IR/Raman Molecular Fingerprint Matcher

Unknown Sample — Bond Mix

O–H stretch 30%

C–H stretch 30%

C=O stretch 10%

C=C aromatic 10%

N=O (NOx) 10%

S=O (SOx) 10%

Detector noise 0.04

Reference Library

Choose a specific reference to compare it directly, or leave on Auto to see whichever library compound best explains the sample spectrum.

Match Result

Best-match compound

–

Confidence

–

Runner-up

–

How it works

Every chemical bond behaves like a tiny spring: two atoms of reduced mass μ connected by a bond of force constant k vibrate at a characteristic wavenumber. Light in that band gets absorbed, producing a peak in the sample's IR/Raman spectrum:

ṽ = 1/(2πc) · √(k/μ),   μ = m₁m₂/(m₁+m₂)
I(ṽ) = Σᵢ Aᵢ · γᵢ² / ((ṽ−ṽᵢ)² + γᵢ²)   (Lorentzian peaks)

Heavier atoms or weaker bonds vibrate slower (lower ṽ): O–H ≈ 3400 cm⁻¹, C–H ≈ 2950 cm⁻¹, C=O ≈ 1720 cm⁻¹, aromatic C=C ≈ 1600 cm⁻¹, N=O ≈ 1560 cm⁻¹, S=O ≈ 1150 cm⁻¹ — a fingerprint no two unrelated molecules share exactly.

Bond-mix sliders — set what fraction of each bond type is present in the unknown sample; the spectrum is built by summing a Lorentzian peak per bond type, scaled by its fraction.
Detector noise — adds random jitter to the measured spectrum, the way a real spectrometer's shot noise and baseline drift blur weak peaks.
Matching — the noisy sample spectrum is compared against every reference compound's known fingerprint via cosine similarity; the closest match wins, and confidence is its similarity margin over the runner-up.

Real-world relevance: this is exactly how environmental labs identify pollutants — an unknown air, water or soil sample's IR/Raman spectrum is matched against a library of known-compound references (NOx, SOx, VOCs, PAHs) to identify contamination without needing to isolate or culture anything.

Sample (measured) Reference (match)