HSP priority list
From OptoelectronicsWiki
Contents
Agenda & Priority List[edit]
Back to Process Hybrid Silicon.
Priority list[edit]
- Process development:
- Etch-back module (Jock, Jared)
- Need to converge on process follower (Excel?) -> start with Si etch / VC etch / bonding
- III/V epi design
- Finalize design
- Order epi
- III/V SOA design
- III/V SOA mask design (UCSB-E-Phi-dev-2 @ UCSB E-Phi Runs)
- Issue: Epi will arrive > 3 months -> test process follower with OPSIS run?
- SOI passives run fabrication
Issues brought up in meetings[edit]
04/20/2012[edit]
03/02/2012[edit]
- Sudha - Removal of organics(/polymer) after RIE#2 etch: Gasonics + 10sec HF etch + Gasonics.
- Bonding (Jon P.) - Could get some tips/demo from Aurrion:
- PR while cleaving
- Minimal handling
20/01/2012[edit]
- AR (Sudha): Process follower to Jon P. & Martijn (incl. wafer details: who, where, version. etc.)
- Epi order: 3 2" wafers per epi design, 2mm spacing between dies
- The Epi of the standard landmark LD design with GRIN-SCH and 6 pairs of 5nm/9nm well/barrier has a low refractive index --> go to 7 pairs of 6nm/9nm well/barrier and revised GRIN-SCH for a higher refractive index.
13/01/2012[edit]
- Sudha/Yongbo.: should we undercut to define the current channel?
- (Martijn) high risk
- REFERENCE EPI: centered QW, similar to past epi + top etch-stop layer + InGaAsP contact helper(?)
- HIGH RISK EPI: e-block layers, GRINSCH
- Jon P.: Resist difficult to remove after Si-doping implants
- can remove with long strip, pirhana (hot)
- long-term: move to thin SiO2 underlayer (implant through this) and remove with BHF
- Jared H.: edge bead (may not be an issue, esp. for 4" wafer processing - use stayout of 5-10mm)
- for small samples, can surround with additional chips during spin
06/01/2012[edit]
- Jon P.: Resist nonuniformity for multiple-die exposure -> >= 1-cm stay-out area?
- Jock: Grating module works for etch depths < 120 nm
- Bandgap issue: PL peak is not equal to lasing wavelength, and lasing wavelength is not equal to SOA gain maximum
- First have nice gain spectrum before we discuss further