Microtechnology/Etching Processes

Etchants

Wet Etchants

Etchants

Etchant	Formula	CAS no	Etches	Possible damage
Acetic Acid	CH3COOH(l)	?	GaAs; Pb; Ti	?
Hydrochloric Acid	HCl(38%, aq)	?	Al; Cr; Cu; Fe2O3; Ga; GaN; In; Fe; Pb; Ni; NiO, Ni2O3; Sn;SnO2; Ti; Zn	GaAs;
Hydrofluoric Acid	HF(49%, aq)	?	GaAs; Ni; SiO2; Ti; Al2O3	PMMA
Nitric Acid	HNO3(70% aq)	?	C; Cu; GaAs; In; Fe; Pb; Ni; Ag; Pd; Pt; Sn; Ti; Zn; ZnO	?
Phosphoric Acid	H3PO4(85% aq)	?	Al; Cu; GaAs; GaN; Fe; Ni; SiN; ZnO	?
Potassium Hydroxide	KOH(s/aq)	?	Al; C; Cu; Ag; GaAs; Si; Ti	?
Sodium Hydroxide	NaOH(s/aq)	?	Al; Cu; Ag; Ti; Si; GaAs; GaN	?
Sulfuric Acid	(96%, aq)	?	C; Cu; GaAs; Fe; Pb; Ni; Ti	?

