HoloVAM: Holographic Volumetric Additive Manufacturing

• Holographic Volumetric Additive Manufacturing (HoloVAM) is a light-based 3D printing method that employs phase-only holography and tomographic design to enable parallel, high-resolution curing across an entire resin volume.
• It integrates a minimal optical architecture with advanced spatial light modulators and iterative computational optimization to achieve photon efficiencies of 20–30% and feature resolutions between 75–110 μm.
• Innovative resin photochemistry and dual-color photoinhibition facilitate precise dose control and material conversion, resulting in compact, high-throughput fabrication platforms.

Holographic Volumetric Additive Manufacturing (HoloVAM) is a light-based three-dimensional fabrication technique that exploits computer-generated holography (CGH) to deposit energy throughout an entire photopolymerizable volume, enabling parallel production of complex mesoscale objects with high resolution, superior photon efficiency, and reduced optical system complexity. HoloVAM unifies tomographic light field design, phase-only modulation, and advanced point spread function (PSF) engineering to focus dose distributions, overcoming core limitations of traditional amplitude-modulated volumetric additive manufacturing (VAM). Innovations in system architecture, computational optimization, and resin photochemistry have positioned HoloVAM as a principal route toward high-throughput, compact, and application-flexible volumetric fabrication platforms (Madsen et al., 5 Dec 2025, Álvarez-Castaño et al., 24 Jan 2024, Álvarez-Castaño et al., 3 Jun 2025, Li et al., 28 Jan 2024, Wang et al., 2023).

1. Optical and System Architectures

HoloVAM is realized through a phase-only spatial light modulator (SLM) or phase-only light modulating platform (PLM), eliminating complex imaging optics and relay lens architectures required in conventional projector-based VAM systems. The canonical system comprises:

• A single-mode 405 nm laser source, expanded and collimated to illuminate the SLM or PLM.
• A phase-encoded hologram displayed on the SLM or PLM, combining CGH for the tomographic pattern and a digital lens in the phase domain to place the SLM in the Fourier plane of the resin volume.
• Free-space light propagation directly into a rotating vial of photopolymeric resin, requiring no relay lenses, index-matching baths, or refractive optics.
• Synchronous rotation of the cylindrical vial on a motorized stage; at each rotation angle θ, a corresponding hologram encoding the projection for that angle is displayed.
• Photon-efficient elements: SLMs (e.g., Holoeye PLUTO-2.1-UV) or MEMS-based PLMs (e.g., DLP6750 EVM) providing ≥92% fill factor, high phase quantization, and compatibility with UV/visible wavelengths.
• Photon efficiency exceeding 20–30% via phase-only encoding, an order-of-magnitude improvement over amplitude-based engines (~1–2%) (Madsen et al., 5 Dec 2025, Álvarez-Castaño et al., 3 Jun 2025).

This minimal optical train (laser → SLM → free-space → resin vial) supports a mechanically compact enclosure (< 20 cm³) and high alignment tolerance (Madsen et al., 5 Dec 2025).

2. Computational Holography and Dose Engineering

At the core of HoloVAM is the generation of tomographic light fields tailored to the three-dimensional target geometry. The workflow comprises:

• Voxelization of the desired 3D target and computation of the Radon transform over discrete projection angles to yield the sinogram.
• Partition of each projection into an array of tiles (the HoloTile approach), with each tile phase-encoded for its local projection. The tiles are synthesized using iterative CGH (e.g., Gerchberg–Saxton) or direct-search optimization, yielding phase maps that, upon Fourier propagation, create the desired lateral and axial intensity distributions (Álvarez-Castaño et al., 24 Jan 2024).
• PSF shaping to extend the axial envelope and confine cure to specified regions. This is commonly achieved by superposing a Bessel-like phase mask

$φ_{PSF}(r') = k R \sin\alpha · r'$

so that

$PSF(r,z) ∝ |J_0(k r \sin\alpha)|^2$

thereby producing an axially elongated, sidelobe-suppressed intensity profile over depth $Z_D \simeq R/\tan{\alpha}$ (Madsen et al., 5 Dec 2025, Álvarez-Castaño et al., 24 Jan 2024).

• Dose accumulation—the time-integrated intensity at each voxel—determines material conversion via a photopolymerization threshold $D_{th}$:

$D(x,y,z) = \sum_{\theta} I(x,y,z;\theta)\,\Delta t$

Cure occurs where $D(x,y,z) \geq D_{th}$.

• Full simulation and loss-based optimization frameworks (e.g., in PyTorch) are used to co-optimize phase (and, in multi-beam systems, amplitude) across all beams or time steps, incorporating forward models of optical propagation and nonlinear material response. Fully coupled optimization achieves lower final errors and increased energy efficiency compared to decoupled projection (Li et al., 28 Jan 2024).
• Compensation for cylindrical refraction at air–glass and glass–resin interfaces is accomplished through pre-warping of projection data, with forward models conforming to Snell’s law (Madsen et al., 5 Dec 2025).

