Adsorbent Coatings for Adsorption Heat Transformation: From Synthesis to Application

Abstract

In recent years, growing energy demands and environmental pollution caused by the extensive use of fossil fuels have inspired considerable research interest in adsorptive heat transformation (AHT). This technology offers effective utilization of low-grade solar or waste thermal energy for cooling and heating with low environmental impact. Increasing the AHT power is a keystone for further development and dissemination of this emerging technology. The AHT power is mainly determined by ad/desorption dynamics, which is significantly hindered by slow heat transfer between the adsorbent and heat exchanger. Shaping the adsorbent bed as a coating on the heat exchanger surface is considered an effective route to enhance heat transfer and increase the AHT power. In this review, the technology of adsorbent coating for AHT is comprehensively surveyed, including coating synthesis, adsorption dynamics, and use in real AHT devices. The advantages of the coated bed configuration are considered, and its challenges are outlined. Finally, recommendations for better organization of the coating’s structure for rational control of the relative contributions of heat and mass transfer are considered.

1. Introduction

According to the BP Statistical review of world energy [1], primary energy demand increased in 2021 by 5.8%. After the temporary reduction in 2019 due to the COVID-19 pandemic, it exceeded the pre-pandemic level of 2019 by 1.3%. Although this growth was mainly due to an increase in renewable energy, fossil fuel still accounted for 82% of primary energy consumption, resulting in the growth of carbon dioxide emission to 33.9 Gt_CO₂ (+5.9%). Thus, “the world remains on an unsustainable path” [1]. Thermal energy for cooling/heating is one of the largest energy sectors, which accounts for about half (51%) of global final energy demand [2]. For these reasons, in recent years, research related to the utilization of renewable energy for cooling/heating has become one of the hottest topics.

The AHT includes Adsorption Chillers (ACs), Heat Pumps (AHPs), and Heat Storage (AHS). It allows low-temperature renewable and waste heat to be stored and converted into a form suitable for cooling/heating. Due to low electricity consumption and the use of fluids with low environment impact (water, alcohols, ammonia), AHT is considered an energy and environment-saving alternative to compression chillers and heat pumps. Wide AHT dissemination will contribute to achieving carbon-neutral targets, claimed by the mainstream countries [3,4,5]. Dawoud [6] has compared gas-fired adsorption heat pumps with conventional technologies (high- and low-temperature non-condensing and condensing boilers) in terms of primary energy consumption, renewable energy share, and CO₂ emission reduction. It was shown that an AHP providing heating and domestic hot water for a 150 m² building allowed the primary energy consumption and CO₂ emission to be reduced by 39%, with a 21% renewable energy share.

The comparison of AHT with conventional compression technologies revealed that the main drawback of ACs and AHPs is their large size, which is related to low specific power (SP), and high equipment cost [5]. The high cost of about 1000 Euro/kW is compensated by lower operating costs and saves 60–80% of electricity [7]. The first adsorptive chillers of 50–1000 kW power entered the market as early as 1986 from two Japanese companies, Nishiyodo Kuchouki [8] and Mycom [9]. Nowadays, the AC market is mainly concentrated in Germany due to favorable national funding schemes and high electricity costs. ACs are produced by a number of companies, e.g., Fahrenheit [10], Invensor [11], Mitsubishi [12], and others. By 2018, nine European and four international companies offered ACs to the market; however, so far, their share is small as compared with conventional compression chillers. Therefore, there is still much room for AC improvement, one of which is addressed in this review.

The AHT is based on the reversible adsorption of the working fluid on the adsorbent (Figure 1a), as described in detail elsewhere [3]. A typical closed AHT cycle driven by temperature variations comprises two isobaric stages (adsorption 4-1 and desorption 2-3) and two isosteric stages (heating 1-2 and cooling 3-4) (Figure 1b) [13]. Two important performance indexes of the AHT unit are the Coefficient of Performance (COP):

COP = Q_us/Q_re = Q_us/(Q_d + Q_s),

(1)

and the Specific Power (SP):

SP = Q_us/(t_c m_ad),

(2)

where Q_us is the useful heat; Q_re is the external heat supplied, equal to the sum of the desorption heat Q_d and sensible heat Q_s for heating the adsorbent during stages (1–2) of isosteric heating and (2–3) of isobaric desorption (Figure 1b); t_c is the cycle duration; m_ad is the adsorbent mass. Under AC mode, the target cooling effect (Q_ev) is produced in the evaporator, and Q_us = Q_ev. Under AHP mode, the heat released in the condenser (Q_con) and adsorber (Q_ad) during the adsorption stage are used for heating, Q_us = Q_ad + Q_con. Since the ad/desorption stages are much longer than isosteric heating/cooling, ad/desorption dynamics is the crucial factor that defines the SP of an AHT unit. The adsorption dynamics are governed by coupled heat and mass transfer (HMT) in an Adsorbent (Ad)—Heat Exchanger (HEx) unit (Ad-HEx). A number of factors, including the adsorbent bed configuration, the diffusivity of the refrigerant, bed thermal conductivity, and heat transfer between the Ad and HEx, determine HMT in Ad-HExs for AHT [3,14,15].

A simple and usable Ad-HEx configuration, referred to as a granulated bed, comprises loose adsorbent grains placed between the metal fins of the HEx (Figure 2a). Such beds of a small thickness are characterized by fast intergrain mass transport due to voids between the grains, which act as transport pores. For instance, for loose grains (0.4–0.9 mm size) of a composite sorbent LiCl/MWCNT loaded into a finned flat-tube HEx, it was shown that the water sorption dynamics did not depend on the bed thickness H at H < 30 mm [17]. Hence, the intergrain diffusion did not affect adsorption dynamics. At further increasing bed thickness, the resistance to mass transfer in the bed rises, which can decelerate adsorption dynamics. Thus, ethanol adsorption on loose grains of activated carbon SRD 1352/3 in a finned flat-tube HEx was decelerated at H ≥ 40 mm due to the increased intergrain mass transfer resistance [18]. It should be noted that the bed thickness, at which the transition between the intergrain and intragrain mass transfer regimes occurs, depends on many factors, such as the adsorbent nature, its grain size, refrigerant pressure, etc.

The poor heat transport in granulated beds (Figure 2a) is deemed to be the main obstacle to obtaining high heat conversion power in AHT systems. The reason is a small heat conductance between loose adsorbent grains as well as between the grains and the HEx surface through which heat is supplied/removed. It was reported for the adsorption of gases on various adsorbents [14] and for vapor adsorption in adsorption systems for heat conversion [15,22].

An efficient way to overcome this obstacle is heat transfer enhancement by preparing a compact adsorbent layer consolidated with the HEx surface (Figure 2b–d). The first attempts to synthesize compact adsorbent beds for AHT date back to the end of the last century. Guilleminot et al. [23] studied consolidated adsorbent beds composed of zeolite A, silico-aluminate binder, and metal foam as heat conductive additives. Knocke and co-authors studied zeolite coatings of 0.2–3.5 mm thickness prepared on a metal surface with a metal matrix and pore-forming additives to improve heat and mass transfers [24,25,26]. The authors of [27] compacted a zeolite bed by oxidative polymerization of aniline onto the zeolite grains. A grown, thin, heat-conductive film increased the bed’s thermal conductivity by a factor of 4. Then, the bed density was increased by mechanical pressing, which gave additional heat transfer enhancement by 30%. Already the first studies have shown the promise of using coatings to improve the heat conductivity in AHT units (see Section 3 for more detail).

However, they also revealed the flip side of the coin of the heat transfer enhancement in coatings which is the reduction of mass (vapor) transport, which can become a new process limiting the adsorption rate [14,15]. For instance, it was shown that despite the acceleration of heat transfer in the zeolite coating on copper foam, the overall characteristic time of propane adsorption was increased from 15 to 50 s due to the decrease in the bed’s permeability [23]. Consequently, a smart compromise between the heat and mass transfers is needed for the optimization of the Ad-HEx configuration to accelerate adsorption and enhance the SP of AHT units (see Section 4).

