Parallel-Plate Reactive Ion Etcher (RIE)

1. Theory

• Definition: A parallel-plate reactive ion etcher places the wafer on the smaller, RF-powered electrode so that ions accelerated across the sheath combine with reactive radicals to etch directionally (MIT; KTH).
• Plasma: About $10^{15}$ cm${}^{-3}$ neutrals, 1 to 10 per cent radicals, and $10^8$ to $10^{12}$ cm${}^{-3}$ ions and electrons (KTH).
• Ion energy: 100 to 700 eV in RIE mode against 10 to 100 eV in plasma mode, set by the electrode area ratio (KTH; Purdue).
• Ion-enhanced etching: Ions break bonds, help form products and sputter them away, giving anisotropy with chemical selectivity (MIT; KTH).
• Chemistry: Fluorine for Si, carbon-rich fluorocarbons for selective oxide, HBr for gates, chlorine for Al (MIT; Purdue; KTH; Berkeley).

2. Working Principle

• Pressure: 10 to 100 mTorr; lower pressure lengthens the mean free path and raises anisotropy (MIT; KTH).
• Self-bias: The smaller electrode takes the larger voltage, $V_1/V_2=(A_2/A_1{)}^m$ with $m$ usually 1 to 2 (Purdue).
• Recipes: Example parallel-plate recipes for oxide, nitride, Al and W (Berkeley).
• Endpoint: Plasma emission tracks the clearing of a film (Optical Emission Spectroscopy Endpoint (OES) page).

3. Applications

• Pattern transfer: Oxide contacts, nitride, poly gates and metal lines (MIT; KTH; Berkeley).
• Gate etch: HBr/O${}_2$ stopping on 3 nm oxide (KTH).
• Back-sputter and descum: Physical cleaning in the same chamber type (my reading).

Plasma and Ion Energy

Name	Formula	Related Formulae	Variables/Constants
Plasma composition	$n_{neutral}\approx 10^{15}\mathrm{c}{\mathrm{m}}^{-3},n_i=n_e=10^8\text{to}10^{12}\mathrm{c}{\mathrm{m}}^{-3}$	$\text{radicals}=1\text{to}10\%\text{of neutrals}$	$\begin{array}{r}KTH:high-densityplasmasreach10^{11}to10^{12}\\ \mathrm{c}{\mathrm{m}}^{-3}ionsat1to10\mathrm{mTorr}\end{array}$ $\begin{array}{r}(computed:at10^{10}\mathrm{c}{\mathrm{m}}^{-3}theionizedfraction\\ is10^{-5})\end{array}$
Sheath energy	$E_i=10\text{to}100\mathrm{eV}\text{(plasma mode)}$	$E_i=100\text{to}700\mathrm{eV}\text{(RIE mode)}$	$\begin{array}{r}KTH;MIT:smallerpoweredelectrodegives\\ agreatervoltagedropoverthewafer,\\ moreanisotropy,morephysicaletchand\\ lessselectivity;higherpressurereduces\\ thephysicalpart\end{array}$ $\begin{array}{r}(myreading:RIEputsthewaferonthe\\ smallpoweredelectrode,plasmamodeon\\ thelargegroundedone)\end{array}$
Electrode-area law	$\frac{V_1}{V_2}={\left(\frac{A_2}{A_1}\right)}^m$	$m=4\text{(simple theory)},m=1\text{to}2\text{(usual)}$	$\begin{array}{r}Purdue:largerelectrodeusuallythe\\ chamberwall;NPTEL:fourth-power\\ dependence;DCandRFMagnetronSputter\\ Systempage\end{array}$ $\begin{array}{r}(computed:anarearatioof3gives81for\\ m=4but3to9form=1to2)\end{array}$

