Joule Heist search

topics · updated 2026-06-19

Adsorbent Bed Engineering — Geometry, Coatings & TPMS

confidence: medium volatility: warm verified: 2026-06-19fresh

How the adsorber's physical bed — granular, finned, coated, metal-foam, or 3D-printed TPMS — governs specific cooling power. Every conductivity-boosting move buys heat transfer at the cost of mass transfer, inert mass, or manufacturability; TPMS (Lidinoid +17.5% SCP) and conductive additives (+40–65% SCP) are the frontier, bounded by a coating-thickness-squared penalty and a granular-beats-coated-on-volume counterexample.

The sorbent material sets the ceiling on adsorption cooling; the bed — how that sorbent is packed against the heat-exchanger surface — sets how close you get to it. The governing tension is heat transfer vs mass transfer: anything that speeds heat into/out of the sorbent (fins, coatings, metal foam, metal additives) tends to throttle vapor transport, add inert thermal mass, or hurt manufacturability. This article maps the bed-design space and the quantified tradeoffs, extending the brief bed notes in Limitations & Mitigations.

The bed-design taxonomy

A 2024 Renewable & Sustainable Energy Reviews survey catalogs the full spectrum — multibed, multistage, finned, coated, metal-additive, aerogel, metal foam, TPMS, and fluidized beds — and tabulates their enhancement figures. Headline numbers:

Bed approachEffectCost / limit
Metal-additive (Al/Cu/graphite mixed in)up to +40 % SCPadds non-sorbing inert mass → caps COP
Metal foam filled~+50 % adsorption ratepermeability loss, cracking, pressure-drop plateau
Coated (direct-crystallized)~3.5× gravimetric SCP vs granularthickness² heat+mass penalty; binder problem
TPMS (printed lattice)+12.4 % SCP vs finned (Lidinoid +17.5 %)printability/cost; geometry-dependent
Granularcheap, mass-transfer-friendlylow conductivity; but can win on volume

The review’s own conclusion: the best bed combines a composite adsorbent + surface coating + bed-geometry modification — which is why this article pairs with Composite Salt Sorbents.

The coating-thickness-squared law

Coated beds (sorbent crystallized directly onto the metal) give the highest gravimetric SCP — one study measured 0.33 mm coating → 456 W·kg⁻¹ vs 130 W·kg⁻¹ for a finned tube with loose 1.4–2.4 mm pellets (~3.5×). But you cannot just coat thicker: coating thickness adds both a heat-transfer resistance (fluid → sorbent) and a mass-transfer resistance (sorbent → desorbed vapor), and both scale with the square of thickness. That quadratic penalty forces a thin optimum (~0.3–0.5 mm).

The counterexample that matters: coated beds don’t universally win. In one configuration the coating’s volumetric power was actually lower — 93 kW·m⁻³ vs 212 kW·m⁻³ for a granular bed — even while winning on gravimetric SCP. So the “best” bed depends on whether you’re constrained by mass or by volume.

Conductive additives

Mixing a thermally conductive phase into the bed is the cheapest lever. A 2025 open-access Energies study quantified aluminium, copper, and graphite at 5/15/25 wt% in silica gel; a separate experiment found 20 % graphite flake in a coated bed → +65 % SCP and +17 % COP. The catch is the same inert-mass penalty: every gram of non-sorbing metal raises conductivity but depresses COP, so there’s an interior optimum, not monotonic gain. The upper bound of what a well-engineered coating can do shows in an automotive AQSOA-Z02/water study, where a binder-additive-coated tube (BACT) reached ~1875 W·kg⁻¹ SCP at a 400 s cycle off 90 °C coolant — far above the granular benchmark, at the cost of fabrication complexity. A parallel often-overlooked lever is the evaporator: low-pressure water evaporators (flooded, capillary-assisted, falling-film) can reach heat-transfer coefficients up to ~7840 W·m⁻²K⁻¹, and since the evaporator sets the chilled-water approach, it co-limits SCP alongside the bed.

TPMS: which lattice, and why 3D printing

Triply periodic minimal surfaces (gyroid, diamond/Schwarz-D, primitive/Schwarz-P, I-WP, lidinoid) are smooth, periodic lattices with very high surface-area-to-volume ratio. A 2023 CFD study modeled all five as metal adsorber networks vs conventional fins:

GeometryCyclic SCP (W·kg⁻¹)vs fins (357.4)Kinetics
Lidinoid401.6+12.4 % (+17.5 % at own optimal time)slower
Diamond393.6+10 %faster
Gyroid379.3+6 %faster
Primitive—−6.9 %faster

So Lidinoid wins on SCP (its geometry maximizes the metal-to-porous-media conductivity boost) while gyroid/diamond have faster kinetics, and primitive can actually lose — not every TPMS beats fins. Independent heat-sink data adds the flow side: gyroid has ~8 % lower thermal resistance and lower pressure drop than diamond (optimal porosity 0.72–0.77), while diamond cools hotspots better at higher hydraulic cost. TPMS only beats fins at medium/high porosity (0.5–0.8); higher porosity lowers SCP.

The enabler is metal additive manufacturing — gyroid adsorbers have been printed from AlSi10Mg (SLM/DMLS). Plastic-printed silica-gel/water modules are projected to cut manufacturing cost 50–75 % over the next decade, and the EU DYMAN project (see Commercial Adsorption Chillers) is pursuing exactly this — 3D-printed sorbent heat exchangers — as the path to cheaper low-temperature machines. No TPMS-specific COP has been published yet (a gap).

See Also

Sources