Zeolite — Adsorbent Families, Numbers & the Low-Regeneration Pivot

Zeolite is one of the core solid sorbents of Adsorption Cooling — a crystalline porous solid that binds a refrigerant vapour (almost always water) on its internal surface, then releases it when heated. It offers higher uptake, specific cooling power, and thermal stability than silica gel, but classic grades pay for it with a high regeneration temperature that locks out low-grade heat. This article is the sorbent-side companion to Performance & Numbers and Composite Salt Sorbents — what zeolite is, the families, the numbers, and why the action has moved to aluminophosphates.

Zeolite belongs unambiguously to the adsorption side of the sorption split: a solid surface sorbent, never a liquid bulk one — see Adsorption vs Absorption. The refrigerant is almost always water (R718); ammonia and methanol pairs exist but are niche, and every reviewer’s pair-selection criteria (Demir: high latent heat, non-toxic, stable, high capacity and conductivity, cheap, physically reversible) point back at water.

Material families

Family	Examples	Water uptake (kg·kg⁻¹)	Regen / desorption temp	Role
Classic aluminosilicate	13X (NaX, FAU), 4A (LTA), NaY	0.22–0.30	high, ~150–250 °C	High uptake and stability; the historic workhorse, but too hot for waste heat — though Elsheniti notes it can be fired directly by engine exhaust
Natural zeolites	mordenite, clinoptilolite	0.11–0.12	high	Cheap, low capacity; mostly of academic interest
Silico-/alumino-phosphates	SAPO-34, AlPO-LTA, AQSOA Z01/Z02/Z05, EMM-8	~0.22–0.28	low, ~60–90 °C	The low-regeneration pivot — high COP on low-grade solar/waste heat
Other zeotypes studied	ZSM-5 (MFI)	lower	high	Tested for solar adsorption; underperforms SAPO-34 (Yuan)

The distinction that matters is classic aluminosilicate vs aluminophosphate, and it is mechanistic, not cosmetic (see below). The Zotero registry also holds the broader zeolite literature this article does not digest in full — Pérez-Botella’s 2022 Chemical Reviews survey of zeolites in adsorption processes, Corma’s 2003 catalysis review, Freni’s 2015 monograph on zeolite-based coatings for adsorption heat pumps, and Boruta’s 2021 state-of-the-art on working pairs and coated beds.

The numbers

Property	Value	Notes
Water uptake, 13X	0.30 kg·kg⁻¹	4A: 0.22; mordenite/clinoptilolite: 0.11–0.12 (Demir, Table 3)
Heat of adsorption	3300–4200 kJ·kg⁻¹ (Elsheniti); 4400 kJ·kg⁻¹ for 13X and 4A (Demir)	High — the root of the high regen temperature and lower COP; silica gel is ~2500–2800, activated carbon 1800–2000
Energy density, 13X / 4A	~1290 / 1250 kJ·kg⁻¹	Working-pair table, 30–350 °C operating range (Demir)
Desorption range	70–250 °C	Classic grades effectively need ~150–200 °C (Elsheniti; Kumar: “zeolite-water needs ~200 °C”)
Max cooling COP (classic zeolite/water)	~0.30	Anyanwu & Ogueke theoretical maximum; still the best of the classic pairs for A/C duty vs AC/methanol 0.16 and AC/ammonia 0.19 (Al-Yasiri)
Single-bed experimental	COP ~0.25, SCP up to 200 W·kg⁻¹	5 kW, evaporating at 6.5 °C, 1320 s cycle (Kumar)
Mobile double-bed zeolite/water	COP 0.4, SCP 600 W·kg⁻¹	High SCP, compact (Kumar; [[references/adsorption-performance
SAPO-34 benchmark vs MOF	SCP 310 W·kg⁻¹ (SAPO-34) vs 440 (CPO-27(Ni))	130 °C desorption, 900 s cycle (Elsheniti)
Thermal-wave two-bed	COP ≈ 1.0	Internal heat recovery (Elsheniti); heat-recovery cycles add ~25 %, heat+mass recovery >10 % more
Cascade (zeolite/water → AC/methanol)	COP 1.06	Bridges a large temperature gradient (Dous & Meunier, via Demir; Wolak)
Heat + mass recovery on 13X/CaCl₂ composite	COP +125 % vs basic cycle	Tso et al., via Al-Yasiri

