The file is already merged — one workbook (`Book1`/`Sheet1`) with 15 power columns (5–100 µW) plus the eV axis. No Si/reference folder exists, so no background subtraction is needed. Suggested fit range is 1.23–1.45 eV.

Per protocol I must despike before any fit. Doing that now.



Despike cleaned 130 cosmic-ray spikes across all 15 columns; data written back so subsequent fits use cleaned data. The main exciton sits near ~1.32–1.33 eV, inside the suggested 1.23–1.45 eV window.

Now running the full power-dependence analysis: Lorentzian fit on every spectrum, extract intensity + xc per power, build the two panels (intensity-vs-power and xc-vs-power), and apply the allometric (y = a·P^b) fit.



![Power Fit]([file link removed])

**Pipeline executed:** inspect → despike (130 cosmic-ray spikes removed, σ=4, written back) → Lorentzian fit on all 15 merged spectra (window 1.23–1.45 eV) → extract intensity + xc per power → build both panels → allometric fit.

**Per-spectrum Lorentzian fits** — all 15 columns converged at **HIGH** quality (R² = 0.887–0.984, median 0.981; SNR 7.6–24):

| Power (µW) | xc (eV) | FWHM (meV) | Intensity (counts) |
|---|---|---|---|
| 5 | 1.3206 | 62.4 | 98.8 |
| 20 | 1.3278 | 62.7 | 353.8 |
| 50 | 1.3328 | 59.0 | 818.3 |
| 80 | 1.3345 | 56.3 | 1264.7 |
| 100 | 1.3356 | 54.4 | 1714.6 |

**Panel 1 — Peak intensity vs power (allometric fit I = a·P^b):**
- a = 16.28 ± 1.81
- **b = 1.0048 ± 0.0261**, R² = 0.995
- Interpretation: **~linear (excitonic)** — intensity scales proportionally with excitation power over 5–100 µW (no saturation / biexciton regime).

**Panel 2 — Peak position (xc) vs power:**
- xc **blueshifts** monotonically from 1.3206 eV (5 µW) to 1.3356 eV (100 µW) — a total shift of ~15 meV.
- Trend slope ≈ **+5.0 meV per decade of power**; FWHM narrows 62 → 54 meV across the range (consistent with reduced carrier heating / band-gap renormalization at higher power).

📥 Full per-point data: [power_fit_results.csv]([file link removed])

RESULT: b = 1.0048 ± 0.0261