Skip to content. | Skip to navigation

Sections
Personal tools
You are here: Home / Process / Community Service Projects / Cleanliness & Contamination / TXRF analysis of gold-contaminated wafers

TXRF analysis of gold-contaminated wafers

A recent survey of the SNF community indicates that there is some interest in moving one of our DRIE tools, stsetch, into the Contaminated material group. Before doing so, we would like to quantify how much gold would be exposed to the plasma if a contaminated wafer were etched. This page describes the purpose, procedure and results of this experiment.

Purpose

In a recent survey of the SNF community, 56% of respondents stated they are willing to move one of our DRIE tools, stsetch, into the Contaminated material group. However, this decision has widespread consequences, particularly for those that use stsetch for clean processing. Thus, more supporting data is required. As a starting point, Mary Tang has suggested an experiment, described below, which aims to quantify the amount of gold exposed to the plasma when contaminated wafers are etched. The experiment relies on a technique known as Total Reflection X-ray Fluorescence (TXRF), which is the industry standard for analysis of metallic surface contamination on semiconductor wafers.

 

Procedure

 

Samples

Two samples will be prepared, both consisting of a Si wafer and 2000 A of evaporated gold. The gold will be patterned by lift-off for the first sample and etch back for the second. The starting material for both samples will be prime-grade Si wafers (N/Phos, 1-10 ohm-cm) cleaned at wbnonmetal. The mask(s) for patterning is to be determined. Ideally, two masks of opposite polarity are used, one for lift-off and the other for etch back, so the resulting patterns are identical.

 

Lift-off (adapted from Vikram Mukundan's process)

  1. HMDS prime (yes).
  2. Spin LOL2000 at 3000 rpm for 60 s (headway).
  3. Bake for 5 min at 150 C (litho hotplate).
  4. Spin 1.6 um Shipley 3612 (svgcoat).
  5. Expose 1.3 s (karlsuss2).
  6. Develop (svgdev).
  7. Etch native oxide, 6:1 BOE for 30 s (wbnonmetal).
  8. Evaporate 2000 A of Au (innotec).
  9. Strip in Microposit Remover 1165 (wbsolvent).
  10. Rinse with IPA and dry (wbsolvent).

 

Etch Back (adapted from Nahid Harjee's process)

  1. Etch native oxide, 6:1 BOE for 30 s (wbnonmetal).
  2. Evaporate 2000 A of Au (innotec).
  3. HMDS prime (yes).
  4. Spin 1.6 um Shipley 3612 (svgcoat).
  5. Expose 1.3 s (karlsuss2).
  6. Develop (svgdev).
  7. Etch Au with wet Au etchant (wbgeneral).
  8. Strip resist with acetone (wbsolvent).
  9. Rinse with IPA and dry (wbsolvent).

 

Analysis

TXRF will be performed on bare Si at 5 locations on each wafer (top, bottom, left, right, center). The analysis yields surface concentration of gold (atoms/cm2).

 

Results

To be reported.

 

Document Actions