Atomic Force Microscope

Role in QLT Fabrication

Why Surface Roughness Matters for Quantum Photonics

The AFM provides 3D topographic imaging at sub-nanometer vertical resolution ● the ONLY tool that can directly measure waveguide surface roughness at the level that affects optical propagation loss. For QLT's room-temperature quantum photonic chip, surface roughness is the dominant loss mechanism.

Waveguide scattering loss is dominated by sidewall and surface roughness:

SCATTERING LOSS EQUATION:

α_scatter ∝ (σ² / λ²) × (n₁² - n₂²)²

where:
σ  = RMS surface roughness
λ  = operating wavelength (1550 nm)
n₁ = core refractive index (SiN = 1.99)
n₂ = cladding refractive index (SiO₂ = 1.45)

Numerical impact for SiN waveguides at 1550 nm:

σ = 0.1 nm → loss ≈ 0.05 dB/cm  (state-of-the-art)
σ = 0.3 nm → loss ≈ 0.5 dB/cm   (LIGENTEC spec ● our target)
σ = 0.5 nm → loss ≈ 1.4 dB/cm   (marginal for quantum applications)
σ = 1.0 nm → loss ≈ 5.0 dB/cm   (unacceptable ● 10× too high)

For quantum OPC at 1550 nm, we need propagation loss < 1 dB/cm
→ surface roughness MUST be < 0.5 nm RMS
→ the AFM is the ONLY instrument that can distinguish 0.3 nm from 0.5 nm

AFM vs. Other Metrology Tools

Conclusion: AFM is irreplaceable for sub-nm roughness QC. The optical profilometer (#15) handles routine roughness screening; the AFM provides definitive measurements and arbitrates when results are borderline.

Specific Measurements for QLT

AFM in the QLT Metrology Hierarchy

ROUGHNESS MEASUREMENT DECISION TREE:

START: Need roughness measurement
│
├── Routine screening (Sa > 1 nm expected)?
│   └── Use OPTICAL PROFILER (#15) ● fast, non-contact, 3D areal
│
├── Quantum-grade roughness (Sa < 1 nm target)?
│   └── Use AFM (#14) ● definitive sub-nm measurement
│
├── Soft film (As₂S₃, PVDF-TrFE)?
│   ├── Screening → OPTICAL PROFILER (#15) ● zero contact risk
│   └── Definitive → AFM (#14) in tapping mode (minimal force)
│
├── Step height measurement?
│   ├── Primary → STYLUS PROFILOMETER (#07) ● absolute reference
│   └── Cross-check → OPTICAL PROFILER (#15) or AFM (#14)
│
└── Refractive index + thickness?
└── ELLIPSOMETER (#16) ● only tool for n,k measurement

Recommended Configuration

Process Integration

System Setup and First Scan

STARTUP PROCEDURE (AFM Workshop TT-2):

STEP 1:  Power on controller (Ebox) and PC
STEP 2:  Launch AFM software
STEP 3:  Mount probe (tapping mode silicon cantilever)
├── Place probe chip in probe holder using tweezers
├── Secure with spring clip
└── Insert holder into AFM stage
STEP 4:  Align laser on cantilever using video microscope
├── View cantilever through microscope
├── Adjust laser XY position until beam is centered on cantilever tip
└── Adjust detector position to center the reflected beam
STEP 5:  Mount sample
├── Affix sample to magnetic sample puck using double-sided tape
├── Place puck on magnetic stage
└── Navigate to area of interest using video microscope
STEP 6:  Frequency tune (tapping mode)
├── Software sweeps frequency to find cantilever resonance (~300 kHz typical)
├── Set drive amplitude to achieve ~80% free amplitude
└── Set setpoint to ~70% of free amplitude (gentle tapping)
STEP 7:  Approach
├── Use coarse approach (motorized) to bring probe within ~100 μm of surface
├── Initiate automated approach → probe engages surface
└── Verify stable feedback signal
STEP 8:  Scan
├── Set scan parameters (area, speed, resolution)
├── Start scan → image builds line by line
└── Monitor real-time for artifacts (probe tip contamination, feedback instability)
STEP 9:  Save and analyze
├── Save raw image data
├── Open in Gwyddion for quantitative analysis
└── Export roughness values and images for QC report

Total setup time (first scan of session): ~30 minutes
Subsequent scans: ~15-20 minutes each (probe already aligned)

Calibration Procedures

WEEKLY Z-AXIS CALIBRATION:

1. Mount NIST-traceable step-height standard (e.g., 100 nm step)
2. Scan across step edge in tapping mode
3. Measure step height in analysis software
4. Expected: 100 ± 2 nm
5. If > 2 nm deviation: adjust Z calibration coefficient in software
6. Log result

WEEKLY ROUGHNESS BASELINE:

1. Cleave fresh muscovite mica sheet
2. Mount on sample puck
3. Scan 2 × 2 μm area in tapping mode
4. Calculate Sq roughness → should be < 0.1 nm (atomically flat)
5. If Sq > 0.15 nm: probe tip is contaminated or damaged → replace probe
6. Log result

MONTHLY XY CALIBRATION:

1. Mount calibration grid (e.g., 10 μm pitch, 100 nm deep)
2. Scan 50 × 50 μm area
3. Verify grid pitch measures 10.0 ± 0.1 μm
4. If > 1% deviation: adjust XY calibration coefficients
5. Log result

Vendor Options & Pricing

New System Cost ● AFM Workshop TT-2 Configurations

Used/Refurbished Market

Used market sources: LabX (labx.com), Used-Line (used-line.com), Capovani Brothers, eBay (for educational units)

Our Budget Recommendation

NEW (fastest delivery, best value for QLT):

• AFM Workshop TT-2 (50 μm scanner + acoustic enclosure): ~$53,000
• Add vibration isolation table: ~$5,000
• Total: ~$58,000; lead time 2–4 weeks
• US-made (Signal Hill, CA); excellent technical support; fast delivery
• Z noise < 150 pm ● adequate for all QLT surface roughness measurements

REFURBISHED (budget):

• Bruker Dimension 3100 (refurb): $18,000–$25,000
• Add isolation table + enclosure: ~$8,000
• Total: ~$31,000; lead time 2–4 weeks
• Excellent used market availability; NanoScope software well-documented
• May need scanner recalibration ($2,000 service)

Facility Requirements

Vibration ● THE CRITICAL REQUIREMENT

Acoustic Environment

Environment

Safety & Handling