Two cycle-level numbers frame the whole table: typical thermal COPs of adsorption systems are ~0.6 (Elsheniti), and a basic single-bed intermittent cycle — the kind most solar zeolite refrigerators used — can fall below 0.1 because of the bed’s large temperature swing (Wolak). Zeolite’s high heat of adsorption makes it the pair most exposed to that swing.

The central drawback — and the aluminophosphate pivot

Classic zeolites bind water by strong electrostatic interaction with framework cations. That gives a high heat of adsorption and therefore a high regeneration temperature (~150–250 °C) — which disqualifies low-grade waste heat and low-temperature solar, the very niche adsorption cooling exists to serve. Kumar’s feasibility study makes the point bluntly: activated carbon/methanol and silica gel/water suit sources below 100 °C; zeolite/water needs ~200 °C.

The fix is the aluminophosphate (AlPO/SAPO) class, which binds water through a hydrogen-bonded water-cluster mechanism instead of strong electrostatics. The result is an S-shaped (Type V) isotherm — a sharp uptake step over a narrow humidity window — and a much lower isosteric heat, so the material desorbs at low temperature yet still delivers high uptake:

EMM-8 (Liu et al. 2022, Nature Communications) is a low-cost zeolite-like aluminophosphate with SFO topology and 12-membered-ring channels (~0.7 nm) — larger than the ~0.4 nm 8-ring windows of SAPO-34 and AlPO-LTA, hence weaker hydrophilicity and easier desorption. Built from cheap precursors (pseudo-boehmite, phosphoric acid) with a BET area of ~880 m²·g⁻¹, it takes up 0.283 g·g⁻¹ at 25 °C with its step at P/P₀ = 0.15–0.17, has an isosteric heat of only 46.76 kJ·mol⁻¹ — barely above water’s 44 kJ·mol⁻¹ heat of vaporisation — and regenerates at 65 °C. Cooling COP reaches a record 0.85 at a 63 °C driving temperature (5 °C evaporator / 30 °C condenser), beating AlPO-LTA (0.75), MIP-200 (0.78), KMF-1 (0.75), and Co-CUK-1 (0.83); as a heat pump it gives COP 1.75 at an 82 °C drive, with 5–15 °C lower driving temperatures than the reference materials. It survived 30 cycles with no measurable uptake loss, 24 h in boiling water, and heating to 700 °C, and was scaled to ~100 g batches. Bed-relevant physical data: pellet packing density 0.93 g·mL⁻¹, thermal conductivity 0.082–0.37 W·m⁻¹K⁻¹ — the low-conductivity problem below applies to it too.
Mitsubishi Chemical AQSOA (Z01/Z02/Z05, aluminophosphate zeolites) is the commercial OEM adsorbent enabling low-regeneration (<90 °C) machines; it is supplied into chillers rather than sold as one (market report). Z02 is the pair in the automotive coated-bed work in Adsorbent Bed Engineering.
SAPO-34 beats ZSM-5 in a sun-tracking parabolic-trough rig charged with 3.17 kg of 5.7 mm granules (Yuan): at 1800 s adsorption it reached the campaign’s best COP 0.169 (SCP 78.9 W·kg⁻¹), and at 600 s its SCP peaked at 169.7 W·kg⁻¹ (COP 0.122) — the two headline numbers come from different adsorption times, because SCP and COP move oppositely with cycle length, and more than two-thirds of the cooling arrives in the first 600 s. ZSM-5 actually showed the larger bed-temperature rise (better heat transfer) but worse desorption, so SAPO-34 won on both metrics at every time. A SAPO-34/water evacuated-tube unit reached COP 0.575 / EER 5.8 at a 75 °C drive for 10 kW (Al-Yasiri). Ammann et al. (IBM/ETH) engineered 75 µm directed mass-transfer channels into SAPO-34 coatings and doubled the water-sorption rate at identical adsorbent mass — water uptake in SAPO-34 coatings is explicitly mass-transfer-limited — see Fractal / Hierarchical Bed Geometry.