Etch Chemistry and Selectivity

Name	Formula	Related Formulae	Variables/Constants
Silicon in fluorine	$4F+Si\to Si{F}_4\uparrow$	$C{F}_4+O_2:O\text{scavenges}C{F}_x,F\uparrow$	$\begin{array}{r}MIT:O_{2}reducesFrecombinationwithC{F}_3\\ andC{F}_2,toomuch{O}_{2}oxidizesSi;KTH:C{F}_4/O_2\end{array}$ $\begin{array}{r}(myreading:theratepeaksatmoderate{O}_2,\\ thenfallsasthesurfaceoxidizes)\end{array}$
Selective oxide etch	$4F+Si{O}_2\to Si{F}_4+O_2$	$CH{F}_3:r_{Si{O}_2}\gg r_{Si}$	$\begin{array}{r}MIT:fluorine-richoxideetchistoo\\ isotropicandunselective,soFisreduced\\ andCincreased;Purdue:HtiesupFand\\ promotesTeflon-likepolymer,CH{F}_{3}etches\\ Si{O}_{2}fastandSislowly\end{array}$ $\begin{array}{r}(myreading:polymerformsonSibutis\\ consumedbytheoxygenreleasedfromSi{O}_2)\end{array}$
Gate etch selectivity	$Si+Br\to SiBr\uparrow ,\text{sidewall SiOBr}$	$O_2:2\to 5\%\Rightarrow r_{Si{O}_2}:3.7\to \approx 0nm/min$	$\begin{array}{r}KTH:HBr+O_{2}polygateetch,26\mathrm{nm}gates\\ stoppingon3\mathrm{nm}oxide\end{array}$ $\begin{array}{r}(computed:at3.7nm/minthe3\mathrm{nm}oxide\\ wouldbegoneinunder1\min ofover-etch)\end{array}$

Products, Anisotropy and Loading

Name	Formula	Related Formulae	Variables/Constants
Volatile products	$T_b:Si{F}_4=-86,SiC{l}_4=57.6,W{F}_6=17.5{}^{\circ }C$	$AlC{l}_3:177.8{}^{\circ }C\text{(sublimes)}$	$\begin{array}{r}MIT:etchby-productsshouldhavelow\\ boilingpoints;Berkeley:Aletchedin\\ BC{l}_3/C{l}_2/N_2/CHC{l}_{3}at250\mathrm{mTorr},WinS{F}_{6}at\\ 100\mathrm{mTorr}\end{array}$ $\begin{array}{r}(myreading:chlorineisusedforAl\\ becausealuminiumfluorideisnotlisted\\ asvolatile)\end{array}$
Anisotropy mechanisms	$S_{phys}\approx 1,S_{chem}\approx 0.01$	$\text{ions: break bonds, form products, sputter products}$	$\begin{array}{r}S:stickingcoefficient;MIT:physical\\ etchingenhancesanisotropy,chemical\\ etchingmoreisotropic;KTH:moreionsor\\ moresidewallinhibitorimprove\\ directionality\end{array}$ $\begin{array}{r}(myreading:low-stickingradicalsreach\\ sidewalls,soinhibitorsmustprotect\\ them)\end{array}$
Loading	$\text{macro-loading: wafers},\text{micro-loading: open area}$	$p_{RIE}=10\text{to}100\mathrm{mTorr}$	$\begin{array}{r}MIT:toomanywafersorlargeunmasked\\ areasconsumeetchant;Purdue:local\\ loadinginCC{l}_2{F}_2/O_{2}RIE;OpticalEmission\\ SpectroscopyEndpoint(OES)pagefor\\ endpoint\end{array}$ $\begin{array}{r}(myreading:loadingchangestheetch\\ time,soendpointdetectionbeatsfixed\\ timing)\end{array}$

Important Questions

Ques: Why does RIE etch anisotropically when a chemical plasma etch does not?

