Part II, Chapter 2

Light Forces on Atoms

Photon momentum transfer creates two fundamental forces: radiation pressure (scattering force) and the optical dipole force.

2.1 Scattering Force (Radiation Pressure)

When an atom absorbs a photon, it receives a momentum kick ℏk in the direction of the laser beam. Spontaneous emission is isotropic, so over many cycles the recoil from emission averages to zero. The net force is the scattering force — directed along the laser beam.

Derivation 1: Scattering Force from Photon Momentum Transfer

Step 1. Each absorbed photon transfers momentum ℏk to the atom. The scattering rate is R_sc = Γρ_ee, where ρ_ee is the steady-state excited population.

Step 2. Using the steady-state result from the optical Bloch equations:

$$\rho_{ee} = \frac{s_0/2}{1 + s_0 + (2\delta/\Gamma)^2}, \qquad s_0 = \frac{I}{I_{\text{sat}}}$$

Step 3. The scattering force is F = ℏk × R_sc:

$$\mathbf{F}_{\text{scat}} = \hbar\mathbf{k}\frac{\Gamma}{2}\frac{s_0}{1 + s_0 + (2\delta/\Gamma)^2}$$

The maximum force (at saturation, on resonance) is F_max = ℏkΓ/2, producing a maximum deceleration a_max = ℏkΓ/(2m). For sodium: a_max ≈ 10&sup5; g!

2.1.1 Radiation Pressure Applications

The scattering force is used for:

• Zeeman slowers: A spatially varying magnetic field keeps atoms in resonance as they decelerate, producing a slow atomic beam
• Optical molasses: Counter-propagating beams create a velocity-dependent damping force
• Atomic fountains: Launching cold atoms upward for precision spectroscopy

2.2 Dipole Force (Gradient Force)

The dipole force arises from the interaction of the induced atomic dipole moment with the gradient of the light intensity. Unlike the scattering force, it is conservative and can trap atoms.

Derivation 2: Dipole Force from AC Stark Shift

Step 1. The atom-light interaction shifts the atomic energy levels. For large detuning |δ| >> Γ, the ground state energy shift (AC Stark shift / light shift) is:

$$U_{\text{dip}} = -\frac{3\pi c^2}{2\omega_0^3}\frac{\Gamma}{\delta}I(\mathbf{r}) = \frac{\hbar\Omega^2}{4\delta}$$

Step 2. The dipole force is the gradient of this potential:

$$\mathbf{F}_{\text{dip}} = -\nabla U_{\text{dip}} = \frac{3\pi c^2}{2\omega_0^3}\frac{\Gamma}{\delta}\nabla I(\mathbf{r})$$

Step 3. The sign of the detuning determines the trapping behavior:

Red detuning (δ < 0): U_dip < 0 at intensity maxima. Atoms attracted to high intensity → optical traps, optical lattices.

Blue detuning (δ > 0): U_dip > 0 at intensity maxima. Atoms repelled from high intensity → hollow beam guides, optical barriers.

2.2.2 Optical Trapping

A tightly focused red-detuned laser beam creates a 3D optical trap (optical dipole trap or optical tweezers). The trap depth is:

$$U_0 = \frac{3\pi c^2}{2\omega_0^3}\frac{\Gamma}{|\delta|}I_0$$

Typical depths are 1-10 mK for neutral atoms. The photon scattering rate in the trap scales as Γ_sc ∝ I/δ², while the trap depth scales as U ∝ I/δ. Large detuning with high intensity minimizes heating while maintaining deep traps.

2.3 Dressed Atom Picture

The dressed atom model treats both the atom and the light field quantum mechanically. The atom-photon system has eigenstates (dressed states) that are superpositions of |g, n+1〉 and |e, n〉, where n is the photon number.