Resources

Wikipedia

• Wet etching
• Chemical etching

Wet Etch Overview Table

Wet Etches Overview Table

Material	Etchant mixture and ratios	Etch rate	Also etches	Doesn't etch	Notes	Link	Refs
Aluminium (Al)	H3PO4(85%, aq)@120°C	?	SiN	?	Must be heated	Metal etches
Aluminium (Al)	19 H3PO4(85%, aq):1 acetic acid (l, H3COOH):1 HNO3(70%, aq):2 H2O|}	40Å/s=240nm/min	SiN, M	SiO2, Si, PR	Probably at room temp comparing to similar etch below.	?
Aluminium (Al)	16 H3PO4(85%, aq) : 1 acetic acid (l, H3COOH) : 1 HNO3(70%, aq) : 2 H2O	200nm/min@25°C, 600nm/min@40°C	SiN, M	SiO2, Si, PR	?	?
Aluminium (Al)	1 NaOH (s): 1H2O	?	?	?	can be used at 25°C, heated makes it faster.	?
Aluminium Nitride (AlN)	H3PO4(85%, aq)@85°C	30nm/min	Ni, Ti	Mo	?	?	?
Aluminium Oxide (Al2O3)	1 NH4OH(30%, aq) : 1 H2O2(30%, aq) : 3 H2O @80°C	~30-40 nm/min @80oC	Al, Poly, PR, Mo	SiO2, SiN, Si, M	?	?
Aluminium Oxide (Al2O3) by ALD	H3PO4 (85%)	~40-50 nm/min @ 70oC	InGaZnO4 (IGZO)	Au, Mo, PR
Carbon (C)	H3PO4(85%, aq): CrO3(s) : NaCN(s)	?	SiN	SiO2, Si, PR	weight/vol ratios? and what's the chemistry in this reaction and the specific kinds of carbon being etched?	?
Chromium (Cr)	2 KMnO4(s) : 3 NaOH(s) : 12 H2O	?	Al	SiO2, SiN, Si, M, PR	?	?
Chromium (Cr)	3HCl(38%, aq) : 1Water	?	?	?	?	?
Chromium (Cr)	1HCl(38%, aq) : 1Glycerin	?	?	?	?	?
Copper (Cu)	30% FeCl3(s)	?	Ni	SiO2, SiN, Si, M, PR	?	Metal etches
Copper (Cu)	5 HNO3(70%, aq) : 1 H20	?	?	?	?	?
Copper (Cu)	Ammonium Persulphate	?	?	?	?	?
Gallium Arsenide (GaAs)	5%vol Br(l) in CH3OH(l)	anisotropic	Fe	SiO2, SiN, Si, M	Br reacts with methanol producing HBr -be careful!	?
Gallium Arsenide (GaAs)	1 NH4OH(30%, aq) : 1 H2O2(30%, aq)	anisotropic	Al, Ag, Poly	SiO2, SiN, Si, M	?	?
Gold (Au)	115g KI : 65g I : 100ml Water	?	Fe	SiO2, SiN, Si, M, PR	Alternative recipe 1 I2(s) : 2 KI(s): 10 H2O	Metal etches
Gold (Au)	KCN(s)	?	Ag, Cu	Al2O3, SiO2, SiN, Si, M, PR	?	Metal etches
Gold (Au)	Aqua Regia (3 HCl(38%, aq) : 1 HNO3(70%, aq))	?	Etches all metals	?	discard after use	Metal etches
Iron (Fe)	1 I2(s) : 2 KI(s): 10 H2O	?	Au	SiO2, SiN, Si, M, PR	?	Iron also etched by 1HCl:1H2O; 1HNO3:1H2O.
Lead (Pb)	Acetic Acid (l, H3COOH): H2O2(30% aq)	?	?	?	dissolves solder connections. Etchant can also be used diluted eg. 5 times in water.	?
Molybdenum (Mo)	1 HCl(38%, aq):1 H2O2(30% aq)	100nm/min@18°C	?	PR	?	?
Molybdenum (Mo)	1 H2SO4(96%, aq) : 1 HNO3(70%, aq) : 1 Water	?	?	?	?	?
Nichrome (NiCr)	H2SO4(96%, aq)@100°C	?	?	?	?	?
Nichrome (NiCr)	1 HCl(38%, aq) : 1 HNO3(70%, aq) : 3 Water	?	?	?	?	?
Nickel (Ni)	30% FeCl3(s) in H2O	?	Cu	SiO2, SiN, Si, M, PR	?	Metal etches
Nickel (Ni)	5 HCl(38%, aq) : 1 HNO3(70%, aq)	?	?	?	?	?
Nickel (Ni)	5 HF(49%, aq) : 1 HNO3(70%, aq)	?	?	?	?	?
Palladium (Pd)	Aqua Regia (3 HCl(38%, aq) : 1 HNO3(70%, aq))	?	Etches all metals	?	discard after use	Metal etches