Material selection is not obvious. Tatlier’s 2023 paper — “Selection of a favorable zeolite for solar adsorption cooling: how straightforward is it?” (registry metadata only; not ingested in full) — flags the problem in its title, and the EMM-8 work shows the mechanism: what matters is where the isotherm step falls relative to the cycle’s evaporator, condenser, and drive temperatures, not the headline uptake or surface area.

Engineering constraints

Poor heat and mass transfer. Zeolite beds have low thermal conductivity; Demir names it, with intermittency and vacuum, as one of the three core problems of adsorption heat pumps. Mitigations follow the standard bed playbook — coated beds (adsorbent on tubes, fins, or metal foam for fast heat and mass transfer; InvenSor’s direct-coated zeolite beds are the commercial example, and Demir cites copper-foam and hydrothermally grown zeolite supports), plus a 3-D graphene binder for zeolite 13X that improves conductivity, vapour transport, and uptake kinetics (Gildernew 2022, registry). Yuan’s tubular bed shows the same logic in a solar receiver: an axial copper cooling channel plus a reticular mass-transfer channel inside a black-chrome-coated absorber tube. See Adsorbent Bed Engineering.

Vacuum operation. Water/zeolite runs under high vacuum (water must boil at chilled-water temperature), an engineering and maintenance burden that Kumar’s rig identified as its main reliability limit. NH₃/zeolite avoids the vacuum but runs at 4–11× ambient pressure; a binary NH₃+H₂O fluid can run near ambient (Wang & Zhu, via Demir).

Cycle / hydrothermal stability. Repeated hydration–dehydration can degrade frameworks. Fraunhofer ISE’s cycling study — silica gels, zeolites, AlPO/SAPO, and MOFs under continuous water-vapour cycling — finds degradation is front-loaded (most loss in the first cycles) and singles out the Cu-BTC MOF, which lost 40 % of capacity within 15 cycles, as the fragile end; EMM-8’s own data (30 cycles, boiling water, 700 °C) put aluminophosphates on the robust end. Early-cycle screening is the recommended gate.

Collector burden. Among closed solar-cooling cycles adsorption needs the largest collector area — 3.49 m² per kW vs 2.77 for absorption (Henning, via Al-Yasiri) — which is a system cost zeolite’s high drive temperature makes worse unless the aluminophosphate route is taken.

What actually ships

Every commercial adsorption chiller still runs on silica gel or zeolite with a water refrigerant — not lab MOFs: the USD 215.4 M (2025 → 306.8 M by 2035, 3.6 % CAGR) market analysis contains zero MOF mentions, and demand is industrial waste-heat recovery (chemicals, food, oil & gas) led by Hitachi, Johnson Controls/YORK, Fahrenheit, Bry-Air, Thermax, and Kawasaki. See Commercial Adsorption Chillers.