In this review, the recent advancements in the adsorbent coating technology for AHT are comprehensively surveyed, including the coating synthesis (Section 2), the influence of the coating parameters on the heat and mass transfer characteristics (Section 3) and adsorption dynamics (Section 4), and real AHT units employing adsorbent coatings (Section 5). Although the coating synthesis routes are similar for closed and open AHT systems, the adsorption dynamics and system configurations differ significantly. In this review, we focused on closed AHT systems, whose advantages and challenges will be outlined for various coating configurations. Coated HExs for open air-conditioning and desiccant systems are out of this review and surveyed, e.g., in [28]. Finally, we discuss some prospects for designing the structured adsorbent coatings, optimized for the enhancement of both heat and mass transfer.

2. Synthesis Routes

To date, three basic approaches for the preparation of coated Ad-HExs units have been developed, namely, a post-synthetic fabrication of binder-based consolidated adsorbent bed (Figure 2b), adsorbent coating on the HEx surface (Figure 2c), and in situ growth of binder-free adsorbent coating (Figure 2d). All the methods are employed for a large number of adsorbents with different chemical natures. Their main features are described below.

2.1. Consolidated Adsorbent Bed

Consolidated adsorbent bed configuration is accommodated between fins/channels of the HEx (Figure 2b). To prepare the consolidated bed, a paste composed of adsorbent grains and binder-based suspension is deposited on the HEx surface to form quite a thick layer, followed by thermal drying [19]. To further enhance the thermal contact between the grains and between the grains and the metal surface, the layer can be additionally compressed [29]. Heat-conductive additives such as expanded graphite [29] and metal ribs [30] can also be used. The consolidated adsorbent beds are characterized by the largest thickness of up to several millimeters. For this reason, slow mass-transfer in the bed could strongly hinder adsorption dynamics. Typically, common polymers (polyvinyl alcohol (PVA) [29]), clays (bentonite [19], expanded graphite (EG) [31]) are used as binders. The binder can partially block the pore space, which reduces both adsorption uptake and dynamics. Pal et al. used polymerized ionic liquids (PIL) as a binder for the preparation of a consolidated bed of activated carbon Maxsorb III [32] that allows the specific surface and pore volume of the composite to be increased by 11% and 18%, respectively, compared with PVA as a binder. Owing to the affinity of the ionic liquid to ethanol, the adsorption uptake of the composite was 22% higher than that of the parent Maxsorb III, while the thermal conductivity was 85% higher. Similarly, the consolidated bed composed of Maxsorb III and ionic liquid demonstrated 14% higher specific volumetric CO₂ adsorption than the Maxsorb III–PVA consolidated bed [33]. EG was shown to serve as both a binder and a heat transfer additive [31].

In sum, the consolidated adsorbent bed is relatively easy to implement; however, there are some inherent issues that limit its practical use for AHT:

• a small amount of binder does not provide the proper strength of the bed; its larger amount can reduce the bed adsorption capacity;
• high density and large thickness of the bed, especially synthesized by compressing method, can reduce mass transfer.

2.2. Binder-Based Adsorbent Coatings

For the fabrication of binder-based adsorbent coating, the adsorbent powder is mixed with a binder and a solvent to obtain a slurry, which can be further deposited on the HEx surface by several methods [34]. The binder-based coating is a mature technique widely applied to obtain films of various adsorbents, including zeolite SAPO-34 [7], MOFs—aluminum fumarate [35], MIL-101(Cr), and HKUST-1 [36], MIL-160 [37], MIL-100(Fe) [38], composite sorbents LiCl/(Wakkanai siliceous shale) [39], etc. Various adhesive compounds of organic (PVA [35,40], polyvinylpyrrolidone [40,41], hydroxyethyl cellulose [35], silicones [36,42], silanes [20,43], gelatine [40], etc.) and inorganic (bentonite [44], sepiolite [45], pseudoboehmite [46], silica sol [38,47], etc.) nature can be used as binders. The binder serves both to bond the adsorbent particles with each other and to provide good adhesion between the adsorbent coating and the metal surface. Organic binders usually ensure better adhesion, mechanical strength, and hydrothermal stability of the film, while inorganic ones enable better thermal conductivity [47]. The metal substrate is preliminarily degreased in alkalis, acids, acetone, or alcohols. The substrate surface can also be roughened to increase the adhesion between the metal and the coating. Various additives, like carbon fibers, can be used to reinforce the layer [44]. Below, the main synthetic routes to prepare the binder-based coatings on the HEx surface are briefly described.

Dip-coating is a conventional technique used to coat surfaces with various geometries, including finned tubes, finned flat tubes, and microchannel HExs. In this method, the preliminary cleaned and treated HEx is dipped into the suspension of adsorbent in the binder solution, taken out, and dried to obtain a uniform film on the HEx surface. Due to the excess of the suspension at the HEx bottom, the coating there is usually thicker than at the top. The coating thickness depends mainly on the suspension viscosity and the speed at which the substrate is taken out of the suspension; thorough control of these parameters is required. To increase the film’s thickness, successive dipping/drying cycles are applied. Such coatings are characterized by moderate thicknesses from ≈15 μm to ≈1 mm [35,43,44].

A binderless MIL-101(Cr) coating was prepared by the dip-coating method using Cu foam as substrate [48]. MIL-101 powder was uniformly deposited in the three-dimensional pores of copper foam with a volume fraction of up to 90%. Vivekh et al. developed composite polymer adsorbents based on LiCl embedded in a polymer matrix, such as PVA [49] and sodium polyacrylate [50]. The composite sorbents were coated on aluminum finned-tube HEx by the dip-coating method to reach the layer thickness of 0.20–0.25 mm. The polymer served both as a matrix for LiCl loading and a binder that ensures the adhesion between the coating and the HEs surface. The water vapor uptake involves several mechanisms:

• adsorption of water on the polymer surface;
• LiCl interaction with water vapor resulting in the LiCl deliquescence and water vapor absorption;
• repulsive forces between ions cause the polymer swelling that circumvent the solution leakage. Such a combination of different mechanisms allows a superior sorption capacity of 1.8–2.5 g/g at RH = 80%.

Drop-coating is often used to coat small surfaces [36,51]. An adsorbent–binder–solvent suspension is poured on the substrate surface as drops and dried without further spreading. The drop-coating can also be applied for larger HExs; however, only ones having quite simple geometry. In this case, the HEx is placed in contact with liquid slurry falling from above, followed by drying. The main advantage of this method over dip-coating is a smaller amount of slurry needed, which can reduce the cost of the coating process. The coating thickness is strongly affected by the suspension viscosity, so this parameter has to be precisely controlled. Multiple coatings can be applied to increase the film thickness. A moderate coating thickness in the range of 100–800 μm is typically achieved [36,52,53]. The coatings formed using this method are often characterized by uneven thickness because of inconsistent drying conditions and uneasy control.

Spin-coating is a modification of the simple drop-coating technique, in which slurry dropped on the support is distributed by centrifugal force. It allows for obtaining a more uniform film but can only be used on flat surfaces.

Electrospinning is a technique for the synthesis of porous microfibres, which can be used for the immobilization of the adsorbent/polymer composites on the substrate surface [54]. An adsorbent/ polymer precursor solution is loaded into a syringe and spun by means of the action of electrostatic forces on an electrically charged jet to form the fibers on the support surface. It can be used for coating HExs of various geometries. The polymer binder enables good contact between the adsorbent grains and the support, which is expected to improve heat transfer. The morphology of coatings in the form of the adsorbent grains embedded in polymer microfibres with large transport pores between them promotes fast vapor transfer to the adsorbent grains. Coatings composed of (silica gel)/polyacrylonitrile [54] and SAPO-34/polyacrylonitrile composites [55] on metal support were prepared. It was shown that they were mechanically stable and possessed good hydrothermal stability under conditions of typical AHT cycles. The electrospinning process and polyacrylonitrile binder did not deteriorate the adsorption ability of the adsorbents.

