Stokes and Anti-Stokes Raman Lines

Introduction

A Raman spectrum is often plotted as intensity versus shift from a laser line, not versus the actual color of every outgoing photon. A vibrating molecule can take energy from a photon, creating a Stokes line, or give vibrational energy to a photon, creating an anti-Stokes line. Both shifts can identify the same vibrational mode. Their differing intensities also reveal something about how many molecules occupied an excited vibrational state before scattering. Reading the axes and energy direction correctly is the first step toward using that information.

Core explanation

Let the excitation photon have wavenumber ν̃₀ and a vibrational mode have wavenumber ν̃ v. In an ideal single-quantum Stokes event, the molecule ends one vibrational level higher, so the scattered photon has ν̃ S = ν̃₀ − ν̃ v. In the corresponding anti-Stokes event, the molecule starts in a vibrationally excited level and ends lower, giving the photon energy: ν̃ AS = ν̃₀ + ν̃ v. Rayleigh scattering is elastic and appears at ν̃₀. Since photon energy is hcν̃, lower wavenumber means lower photon energy and longer wavelength. IUPAC's anti-Stokes Raman definition explicitly says that the scattered radiation has greater energy than the exciting radiation. The matching Stokes line has less.

The laser line and the two Raman lines are symmetric in wavenumber shift for a simple mode, but not in wavelength measured in nanometers. Wavelength and wavenumber have a reciprocal relation, ν̃ = 1/λ when units are consistent. A 1000 cm⁻¹ shift does not correspond to the same number of nanometers on each side of a laser wavelength. Most Raman charts display positive shift magnitudes for Stokes peaks, and some instruments show anti-Stokes values with negative sign. Others place anti-Stokes on the positive side using a different axis convention. Always read the axis label before deciding whether an observed peak gained or lost photon energy.

At thermal equilibrium, most molecules occupy low vibrational levels when the mode energy substantially exceeds k BT. Stokes scattering can start from the vibrational ground state, while the simplest anti-Stokes event requires an initially excited molecule. The population ratio for two harmonic levels is approximately N₁/N₀ = exp(−hcν̃ v/k BT), ignoring degeneracy differences. Thus anti-Stokes features are often weaker at room temperature and grow relative to Stokes features as temperature rises. A primary Raman thermometry study uses this intensity difference to infer local temperature and discusses the need for calibration.

The intensity ratio is not exactly the raw Boltzmann population ratio. Raman scattering strength has a frequency dependence, commonly approximated by a scattered-frequency-to-the-fourth-power factor for comparable conditions. Detector response, optical filters, collection geometry, absorption, resonance enhancement and local heating can also favor one side. A simplified corrected ratio for a mode may be written I AS/I S ≈ [(ν̃₀+ν̃ v)/(ν̃₀−ν̃ v)]⁴ exp(−hcν̃ v/k BT), multiplied by any relevant calibration factors. One must not use this approximation when the sample is driven far from thermal equilibrium, when fluorescence overlaps a sideband or when resonance conditions make the cross sections differ unexpectedly. The correction is especially important for precise thermometry.

Stokes and anti-Stokes line positions give complementary consistency checks. If a band at shift +800 cm⁻¹ on the Stokes side has a matching anti-Stokes band at −800 cm⁻¹ under a signed shift convention, both can belong to the same mode. Their common shift is a vibrational energy, not two separate bond vibrations. Line width may differ with experimental signal-to-noise, but the mode assignment should be coherent. Strong laser heating can raise the local vibrational population and therefore the anti-Stokes signal, so an apparently high anti-Stokes/Stokes ratio may reveal measurement perturbation rather than the sample's undisturbed temperature.

The physical scattering process passes through a short-lived virtual-state description, not necessarily a long-lived molecular electronic excited state. Raman activity depends on change in polarizability during the vibration, whereas infrared absorption depends on change in dipole moment. The next page compares their selection rules. For the present energy accounting, one photon enters, one leaves and the molecule's vibrational energy changes by the difference. This conservation check works regardless of how a plotting program labels the horizontal axis.

Step-by-step reasoning

1. Convert excitation wavelength to wavenumber if necessary, using ν̃₀ = 1/λ with consistent length units. 2. Identify whether the outgoing photon has lower, equal or higher energy than the excitation photon. 3. Subtract ν̃ v for Stokes, add ν̃ v for anti-Stokes, or use ν̃₀ for Rayleigh. 4. Translate the shift into wavelength only after completing the wavenumber calculation. 5. For intensity questions, estimate the excited-state population with a Boltzmann factor and then consider frequency and instrument corrections. 6. Check whether heating, resonance or non-equilibrium excitation invalidates simple thermal-population reasoning.

Visual explanation

Draw a ground electronic state with vibrational levels v = 0 and v = 1, plus an unlabeled virtual-state region above them. A Stokes arrow begins at v = 0 and ends at v = 1 after scattering; its outgoing photon arrow is shorter than the incoming photon arrow. An anti-Stokes arrow starts at v = 1 and ends at v = 0; its outgoing photon arrow is longer. Put Rayleigh beside them with equal arrows. Under the diagram, plot three lines on an absolute wavenumber axis around ν̃₀ and then redraw them on a signed shift axis to show how axis convention changes their labels but not the physics.