Derivation 3: Dressed State Energies

Step 1. The uncoupled states |g, n+1〉 and |e, n〉 are nearly degenerate (separated by detuning δ). The Jaynes-Cummings Hamiltonian couples them with strength Ω_n = Ω√(n+1)/2.

$$H_n = \hbar\begin{pmatrix} -\delta/2 & \Omega_n/2 \\ \Omega_n/2 & \delta/2 \end{pmatrix}$$

Step 2. Diagonalizing this 2×2 matrix:

$$E_\pm(n) = (n + \tfrac{1}{2})\hbar\omega \pm \frac{\hbar}{2}\sqrt{\delta^2 + \Omega_n^2}$$

The dressed states are split by the generalized Rabi frequency. In an inhomogeneous field (standing wave), the position-dependent Rabi frequency creates a spatially varying energy landscape — the basis of the dipole force in the quantum picture.

2.4 Optical Molasses

Two counter-propagating red-detuned laser beams create a velocity-dependent damping force that slows atoms — optical molasses. This is the foundation of Doppler cooling.

Derivation 4: Velocity-Dependent Force in Optical Molasses

Step 1. An atom moving with velocity v sees Doppler-shifted frequencies. For counter-propagating beams at frequency ω, the effective detunings are:

$$\delta_\pm = \delta \mp kv$$

Step 2. The total force is the sum of scattering forces from both beams:

$$F = F_+(v) + F_-(v) = \hbar k \frac{\Gamma}{2}\left[\frac{s_0}{1 + s_0 + (2(\delta - kv)/\Gamma)^2} - \frac{s_0}{1 + s_0 + (2(\delta + kv)/\Gamma)^2}\right]$$

Step 3. For small velocities (kv << Γ), expand to first order in v:

$$F \approx -\beta v, \qquad \beta = -8\hbar k^2 s_0 \frac{\delta/\Gamma}{(1 + s_0 + (2\delta/\Gamma)^2)^2}$$

For δ < 0 (red detuning), β > 0: a viscous damping force. The atom experiences friction proportional to its velocity — hence "optical molasses."

2.5 Optical Lattices

Two counter-propagating laser beams form a standing wave with a periodic intensity pattern. The dipole potential creates a 1D lattice with period λ/2. Three orthogonal standing waves create a 3D optical lattice — an artificial crystal of light trapping atoms at each site.

Derivation 5: Band Structure in an Optical Lattice

Step 1. The potential from a 1D standing wave is:

$$V(x) = -V_0 \cos^2(kx) = -\frac{V_0}{2}[1 + \cos(2kx)]$$

Step 2. The Schrödinger equation becomes the Mathieu equation:

$$-\frac{\hbar^2}{2m}\frac{d^2\psi}{dx^2} - \frac{V_0}{2}\cos(2kx)\psi = (E + V_0/2)\psi$$

Step 3. By Bloch's theorem, solutions have the form ψ = e^iqxu_q(x) with u_q periodic. The energy spectrum forms bands separated by gaps:

For deep lattices (V₀ >> E_R = ℏ²k²/2m), the bands are narrow and atoms are tightly localized at lattice sites. The tunneling rate between sites is J ∝ exp(-√(V₀/E_R)). This system realizes the Bose-Hubbard model, enabling quantum simulation of condensed matter physics.

2.6 Optical Tweezers: From Atoms to Biology

Optical tweezers use a tightly focused laser beam to trap and manipulate objects ranging from single atoms to biological cells. The physics depends on the object size relative to the wavelength.

2.6.1 Rayleigh Regime (atoms, nanoparticles: d << λ)

For particles much smaller than the wavelength, the dipole approximation applies. The gradient force is proportional to the polarizability α and the intensity gradient:

$$\mathbf{F}_{\text{grad}} = \frac{\alpha}{2}\nabla|\mathbf{E}|^2 = \frac{\alpha}{2\epsilon_0 c}\nabla I$$

For atoms, α is the AC polarizability derived from the dipole potential. For dielectric nanoparticles, α = 3ε₀V(ε_r-1)/(ε_r+2).

2.6.2 Mie Regime (cells, microspheres: d >> λ)

For large objects, the ray optics picture applies. Refraction of light rays through the particle transfers momentum, creating a net force toward the beam focus. The trapping force depends on the numerical aperture of the focusing objective and the refractive index contrast between the particle and the medium.