Spray-coating is similar to the drop-coating technique, but the suspension is sprayed on the HEx surface by a spray gun. This method requires a lesser amount of the suspension compared with common dip-coating, prevents blockage of the fin space, and allows uniform coatings to be obtained. Spray-coating can be applied for both flat plates and finned-tube or channel HExs [56]. However, the use of this method can be restricted by the depth of the suspension flow inside the channels, which can result in the formation of an irregular and non-homogeneous film [34]. Usually, a slurry containing the adsorbent, binder, and solvent is used for spaying [7,57]. Ge et al. [56] developed a binderless spray-coating method, where the suspension of fine aluminum-based MOFs (MIL-96 and MIL-100) powder in water was sprayed on the HEx surface. Zhu et al. [57] used an innovative electrostatic spray-coating method to prepare a vehicle radiator coated by a composite CaCl₂/zeolite. Using an electrostatic field along with the spraying technique affords the following advantages: evenly distributed and strongly adhered coating and a well-controlled film of 10 to 500 μm thickness.

Thus, binder-based coating is a common, scalable, and inexpensive technique for the preparation of compact Ad-HExs. The coating thickness can be varied over a wide range which affords to control HMT in the coating. The main drawback of binder-based coatings is the possible deterioration of their adsorption properties due to:

• lower adsorbent content in the coating;
• partial blockage of pores by the binder deposited on the grain surface, which can reduce the adsorption capacity and hinder mass transfer. The first effect is minor at a small binder content of 5–10%; however, the content might not be sufficient for good adhesion and coating strength.

The latter strongly depends on the binder viscosity and adhesion, adsorbent porosity, etc. Kummer et al. [36] described the fabrication of HKUST-1 and MIL-101(Cr) coatings with a silicone binder. The HKUST-1 methanol adsorption capacity was not affected by the binder, while that of MIL-101(Cr) was slightly reduced (by 10%). Li et al. studied various binders for preparing silica gel coatings [45]. Using epoxy and gelatine binders (ca. 10 wt.%) resulted in a strong reduction of the specific surface area from 680 m²/g for pure silica to 48 and 286 m²/g, respectively. Hydroxyethyl cellulose enabled a moderate decrease in the specific pore volume and surface area. Thus, the proper selection of the binder and its content is a crucial factor for the synthesis of adsorbent coatings with optimized HMT and adsorption ability.

2.3. In-Situ Synthesized Coatings

In-situ coating synthesis implies direct crystallization of the adsorbent grains on the HEx metal surface. Since a pure adsorbent is grown on the HEx surface, this method does not require binders, which allows for avoiding the reduction of the adsorption capacity of the coating due to the pore space blockage. Such coatings are characterized by perfect contact of the adsorbent crystals with the HEx surface, which results in a great reduction in the thermal contact resistance. The in-situ synthesis can be applied for the coating of crystalline adsorbents such as zeolites [58,59], pure or metal-substituted aluminophosphates [59,60], and MOFs [61,62]. Direct coating synthesis implies the creation of conditions for preferential nucleation and growth of the adsorbent crystals on the metal support surface but not in the bulk of the solution. This can be achieved in several ways, including various methods of seeding, substrate heating, and thermal gradient.

Bonacorsi et al. [59] used the preliminary seeding of the metal substrate with the nuclei of two zeolites, Z and 4A, by immersing the substrate in the colloid suspension of the precursor of submicron size. Then, the substrate was placed on the bottom of the reactor, the reactant solution was poured into the reactor, and the synthesis was carried out. Under the synthesis conditions, the solution pH was higher than the point of zero charge, and the metal surface (copper and stainless steel) in the aqueous solution was covered by negatively charged M-O⁻ groups. This induced the specific adsorption of Na⁺ ions from the reaction mixture and enabled the nucleation of zeolite crystals on the surface. The preliminary seeding of the substrate reduced the nucleation time and accelerated the zeolite crystal growth on the surface against the solution bulk. For SAPO, the favorable conditions for the surface crystal growth were reached when using aluminum as substrate due to the formation of a stable oxide/hydroxide layer on the surface, which promotes preferential surface nucleation and crystal growth.

Substrate heating is another method for direct zeolite coating. In this method, the substrate immersed in the reaction mixture is heated directly while the mixture is maintained cold. The temperature difference promotes surface zeolite crystallization and slows down the reaction in the solution. Schnabel et al. [58] synthesized zeolite A and X coatings on stainless steel plates by the substrate heating method. The film thickness of 38 and 230 μm was achieved for zeolite A and X, respectively. The coatings showed good mechanical stability and remained intact after long kinetic measurements.

Jeremias et al. [61] synthesized an HKUST-1 (copper trimesate) coating on a Cu substrate using a thermal gradient method. The substrate was immersed into the reactant solution (copper salt and trimesic acid in dimethylformamide) and heated to the reaction temperature while the bulk solution was intensively chilled. Thus, the synthesis conditions are created only near the substrate surface, which enables the surface crystallization of the MOF. The coating consisted of microporous crystals of less than 10 μm size with macropores of 30–50 μm size between them. The film of 100 μm thickness showed good mechanical stability. Homogeneous coatings were prepared on a plate aluminum support and 3D structures (foams, fibers).

The so-called tailored partial support transformation method involves the crystallization of zeolites (AlPO-5, SAPO-34, SAPO-18) on the surface of the aluminum substrate, which serves as both a support for the coating and a source of aluminum for the zeolite syntheses [63]. Other reactants and the structure-directing template are supplied from the solution. The synthesis is carried out under hydrothermal conditions. Coatings with high mechanical stability and remarkable water adsorption capacity (up to 0.2 g_water/g_composite) were obtained. A similar method was used for the preparation of Al-based MOFs (MIL-96 and MIL-100) coatings on an aluminum substrate [56].

The thickness of the in situ synthesized coatings is determined by two competing processes, namely, crystal growth at the liquid/substrate interface and simultaneous crystal nucleation and accretion in the bulk solution. At increasing reaction duration, the nucleation and crystal growth start in the solution bulk, which competes with the surface crystal growth [59]. Some efforts are made to hamper the nucleation and crystal growth in the solution or to accelerate the nucleation and growth at the interface (e.g., selective heating of the metal substrate, preliminary substrate treatment, aluminizing, pre-seeding of the metal surface) [58,61,64]. For this reason, the thickness of the coating obtained by the in situ synthesis is essentially smaller as compared to other methods and usually does not exceed 100–300 μm. The directly synthesized coatings are attached to the substrate by chemical bonds, which results in good adhesion between the surface and the film and reduces the contact thermal resistance [58,61]. Simultaneously, the absence of a binder enables higher uptake and faster mass transfer.

In sum, both binder-based and binderless coatings possess their own advantageous and drawbacks. Using a binder, it is possible to obtain durable coatings with a thickness variable in a wide range. The main drawback of these coatings is the reduction of the adsorption capacity due to binder additives and partial blockage of the adsorbent pores. The direct binderless synthesis of zeolites and MOFs circumvents the adsorption ability reduction. However, it is more complicated and expensive and results in a limited coating thickness. On the one hand, a small coating thickness leads to high values of the S/m_ad ratio that promotes rapid adsorption. On the other hand, it enlarges the ratio m_HEx/m_ads of the HEx mass to the adsorbent mass and, consequently, reduces the COP due to the rising impact of the sensible heat Q_s (Equation (1)).

