The research, conducted by Melkon Tatlier of the Department of Chemical Engineering at Istanbul Technical University, introduces a concept called desorption-driven sorption cooling and analyzes a representative implementation known as a variable-volume adsorption cooling, or VVAC, system. The core idea is elegant. In every adsorption cycle, desorption of the working fluid from the adsorbent is an endothermic process, meaning it absorbs heat. In conventional adsorption chillers, that heat of desorption is wasted at high temperature because the adsorbent must be heated to release the refrigerant. Tatlier’s proposal flips this arrangement: by reducing the partial pressure of water vapor in the adsorber without raising the temperature, desorption can be induced near ambient conditions, and the endothermic heat it absorbs becomes the useful cooling effect itself.

The mechanism behind the VVAC system is straightforward in principle. The adsorber, filled with a zeolite adsorbent saturated with water vapor, begins the cycle at an initial volume and pressure. Expanding the adsorber volume mechanically lowers the vapor pressure, shifting the adsorption equilibrium toward lower loadings and forcing water out of the zeolite. Because desorption absorbs heat, the surroundings cool down, and that cooling can be delivered to a heat-transfer fluid circulating through heat exchanger tubes inside the adsorber. Compressing the volume back to its original size raises the pressure, promotes re-adsorption, and releases the heat of adsorption, which could serve as a heating effect. Running two or more adsorbers out of phase would allow continuous production of both cooling and heating.

Zeolites were chosen as the adsorbents for a specific thermodynamic reason. Materials such as NaX and NaY exhibit steep adsorption isotherms and large adsorption enthalpies, so even a modest change in vapor pressure produces a substantial change in equilibrium water loading. Tatlier modeled the system using Langmuir isotherm equations fitted to published adsorption data for both zeolites, combined with the ideal gas law to track the water distributed between the gas phase and the adsorbed phase. The calculations assumed isothermal operation at 298 K and instantaneous adsorption equilibrium, deliberately idealized conditions that the author emphasizes represent an upper-bound estimate of the equilibrium desorption response rather than the transient behavior of a real device.

The simulation procedure was iterative. Starting from an initial pressure and volume, the model calculated the gas-phase water content and the equilibrium loading on the zeolite. Expanding the volume produced a new, lower pressure, which in turn shifted the equilibrium loading and drove desorption. The water released from the zeolite then increased the gas-phase amount, partially raising the pressure again, which triggered further re-equilibration. This feedback loop between pressure and loading produced a striking dynamic: the partial pressure and the water content of the zeolite oscillated in successive desorption and adsorption steps, with the oscillations either decaying progressively toward a periodic steady state or persisting at nearly constant amplitude, depending on the operating conditions.

The two key parameters governing performance were the initial water vapor pressure and the ratio of the final to the initial adsorber volume. Lower initial pressures and larger expansion ratios both promoted greater desorption. For zeolite NaX at an initial pressure of 0.5 kPa, increasing the volume ratio from 10 to 50 raised the amount of water desorbed from about 0.13 to 0.25 grams per gram of zeolite. At an initial pressure of 3 kPa, the same increase in volume ratio lifted desorption from roughly 0.040 to 0.14 grams per gram. Zeolite NaY, which is more hydrophobic than NaX and therefore holds less water at a given vapor pressure, desorbed about 10 to 35 percent more water under comparable conditions, suggesting that adsorbents with lower water affinity at the operating pressure may offer greater cooling potential in this framework.

The dynamic regimes themselves carry practical implications. Under some combinations of pressure, volume and adsorbent mass, the system approached a periodic steady state in which progressively smaller amounts of water were exchanged between the gas and adsorbed phases at each step. Under other conditions, notably lower initial pressures, the oscillations persisted, meaning repeated desorption waves could continue to deliver cooling over successive cycles. The author attributes this behavior to the coupling between adsorber volume and vapor inventory: expansion lowers the pressure, while desorption raises it, and the relative strength of these opposing effects determines whether the oscillations decay or persist. The model also showed that increasing the zeolite mass, from 100 to 500 grams, preserved the overall behavior while increasing total water transfer, although the minimum adsorber volume needed to keep the vapor below saturation and avoid condensation grew accordingly.

The thermodynamic accounting is where the concept becomes genuinely provocative. Because the pressure reduction is achieved by changing the adsorber volume rather than compressing a large external gas inventory, the minimum reversible work required to initiate the expansion is small. Tatlier calculated ideal coefficients of performance of approximately 45.5 for a volume ratio of 10 and 96.5 for a ratio of 50, using an initial pressure of 3 kPa and 100 grams of NaX zeolite. These numbers are not directly comparable to practical system COPs, and the author is explicit about this: like a Carnot limit, the ideal COP represents a theoretical upper bound that ignores mechanical losses, finite adsorption kinetics, diffusion resistance, heat transfer limitations, leakage and auxiliary energy consumption. Introducing an overall mechanical efficiency of 50 percent, for example, would halve the ideal value. The more meaningful takeaway, the study argues, is that a relatively small mechanical perturbation of the adsorber can trigger the release of a much larger amount of thermal energy stored in the adsorption equilibrium.

For context, pressure-driven adsorption cooling has been explored before, most recently using CO2 and metal-organic framework working pairs, including hybrid compression-adsorption cycles and experimental demonstrations with MOF-coated heat exchangers. A continuous pressure-driven CO2/MIL-101 system has reported practical COP values of 0.75 to 1.4. Those systems, however, rely on compressors and auxiliary components to generate the pressure swing, whereas the VVAC concept requires only the work associated with changing the adsorber volume itself. The author notes that direct numerical comparison between the two approaches should be made with caution, since the idealized values in the new study serve as indicators of thermodynamic potential rather than predictions of achievable device performance.

The study is intentionally limited to a simplified equilibrium analysis, and significant questions remain before any practical machine could be built. The large volume ratios investigated are parametric conditions exploring the potential operating range, not requirements for a real device, and the actual mechanical implementation would face challenges of piston displacement, dead volume, actuator requirements and durability. The model also does not capture the time scale of the predicted oscillations, since it assumes instantaneous equilibrium rather than explicit heat and mass transfer kinetics. Future work, the author suggests, should investigate gradual or continuous pressure variation with simultaneous desorption and volume change, coupled with detailed heat and mass transfer analyses and experimental validation. Beyond the variable-volume approach, the broader concept of desorption-driven sorption cooling could in principle be realized through humidity swings, membrane-assisted vapor removal or other low-work methods for reducing adsorbate partial pressure, opening a path toward cooling systems that harvest the heat of desorption directly, near ambient temperature, with minimal mechanical input.

Subject of Research: Desorption-driven sorption cooling: a thermodynamic analysis of a variable-volume implementation

Article Title: Desorption-driven sorption cooling: a thermodynamic analysis of a variable-volume implementation

Article References: Desorption-driven sorption cooling: a thermodynamic analysis of a variable-volume implementation. (n.d.). https://doi.org/10.3389/fther.2026.1952638

Image Credits: AI Generated

DOI: 10.3389/fther.2026.1952638

Keywords: Desorption-driven, sorption, cooling, thermodynamic, analysis, variable-volume, implementation, scientific research