Because ions accelerated across the sheath drive the reaction downward. MIT explains that placing the wafer on the smaller electrode gives a greater voltage drop above it and more energetic ions, increasing anisotropy and the physical part of the etch but lowering selectivity. KTH gives 100 to 700 $\mathrm{eV}$ in RIE mode against 10 to 100 $\mathrm{eV}$ in plasma mode. MIT lists three ways ions enhance etching: breaking bonds so neutrals react more readily, helping form volatile products and sputtering products away. It also notes that physical etching has a sticking coefficient near 1 and chemical etching near 0.01, which spreads chemical attack onto sidewalls. KTH adds that directionality improves with more ion bombardment or more sidewall inhibitor. Reference: MIT 6.152J Micro/Nano Processing Technology, lecture 17 ‘Etching’ (2005): wet chemical, isotropic and selective against dry physical, anisotropic and not selective by itself, wet etch bias and HF attack of SiO2 driving the move to dry etching, parallel-plate plasma mode with a larger wafer electrode giving weaker, more uniform bombardment and higher pressure reducing the physical part, RIE with a smaller etch electrode giving a greater voltage drop, more anisotropy, more physical etch and less selectivity, RIE at 10 to 100 mTorr with lower pressure lengthening the mean free path, high-density plasmas at 1 to 10 mTorr, CF4 giving F and 4F + Si -> SiF4, O2 combining with CF3 and CF2 to reduce F recombination but oxidizing Si in excess, oxide etching 4F + SiO2 -> SiF4 + O2 too isotropic and unselective unless F is reduced and C increased, three ion-enhancement mechanisms (bond breaking, product formation, product sputtering), sticking about 1 for physical and 0.01 for chemical etching, by-product boiling points SiCl4 57.6 C, SiF4 -86 C, WF6 17.5 C, AlCl3 177.8 C (sublimes), macro-loading and micro-loading (fetch-tool summary); KTH Royal Institute of Technology, Lecture 7 ‘Etching’: about 1e15 cm-3 neutrals with 1 to 10 per cent radicals and 1e8 to 1e12 cm-3 ions and electrons, sheath 10 to 100 eV in plasma mode and 100 to 700 eV in RIE mode, RIE at 10 to 100 mTorr, high-density plasmas 1e11 to 1e12 cm-3 at 1 to 10 mTorr, HBr + O2 for poly gates with SiBr volatile and SiOBr sidewall passivation, raising O2 from 2 to 5 per cent dropping SiO2 rate from 3.7 nm/min to about 0 for 26 nm gates stopping on 3 nm oxide, ion-enhanced etching anisotropic and selective, directionality from more ions or more sidewall inhibitor, smaller electrode giving larger voltage to the plasma (fetch-tool summary).

Ques: How does electrode area set the ion energy, and which exponent is more likely?

Through $V_1/V_2=(A_2/A_1{)}^m$; an exponent of 1 to 2 is more likely in practice. Purdue gives $m=4$ in simple theory but states that it is usually 1 to 2, with the larger electrode being the chamber wall. NPTEL states only the fourth-power dependence. For an area ratio of 3, $m=4$ predicts a voltage ratio of 81, while $m=1$ to 2 gives 3 to 9 (computed). The Purdue statement is more likely right for real reactors because it reports the usual measured range, while NPTEL repeats the idealised value (my reading). Either way, the smaller electrode takes the larger sheath voltage, which is why the wafer sits on the small powered electrode (MIT). The same physics sets target bias on the DC and RF Magnetron Sputter System page. Reference: Purdue ECE 695Q, lectures 46 and 47 (nanoHUB, April 2016): electrode voltage-area relation V1/V2 = (A2/A1)^m with m = 4 in simple theory but usually 1 to 2, the larger electrode usually the chamber wall, hydrogen tying up F and promoting Teflon-like polymer, CHF3 etching SiO2 fast and Si slowly, C4F8 for polymerization in deep RIE, high-density plasma at 1 to 10 mTorr with bias below 1 kV and ion energy set independently of density, local loading in CCl2F2/O2 RIE (fetch-tool summary); NPTEL course 104103019, Module 4, Lecture 25 ‘Electrode-size effects’: fourth-power dependence of sheath-voltage ratio on electrode area, large grounded electrode raising the target sheath potential (fetch-tool summary); MIT 6.152J Micro/Nano Processing Technology, lecture 17 ‘Etching’ (2005): wet chemical, isotropic and selective against dry physical, anisotropic and not selective by itself, wet etch bias and HF attack of SiO2 driving the move to dry etching, parallel-plate plasma mode with a larger wafer electrode giving weaker, more uniform bombardment and higher pressure reducing the physical part, RIE with a smaller etch electrode giving a greater voltage drop, more anisotropy, more physical etch and less selectivity, RIE at 10 to 100 mTorr with lower pressure lengthening the mean free path, high-density plasmas at 1 to 10 mTorr, CF4 giving F and 4F + Si -> SiF4, O2 combining with CF3 and CF2 to reduce F recombination but oxidizing Si in excess, oxide etching 4F + SiO2 -> SiF4 + O2 too isotropic and unselective unless F is reduced and C increased, three ion-enhancement mechanisms (bond breaking, product formation, product sputtering), sticking about 1 for physical and 0.01 for chemical etching, by-product boiling points SiCl4 57.6 C, SiF4 -86 C, WF6 17.5 C, AlCl3 177.8 C (sublimes), macro-loading and micro-loading (fetch-tool summary); DC and RF Magnetron Sputter System page of this vault.