Platinum (Pt)	8 HCl(38%, aq) : 1 HNO3(70%, aq)	use @70°C	Etches all metals	?	Age for one hour before use!, Discard after use	Metal etches
Platinum (Pt)	3 HCl(38%, aq) : 1 HNO3(70%, aq): 4 H2O	use @95°C	Etches all metals	?	Discard after use	Metal etches
Polymers	5 NH4OH(30%, aq) : 1 H2O2(30%, aq) @ 120°C	Al	SiO2, SiN, Si, M	Polymers such as wax, photoresist, epoxy...	?	?
Polysilicon (PolySi)	50 HNO3(70%, aq): 20 H2O (l): 1 HF (49% aq)	540 nm/min @ 25°C	?	?	Remove oxide first by a HF based etch.	?
Polysilicon (PolySi)	3 HNO3(70%, aq): 1 HF (49% aq)	4.2 micron/min	?	?	Remove oxide first by a HF based etch.	?
Residual resist and polymers residues	Piranha (acidic or basic)	Al	SiO2, SiN, Si	Polymers such as wax, photoresist, epoxy...	Piranha forms explosives if mixed with volatile organic compounds. It tends to form 'burnt' carbon residues if too thick residues are present on the sample.	Piranha
Silica (SiO2) Thermally grown	Buffered HF 6vol NH4F: 1vol HF	120 nm/min @ 25°C	?	?	?	Silicon oxide etch
Silica (SiO2)	1 HF(49%, aq) : 5 NH4F(40%, aq) : 5 H2O (BOE)	20 Å/s = 120nm/min @ 25°C	M	SiN, Si	?	Silicon oxide etch
Silica (SiO2) Thermally grown	1 HF(49%, aq) : 10 Water	20-30 nm/min @ 25°C	PMMA;poly n+ Si ~1nm/min; Stoic. SiN 1nm/min; Low stress SiN ~3Å/min; Ti 1µm/min; Al ~.2-1µm/min;	Si; undoped poly Si;	?	Silicon oxide etch
Silica (SiO2) Thermally grown	1 HF(49%, aq) : 100 Water	1.8 nm/min @ 25°C	?	?	?	Silicon oxide etch
Silica (SiO2) Thermally grown (wet oxide)	49% HF	1.8-2.3 microns/min @ 25°C	Stoic. SiN 14nm/min; Low stress SiN ~50nm/min;W <5nm/min; Ti >1µm/min; Al ~4nm/min;	Si,PolySi	?	Silicon oxide etch
Silica (SiO2) CVD	1 HF(49%, aq) : 10 Water	?	?	?	?	Silicon oxide etch
Silicon	64 HNO3(70%, aq) : 3 NH4F(40%, aq) : 33 H2O	100 Å/s	M	SiN, PR	Isotropic etch	?
Silicon (Si)	2 HF(49%, aq): 2 HNO3(70%, aq) : 1 Water	?	?	?	?	Silicon oxide etch
Silicon (Si)	3 HF(49%, aq): 5 HNO3(70%, aq) : 3 acetic acid (l, H3COOH)	?	?	?	?	Silicon oxide etch
Silicon (Si)	NaOH in water	?	?	?	Use almost saturated solution near boiling point	Silicon oxide etch
Silicon nitride (Si3N4)	H3PO4(85%, aq)	6.5 nm/min @ 25°C	Al;	?	use reflux at 180°C. Dry etching is often better.	Silicon nitride etch
Silver (Ag)	1 NH4OH(30%, aq) : 1 H2O2(30%, aq)	quick etch	Al, Poly	SiO2, SiN, Si, M	?	?
Silver (Ag)	1 HNO3(70%, aq) : 1 Water	?	?	?	?	?
Tantalum (Ta)	2 HF(49%, aq): 2 HNO3(70%, aq) : 5 H2O	?	?	?	?	?
Tin (Sn)	2 HClO4(85%, aq) : 7 acetic acid (l, H3COOH)	?	Pb, Ti	SiO2, SiN, Si, PR	?	?
Tin (sn)	1 HF(49%, aq) : 1 HNO3(70%, aq)	?	?	?	?	?
Tin (sn)	H2SO4(96%, aq)	?	?	?	use at 80°C	?
Titanum (Ti)	1 HF(49% aq) : 30 H2SO4 (96% aq): 69 Water	?	?	?	use at 70°C	?
Titanium-Tungsten (TiW)	H2O2(30%, aq)	5 nm/min @ 25°C	?	?	?	Metal etches
Tungsten (W)	1 HF(49%, aq) : 2 HNO3(70%, aq)	?	?	?	?	Metal etches
Vanadium (Va)	1 HF(49%, aq) : 1 HNO3(70%, aq) : 1 water	?	?	?	?	?