3. Effect of Compact Bed Parameters on Adsorption Dynamics

Commonly, the interaction of vapor molecules with an adsorbent surface is very fast [14], and the overall adsorption rate is determined by HMT in the Ad-HEx unit [22].

Table 1. Characteristics of HMT for different adsorbent beds. Thermal conductivity (λ), wall heat transfer coefficient (h), gas permeability (A), thickness (l).

Adsorbent/ Binder/Additive	λ, W/(m K)	h, W/(m2 K)	A, m2	Remarks	Ref.
Granulated beds
Zeolites	0.056–0.12	20–45	10−9–10−11	-	[23,27,65]
AQSOA materials	0.1–0.2	-	-	-	[66,67]
Silica gel	0.11–0.26	-	-	-	[31,45]
Active carbon	0.14–0.17	-	-	-	[31]
Maxsorb III	0.066	-	-	-	[32,33]
Al-fumarate	≈0.12	-	-	-	[62]
MIL-101 (Cr) powder	0.06	-	-	-	[48]
Binder-based adsorbent coatings
Zeolite 13X/polyaniline	0.15–0.26	-	-		[27]
Zeolite/polymer binder		230	-	l = 0.7 mm	[68]
AQSOA Z01/organic binder	0.37	-	-	l = 0.3 mm	[67]
Silica/polyvinylpyrrolidone	0.26	-	-	-	[41]
Ti-substituted SAPO-34/SilRes binder		200–500	-	l = 0.16-0.47 mm	[42]
SAPO-34/clay binder		600–700	-	l = 0.38–11.1 mm	[44]
LiCl-silica composite/glue, silica sol binder	3.2–5.8	-	-	mad/S = 0.4–0.48 kg/m2	[47]
MIL-101(Cr)/Cu foam	0.38–0.86	-	-	Dip-coating without binder	[48]
Consolidated beds
Maxsorb III/PVA/EG	0.15–0.74	-	-	Compressing, l = 15–17 mm	[29]
Maxsorb III/PVA	0.099	-	-	-	[32,33]
Maxsorb III/PIL	0.122	-	-	Increased specific surface area, pore volume, adsorption capacity as compared to PVA	[32,33]
Zeolite 13X, 4A/silico-aluminate gel	0.36	45	-	Pressing, l = 25 mm	[23]
Zeolite 13X, 4A/silico-aluminate gel/Cu, Ni foam	1.7–8.3	110–180	2 × 10−13	Pressing, l = 25 mm, the Cu effect is stronger than Ni	[23]
Silica gel/EG	10–20	-	(3–40) × 10−12	Pressing, l = 4 mm,	[31]
Zeolite CBV 901 (Y)/ pseudoboehmite	0.4	55	10−13–10−12	l = 5 mm, effect of pore-forming additives	[69]
Zeolite/EG	5–10	500–1000	2 × 10−13		[65]
In-situ synthesized coatings
HKUST-1/Cu support	1.2–1.4	(3.5–3.8) × 104	-	l = 0.1 mm	[61]
Zeolites 4A, Y, SAPO-34/ Al, stainless steel support	0.15	>1000	10−8	l = 0.02–0.08 mm
SAPO-34/Al fibers	8	>1000	-	l = 2–4 mm *	[70]
SAPO-34/Al foam	14	>1000	10−8	l = 2–5 mm *	[70]
Al-fumarate	0.31–0.33	≈2000	-	l = 0.28 mm	[62]

* Thickness of the composite Ad/(metal foam).

shows that the transition from granulated beds to coatings leads to better heat conductance and worse vapor permeability.

The overall effect of bed compaction on the adsorption dynamics is, however, unclear. Additional uncertainty is related to the facts that

• heat and mass transfer in porous media are strongly coupled with each other since adsorption releases a large amount of heat, and
• the macroscopic heat transfer parameters in Table 1 (h and λ) were measured under quasi-equilibrium conditions and may differ from those in real AHT systems [71].

Both of these issues are discussed below. In our opinion, the most reliable is direct experimental detection of the acceleration effect. This Chapter summarizes literature data on the dynamics of water and methanol adsorption in coatings/(consolidated beds) prepared for AHT applications. The aim of this analysis is to understand how the coating parameters affect the adsorption rate, correctly compare the dynamics in coatings and granulated beds, and reasonably formulate their advantages and shortages.

To our knowledge, the kinetics of adsorption in coatings on a metal support was first investigated by Knocke and co-authors, both numerically and experimentally [24,25,26]. The water vapor adsorption in a zeolite layer consisting of loose grains (layer thickness 3.5 mm, grain diameter 1–2 mm) was compared with three coated configurations prepared only with a binder, with the pore-forming agents (melamine, tartaric acid) to help mass transport, and, finally, with a honeycomb metal matrix to intensify heat transfer. More detailed experimental studies on the dynamics of adsorption in coatings have been carried out this century by the Large Pressure Jump (LPJ) and Large Temperature Jump (LTJ) methods. The features of such measurements are considered below. Then, the effect of various factors (layer thickness, binder content, contact resistance) on the adsorption kinetics in coatings is surveyed. Finally, a new experimental technique for analyzing the relative contributions of heat and mass transfers to the overall adsorption dynamics is considered. Examples are given of how this tool makes it possible to indicate the limiting stage of adsorption and accelerate it by means of a special arrangement of the coating.

3.1. Features of Experimental Studying Adsorption Dynamics in Compact Layers for AHT Units

Usually, adsorption kinetics is studied under quasi-equilibrium conditions by a small deviation of the system from the equilibrium, recording the system response and analyzing it, e.g., by the so-called Isothermal Differential Step method [14]. The basic heat transfer parameters—the thermal conductivity λ and the wall heat transfer coefficient h—are also measured at quasi-equilibrium conditions. However, in a real AHT unit, the adsorption conditions are quite different, as the adsorption is initiated either by a large jump in vapor pressure over the layer (pressure-initiation) or a large drop in temperature of the metal support, on which the layer is mounted (temperature-initiation). Both result in a large deviation of the system from adsorption equilibrium so that the actual values of λ and h can differ from the quasi-equilibrium ones [71]. In this case, direct application of the LTJ/LPJ method is recommended. For the correct study, the applied pressure jump or temperature drop has to be as close as possible to those in the studied AHT cycle. It was this idea that formed the basis of the LTJ/LPJ approach [15]. In the first studies on adsorption dynamics in coatings, the LPJ method was used in a different way. The applied pressure jump was usually between vacuum and high final vapor pressure (up to 50 mbar), which was chosen arbitrarily and did not correspond to any particular AHT cycle [68].

The correct dynamic comparison of compact and granulated beds needs not only proper choice of the initial and final temperatures and vapor pressures but also fixing similar geometrical characteristics of the bed’s configurations. Since many significant bed parameters can change at once, it is convenient to search for universal parameters that characterize the bed in more general terms. The authors of [72], following ref. [73], suggested comparing compact and granulated beds at equal values of the ratio S/m = <heat transfer surface area>/<adsorbent mass>. Further application of this approach [74,75,76] confirmed that it is very useful for correct elucidation when coatings provide better dynamics than common beds composed of loose adsorbent grains.

3.2. General Remarks on HMT in Compact Adsorbent Beds

Here we briefly discuss some useful issues of how HMT can affect adsorption dynamics in compact adsorbent beds. The models considered below can be used to determine the transport mechanism that controls the overall adsorption rate and to qualitatively analyze the crossover between limiting transport regimes.

3.2.1. Mass Transfer in Porous Adsorbent Beds

The mass transfer resistance of the porous layer is associated with the vapor transport to the adsorbent surface through the pore system. If the layer is obtained by gluing adsorbent grains, which have an intrinsic porosity, two pore subsystems can be distinguished:

• a system of macro/mesopores that provide vapor delivery to the outer grain surface (transport pores) and
• meso/micropore system, through which the vapor reaches the adsorption sites inside the pores, where it is adsorbed (reaction pores) [14].