Ques: How is oxide etched selectively to silicon?

By starving the plasma of fluorine and adding carbon. MIT notes that fluorine-rich oxide etching, $4F+Si{O}_2\to Si{F}_4+O_2$, is too isotropic with poor selectivity to Si. Its remedy is to reduce fluorine production and increase carbon. Purdue explains that hydrogen ties up fluorine and promotes Teflon-like polymerization, and that CHF${}_3$ etches SiO${}_2$ fast and Si slowly. In my reading, polymer then builds up on exposed silicon, while on oxide the released oxygen consumes it, so etching continues only on the oxide. Berkeley’s parallel-plate oxide recipes mix CF${}_4$ with CHF${}_3$ and helium, for example 90, 30 and 120 $\mathrm{sccm}$ at 850 $\mathrm{W}$ in its LAM 2. Reference: MIT 6.152J Micro/Nano Processing Technology, lecture 17 ‘Etching’ (2005): wet chemical, isotropic and selective against dry physical, anisotropic and not selective by itself, wet etch bias and HF attack of SiO2 driving the move to dry etching, parallel-plate plasma mode with a larger wafer electrode giving weaker, more uniform bombardment and higher pressure reducing the physical part, RIE with a smaller etch electrode giving a greater voltage drop, more anisotropy, more physical etch and less selectivity, RIE at 10 to 100 mTorr with lower pressure lengthening the mean free path, high-density plasmas at 1 to 10 mTorr, CF4 giving F and 4F + Si -> SiF4, O2 combining with CF3 and CF2 to reduce F recombination but oxidizing Si in excess, oxide etching 4F + SiO2 -> SiF4 + O2 too isotropic and unselective unless F is reduced and C increased, three ion-enhancement mechanisms (bond breaking, product formation, product sputtering), sticking about 1 for physical and 0.01 for chemical etching, by-product boiling points SiCl4 57.6 C, SiF4 -86 C, WF6 17.5 C, AlCl3 177.8 C (sublimes), macro-loading and micro-loading (fetch-tool summary); Purdue ECE 695Q, lectures 46 and 47 (nanoHUB, April 2016): electrode voltage-area relation V1/V2 = (A2/A1)^m with m = 4 in simple theory but usually 1 to 2, the larger electrode usually the chamber wall, hydrogen tying up F and promoting Teflon-like polymer, CHF3 etching SiO2 fast and Si slowly, C4F8 for polymerization in deep RIE, high-density plasma at 1 to 10 mTorr with bias below 1 kV and ion energy set independently of density, local loading in CCl2F2/O2 RIE (fetch-tool summary); Marvell NanoLab (UC Berkeley) lab manual chapter 1.9 ‘VLSI etchants’: Technics-c SF6 12.9 sccm + He 21.0 sccm at 100 W removing poly isotropically at about the nitride rate, Technics-c CF4 10 + CHF3 5 + He 10 sccm at 200 W, LAM 1 CCl4 130 + O2 15 + He 130 sccm at 280 mTorr and 300 W, LAM 2 CF4 90 + CHF3 30 + He 120 sccm at 2.8 torr and 850 W with a 0.38 cm gap, LAM 3 Al BCl3 50 + N2 50 + Cl2 30 + CHCl3 20 sccm at 250 mTorr and 250 W, LAM 3 W SF6 20 sccm at 100 mTorr and 125 W (fetch-tool summary).