2.6.3 Optical Tweezer Arrays for Quantum Computing

Modern experiments use spatial light modulators (SLMs) or acousto-optic deflectors (AODs) to create programmable arrays of hundreds of optical tweezers, each holding a single atom:

• Loading: Random loading from a MOT (50% fill probability per site)
• Rearrangement: Defect-free arrays created by moving atoms with movable tweezers
• Entangling: Rydberg excitation creates strong long-range interactions for two-qubit gates
• Readout: Fluorescence imaging resolves individual atoms in each tweezer

Groups at Harvard/MIT (Lukin), Caltech (Endres), and Atom Computing have demonstrated arrays with >1000 atoms, achieving >99.5% two-qubit gate fidelity using Rydberg interactions.

2.7 Forces in Standing Waves: The Dressed Atom Picture

In a standing wave, the dressed state energies vary spatially with the local Rabi frequency Ω(x) = Ω₀ sin(kx). This creates a periodic potential landscape for the dressed atom.

2.7.1 Dressed State Potentials

The two dressed states have position-dependent energies:

$$E_\pm(x) = \pm\frac{\hbar}{2}\sqrt{\delta^2 + \Omega_0^2\sin^2(kx)}$$

At blue detuning (δ > 0), atoms in the lower dressed state are attracted to intensity nodes (V-shaped potential wells). At red detuning (δ < 0), atoms are attracted to antinodes.

2.7.2 Channeling and Diffraction

Atoms passing through a standing wave experience the dipole force as a phase grating. The periodic potential causes Bragg diffraction of the atomic de Broglie wave:

$$p_n = p_0 + 2n\hbar k, \qquad n = 0, \pm 1, \pm 2, \ldots$$

This is the Kapitza-Dirac effect — diffraction of matter waves by a standing light wave, the inverse of optical diffraction by a material grating. First observed with electrons by Freimund et al. (2001) and with atoms by Gould, Ruff, and Pritchard (1986).

2.7.3 Bloch Oscillations

Atoms loaded into the lowest band of an optical lattice and subjected to a constant external force F (gravity or magnetic gradient) undergo Bloch oscillations: periodic oscillations in momentum space with frequency ω_B = Fd/ℏ, where d = λ/2 is the lattice period. These oscillations have been used to measure gravity with atoms (g = 9.80 m/s²) and the fine structure constant.

2.8 Atom Optics

Light forces enable the construction of optical elements for atomic de Broglie waves —atom optics. Just as glass lenses and mirrors manipulate light, laser fields can focus, reflect, split, and diffract atomic beams.

Atom Optics Toolkit

Atom Mirror

A blue-detuned evanescent wave above a glass surface creates a repulsive barrier that reflects atoms like a mirror. Used in atom trampoline experiments and surface probes.

Atom Lens

A focused Gaussian beam creates a harmonic potential that acts as a thin lens for atoms. Atom lithography uses standing waves to focus atoms onto substrates with nanometer resolution.

Atom Beam Splitter

Bragg diffraction from a standing wave coherently splits an atomic beam into two momentum states, forming the basis of atom interferometers. Raman transitions between hyperfine states provide an alternative beam splitter mechanism.

Atom Waveguide

Hollow-core optical fibers or blue-detuned Laguerre-Gaussian beams guide atoms along channels. Integrated atom-optical circuits on atom chips combine magnetic and optical elements for compact quantum sensors.

Applications

Optical Tweezers

Arthur Ashkin's invention of optical tweezers (Nobel Prize 2018) uses a tightly focused laser to trap and manipulate microscopic objects from atoms to biological cells. Modern tweezer arrays trap hundreds of individual atoms for quantum computing.

Quantum Simulation

Ultracold atoms in optical lattices simulate condensed matter systems — the Bose-Hubbard model, Mott insulator transitions, and antiferromagnetic ordering have been directly observed. This is Feynman's vision of quantum simulation made real.

Atomic Beam Deceleration

Zeeman slowers use the scattering force with a spatially varying magnetic field to decelerate thermal atomic beams from ~500 m/s to ~10 m/s, providing slow atoms for loading into magneto-optical traps.

