anisotropic sputter etching of ZnSe/GaAs in MRC tool

Kristiaan De Greve kdegreve at
Thu Apr 5 23:27:43 PDT 2007

Dear Madam or Sir,

after having discussed with Jim McVittie about the possibility of doing
anisotropic etching/sputter etching-ion milling at Stanford, I would like
to file a request to use the MRC Argon plasma.

The basic thing I am looking for is anisotropic etching of ZnMgSe (up to 18%
Mg) MBE grown epilayers on GaAs substrates, in order to make freestanding
mesas and microdisks. Mesas and microdisks (resp. sizes: 50-200 nm and ~1
um) are lithographically defined by e-beam writing on negative e-beam
resist layer of ~ 200 nm (MAN 2403).

In the official SpecMat form:

1) name: Kristiaan De Greve, kdegreve in coral, 650 283 0803 (cell),
kdegreve at, EE PhD student in Yoshi Yamamoto's group.

2) * material: zincselenide(ZnSe)/zincmagnesiumselenide(ZnMgSe, ~18% Mg)
heterostrocture, epilayer (~80nm) on GaAs substrate, small pieces of 3x4mm;
covered with developed MAN2403 negative e-beam resist
   * main hazards: - GaAs: toxic when inhaled/swallowed, possibly
carcinogenic, for the MSDS
                   - ZnSe: toxic when inhaled, swallowed, for the MSDS

   * storage group: it's not really a chemical, I don't see any category
under which it would fall

3) 'supplyer': MBE grown at the university of Paderborn, Germany, by Dr.
Alex Pawlis in the group of prof. Klaus Lischka (apawlis at

4) reason for request: in order to isolate small ZnSe mesa structures, we
need a dry etching (preferrably) or sputter etching process that can
anisotropically transfer patterns written into negative e-beamresist into
the ZnSe/GaAs; the mesa height is about 100 nm, i.e. into the GaAs

After talking with Jim McVittie, the option of sputter etching was withheld
(the alternative, dry etching in the PQuest - a tool initially purchased by
the Yamamoto group! - meeting resistance from GaAs III-V laser processors
who fear contamination of Zn into their processes). As the Ar plasma is
essentially non reactive, this tool has been used in the past for ion
milling/sputter etching of all kinds of materials, including Zn compounds.

No other tools are available for anitropic etching of these ZnSe epilayers;
wet etching loses quite some resolution, and the MRC plasma, although
second to the PQuest, would be of tremendous help with our experiments.

5) process flow: - spin coating of MAN 2403 resist, 3000 rpm 30s, curing
                   @90deg C, 1' @Ginzton cleanroom (3x4 mm pieces)
                 - e-beam writing in Raith 150, 360 uC dose, 10 kV
                   acceleration voltage
                 - development of MAN resist in MaD532, 30s, rinse with DI
                 - sputter etching in MRC, negative e-beam resist as mask
                 - resist stripping in MAN resist stripper (remover 1165)
                   in Ginzton cleanroom

6) amount and form: 3x4 wafer pieces, 3-4 pieces in a run

7) storage: in sample holders and subsequently vacuum chambers in Ginzton

8) disposal: these samples are actively measured and remeasured (optical,
cryogenic experiments); left-over pieces are disposed of as hazardous waste
in Ginzton labs.

Many thanks and best regards,

Kristiaan De Greve.

More information about the specmat mailing list