Ques: How does oxygen change fluorine-based silicon etching?

It raises the rate up to a point, then stops it. MIT explains that CF${}_4$ supplies F atoms that etch silicon as $4F+Si\to Si{F}_4$. Added O${}_2$ combines with CF${}_3$ and CF${}_2$ fragments and reduces their recombination with F, so more F is available, but too much O${}_2$ oxidizes the silicon. KTH lists CF${}_4$/O${}_2$ as a standard chemistry. For gates, KTH describes HBr with O${}_2$: silicon leaves as volatile SiBr, and SiOBr passivates the sidewalls. Raising O${}_2$ from 2 to 5 per cent cut the SiO${}_2$ rate from 3.7 $nm/min$ to about zero, allowing 26 $\mathrm{nm}$ gates to stop on 3 $\mathrm{nm}$ of oxide. At 3.7 $nm/min$ that oxide would vanish in under a minute of over-etch (computed). Reference: MIT 6.152J Micro/Nano Processing Technology, lecture 17 ‘Etching’ (2005): wet chemical, isotropic and selective against dry physical, anisotropic and not selective by itself, wet etch bias and HF attack of SiO2 driving the move to dry etching, parallel-plate plasma mode with a larger wafer electrode giving weaker, more uniform bombardment and higher pressure reducing the physical part, RIE with a smaller etch electrode giving a greater voltage drop, more anisotropy, more physical etch and less selectivity, RIE at 10 to 100 mTorr with lower pressure lengthening the mean free path, high-density plasmas at 1 to 10 mTorr, CF4 giving F and 4F + Si -> SiF4, O2 combining with CF3 and CF2 to reduce F recombination but oxidizing Si in excess, oxide etching 4F + SiO2 -> SiF4 + O2 too isotropic and unselective unless F is reduced and C increased, three ion-enhancement mechanisms (bond breaking, product formation, product sputtering), sticking about 1 for physical and 0.01 for chemical etching, by-product boiling points SiCl4 57.6 C, SiF4 -86 C, WF6 17.5 C, AlCl3 177.8 C (sublimes), macro-loading and micro-loading (fetch-tool summary); KTH Royal Institute of Technology, Lecture 7 ‘Etching’: about 1e15 cm-3 neutrals with 1 to 10 per cent radicals and 1e8 to 1e12 cm-3 ions and electrons, sheath 10 to 100 eV in plasma mode and 100 to 700 eV in RIE mode, RIE at 10 to 100 mTorr, high-density plasmas 1e11 to 1e12 cm-3 at 1 to 10 mTorr, HBr + O2 for poly gates with SiBr volatile and SiOBr sidewall passivation, raising O2 from 2 to 5 per cent dropping SiO2 rate from 3.7 nm/min to about 0 for 26 nm gates stopping on 3 nm oxide, ion-enhanced etching anisotropic and selective, directionality from more ions or more sidewall inhibitor, smaller electrode giving larger voltage to the plasma (fetch-tool summary).

Ques: Why is aluminium etched with chlorine rather than fluorine?