Optical Lattice Clocks

Atoms trapped at the nodes of a "magic wavelength" optical lattice experience zero differential light shift. This enables the most precise clocks ever built, with fractional uncertainties reaching 10&supmin;¹&sup8;.

2.9 Comparison of Scattering and Dipole Forces

The two light forces have complementary properties suited for different applications.

Scattering Force Properties

• Origin: Absorption followed by spontaneous emission
• Direction: Along the laser beam direction
• Scaling: F ∝ I/(1 + I/I_sat) — saturates at ℏkΓ/2
• Dissipative: Yes — involves irreversible spontaneous emission
• Applications: Laser cooling, Zeeman slowing, optical molasses, MOT

Dipole Force Properties

• Origin: Interaction of induced dipole with intensity gradient
• Direction: Along the intensity gradient
• Scaling: F ∝ I/δ — does not saturate
• Conservative: Yes — derivable from a potential
• Applications: Optical traps, tweezers, lattices, waveguides

The figure of merit for dipole traps is the ratio of trap depth to scattering rate:

$$\frac{U_{\text{dip}}}{\Gamma_{\text{sc}}} = \frac{\hbar\delta}{\Gamma}$$

This motivates using far-detuned traps (|δ| >> Γ) with high intensity. At a magic wavelength, the AC Stark shifts of ground and excited states are equal, leaving the transition frequency unperturbed — essential for optical lattice clocks achieving 10&supmin;¹&sup8; fractional uncertainties.

2.10 Advanced Trapping Geometries

Beyond simple focused beams and standing waves, sophisticated trapping geometries enable new physics and applications.

2.10.1 Ring Traps and Toroidal Potentials

Laguerre-Gaussian beams with orbital angular momentum or painted potentials create ring-shaped traps. BECs in ring traps support persistent currents — quantized flow around the ring that persists without dissipation. These systems are atomic analogs of SQUIDs and have been used to study superfluid hydrodynamics and the Sagnac effect for inertial sensing.

2.10.2 Box Traps

Flat-bottomed "box" traps are created using blue-detuned sheet beams for the walls and a digital micromirror device (DMD) for arbitrary potential shaping. Unlike harmonic traps, box traps provide uniform density — simplifying comparison with homogeneous many-body theory. Gaunt et al. (Cambridge, 2013) first loaded a BEC into a box trap.

2.10.3 Programmable Potentials

Spatial light modulators (SLMs) and DMDs project arbitrary 2D potential landscapes for atoms. Combined with the time-averaging technique (painting potentials with rapidly scanned beams), virtually any trapping geometry can be realized — from double wells to lattices with arbitrary disorder patterns to ring lattices for studying topological physics.

Key Equations Summary

Scattering Force:

$$\mathbf{F}_{\text{scat}} = \hbar\mathbf{k}\frac{\Gamma}{2}\frac{s_0}{1 + s_0 + (2\delta/\Gamma)^2}$$

Dipole Potential:

$$U_{\text{dip}} = -\frac{3\pi c^2}{2\omega_0^3}\frac{\Gamma}{\delta}I(\mathbf{r})$$

Molasses Damping Force:

$$F \approx -\beta v, \qquad \beta = -8\hbar k^2 s_0 \frac{\delta/\Gamma}{(1 + s_0 + (2\delta/\Gamma)^2)^2}$$

Dressed State Splitting:

$$E_\pm = \pm\frac{\hbar}{2}\sqrt{\delta^2 + \Omega^2}$$

Recoil Energy:

$$E_R = \frac{\hbar^2 k^2}{2m}, \qquad T_R = \frac{E_R}{k_B}$$

Photon Scattering Rate in Dipole Trap:

$$\Gamma_{\text{sc}} = \frac{3\pi c^2}{2\hbar\omega_0^3}\left(\frac{\Gamma}{\delta}\right)^2 I$$

Historical Context

1619 — Kepler: Proposed that comet tails point away from the Sun due to radiation pressure — the first suggestion of light forces on matter.