References:

• table of etchants on http://grover.mirc.gatech.edu/processing/Etchants.pdf
• table of etchants on http://www.siliconfareast.com/etch_recipes.htm
• See also the two papers 'Etch Rates for Micromachining Processing' Part I and II in JOURNAL OF MICROELECTROMECHANICAL SYSTEMS by Kirt R. Wiliams et al.
• Extensive table with etch rates on http://www.eng.utah.edu/~gale/mems/etch%20rates.pdf

Wet Etch Compatibility Chart

Wet Etch Compatibility Chart (M metals; PR Cured photoresist; Poly Polymers)

Etchant	49% HF	BOE	KOH	H3PO4	KI gold etch
Si undoped	Yes	Yes	No	Yes
Si doped
Poly Si
Therm SiO2
PEVCD SiO2
SiN
Au
Al
Ni
PMMA

Capillary Effects

The surface tension of a wet etch solution can often make problems with movable parts in microchips and droplet formations that leave residues.

Avoiding capillary effects

• Use dry/gas etching instead of wet (eg. HF vapor instead of HF solutions)
• Use critical point drying
• Quickly take from wet etchant into a water rinse bath for a thorough rinse and then quickly into an ethanol bath and then dry - the low surface tension of ethanol reduces capillary effects when drying.
• use silicon structures with an oxide sacrificial layer - the very hydrophobic hydrogen terminated silicon after a HF wet etch will avoid capillary effects.

Silicon KOH Etch

Potassium hydroxide (KOH) is an anisotropic wet etch that preferentially etches the 100 planes of Si and almost doesn't attack the 111 planes. This leads to a V shaped pyramidal holes in Si 100 from square openings in the etch mask, with side edges at a 54.7deg angle from the surface. The etch rate does not depend on As, P, Sb dopants, but too high B doping will reduce the etchrate in the 110 direction. The etchrate in the 100 and 110 direction can be varied by adding isopropanol to the solution.

Overall reaction: Si + 2OH^- + 4H₂O -> Si(OH)₂⁺⁺ + 2H₂ + 4OH^-

KOH% determination: KOH (%) = KOH dry mass(g) / solvents(ml)

Recipe for a typical 30% KOH/Isopropanol etch solution:

• 70g KOH pellets dissolved in 190mL DI water (use heat and/or ultrasound to dissolve quickly)
• Add 40mL Isopropanol
• The etch rate for should be about 1 micron/minute at 80°C

KOH etch masks can be made from silicon nitride or silicon oxide (though SiO₂ is slowly etched by KOH)

See also

• http://www.virginiasemi.com/pdf/siliconetchingandcleaning.pdf
• 'Review: The effect of alcohol additives on etching characteristics in KOH solutions' Sensors and Actuators A 101 (2002) 255–261 by Irena Zubel

Silicon Oxide Etch (HF, BHF, BOE)

NB: HF is very dangerous, it diffuses quicker than anything you can add to try and reduce the damage, so the key is NOT to get in touch with it in the first place. It attacks the calcium in your bones, the nerves, and the blood vessels - and importantly, it does not burn like other acids!! Wear lab coats, eye protection and 6h gloves, and work in a closed fume hood.

Fundamental reaction SiO₂ + 6HF -> H₂SiF₆+ 2H₂O

Etchants:

• HF or Hydrofluoric acid is a highly corrosive and toxic solution of hydrogen fluoride in water.
• Buffered Hydrofluoric Etch (BHF) or Buffered Oxide Etch (BOE) is a mixture of ammonium fluoride and hydrofluoric acid with a more controlled etch rate of silicon oxide.
• ammonium fluoride containing etches give silicon surfaces with an atomically smoother surface than HF, ammonium fluoride solutions can also be used to make atomically flat surfaces. See Appl. Phys. Lett. vol 56 p. 656 1990 by Higashi.

An oxide etch is often used to remove the impurity containing native oxide layer of wafers before contamination sensitive processes. BOE has a more controllable oxide etch rate than HF (the pH is stabilized by the buffer) but also etches Si slowly and the higher pH in BOE can cause metal precipitation, so for clean processes or thin underlying Si layers a HF etch is preferable.

• 49% HF is used for fast removal of oxide
• BOE gives a slower removal of oxide, but can extend the lifetime of a photoresist mask. Etch rate typically 1000-2500 Å/min.
• Diluted HF etches - say 5% HF - is used for removal of native oxide in about 30 seconds. The surface becomes highly hydrophobic.
• HF/HCl or HF/Glycerin mixtures can be used to make less rough surfaces when thinning oxide layers
• HF mixed with isopropanol can be used to increase the wetting properties of the solution to better etch into narrow pores.

Etch rates vary depending on on oxide quality (eg. wether its wet furnace grown or PECVD)

Recipe for Buffered Oxide Etch or Buffered Hydrofluoric Etch

• Prepare the 40% NH₄F solution, eg 40g NH₄F in 60mL water.
• 6 parts 40% NH₄F and 1 part 49% HF - HF etches glass so use plastic beakers! Add HF into NH₄F instead of NH₄F into HF.