Because the product must be volatile. MIT states that etch by-products should have low boiling points and lists SiF${}_4$ at -86 ${}^{\circ }\mathrm{C}$, SiCl${}_4$ at 57.6 ${}^{\circ }\mathrm{C}$, WF${}_6$ at 17.5 ${}^{\circ }\mathrm{C}$ and AlCl${}_3$ subliming at 177.8 ${}^{\circ }\mathrm{C}$. Aluminium fluoride does not appear among the volatile products, so in my reading fluorine would leave a non-volatile layer on aluminium. Berkeley’s LAM 3 aluminium recipe accordingly uses BCl${}_3$ 50, N${}_2$ 50, Cl${}_2$ 30 and CHCl${}_3$ 20 $\mathrm{sccm}$ at 250 $\mathrm{mTorr}$ and 250 $\mathrm{W}$. Tungsten, whose WF${}_6$ is volatile, is etched in SF${}_6$ at 100 $\mathrm{mTorr}$ and 125 $\mathrm{W}$. Berkeley also follows Al etching with a CF${}_4$/O${}_2$ step in the airlock (my reading: to remove residual chlorine). Reference: MIT 6.152J Micro/Nano Processing Technology, lecture 17 ‘Etching’ (2005): wet chemical, isotropic and selective against dry physical, anisotropic and not selective by itself, wet etch bias and HF attack of SiO2 driving the move to dry etching, parallel-plate plasma mode with a larger wafer electrode giving weaker, more uniform bombardment and higher pressure reducing the physical part, RIE with a smaller etch electrode giving a greater voltage drop, more anisotropy, more physical etch and less selectivity, RIE at 10 to 100 mTorr with lower pressure lengthening the mean free path, high-density plasmas at 1 to 10 mTorr, CF4 giving F and 4F + Si -> SiF4, O2 combining with CF3 and CF2 to reduce F recombination but oxidizing Si in excess, oxide etching 4F + SiO2 -> SiF4 + O2 too isotropic and unselective unless F is reduced and C increased, three ion-enhancement mechanisms (bond breaking, product formation, product sputtering), sticking about 1 for physical and 0.01 for chemical etching, by-product boiling points SiCl4 57.6 C, SiF4 -86 C, WF6 17.5 C, AlCl3 177.8 C (sublimes), macro-loading and micro-loading (fetch-tool summary); Marvell NanoLab (UC Berkeley) lab manual chapter 1.9 ‘VLSI etchants’: Technics-c SF6 12.9 sccm + He 21.0 sccm at 100 W removing poly isotropically at about the nitride rate, Technics-c CF4 10 + CHF3 5 + He 10 sccm at 200 W, LAM 1 CCl4 130 + O2 15 + He 130 sccm at 280 mTorr and 300 W, LAM 2 CF4 90 + CHF3 30 + He 120 sccm at 2.8 torr and 850 W with a 0.38 cm gap, LAM 3 Al BCl3 50 + N2 50 + Cl2 30 + CHCl3 20 sccm at 250 mTorr and 250 W, LAM 3 W SF6 20 sccm at 100 mTorr and 125 W (fetch-tool summary).

Ques: What is loading, and how is it handled?