The relative contributions of the intergrain and intragrain mass transfer resistances in the adsorbent bed composed of grains of size d_g (Figure 3a) can be briefly characterized by the ratio α = τ_inter/τ_intra of the appropriate characteristic times. In its turn, these times are determined by the pore size and the vapor diffusivity D. The time τ_inter = L²/D_inter gives the time scale of vapor diffusion through the layer of thickness L (here D_inter is the vapor diffusivity in the intergrain transport pores). The time τ_intra = R²/D_intra describes the vapor transport inside the primary grains of the average radius R = d_g/2 (here, D_intra is the vapor diffusivity in the subsystem of the reaction pores) [77].

The limiting case of α = (L²/D_inter)/(R²/D_intra) >> 1 is realized for tiny adsorbent grains (small d_g) or/and thick layers (large L) (Figure 3a). As the equilibrium inside the grains is fast, the adsorption is kinetically controlled by the vapor diffusion in the layer voids. This leads to the formation of an adsorption front and its gradual propagation inside the layer (Figure 3b). In the opposite limit α << 1 (large grains or/and thin layers), the vapor pressure inside the intergrain voids is constant and equal to that above the layer. As a result, the adsorbate concentration is almost uniform within the entire layer (Figure 3d), and the water adsorption by the single grain determines the adsorption rate.

It is worth mentioning that the above estimation assumes isothermal conditions and ignores the relative vapor populations in inter- and intragrain pores [14]. However, it can be used for a qualitative illustration of a crossover between the two limiting regimes, which occurs at α ~ 1 (Figure 3c). In practice, this crossover can be intently managed by varying the layer parameters, as demonstrated in ref. [77]. For instance, the quasi-uniform water adsorption mode transforms into the front-propagation mode when the grain size reduces from 0.25–0.5 mm to 0.04–0.056 mm (Figure 4). The binder content C is another tool to manage the crossover between limiting transport modes. Its decrease from 20 to 2.5 wt.% causes the gradual transformation of the front mode to the uniform adsorption mode [77] (not presented).

The contribution of intergrain vapor diffusion can be enhanced by using various pore-forming additives such as melamine, tartaric acid, ammonia bicarbonate, etc., which can intently create transport pores inside the compact adsorbent bed. These organic components, being added to the mixture of adsorbent, water, and binder, are then decomposed and removed during heating at high temperatures. This can lead to a significant increase in the macropore fraction and an enhancement of the mass transfer inside the modified compact bed compared to the reference non-treated bed [24].

Another type of pore-forming agent is presented by pyrolysable additives, which burn out during firing at high temperatures. Examples of such pore-forming agents are wheat particles, starch, PMMA, poppy seed, yeast, and sawdust [78,79,80]. They are used for preparing porous solids with controlled microstructure for a variety of applications such as membrane separation, thermal insulation, catalyst, etc. However, for AHT, the potential of this approach has not yet been fully revealed and requires systematic research.

3.2.2. Heat Transfer to and Inside the Adsorbent Bed

As noted above, the intensification of heat transfer was one of the main motivations for using the adsorbent in the coating form. Indeed, poor heat transfer in the beds composed of loose grains is deemed to be a crucial factor for heat transformers based on solid sorption technology, and its improvement is a key point [22].

The overall heat transfer coefficient A in the system “heat transfer fluid—metal support — adsorbent”:

1/A = 1/h_f-m + l_m/λ_m + 1/h_m-Ad + l_Ad/λ_Ad

(3)

depends on four parameters: the heat transfer coefficients h_f-m and h_m-Ad on the fluid and adsorbent sides of the support and the thermal conductivities λ_Ad and λ_m of the adsorbent bed and metal, respectively [22]. Here l_m and l_Ad are the thicknesses of the support and the adsorbent bed. For complex HExs, more elaborated expressions are used, which account for the efficiency of heat transfer through the metal HEx skeleton (see [81] for straight fins of rectangular profile, means for a common finned flat-tube HEx).

The dependence of the thermal conductivity λ_Ad of the adsorbent bed on its porosity was studied under quasi-equilibrium conditions in many works and summarized in many books (see [82,83,84,85]). In a compact bed, the volume of voids is significantly reduced compared to a granulated one. This leads to an essential increase in the bed’s thermal conductivity since heat is better transferred through the solid phase than through the gas (vapor) phase. According to the literature, the thermal conductivity of granulated beds commonly ranges from 0.05 to 0.25 W/(m K) (Table 1) [23,32,33,66,67]. For adsorbent coatings and consolidated beds shaped with organic binders, it raises to 0.1–0.4 W/(m K) [27,32,33,83]. The coating, prepared of LiCl/silica composite with a commercial glue and silica sol as the binders, showed thermal conductivity as high as 3.2–5.8 W/(m K). In situ growth of adsorbent on the surface of the metal support ensures the intimate contact between the adsorbent crystals and the metal surface, which offers a significant growth of thermal conductivity and, especially, of the wall heat transfer coefficient, which exceeds 1000 W/(m² K) for MOF HKUST-1, zeolite and SAPO-34 coatings [61,70].

In contrast, the vapor is better transferred over voids between the grains, so it worsens in the compact bed. Indeed, the bed permeability reduces by several orders of magnitude—from (10⁻¹¹–10⁻⁹ m²) to (10⁻¹³–10⁻¹² m²) for granulated and consolidated beds, respectively (Table 1). This can lead to a crossover from heat transfer to mass transfer limitation mode [19,30]. For the overall acceleration effect, a smart trade-off is needed between improving heat transfer and worsening mass transfer to obtain a positive overall effect, that is, acceleration of adsorption.

For this reason, the choice of binder and its content is a subject of many studies. The most popular binders are siloxanes [52], silicones [86], pseudoboehmite [46,69], polyaniline [69,87], polytetraftorethylene [46,88], silanes [43], mineral clays (kaolin, montmorillonite, attapulgite, bentonite, sepiolite, etc.) [19,44,45,46], polyvinylpyrrolidone [41], epoxy resin [89], polyvinylalcohol [35], polyethylene glycol [45], carbonized PVA [17], silica sol [38,47], as well as materials of natural origin, such as corn-flour, gelatin, hydroxyethyl cellulose [45], alginates [90], etc.

Another way to improve heat transfer in a compact adsorbent bed is the use of additives with enhanced heat transfer properties, such as metal powder [91,92], foam [23,93], chips [93], expanded graphite [46], graphite flakes [94], SiC [46], Si₃N₄ [46], carbon microfibers [66], etc. These additives are deemed to improve heat transfer to and between the adsorbent grains. Indeed, the consolidated bed composed of zeolite A with silico-aluminate binder had a thermal conductivity of 0.36 W/(m K) [23]. Embedding copper and nickel foams into the bed gave the gain for thermal conductivity to 8.0 and 1.7 W/(m K), respectively. The wall heat transfer coefficient also increased from 20 W/(m² K) for granulated beds to 110–180 W/(m² K) for coatings/(consolidated beds) (Table 1) [23]. Such a great acceleration of heat transfer parameters is shown to be due to the associative effect of both adsorbent consolidation and metallic foam additive. However, an essential effect was observed only at a large content when the additive forms continuous heat-conducting chains, which significantly increases the thermal conductivity of the layer in the direction of the chain due to the percolation effect [95]. However, the high content of the inert component can lead to a strong, undesirable decrease in COP [96,97,98].

For expanded graphite, a special effect is observed when external pressure is applied to the bed. It forces the heat-conducting graphite sheets to line up in a direction perpendicular to the applied pressure [99,100]. As a result, the bed thermal conductivity in this direction can significantly increase, which makes it possible to selectively enhance the heat transfer in the desired direction. This is usually the direction perpendicular to the HEx plate.