InvenSor GmbH — direct-coated zeolite + water in the 10–30 kW class. LTC 10 e plus: 10 kW, COP 0.52 nominal / 0.75 max, hot water 45–100 °C (72 °C nominal), chilled water 10–25 °C. LTC 30 e plus: 10–35 kW, COP up to 0.75, drive from ~50 °C. HTC 18 plus: 18 kW tuned for high-ambient (>40 °C) recooling, drive from 85 °C, COP ~0.5 — the catalogue illustration that ambient temperature, not just drive temperature, sets zeolite COP. (Datasheets are image-based and distributor-hosted; figures come from a case study and snippets.)
Mayekawa — >70 kW zeolite/water units; also historically Nishiyodo, Macom, and SorTech on silica gel, with locomotive zeolite 13X/water coolers among the vehicle applications (Wolak).
Mitsubishi Chemical AQSOA — OEM aluminophosphate adsorbent (not a chiller).
OxiCool Inc. (Barot et al., US Patent 9,765,998 B2, 2017) — a water-refrigerant two-bed cooler for truck sleeper cabins, desorbed by engine exhaust or a fuel-fired heater, with zeolite as the primary named adsorbent. Its claimed novelty is a “winterization” configuration: near freezing, the controller adsorbs the water into the zeolite’s crystalline structure so that no bulk liquid can form ice — freeze protection without antifreeze — and the two alternating beds give substantially continuous cooling that a single bed cannot.
Façade integration: CoolSkin (Univ. Stuttgart, SFB 1244) embeds a vacuum zeolite/water adsorber behind selective-coated absorber sheet and double glazing as a switchable solar collector, regenerating by day (adsorber ~100 °C) and cooling by night: ~54 W·m⁻² of façade sustained for 12 h in the reference case, ~100 W·m⁻² peak, up to 150 W·m⁻² in a parameter-optimised best case — against a demand of ~60 W·m⁻² residential / ~100 W·m⁻² commercial. Every aesthetic modification tested (folded absorber, ETFE cushions or vacuum panels) reduced cooling power (Greiner).
Historically several firms — SunMark, BLM, Zeopower, EG Solar, Solaref — failed to commercialise intermittent solar zeolite/water units (Wolak), the low-COP single-bed cycle being the likely reason.

Sources

Liu et al. — Ultralow-temperature water-based sorption refrigeration (EMM-8 aluminophosphate) — SFO 12-ring structure, 0.283 g·g⁻¹, 46.76 kJ·mol⁻¹, COP 0.85 at 63 °C, stability, conductivity
Elsheniti — Adsorption refrigeration technologies — 70–250 °C range, 3300–4200 kJ·kg⁻¹, SAPO-34 vs CPO-27(Ni) SCP, thermal-wave COP ≈ 1.0, typical COP ~0.6
Kumar — Feasibility of adsorption cooling — ~200 °C requirement, single-bed and mobile double-bed benchmarks, vacuum as reliability limit
Al-Yasiri — Solar-powered cooling for buildings — Anyanwu & Ogueke COP 0.30, SAPO-34/water COP 0.575, Tso +125 %, 3.49 m²/kW collector area
Yuan — Solar adsorption with concentrated collector (SAPO-34 vs ZSM-5) — COP 0.169 / SCP 169.7 at different adsorption times, tubular receiver bed, ZSM-5 heat-transfer paradox
Greiner — CoolSkin façade-integrated adsorption cooling — vacuum zeolite/water façade, 54 / 100 / 150 W·m⁻², aesthetic-modification penalty
Barot / OxiCool — Adsorption cooling systems patent (US 9,765,998 B2) — winterization in the crystal structure, two-bed continuity, exhaust-driven desorption
Demir — Adsorption heat pump problems & solutions — working-pair table (13X, 4A, mordenite), three core problems, cascade COP 1.06, vacuum vs NH₃ pressure, coated-bed supports
Wolak — Cooling effect of adsorption/desorption — single-bed COP <0.1, commercial deployments (Mayekawa, locomotive 13X), failed intermittent vendors
Fraunhofer ISE — Cycle stability of sorbents — front-loaded hydrothermal degradation across zeolites, AlPO/SAPO, MOFs
Ammann — SAPO-34 directed mass-transfer channels — 75 µm channels, 2× sorption rate at identical mass, mass-transfer-limited coatings
InvenSor LTC/HTC datasheets — direct-coated zeolite chillers, LTC 10/30 and HTC 18 specs
Adsorption chillers market — USD 215 M market, silica gel/zeolite only, incumbents, AQSOA as OEM adsorbent
Zotero cooling registry — metadata for Pérez-Botella, Corma, Tatlier, Freni, Boruta, Gildernew and the wider zeolite literature