It is the dependence of etch rate on how much material is exposed. MIT distinguishes macro-loading, where too many wafers consume the etchant, from micro-loading, where large unmasked areas consume it locally. Purdue shows local loading in CCl${}_2$F${}_2$/O${}_2$ RIE. Because the radical supply is shared, the rate falls as exposed area rises, so a fixed etch time over-etches lightly loaded wafers and under-etches heavily loaded ones (my reading). Operating at 10 to 100 $\mathrm{mTorr}$ (MIT; KTH) does not remove this. The usual answer is endpoint detection from the plasma emission, as on the Optical Emission Spectroscopy Endpoint (OES) page, followed by a set over-etch (my reading). Reference: MIT 6.152J Micro/Nano Processing Technology, lecture 17 ‘Etching’ (2005): wet chemical, isotropic and selective against dry physical, anisotropic and not selective by itself, wet etch bias and HF attack of SiO2 driving the move to dry etching, parallel-plate plasma mode with a larger wafer electrode giving weaker, more uniform bombardment and higher pressure reducing the physical part, RIE with a smaller etch electrode giving a greater voltage drop, more anisotropy, more physical etch and less selectivity, RIE at 10 to 100 mTorr with lower pressure lengthening the mean free path, high-density plasmas at 1 to 10 mTorr, CF4 giving F and 4F + Si -> SiF4, O2 combining with CF3 and CF2 to reduce F recombination but oxidizing Si in excess, oxide etching 4F + SiO2 -> SiF4 + O2 too isotropic and unselective unless F is reduced and C increased, three ion-enhancement mechanisms (bond breaking, product formation, product sputtering), sticking about 1 for physical and 0.01 for chemical etching, by-product boiling points SiCl4 57.6 C, SiF4 -86 C, WF6 17.5 C, AlCl3 177.8 C (sublimes), macro-loading and micro-loading (fetch-tool summary); Purdue ECE 695Q, lectures 46 and 47 (nanoHUB, April 2016): electrode voltage-area relation V1/V2 = (A2/A1)^m with m = 4 in simple theory but usually 1 to 2, the larger electrode usually the chamber wall, hydrogen tying up F and promoting Teflon-like polymer, CHF3 etching SiO2 fast and Si slowly, C4F8 for polymerization in deep RIE, high-density plasma at 1 to 10 mTorr with bias below 1 kV and ion energy set independently of density, local loading in CCl2F2/O2 RIE (fetch-tool summary); KTH Royal Institute of Technology, Lecture 7 ‘Etching’: about 1e15 cm-3 neutrals with 1 to 10 per cent radicals and 1e8 to 1e12 cm-3 ions and electrons, sheath 10 to 100 eV in plasma mode and 100 to 700 eV in RIE mode, RIE at 10 to 100 mTorr, high-density plasmas 1e11 to 1e12 cm-3 at 1 to 10 mTorr, HBr + O2 for poly gates with SiBr volatile and SiOBr sidewall passivation, raising O2 from 2 to 5 per cent dropping SiO2 rate from 3.7 nm/min to about 0 for 26 nm gates stopping on 3 nm oxide, ion-enhanced etching anisotropic and selective, directionality from more ions or more sidewall inhibitor, smaller electrode giving larger voltage to the plasma (fetch-tool summary); Optical Emission Spectroscopy Endpoint (OES) page of this vault.

References

1. MIT 6.152J Micro/Nano Processing Technology, “Lecture 17: etching,” MIT OpenCourseWare, 2005. Available: https://ocw.mit.edu/courses/6-152j-micro-nano-processing-technology-fall-2005/59b9f31d836355a3380aa156925508bc_lecture17.pdf
2. KTH Royal Institute of Technology, “Lecture 7: etching,” course slides. Available: https://www.kth.se/social/upload/510f795cf276544e1ddda13f/Lecture%207%20Etching.pdf
3. Purdue ECE 695Q, lectures 46 and 47, April 2016, nanoHUB. Available: https://nanohub.org/resources/25359/download/2016.04.20-ECE695Q-L46.pdf and https://nanohub.org/resources/25351/download/2016.04.22-ECE695Q-L47.pdf
4. NPTEL course 104103019, Module 4, Lecture 25, “Electrode-size effects.” Available: https://archive.nptel.ac.in/content/storage2/courses/104103019/module4/lec25/16.html
5. Marvell NanoLab, UC Berkeley, “Chapter 1.9: VLSI etchants,” lab manual. Available: https://nanolab.berkeley.edu/labmanual/chap1/1.09vslietch.pdf
6. DC and RF Magnetron Sputter System page of this vault.
7. Optical Emission Spectroscopy Endpoint (OES) page of this vault.

Related pages in this vault: Plasma Asher (Barrel and Downstream), Wet Etch Bench, Optical Emission Spectroscopy Endpoint (OES), Langmuir Probe, DC and RF Magnetron Sputter System, Inductively Coupled Plasma Etcher (ICP-RIE), Ion Beam Milling System, Atomic Layer Etch (ALE) Tool.