How to Detect Borate Impurities in Silicon Wafer Processing

Overview of Technical Issues:

The detection device insufficiently measures borate impurity concentration in silicon wafers during processing, failing to provide adequate sensitivity or specificity to identify contamination introduced by processing chemicals; this leads to undetected borate-contaminated wafers proceeding through manufacturing, risking device performance degradation and yield loss. The goal is to establish a reliable detection method that can accurately identify and quantify borate impurities to enable effective contamination control in the wafer fabrication process.

Solution directions generated for this problem

Problem Direction 1 :

ImproveDetection signal intensity
VS
ConstraintDetection cycle time

Inspiration 1 : Cross-domain reference

Application Principle: #19 Periodic action
Cross-domain applicability Assess applicability
Fluorescence Imaging Flow Cytometry With Enhanced Image Resolution
Innovative Solution Refine solution

Time-gated pulsed laser excitation with synchronized detection for rapid borate trace analysis

Replace continuous illumination with pulsed high-intensity laser excitation
How to solve :
  • Deploy Q-switched UV laser (266 nm wavelength) with 10 ns pulse width at 10 kHz repetition rate, delivering peak power 50 kW to excite borate-specific fluorescence in short bursts
  • Synchronize time-gated photon counting detector with 2 ns gate window opening 5–15 ns after each laser pulse, capturing borate emission while rejecting silicon matrix background and scattered light
  • Accumulate signal over 30,000 pulses (3-second integration per measurement spot), scanning 25 sites across wafer surface to achieve full-wafer mapping in 4.5 minutes total cycle time
Expected Effect : Sensitivity 8×10⁹ atoms/cm³, SNR 12:1, cycle time 4.5 min
Risk Control :
  • laser pulse energy stability ±3% tolerance required
  • detector gate timing jitter <500 ps
  • borate fluorescence quantum yield variation with surface chemistry

Problem Direction 2 :

ImproveMeasurement sensitivity threshold
VS
ConstraintDetection cycle time

Inspiration 1 : Cross-domain reference

Application Principle: #32 Color changes
Cross-domain applicability Assess applicability
A nanobody specifically recognizing Staphylococcus aureus, its preparation method and application
Innovative Solution Refine solution

Wavelength-selective dual-channel borate detection with real-time spectral subtraction

Exploit borate-specific spectral signature for rapid trace detection
How to solve :
  • Deploy dual-wavelength optical system: primary channel at 208 nm (borate B-O bond absorption peak) and reference channel at 230 nm (silicon matrix baseline)
  • real-time differential measurement isolates borate signal from background in single scan
  • Implement high-throughput spectrometer with 2048-element linear CCD array capturing full spectral range simultaneously
  • integration time 180 seconds per wafer, parallel data acquisition eliminates sequential scanning delay
  • Apply adaptive baseline correction algorithm: reference channel signal auto-scales to match matrix composition variations, subtracts from primary channel in real-time achieving SNR ≥10:1 at 1×10¹⁰ atoms/cm³ without repeated calibrations
Expected Effect : Sensitivity 1×10¹⁰ atoms/cm³, cycle time 4.5 min, SNR 12:1
Risk Control :
  • UV optical component degradation over time
  • spectral drift requiring weekly wavelength calibration
  • CCD dark current noise at extended integration

Problem Direction 3 :

ImproveSignal discrimination capability
VS
ConstraintEquipment operational complexity

Inspiration 1 : Cross-domain reference

Application Principle: #25 Self-service
Cross-domain applicability Assess applicability
Systems and methods for enhanced detection and analyte quantification
Innovative Solution Refine solution

Embedded reference-chip auto-calibration for inline borate detection

Embed reference chip for automated calibration
How to solve :
  • Integrate a borate-doped silicon reference chip (concentration 5×10¹⁰ atoms/cm³, ±5% tolerance) permanently mounted in the detection chamber
  • system auto-measures reference before each wafer scan, establishing real-time baseline without operator calibration
  • Implement dual-beam optical architecture: primary beam targets wafer surface, secondary beam simultaneously reads reference chip
  • firmware performs automatic background subtraction and drift correction in <2 seconds, achieving 10:1 SNR
  • Use hermetically sealed reference module with UV-grade quartz window, NIST-traceable certification valid 12 months
  • automated self-check routine verifies reference signal within ±3% daily, flags degradation for replacement
Expected Effect : SNR 10:1 achieved; zero manual calibration; cycle time <6 min
Risk Control :
  • reference chip degradation over time
  • optical path alignment drift
  • firmware algorithm stability under fab environment variation
Patsnap Eureka Solution