Recipe for BHF/HCl etch for smooth oxide

• add 5mL Buffered Oxide etch as above to 85mL water
• Add 10mL conc. HCl

Etch rate about 1 micron/min at room temperature.

Silicon Nitride Etch

Typical etch rates

• Standard etch in H₃PO₄ 100Å/min at 180°C, 55Å/min at 165°C. Use a short BHF dip first to remove oxynitride layer.
• 10% HF 5000 Å/min
• 1% HF 600 Å/min
• BHF (7:1) 5-20 Å/min

Metal Etches

• table of etchants on http://grover.mirc.gatech.edu/processing/Etchants.pdf
• See also the two papers 'Etch Rates for Micromachining Processing' Part I and II in JOURNAL OF MICROELECTROMECHANICAL SYSTEMS by Kirt R. Wiliams et al; Volume 5 Number 4 December 1996 and Volume 12 Number 6, December 2003. DOI: 10.1109/84.546406 & DOI: 10.1109/JMEMS.2003.820936
• Table with etch rates from Kirt R. Wiliams et al. http://www.eng.utah.edu/~gale/mems/etch%20rates.pdf

Cleaning Methods

There are several standard cleaning procedures. Some use a wealth of dangerous and highly corrosive chemicals. Don't underestimate the power of keeping things clean, and also simple soap rinses before starting on the more dangerous processes will probably increase the quality of the result.

Rinsing

In cleanrooms you often see people rinsing by submerging a wafer into a bath with running water. This has the advantage that the wafers do not dry out and eventually whatever should be washed away will be removed.

But think of how you can get tea leaves out of a tea pot: If you put the pot under running water, it will probably never ever really get completely free from leaves, whereas if you pour as much out as possible, add a little water, empty again and repeat a couple of times, your pot is completely clean with almost no use of water. The math is simple. If you add 1/10 of water pr minute in a large bath you dilution will take place very slowly compared to adding a small amount of water that maybe is gives a 50/50 dilution in one go and thereby rapidly decrease the fraction of original contaminant considerably.

Flushing your wafers with a jet of water probably orders of magnitude more efficient and faster at rinsing than submerging in a bath...

Ultrasound

Ultrasound baths work by standing waves of high frequency sound that at the wave anti-nodes create so high pressure variations that water vapor bubbles form and implode during the sounds pressure cycle. The bubble implosion creates shock waves that knock any loose material off surfaces and also can initiate chemical reactions or cause pitting in soft materials and damage smaller MEMS structures.

Soap

Don't underestimate the power of ordinary soap or more harsh treatments with surfactants such as Triton-X. Especially together with ultrasound.

RCA

A standard wafer cleaning method developed by the RCA corp. It is made from 2 baths:

• RCA1 is a H₂O:NH₄OH:H₂O₂ cleaning of organic residues
• RCA2 is a H₂O:HCl:H₂O₂ etch of metal impurities.

RCA cleaning is often used before furnace processes.

See also

• http://www.virginiasemi.com/pdf/siliconetchingandcleaning.pdf

Piranha

Piranha solutions are often made from sulfuric acid (H₂SO₄) and hydrogen peroxide (H₂O₂). But basic solutions and other recipes are also named piranha. The solution is very corrosive. Use fume hood, lab coat, eye protection and nitrile/6H gloves. To effectively demonstrate to people how dangerous it is, try putting a small droplet on a piece of paper and it instantly is burning black with a hissing sound.

• Acid Piranha (Caro's acid, Sulfuric Peroxide) - for 10mL Piranha, pour 7mL 95% H₂SO₄ into 3mL 30% H₂O₂ -use glass beakers. The mixture heats on mixing to about 80°C

• Peroxydisulfuric etch (also nicknamed 7-up since it bubbles upon heating to 80°C).