It is essential that the compact bed has good thermal contact with the metal surface, through which heat is supplied and removed since the contact resistance makes a significant contribution to the total thermal resistance of the bed. According to estimates of [41], the wall contact resistance is usually equivalent to an adsorbent thickness of 0.2–0.7 mm in the conventional granulated layers, which is comparable with a typical layer thickness in AHT units. Thus, it is expected that the effect of contact resistance on the adsorption dynamics can be essential, primarily in thin layers (see Section 3.3). It should be eliminated, or at least minimized, by synthesizing an adsorbent layer directly on the HEx surface [53,60,63,64,101,102,103]. This approach is especially successful if the layer growth is epitaxial so that there is no interface border. The wall contact resistance can also be significantly reduced if a compact bed is glued to the substrate using a thin film of adhesive with high thermal conductivity (see, e.g., [52,70,72,86]). The latter approach allows a larger mass of adsorbent to be placed per unit of the surface area of HEx than the direct crystallization method.

An important feature of AHT systems is that the adsorption initiation due to a large temperature drop applied can cause essential gradients of temperature between the HEx fins and the adsorbent bed as well as within the bed. If local adsorption equilibrium takes place, large gradients of vapor pressure may appear in the bed. Indeed, the vapor pressure near colder adsorbent sites/spots/grains is lower than near warmer ones. It leads to the diffusion/convection vapor flux from hotter to colder sites [71,104]. As the vapor phase contains the latent heat of adsorption, its transport delivers this heat to colder sites. The heat flux rises with the temperature almost exponentially [71] and can make an essential contribution to the overall heat transfer in porous AHT beds consisting of loose adsorbent grains. The coupling of heat and mass transfer occurs as the water molecules desorbed from hotter sites freely transfer to colder sites, where they adsorb with the release of large adsorption heat. This mode of heat transfer can be called the “heat pipe” mode because it is driven by the concentration gradient of water vapor, as in a common heat pipe [105,106]. Probably, just this mode provides very high coefficients (up to 200 W/(m² K)) of heat transfer between the HEx surface and the thin layer of loose adsorbent grains reported in [72,73,74,75,76]. The excessive reduction of mass transfer in compact beds where loose adsorbent grains are consolidated with a binder can result in dramatic inhibition of this heat transfer mode. As a result, the total heat transport decreases, despite the increase in the steady-state heat conductivity due to the larger density of the compact bed. Again, the texture of the compact layer has to be carefully organized to reach a smart compromise between heat and mass transfers, also taking into account the “heat pipe” mode.

3.3. Effect of the Contact Resistance

As already discussed above, the thermal resistance in the system “metal support–porous bed” is largely associated with the wall contact resistance. One of the first studies that revealed it in AHT-related systems was [26]. This important finding was then confirmed in many works [17,18,44,63,72]. It was shown in [44] that the use of the heat-conductive paste (Thermigrease TG20031, Dr. Dietrich Müller GmbH, Germany) is a more efficient way to reduce the contact resistance than a simple mechanical pressing of the adsorbent layer to the HEx wall using a screw-based system. The authors of [72] showed that a compact layer of activated carbon, not glued to the substrate but simply lying freely on it, adsorbs/desorbs methanol 2.6/3.0 folds slower than the reference layer composed of loose carbon grains. If glued, however, adsorption and desorption became faster than in the reference layer by factors of 1.6 (Figure 5) and 2.0 [72], respectively, so that a perceivable acceleration effect was observed.

It is significant that a compact layer placed on the support without glue adsorbs vapor more slowly than the reference granulated bed. Indeed, the heat conductance through the compact bed increases because its density is much larger than that of the reference bed. Therefore, a larger stationary bed conductance is not always advantageous for improved heat transfer in AHT units. Probably, in the compact bed, the decrease in convective heat transfer (“heat-pipe” mode) can surpass the effect of a higher stationary conductance.

3.4. Effect of the Layer Thickness

The thickness of the compact adsorbent bed is one of the most important factors determining the adsorption dynamics in AHT units [44,53,58,72,101]. Obviously, the adsorption rate decreases in a thicker bed due to an increase in resistance to both heat and mass transfer. Thus, in [44], a microporous SAPO-34 with various thicknesses from 0.38 to 1.1 mm, was compacted with the binder (18.5 wt.% bentonite) and then fixed to an aluminum substrate using a heat conductive paste. The characteristic adsorption time τ_0.5 gradually decreased for thinner layers from 15.0 to 3.3 s. This time appeared to be shorter than that for a reference layer of loose SAPO grains with a thickness of ~0.9 mm (τ_0.5 = 17.3 s). In terms of the ratio S/m, this layer is close to a compact layer with a thickness of 0.61 mm, but in the latter, the process is 2.9 times faster. This again shows that the proper application of binder and adhesive can result in the appreciable acceleration of adsorption. A similar deceleration of adsorption at an increasing layer thickness of 0.08 to 0.76 mm was reported for compact beds of AlPO-18 prepared by using PVA as a binder and glued to the aluminum support [53]. No comparison of the compact layer with the reference granulated one was provided in this paper.

The correct comparison of these layers was first reported in [72] for compact carbon layers studied at fixed ratios S/m (0.4–1.6 m²/kg). The layers were prepared with carbonized PVA as a binder. The effects of the grain size (0.8–1.6 mm) and the layer thickness were studied at the PVA content of 12 wt.%. The most notable findings are: the methanol adsorption dynamics is invariant with respect to the ratio S/m regardless of the grain size (Figure 6a) and a linear relationship between the maximal (initial) specific power and this ratio is observed (Figure 6b). The conclusions are very similar to those made for the reference beds of loose carbon grains without binder; however, the adsorption in the compact beds is faster by a factor of 1.5–3.5 as compared to the reference beds with the same (S/m)-ratio. This is a typical scale of the acceleration effect due to the adsorbent coating. The effect is pronounced for thin beds, as for them, the contribution of the contact resistance is dominant and almost disappeared for thick layers (S/m ≤ 0.5 m²/kg) (Figure 6b).

The dynamics of water sorption on a new composite, “LiCl confined to Multi-Wall Carbon NanoTubes (MWCNT)”, shaped as grains and tablets (with and without PVA as a binder), was studied in [76] by the LTJ method. The temperature variation was typical for AHT cycles. The thickness of the granulated bed and the tablets varied from 0.8 to 2 mm. The tablets were glued to the metal substrate by a thermal paste. It was found that both sorption and desorption were much faster for tablets prepared with PVA than for the two other configurations with the same ratio (S/m) = 2.6 m²/kg. The cooling power increases for thinner layers and is nearly proportional to the (S/m)-ratio and can exceed 10 kW/kg (at conversion 0.7).

It is interesting that refs. [44,53] also reported some indications that the ratio (S/m) is a dynamic invariant, namely, the almost constant average mass flux in Table 3 in [44] and constant initial rate of water adsorption (see Figures 4a and 5a in [53]). Thus, the ratio (S/m) can be recommended as a measure of dynamic perfection of HExs coated with adsorbent just in the same way as it was proposed for granulated AHT beds in [73].

