microfabrication: Front End Processing, Fall 2004, lecture transcript on cleaning and gettering. Available: https://ocw.mit.edu/courses/6-774-physics-of-microfabrication-front-end-processing-fall-2004/1RhMlsWln7tVtc_3WeIdSblNinAm7UP8N_transcript.pdf
3. Georgia Tech IEN Micro/Nano Fabrication Facility, “RCA clean.” Available: https://cleanroom.gatech.edu/articles/rcaclean
4. University of Rochester URnano, “SOP: RCA clean bench.” Available: https://www.rochester.edu/urnano/assets/pdf/sop_rca_clean_bench.pdf
5. Semiconductor History Museum of Japan, “RCA cleaning,” 2022. Available: https://www.shmj.or.jp/english/pdf/em/exhibi2432E.pdf
6. H. T. Kim, G. H. Park, B. J. Cho, J. H. Lee, M. S. Kim, J. Y. Kim and J. G. Park, “Prevention of metal contamination in sub 50 nm SC1 cleaning process,” 228th ECS Meeting, 2015 (abstract read). Available: https://ecs.confex.com/ecs/228/webprogram/Paper57843.html
7. K. K. Christenson and S. M. Smith, “Effects of SC-1 dilution and temperature variations on etch rate and surface haze,” MRS Proceedings, vol. 386, p. 135, 1995 (abstract read). Available: https://www.cambridge.org/core/journals/mrs-online-proceedings-library-archive/article/abs/effects-of-sc1-dilution-and-temperature-variations-on-etch-rate-and-surface-haze/3B21F4233E714DCAAE4E916BB0287AE3
8. Defect Sources and Classification page of this vault.

Related pages in this vault: Defect Sources and Classification, Thermal Oxidation Furnace, Diffusion Furnace, X-Ray Fluorescence Spectrometer (XRF and TXRF), Laser-Scattering Particle Inspection, CMP Polisher, Megasonic Single-Wafer Cleaner, Spin Rinse Dryer and IPA Vapor Dryer, Vapor HF and Plasma Dry Clean Tool, Wet Etch Bench, Plasma Asher (Barrel and Downstream).