1970 — Ashkin: Demonstrated that focused laser beams could accelerate and trap dielectric microspheres, founding the field of optical manipulation.

1975 — Hänsch & Schawlow: Proposed laser cooling of atoms using the Doppler effect with counter-propagating beams.

1985 — Chu et al.: First demonstration of 3D optical molasses, cooling sodium atoms to 240 μK. Nobel Prize 1997.

1993 — Grynberg et al.: Observed atoms trapped in optical lattices, opening the era of quantum simulation with cold atoms.

2002 — Greiner et al.: Observed the superfluid-to-Mott insulator quantum phase transition with atoms in an optical lattice.

Interactive Simulation

This simulation visualizes the scattering and dipole forces, the optical molasses damping mechanism, and the band structure of an optical lattice.

Light Forces: Scattering, Dipole, and Optical Lattice

Python

script.py103 lines

import numpy as np
import matplotlib.pyplot as plt

fig, axes = plt.subplots(2, 2, figsize=(12, 10))
fig.patch.set_facecolor("#0a0a0a")
for ax in axes.flat:
    ax.set_facecolor("#111111")

Gamma = 1.0  # natural linewidth (units)

# Plot 1: Scattering force vs velocity for different detunings
ax1 = axes[0, 0]
v = np.linspace(-5, 5, 500)
k = 1.0
s0 = 1.0

for delta, col, label in [(-0.5, "#a78bfa", "delta=-0.5"),
                           (-1.0, "#c084fc", "delta=-1.0"),
                           (-2.0, "#f97316", "delta=-2.0")]:
    F_plus = s0 / (1 + s0 + (2*(delta - k*v)/Gamma)**2)
    F_minus = -s0 / (1 + s0 + (2*(delta + k*v)/Gamma)**2)
    F_total = (Gamma/2) * (F_plus + F_minus)
    ax1.plot(v, F_total, color=col, linewidth=2, label=label)

ax1.axhline(y=0, color="gray", linestyle="--", alpha=0.3)
ax1.axvline(x=0, color="gray", linestyle="--", alpha=0.3)
ax1.set_xlabel("Velocity (Gamma/k units)", color="white")
ax1.set_ylabel("Force (hbar k Gamma/2 units)", color="white")
ax1.set_title("Optical Molasses Force vs Velocity", color="#c4b5fd", fontsize=14)
ax1.legend(facecolor="#1a1a2e", edgecolor="#7c3aed", labelcolor="white")
ax1.tick_params(colors="white")
for spine in ax1.spines.values():
    spine.set_color("#7c3aed")

# Plot 2: Dipole potential for Gaussian beam
ax2 = axes[0, 1]
x = np.linspace(-3, 3, 300)
y = np.linspace(-3, 3, 300)
X, Y = np.meshgrid(x, y)
w0 = 1.0  # beam waist
I_gauss = np.exp(-2*(X**2 + Y**2)/w0**2)

# Red detuned: attractive
U_red = -I_gauss
im = ax2.contourf(X, Y, U_red, levels=30, cmap="magma")
ax2.set_xlabel("x / w0", color="white")
ax2.set_ylabel("y / w0", color="white")
ax2.set_title("Dipole Trap Potential (red detuned)", color="#c4b5fd", fontsize=14)
ax2.tick_params(colors="white")
for spine in ax2.spines.values():
    spine.set_color("#7c3aed")

# Plot 3: Standing wave potential and band structure
ax3 = axes[1, 0]
x_lat = np.linspace(0, 3, 300)
for V0, col, label in [(2, "#a78bfa", "V0 = 2 Er"),
                        (5, "#c084fc", "V0 = 5 Er"),
                        (10, "#f97316", "V0 = 10 Er")]:
    V_lattice = -V0 * np.cos(2 * np.pi * x_lat)**2
    ax3.plot(x_lat, V_lattice, color=col, linewidth=2, label=label)

ax3.set_xlabel("Position (lambda/2 units)", color="white")
ax3.set_ylabel("Potential (Er units)", color="white")
ax3.set_title("1D Optical Lattice Potential", color="#c4b5fd", fontsize=14)
ax3.legend(facecolor="#1a1a2e", edgecolor="#7c3aed", labelcolor="white")
ax3.tick_params(colors="white")
for spine in ax3.spines.values():
    spine.set_color("#7c3aed")