3.5. Effect of the Binder Content

For binder-based coatings and consolidated beds, the binder content C is another tool to manage the crossover between limiting transport modes. As mentioned above, the binder content decrease from 20 to 2.5 wt.% causes the gradual transformation of the front regime to the uniform adsorption mode [77]. This indirectly indicates that at C = 20 wt.%, the bed is too dense, which slows down mass transfer and negates the benefits caused by increasing the bed’s thermal conductance. A similar conclusion was made in [107] for a layer of activated carbon glued with carbonized PVA. The acceleration effect was revealed only at C = 15 wt.% (Figure 7). An increase in the binder content leads to a gradual slowdown of adsorption: at 20 wt.%, the process rate in the compact layer decreases and becomes close to that in the reference loose grain bed; at 25 wt.%, the adsorption further slows down. Unfortunately, at C < 15 wt.%, the layer was too mechanically weak and could not be tested. Thus, to select the optimal binder content, it is necessary not only to compromise between the increase in heat transfer and the decrease in the mass transfer but also to ensure a good mechanical strength and hydrothermal stability of the layer. A good mechanical strength can be obtained by adding a small amount (1–2 wt.%) of reinforcement fillers, such as carbon fibers [44]. Hydrothermal stability, although very important, is not considered in this review, and recent reviews and books are offered to interested readers [34,70].

The effect of the binder also depends on the binder’s nature and especially on how the binder is located in the layer. It is preferable if the binder is concentrated near the contact points between the grains, thus increasing the heat transfer surface area. In this case, there is a hope of selective enhancing heat transfer without slowing down the supply of vapor to the outer surface of the grains. The effect of grains overlapping, which leads to a heat transfer contact area instead of a contact point, was studied by numerical simulation in [108,109]. The authors of [109] compared a common bed of grains having point contact, grains overlapping with a region of 10% of the grain diameter, and grains with a gap of 10% of the grain diameter (Figure 8a). The overlapping configuration simulates the effect of the binder intently located near the grain contact points. The study was performed for two beds with the same ratio S/m and consisted of two layers of larger grains (0.85 mm diameter) or four layers of smaller grains (0.45 mm diameter).

The overlap geometry was found to result in faster adsorption (Figure 8b) due to more efficient conductive heat transfer through the solid (adsorbent) phase. The acceleration effect reached 39% and was more pronounced for the layer consisting of smaller grains. This was explained by the larger number of thermal contacts in this case. Thus, one may expect that the enhancement of the heat transport intently near contact points could be a very effective tool to improve heat transfer and adsorption rate when using a binder to transform the granulated bed into the compact one. The attempt to realize this recommendation in practice was made in [110] (see Section 3.7).

3.6. Studying the Relative Contributions of Heat and Mass Transfer Resistances

As mentioned above, to optimize the compact adsorbent layer for AHT applications, a smart compromise between the increase in heat transfer and the decrease in the mass transfer is obligatory. In AHT units, heat and mass transports are strongly coupled; therefore, their differentiation and clarification of which transport mechanism is limiting are welcome. A common qualitative approach to differentiate various resistances is to vary the adsorbent grain size in a wide range and reveal changes in transport mechanisms. For example, the “grain size insensitive” mode typical for AHT systems based on granulated beds [15] indicates the sufficiently fast intragrain mass transport and the limitation of adsorption rate by heat transfer.

Recently, a novel experimental method for making such discrimination quantitatively was proposed in [111]. This approach, called a Transport Impedance Analysis (TIA), is based on the measurement of the vapor pressure evolution together with the temperatures of the adsorbent and the HEx surface. It somewhat resembles a driving temperature differences method [112], but to quantify the transport resistances, a new parameter, the equivalent thermal impedance, was proposed. First, the authors verified the new method by studying HMT in simple configurations of adsorbent bed, namely, one layer of spherical silica grains placed on a metal support and a monolayer glued to the support. For the first configuration, heat and mass transports were relatively balanced. As expected, gluing the grains enhanced heat transfer so that mass transport became a limiting process, and smaller grains should be used to harmonize the transports.

Thus, the TIA method has been demonstrated to be useful for distinguishing between heat and mass transports at the adsorbent bed level, including coated beds. In practice, the method is a handy quantitative tool for the optimization of heat and mass transfers in real Ad-HEx units.

3.7. Optimization of Adsorption Dynamics in Compact Layers

Applying the new TIA method for studying the water adsorption dynamics on SAPO-34 coatings of 60 and 460 µm thicknesses revealed that, in all the cases, the dynamics were limited by vapor transport inside the coating [113]. Therefore, the coating structure has to be optimized to improve the mass transport. It can be done, e.g., by introducing uniformly spaced longitudinal channels, and the TIA method allows the optimal distance between the channels to be found [114]. As a result, the adsorption rate doubled at the same ratio S/m when compared to the unstructured coatings. Similarly, the macroscopic mass transport channels were added into the zeolite consolidated bed to accelerate slow vapor diffusion [30]. This enabled an increase in the maximum SP from 0.65 to 0.85 kW/m².

A scalable and cost-effective approach for the preparation of advanced structured SAPO-34 coating based on the bottom-up assembly of colloids directed by magnetic and capillary forces was developed in ref. [110]. First, the authors deposited an aqueous suspension of the adsorbent particles, a binder, and ferrofluid oil droplets on an aluminum support. Then, a magnetic field was applied to the deposited suspension to align the droplets into chains in the direction perpendicular to the support, thus creating channels for efficient mass-transfer inside the coating. During the following drying, the oil droplets with the ferrofluid are soaked into packed zeolite particles by capillary forces bridging them into percolating network of adsorbent particles, which facilitates heat transfer through the coating (Figure 9). The architectured coatings of 200–500 μm thickness were prepared with vapor diffusion improved by a factor of 2–5 and heat transport enhanced by 40–50% relative to a reference non-structured coating. Such balanced mass and heat transfers enhance the power-energy product of the adsorbent coating per unit area by 3.3-fold.

4. Lab-Scale AHT Units Employing Coated and Consolidated Adsorbent Bed

To the best of our knowledge, a closed adsorption air conditioning unit based on a consolidated adsorbent bed was first built and studied by Poyelle et al. in 1998 [65]. The use of an adsorbent bed composed of zeolite grains with a highly conductive matrix of expanded graphite allowed a 50-fold increase in thermal conductivity and a 20-fold improvement in the wall heat transfer coefficient; however, at the expense of the permeability reduction by 3 orders of magnitude. A lab-scale unit based on the consolidated zeolite/EG bed loaded into a finned tube HEx with water as refrigerant showed SCP = 97 W/kg_adsorbent, which was a real improvement compared with a granulated bed (SCP = 20 W/kg). The slow mass transfer in the consolidated bed was partially overcome by introducing arteries for vapor diffusion into the bed. A significant performance improvement was achieved with this structured adsorbent bed: the SCP was increased up to 135 W/kg at an evaporation temperature of 4–30 °C. A prototype of a lab-scale chiller employing a “silica gel–EG” consolidated adsorbent bed applied around copper tube HEx was studied in [115]. Although the SCP = 37 W/kg delivered by the best sample was almost twice that for a silica gel granulated bed (21 W/kg), it remained quite moderate. The temperature of the adsorbent bed measured at various distances from the HEx surface showed no significant difference indicating that heat transfer does not limit the adsorption rate, which is determined by mass transfer in the bed.

Another way to overcome the poor mass transfer typical of consolidated beds relates to using HEx coated by a thin adsorbent film. Bauer et al. [63] studied a lab-scale adsorption chiller based on a thin coating of SAPO-34 synthesized in situ on the aluminum finned-tube HEx surface. Enhanced SCP of ≈560 W/kg_Ad-HEx or volumetric VSCP ≈ 350 W/dm³_Ad-HEx was achieved under operating conditions T_ev/T_con/T_des = 15/30/85 °C at the cycle restricted by 90% conversion.

