Independent amplitude and phase control using a single phase-only SLM

Technical Summary: Independent Amplitude and Phase Control Using a Single Phase-Only SLM

Problem Statement
Liquid-crystal spatial light modulators (SLMs) are standard tools for programmable wavefront engineering but are typically restricted to phase-only modulation. While phase-only devices can control propagation, focusing, and interference, they cannot independently prescribe both the amplitude and phase of an optical field. This limitation leads to uncontrolled phase in reconstructed fields, resulting in speckle, nonuniformity, and artifacts, particularly for applications requiring specific complex-field states such as structured beams (e.g., Bessel or vortex beams) or light-based electron-beam shaping. Existing solutions for full complex-field synthesis often require cascaded phase-only SLMs or double-pass geometries, which increase system complexity, alignment sensitivity, and potential damage thresholds in high-power applications. Furthermore, single-SLM dual-phase schemes often suffer from reduced interference fidelity if the coordinate mapping between the two modulation regions is not perfectly unit-magnified.

Methodology
The authors propose a method to achieve independent amplitude and phase control using a single phase-only SLM by utilizing two spatially separated regions of the device as sequential modulation planes. The experimental setup (Fig. 1) operates as follows:

1. Input Polarization: A linearly polarized beam (45° relative to the SLM's active axis) illuminates the first region (Region A).
2. Amplitude Encoding: Region A imposes a spatially dependent phase retardance ($\phi_A$). This creates a spatially varying polarization ellipticity. A linear polarizer placed in the beam path projects this ellipticity, converting the phase retardance into controllable amplitude modulation. The transmitted amplitude is proportional to $\cos[\phi_A(x_p, y_p)/2]$.
3. Imaging: The modulated field is imaged with unit magnification onto the second region (Region B) of the same SLM using a folded $2f$ system. This ensures a direct one-to-one coordinate mapping between the amplitude and phase masks.
4. Phase Encoding: A quarter-wave plate (passed twice) rotates the polarization onto the SLM's active axis before it reaches Region B. Region B applies a second independently programmable phase pattern ($\phi_B$).
5. Complex Field Synthesis: The output complex field $E_{out}$ is given by:
   $$E_{out}(x_p, y_p) = A_0 \cos\left[\frac{\phi_A}{2}\right] \exp\left\{i\left[\frac{\phi_A}{2} + \phi_B\right]\right\}$$
   To synthesize a target field $E_{pupil} = A_0 \mathcal{A} \exp(iP)$, the phase masks are calculated as:
   • $\phi_A = 2 \arccos(\mathcal{A})$
   • $\phi_B = P - \arccos(\mathcal{A})$
     where $\mathcal{A}$ is the normalized target amplitude ($0 \le \mathcal{A} \le 1$) and $P$ is the target phase.

Key Contributions

• Single-Device Implementation: The work demonstrates a compact platform for full complex-field modulation using only one reflective phase-only SLM, eliminating the need for two separate modulators.
• Unit-Magnification Imaging: By imaging Region A directly onto Region B with unit magnification, the method preserves the coordinate mapping required for high-fidelity interference, avoiding the residual amplitude noise often associated with imperfect imaging in dual-region schemes.
• Direct Prescription: The approach allows for the direct, independent prescription of both radial amplitude and azimuthal phase distributions, rather than relying on iterative phase-retrieval algorithms that leave phase as a degree of freedom.

Experimental Results
The authors validated the system using a 1030 nm laser and a HOLOEYE PLUTO-2.1 SLM.

1. Bessel–Gaussian Beams: The system successfully generated zeroth-order Bessel–Gaussian beams. When a helical phase was added, the central intensity was suppressed, creating an annular profile with an on-axis phase singularity. This confirmed independent control of radial amplitude and azimuthal phase.
2. Longitudinal Propagation: Measured longitudinal ($y-z$) intensity profiles for Gaussian, Laguerre–Gaussian, and helical Bessel–Gaussian beams matched scalar diffraction simulations, confirming that the encoded complex field is preserved not only in the focal plane but also during propagation through the focal volume.
3. Arbitrary Intensity Patterns: The method was applied to generate arbitrary focal-plane intensity distributions (e.g., text, logos, and specific intensity landscapes like helical ramps and Zernike-tilt profiles).
   • Quantitative Performance: The agreement between target and measured intensities was quantified using the normalized correlation coefficient ($F_{corr}$). Results ranged from 0.86 (for sharp-edged text patterns) to 0.96 (for smooth grayscale patterns like a smiley face).
   • Limitations: The authors note that the lowest correlation (0.86) for sharp features is attributed to diffraction, alignment errors, and finite-resolution effects. The amplitude modulation step inherently attenuates the field (efficiency trade-off), and the system is limited by the SLM's pixel pitch, fill factor, and phase calibration.

Significance and Claims
The paper claims to provide a compact, high-power-compatible route to programmable complex wavefront engineering. By enabling independent amplitude and phase control on a single device, the method offers a practical enabling technology for:

• Structured illumination and holography.
• Optical control of free-electron beams, including the generation of electron vortex beams, aberration correction, and light-based electron lenses.

The authors emphasize that while the method involves practical trade-offs (such as reduced efficiency due to polarization projection), it avoids the complexity of multi-modulator setups and maintains the physical separation of amplitude and phase control found in double-plane approaches, all within a single optical component.