3. Materials and Photochemistry

Resin systems for HoloVAM are typically acrylate-based, with photoinitiators such as TPO at millimolar concentrations. Key parameters:

• Acrylate monomer: e.g., dipentaerythritol penta/hexa-acrylate.
• Photoinitiator: TPO at 3 mM.
• Resin preparation: degassed via ultrasonic bath, dispensed into cylindrical glass vials (diameter ≈12 mm).
• Polymerization threshold: $D_{th}$ of a few mJ/cm², enabling full solidification within 30 s at incident laser powers of 20–30 mW in lensless systems (Madsen et al., 5 Dec 2025, Álvarez-Castaño et al., 3 Jun 2025).

Integration of advanced resin chemistries, such as binary photoinhibition (BPI), enables direct control over negative features and contrast, using dual-color photochemistry (e.g., oxygen–lophyl radical pair) to realize dose subtraction at the voxel level. This approach yields sub-50 μm lateral resolution, improved interior hollowness, and minimized surface roughness (Wang et al., 2023).

4. Performance Metrics and Empirical Results

Comparative analysis against amplitude-based tomographic VAM demonstrates:

System	Projection Efficiency (η)	Lateral Resolution	Axial Confinement	Print Time (cm-scale)
DMD-Amplitude	<0.5%	≥160 μm	Poor	1–2 min
DMD-Lee Hologram	≈5–10%	164 μm	Improved	<1 min
PLM-HoloVAM	24% (phase-only)	75–110 μm	80% over 12 mm	20–60 s
Lensless HoloVAM	20–30%	200 μm	<10% off-plane	25–30 s

Fabricated geometries include feature-rich models (e.g., sprockets, lattices, complex organics) at centimeter scale, with gear teeth resolved to ≈200 μm and consistent reproducibility (±0.1 mm global dimensions) (Madsen et al., 5 Dec 2025, Álvarez-Castaño et al., 3 Jun 2025, Álvarez-Castaño et al., 24 Jan 2024). Time-multiplexed hologram strategies further suppress speckle (contrast C<0.33 with N=9, mean intensity averaging) and yield sub-100 μm surface fidelity (Álvarez-Castaño et al., 3 Jun 2025).

5. Multi-Beam, Multi-Wavelength, and Optimization Strategies

Generalizations of HoloVAM employ:

• Multi-beam approaches, where independent phase and amplitude masks for each beam are jointly optimized under a nonlinear coupled exposure model (e.g., multi-wavelength or two-photon absorption regimes). The general forward model decomposes the exposure as a polynomial of per-beam intensity fields:

$$F = \Phi(I_1,\ldots,I_M) = \sum_m C_m^{(1)} I_m + \sum_{m,n} C_{mn}^{(2)} I_m I_n + \cdots$$

with material response $M(F)$ mapping excitation to conversion.

• Time-multiplexed “holo-tomography,” decomposing a single beam into N virtual projections, enables distributed dose accumulation with a shared amplitude variable.
• Fully coupled optimization of all beams and amplitudes via gradient-based algorithms yields 14–47% lower loss and improved energy efficiency than decoupled operation (Li et al., 28 Jan 2024).

6. Photoinhibition, Dose Subtraction, and Contrast Enhancement

Dose subtraction, implemented via dual-color photoinhibition, addresses lateral contrast limitations fundamental to both TVAM and HoloVAM. Binary photoinhibition (BPI) introduces a photo-inhibitory species, permitting an exact mathematical reconstruction of any greyscale 3D pattern by allowing negative as well as positive effective doses:

$$D(\mathbf{r}) = \int |E_{vis}(\mathbf{r},t)|^2 dt - \int |E_{UV}(\mathbf{r},t)|^2 dt \rightarrow G(\mathbf{r})$$

Experimentally, BPI achieves reduced lateral feature widths (down to 54 μm), >98% interior void fraction, and minimized surface roughness (Rq ≈1.2 μm under dual-color, vs 1.8 μm single-color) (Wang et al., 2023). Integration with holographic phase modulation further enhances feature fidelity and enables advanced resin compatibilities.

7. Perspectives and Future Directions

HoloVAM enables a mechanically minimal, photon-efficient platform for volumetric fabrication. Prospective developments include:

• Scaling to sub-100 μm features via higher-resolution SLMs and advanced PSF engineering.
• Extension to biofabrication and microgravity manufacturing by enabling volumetric cure in turbid or cellular media using engineered beam modes and real-time wavefront correction.
• Adoption of multi-wavelength photochemistry, closed-loop adaptive focusing for resin inhomogeneity correction, and tiling of multiple PLMs for large-format fabrication (Madsen et al., 5 Dec 2025, Álvarez-Castaño et al., 3 Jun 2025, Li et al., 28 Jan 2024).
• Theoretical and algorithmic advances, including multi-material voxelization, direct dose-subtracting CGH solvers, and integration with acoustic manipulation for hybrid assembly (Melde et al., 2022).

The performance advances and system simplifications afforded by holographic volumetric additive manufacturing (HoloVAM) have established it as a foundational platform for next-generation benchtop and field-deployable 3D fabrication (Madsen et al., 5 Dec 2025, Álvarez-Castaño et al., 3 Jun 2025, Wang et al., 2023, Li et al., 28 Jan 2024, Álvarez-Castaño et al., 24 Jan 2024).