Real-world analogy

A ball bouncing from a moving object can leave with less energy if it pushes the object into faster motion, or more energy if the object gives energy to the ball. Stokes scattering leaves vibrational energy in the molecule; anti-Stokes takes some away. The analogy does not capture quantum selection rules or virtual electronic states, but it makes the direction of energy transfer memorable.

Real-world example

A materials scientist shines a laser on a semiconductor and observes a strong Stokes phonon peak plus a weaker anti-Stokes partner. Increasing laser power raises the anti-Stokes/Stokes ratio while the rest of the apparatus is unchanged. The scientist checks whether the laser has heated the illuminated spot rather than claiming a new chemical bond. Power-dependent spectra, a detector-response calibration and comparison with an independent temperature measure make the thermometry interpretation more reliable.

Why?

Why is an anti-Stokes line usually weaker than its Stokes partner at ordinary temperatures? Anti-Stokes scattering needs molecules already in the relevant vibrational excited state, and that population is thermally smaller than the ground-state population when hcν̃ v is appreciable relative to k BT. Stokes scattering can begin from the more populated ground state. The ratio can approach unity or change under special frequency and resonance conditions, so “always weaker” is too absolute.

Common misconception

“An anti-Stokes line has a lower-energy photon because it appears on the negative-shift side of a graph.” The negative graph coordinate is only a plotting convention; the anti-Stokes photon has higher energy than the laser. Another error is treating equal shift magnitudes as equal nanometer wavelength differences. A third is calculating temperature from raw intensities without correcting instrument response and the frequency dependence of scattering.

Worked example

An excitation laser has wavenumber ν̃₀ = 20,000 cm⁻¹, and a Raman-active mode has shift ν̃ v = 1000 cm⁻¹. Its Stokes photon has ν̃ S = 19,000 cm⁻¹ and wavelength λ S = 10⁷/19,000 ≈ 526.3 nm. Its anti-Stokes photon has ν̃ AS = 21,000 cm⁻¹ and λ AS = 10⁷/21,000 ≈ 476.2 nm. The laser itself is 10⁷/20,000 = 500.0 nm. The shifts are ±1000 cm⁻¹, but the wavelength changes are +26.3 nm and −23.8 nm, not symmetric. For a first population estimate at 300 K, use k BT/hc ≈ 208.5 cm⁻¹, so N₁/N₀ ≈ exp(−1000/208.5) ≈ 0.0083. A raw intensity ratio need not equal 0.0083 exactly because of frequency and instrument factors.

Quick check

1. Does an anti-Stokes photon carry more or less energy than the excitation photon? Answer: More energy; the molecule loses vibrational energy to the outgoing photon. 2. Why does raising sample temperature commonly increase anti-Stokes intensity relative to Stokes intensity? Answer: More molecules thermally populate the initially excited vibrational state required for anti-Stokes scattering.

Exam focus

Use energy conservation first: Stokes photon energy decreases and anti-Stokes photon energy increases. State whether the horizontal axis is scattered wavenumber, wavelength or Raman shift, and convert reciprocal units carefully. For thermometry, write the Boltzmann population factor but mention frequency, calibration and non-equilibrium corrections. Do not confuse Raman Stokes/anti-Stokes shifts with fluorescence Stokes shifts, which involve a different sequence of electronic excitation and emission.

Advanced insight

The anti-Stokes/Stokes ratio can act as a local probe of vibrational temperature, but “temperature” assumes an approximately thermal distribution of vibrational levels in the sampled region. Under intense optical pumping, individual modes may be driven out of equilibrium with the lattice, giving different effective vibrational and bulk temperatures. In surface-enhanced Raman scattering, local electromagnetic enhancement can differ for incoming and outgoing wavelengths, further altering ratios. Careful thermometry therefore combines calibrated optical response, power extrapolation and checks against photochemical changes.

Summary

Stokes scattering gives energy to a molecular vibration, so the outgoing photon has lower energy; anti-Stokes scattering removes vibrational energy, so the outgoing photon has higher energy. Equal mode shifts are symmetric in wavenumber but not wavelength. Their relative intensities depend strongly on vibrational populations and temperature, with frequency and instrument effects requiring correction. Energy bookkeeping and axis reading prevent most interpretation errors.

Practice questions

1. A laser at 18,000 cm⁻¹ scatters from a 500 cm⁻¹ mode. What are the simple Stokes and anti-Stokes photon wavenumbers? Answer: Stokes is 17,500 cm⁻¹; anti-Stokes is 18,500 cm⁻¹. 2. Which process begins from a vibrationally excited molecule in the simplest one-quantum picture? Answer: Anti-Stokes scattering begins from the higher vibrational level and transfers energy to the photon. 3. Can raw anti-Stokes/Stokes peak-height ratio be used directly as a universal thermometer? Answer: No. Scattered-frequency dependence, detector response, filters, resonance and heating require calibration or correction. 4. Why are two lines with shifts of +700 and −700 cm⁻¹ often assigned to one mode? Answer: They can be its Stokes and anti-Stokes partners, which differ by the same vibrational energy in opposite directions.