This is a piranha etch with Ammoniumperoxydisulfate. Mix 1L 95% H₂SO₄ at 80°C with a table spoon of Ammoniumpersulfate. The solution will start to bubble indicating that it is ready. It can be reused several times (add another spoon of Ammoniumperoxydisulfate every time) until it does not bubble any longer and must be replaced.

• Base Piranha is a 3:1 mixture of ammonium hydroxide (NH₄OH) with hydrogen peroxide.

Piranha solutions etches organic compounds vigorously. It can form explosive compounds if mixed with organic solvents, so be careful and do not work near eg. acetone, ethanol or isopropanol with piranha.

The exothermic heat of mixing can bring solution temperatures up to 120°C and can lead to violent boiling, or even splashing of the extremely acidic solution. Explosions may occur if the peroxide solution concentration is more than 50%. A 30% peroxide in water solution is more reasonable.

Too thick organic contamination can harden up in piranha, so a degrease clean is often made before piranha cleaning.

For instance a sequential cleaning as follows is often used for substrates used for epitaxial growth:

• 2min sonication in Tri-chlor ethylene
• 2min sonication in acetone
• 2min sonication in ethanol
• 2min sonication in millipore water (MPW)
• then rinse in MPW 3 times.
• Piranha etch 6 min.
• Rinse in MPW 3 times.
• Dehydration bake 200-250°C for 30 min.

Dry Etching Overview

Dry etching on Wikipedia

Reactive Ion Etching (RIE, DRIE, ASE)

Reactive ion etching (RIE) Overview Table

Dry Etch Overview Table

Material	Etchant gasses and ratios	Etch rate	Also etches	Doesn't etch/selectivity	Notes	Link	Refs
Indium Phosphide (InP)	CH4 + H2 + Ar ("MHA") Cl2 + Ar + He/N2 ("Halide")			SiO2: ~10, Si3N4: ~3	For MHA, Need O2 polymer removal step For Chlorine Chemistry, ~200degC to equalize etch rate of InP/GaAs alloys
Silicon(Si)	Cl2			SiO2
Silicon Dioxide (SiO2)	CHF3	ICP-RIE: CHF3=40sccm, 0.5 Pa, ICP=900W, Bias=200W, Cooling Helium (backside) = 15degC, 15sccm/700Pa Rate: 220nm/min	SiN	InP, Si		http://www.cleanroom.byu.edu/trion_icp.phtml
	CF4	ICP-RIE: CF4=50sccm, Pressure: 12 mTorr, ICP: 600W, RIE Power: 75, Uniformity: ~3% variation, Etch Rate: 3600A/min
Silicon Nitride (Si3N4)	CF4			InP, Si
Tantala, Tantalum Pentoxide (Ta2o5)
Titania, Titanium Dioxide (TiO2)
Titanium (Ti)	Cl2/Ar			SiO2, Si3N4		Etch rate/quality highly dependent on amount of material being etched and temperature of etch. Proper/controlled heatsinking necessary.

Advanced Silicon Etch (ASE) / Deep Reactive Ion Etch (DRIE)

• DRIE on Wikipedia

Dry Cleaning Methods (Plasma, Ozone)

Plasma Ashing

Plasma Ashing on Wikipedia

Ozone Cleaning

Ozone cleaning on Wikipedia

Sputtering

Sputtering on Wikipedia

Laser Ablation

• Carbon dioxide laser cutting of sub-millimeter structures in plastics
• Eximer laser ablation of materials
• Solid state laser ablation of materials

Gas Etches

• Hydro fluoric gas etching of oxide to avoid collapse of silicon MEMS structures that would happen during drying if etched in aqueous fluoride etchants. See eg. http://www.imec.be/wwwinter/microsystems/SPIEpaper.pdf
• Xenon Difluoride (XeF₂) anisotropic (non-directional) dry-etching of Ge for similar reasons as above.
• Laser activated chlorine gas etch of nanoscale patterns

References

See also notes on editing this book about how to add references Microtechnology/About#How to Contribute.