Specific Process Knowledge/Etch/Wet III-V Etches

From LabAdviser

Unless anything else is stated, everything on this page, text and pictures are made by DTU Nanolab.

All links to Kemibrug (SDS) and Labmanager Including APV and QC requires login.

Feedback to this page: click here

This is a collection of III-V wet-etches; the rates should be tested before use. Make sure you have the proper education in mixing of chemicals before use. Wet etching of III-V materials is primarily to be done in 'Fume hood 07: III-V acids/bases'. Contact information, risk assessment and manual regarding this fume hood can be found in LabManager:

Fume hood 07 Info page in LabManager
It is very important that you dispose off chemicals and process waste according to DTU Nanolab regulations and risk assessments for your work.

'Fume hood 07' in cleanroom D3


For dry etching III-V materials see


InP or GaAs substrate etching

When etching InP or GaAs substrates, both 2” wafers and small pieces, you should collect the used chemicals in a waste bottle. This does not apply for etching epilayers.

There should be two bottles in the fumehood one for HCl used to etch InP substrates and one for citric acid: H2O2 or H2SO4 (10%):H2O2 used to etch GaAs substrates. Once the bottles are full, you should bring them in the basement of building 346 and have a new one in CR. For the GaAs waste you should use empty H2O2 bottles (since they should have a special lid that avoid overpressure), while for InP waste we can use an empty and clean developer bottle.

HCl:H3PO4 etch

HCl(37%):H3PO4(85%) is a selective, anisotropic and slow etching of InP. Very slow rate in InGaAsP. The etch rates depend on the orientation of the features and have not yet been calibrated at DTU Nanolab.

See rates below for InP, PQ(1.n) and InGaAs lattice-matched to InP. The acronym PQ(1.n) denotes a lattice-matched InGaAsP on InP with a band gap corresponding to a wavelength of 1.n um.

The temperature is 22 degC +/- 1 degC.


H2SO4:H2O2:H2O etch

H2SO4(10%):H2O2(30%):H2O is a selective etch of InGaAsP with very low etch rate in InP. The acronym PQ(1.n) denotes a lattice-matched InGaAsP on InP with a band gap corresponding to a wavelength of 1.n um.

The etch rates have not yet been calibrated at DTU Nanolab. The temperature is 22 degC +/- 1 degC.


Concentrated H2SO4

Concentrated H2SO4(98%) is used for deoxidation of InP with a very low etch rate in InGaAsP.

H2SO4(98%) etch rates, nm/min
Etchant InP PQ(1.1) PQ(1.3) PQ(1.5) InGaAs
H2SO4(98%) 13 ? <1    


H3PO4:H2O2:H2O etch

H3PO4(85%):H2O2(30%):H2O is a GaAs/AlGaAs-etch which gives a better surface quality than H2SO4-based etches.



BHF, HF etch

BHF etches SiO2 and partially removes native oxide on InGaAs and InP. Do not use BHF unless you know the dangers involved with this chemical and always use 4H gloves!.

BHF (12.5%), HF etch rates, nm/min
Etchant SiO2 PECVD2 Si3N4 E-beam Ti 1 Al(x)GaAs, x>0.5, AlAs
BHF (12.5%) 283 2, 175+/-25 3 88 4 90-120  
HF:H2O       >10000 5

(1) Ti from Titest.prg on Physimeca. It seems there is no measurable etching during first 10 seconds.
(2) Process SiO2ky2 in PECVD2 (2014-July Luisa Ottaviamo @photonics ).
(3) Process STANDARD in PECVD2 (2014-July Luisa Ottaviano @ photonics).
(4) Process SINSTD in PECVD2 (2014-July Luisa Ottaviano @ photonics).
(5) Appl. Phys. Lett. vol. 51, 2222 (1987).

Note: neither Physimeca or PECVD2 is available anymore. Use the Temescal and the PECVD3/PECVD4 instead.

Citric Acid etch

C6H8O7:H2O2 is a selective etch of GaAs; does not etch AlxGa{1-x}As if x > 0.45.

Solid C6H8O7 is mixed 1:1 by weight with H2O (30%) using magnetic stirring. The solution C6H8O7:H2O is thereafter mixed 4:1 volume ratio with H2O2.

The above C6H8O7:H2O2 solution has an etch-rate of ~360 nm/min in GaAs and < 0.5 nm/min in AlAs @ 25 degC.