# Plot 4: Scattering rate vs trap depth trade-off
ax4 = axes[1, 1]
delta_vals = np.linspace(10, 1000, 500)  # in Gamma units
I_fixed = 1000.0  # fixed intensity in Isat units

# U proportional to I/delta
U_trap = I_fixed / delta_vals
# Gamma_sc proportional to I/delta^2
Gamma_sc = I_fixed / delta_vals**2

ax4_twin = ax4.twinx()
line1, = ax4.plot(delta_vals, U_trap, color="#a78bfa", linewidth=2, label="Trap depth (arb.)")
line2, = ax4_twin.plot(delta_vals, Gamma_sc, color="#f97316", linewidth=2, label="Scattering rate")

ax4.set_xlabel("Detuning (Gamma units)", color="white")
ax4.set_ylabel("Trap depth U (arb.)", color="#a78bfa")
ax4_twin.set_ylabel("Scattering rate (arb.)", color="#f97316")
ax4.set_title("Trap Depth vs Heating Trade-off", color="#c4b5fd", fontsize=14)
ax4.tick_params(axis="y", colors="#a78bfa")
ax4.tick_params(axis="x", colors="white")
ax4_twin.tick_params(colors="#f97316")
lines = [line1, line2]
labels = [l.get_label() for l in lines]
ax4.legend(lines, labels, facecolor="#1a1a2e", edgecolor="#7c3aed", labelcolor="white", loc="upper right")
for spine in ax4.spines.values():
    spine.set_color("#7c3aed")
for spine in ax4_twin.spines.values():
    spine.set_color("#7c3aed")

plt.tight_layout()
plt.savefig("output.png", dpi=150, bbox_inches="tight", facecolor="#0a0a0a")
plt.show()
print("Plot saved: Light forces - molasses, dipole trap, optical lattice, and trade-offs.")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

Conceptual Questions

Q1: Why does red detuning attract atoms to intensity maxima?

Red detuning (δ < 0) means the laser is below resonance. The AC Stark shift lowers the ground state energy where the field is strongest. Atoms are attracted to regions where their energy is minimized — the intensity maxima.

Q2: Why does the scattering force saturate but the dipole force does not?

The scattering force is limited by the maximum scattering rate Γ/2 (each atom can only scatter one photon per excited state lifetime). The dipole force, being conservative and not involving real absorption, has no such limit — it grows linearly with intensity for far-detuned light.

Q3: How do optical lattices simulate solid-state physics?

Atoms in optical lattices experience a periodic potential like electrons in a crystal, but with independently tunable parameters (depth, geometry, interactions). The lattice spacing (~500 nm) is 1000x larger than in solids, making imaging possible. The interaction strength is tunable via Feshbach resonances. This creates a clean, controllable quantum simulator.

Further Topics

Advanced Light-Force Physics

• Cavity optomechanics: The radiation pressure force inside an optical cavity creates a coupling between the cavity field and mechanical motion of a mirror or membrane. This enables ground-state cooling of macroscopic oscillators and quantum-limited position measurements (LIGO, mechanical quantum states).
• Rydberg-mediated forces: Atoms excited to high-lying Rydberg states (n > 50) interact via strong van der Waals forces (~MHz at micron distances). These controllable long-range forces enable fast quantum gates and quantum simulation of spin models.
• Artificial gauge fields: By engineering position-dependent Raman couplings, neutral atoms can experience effective magnetic fields and spin-orbit coupling. This enables simulation of the quantum Hall effect and topological insulators with cold atoms.
• Subwavelength optical lattices: Dark-state engineering and multicolor lattice schemes create potentials with periodicities below λ/2, reaching λ/4 or λ/8. These enable access to higher momentum scales and exotic band structures.