An adsorption chiller using small-scale HEx coated by commercial SAPO-34 (adsorbent mass 0.084 kg) with a silane binder was studied in [20]. A flat finned-tube HEx was coated by an adsorbent film with a uniform thickness of 0.1 mm and the ratio m_HEx/m_ads equal to 6. The chiller delivered a high SCP of 675 W/kg_adsorbent, which exceeds SCP = 498 W/kg obtained for the fixed bed of the same adsorbent granules of 0.6–0.7 mm size loaded to the same HEx with m_HEx/m_ads = 1.96. However, due to the smaller amount of the adsorbent loaded and higher m_HEx/m_ads ratio, the volumetric SCP = 93 W/dm³_HEx and COP = 0.24 were lower than those for the fixed silica gel bed (SCP = 212 W/dm³ and COP = 0.4). The adsorption stability of the coated adsorber subjected to 600 adsorption cycles was successfully verified, showing no reduction in the adsorption ability. McCague et al. [116] compared a chiller based on a commercial finned-tube HEx with FAM-Z02 (SAPO-34) coating by Mitsubishi Plastic (0.33 mm thickness, mass of adsorbent coating 0.80 kg, m_HEx/m_ads = 3.1) with the same HEx loaded with FAM-Z02 pellets of 1.2–2.4 mm size (mass of adsorbent 1.97 kg, m_HEx/m_ads = 1.3). Under operating conditions T_ev/T_con/T_des = 15/30/90 °C, the SCP = 300–456 W/kg and VSCP = 60–90 kW/m³ were obtained for the coated HEx. For fixed beds, the SCP was significantly lower (50–60 kW/kg), whereas the VSCP was much larger (100–150 kW/m³).

An adsorption chiller containing two adsorbers with a new type of HEx based on aluminum sintered metal fiber structures brazed on flat fluid channels was coated with in situ crystallized SAPO-34 (m_HEx/m_ads = 2.6) was tested [112]. A very high VSCP of 190 and 320 kW/m³_Ad-HEx volume was achieved at T_ev/T_con/T_des = 12/32/90 °C and T_ev/T_con/T_des = 19/27/85 °C, respectively. The power was shown to be limited by vapor diffusion in the coating pores. The HEx structure with thinner metal fibers and adsorbent coating was recommended to overcome the mass transfer limitation.

A 1 kW adsorption chiller based on a corrugated microchannel HEx coated with a composite sorbent was developed by He et al. [117]. The sorbent composed of LiCl and low-cost natural mesoporous material Wakkanai Siliceous Shale was loaded into the HEX using epoxy resin to prevent HEx corrosion. An SCP of 0.35–0.41 kW/kg_adsorbent and COP ranged from 0.40 to 0.48 were detected in a simple cooling cycle at operating temperatures T_ev/T_con/T_des = 15/30/80 °C. Using heat recovery allows an increase in the COP up to 0.54.

A laboratory adsorption chiller consisting of a full-scale finned-tube HEx coated with a thin film of aluminum fumarate (Basolite A520) was built and studied in [21]. The Ad-HEx was prepared using the dip-coating method with a polysiloxane-based binder. The film of 300–330 µm thickness was evenly distributed on the HEx surface. The ratio m_HEx/m_ads was equal to 4.67. Under cooling cycle conditions (T_ev/T_con/T_des = 18/30/90 °C), the chiller delivered a peak SCP of 5.88 kW/kg and the average SCP_0.9 = 1.39 kW/kg_adsorbent at conversion 0.9. A large VSCP_0.9 = 101 W/L was achieved, showing a high potential of the aluminum fumarate-coated HEx.

Gkaniatsou et al. [37] developed a scalable method for mass production of another aluminum MOF, namely MIL-160 (aluminum dicarboxylate) and its deposition on the HEx surface using a silicone binder (Figure 10a). A solar heat-driven adsorption chiller utilizing two finned-tube HExs coated with MIL-160 film of 445 ± 95 µm thickness (m_HEx/m_ads = 3.2) was studied (Figure 10b). The SCP ranged from 1.2 kW/kg at T_ev/T_con/T_des = 10/30/85 °C to almost 2.0 kW/kg at T_ev/T_con/T_des = 30/40/95 °C. The feasibility of this chiller employing MOF-160 coating was tested under semi-field outdoor conditions showing stable cold production with a COP of about 0.4.

Adsorbers with a thin adsorbent coating ensure efficient use of the adsorbent and high specific power related to unit adsorbent mass due to good thermal contact between the HEXs surface and the bed. However, their volumetric power is often quite modest as a small amount of adsorbent is deposited on the HEx surface to ensure low mass transfer resistance. Furthermore, a high m_HEx/m_ads ratio inevitably leads to the reduction of both COP and heat storage density due to the rising impact of the sensible heat Q_s related to Ad-HEx inert masses (Equation (1)). To overcome this drawback, more adsorbent has to be placed into the HEx. Three solutions were suggested in [118] (Figure 11):

• increasing the coating thickness that, however, may reduce mass transfer through the adsorbent layer;
• adding adsorbent grains to completely fill a space between fins;
• combination of these two routes.

Following this idea, Sapienza et al. [119] built and studied a hybrid adsorber composed of silica gel loose grains loaded into a finned flat-tube HEx coated with FAM-Z02. The hybrid adsorber was characterized by an acceptable ratio m_HEx/m_ads = 1.85, almost half of that for the coated HEx. The chiller provided the specific cooling power ranging between 7.25 and 9.36 kW/m³ with COP = 0.30–0.35 at T_H = 90 °C, T_L = 15–18 °C, and T_M = 25–28 °C. The hybrid solution allowed a reasonable power increase relative to the reference coated HEx. The same concept was used by Engel [120], who studied an AHS unit based on a hybrid adsorber with commercial FAM-Z02 coating (Mitsubishi Plastics Inc.) filed between the fins with granules of the same adsorbent. The AHS system loaded with the adsorbent of the total mass 3.3 kg (2.5 kg of coating and 0.8 kg of granules) showed an average power of 1.05 and 1.33 kW under cooling and heating modes, respectively, for a period of 30 min.

5. Summary

Nowadays, due to low electricity consumption and the use of environmentally benign working fluids, adsorption heat transformation (AHT) is considered an energy and environment-saving alternative to common compression chillers and heat pumps. Increasing the AHT power is a keystone for the further development and dissemination of this emerging technology. In this review, the shaping of an adsorbent bed as a coating or a consolidated bed attached to the surface of a heat exchanger for enhancement of the AHT-specific power is comprehensively surveyed. Methods for coating/(consolidated bed) synthesis are considered. Then, the specific features of heat- and mass-transfer and adsorption dynamics in coated “adsorbent–heat exchanger” (Ad-HEx) units are analyzed. Finally, real AHT devices employing consolidated beds and coated HExs are described.

Many scalable methods for the synthesis of compact adsorbent beds have been developed so far, such as compression of binder-based consolidated beds, dip-, drop-, and spray-coating with a binder, and direct synthesis of binderless coatings. These methods were applied to prepare compact Ad-HEx units coated with various adsorbents (conventional silica gels and zeolites, innovative aluminophosphates, MOFs, and CSPMs).

This review of the dynamics of water and methanol adsorption on compact adsorbent layers revealed the following features:

• Although coated Ad-HExs are characterized by good conductive heat transfer, their mass transfer resistance is rather high due to the large density of the bed. This leads to quite moderate values of the specific power (less than 150 W/kg).
• The Ad-Hex configuration in the form of a thin coating (100–300 µm) allows reducing mass transfer resistance. The specific power of such AHT units reaches promising values exceeding 1 kW/kg. However, the small coating thickness leads to a large mass ratio “metal to adsorbent” of 3–6, which can significantly lower the unit COP.
• Heat and mass transfers in adsorbent coatings are strongly coupled, so the reduction of mass-transfer can result in the depression of heat transfer and an appropriate deceleration of vapor adsorption.
• A smart compromise between heat and mass transfers can be achieved through the synthesis of structured or architectured coatings. In such coatings, fast heat transfer is ensured due to tight contact between the HEx surface and the coating. The intently introduced macropores or channels for vapor transport enable rapid mass transfer in the coating.

In our opinion, further progress in Ad-HEx configurations can be achieved by the rational design of structured adsorbent coatings with optimized heat and mass transfer.