Assessment of Beaded, Powdered and Coated Desiccants for Atmospheric Water Harvesting in Arid Environments

Abstract

Atmospheric water harvesting (AWH) is a promising alternative to address immediate water needs. Desiccant-based AWH could compete effectively with other commercially available AWH technologies. One of the primary challenges facing desiccant-based AWH is the energy required to desorb the captured water vapor from the desiccant. This work presents a multi-faceted approach targeted explicitly at low-humidity and arid regions, aiming to overcome the limitations of the refrigerant-based AWH system. It includes assessing common desiccants (zeolite, activated alumina, and silica gel) and their forms (beads, powdered, or coated on a substrate). A bench-scale test rig was designed to evaluate different types and forms of desiccants for adsorption and desorption cycles and overall adsorption capacity (g/g), kinetic profiles, and rates. Experimental results indicate that beaded desiccants possess the highest adsorption capacity compared to powdered or coated forms. Furthermore, coated desiccants double the water uptake (1.12 vs. 0.56 g water/g desiccant) and improve adsorption/desorption cycling by 52% compared to beaded forms under the same conditions. Additionally, Brunauer–Emmett–Teller (BET), X-ray diffraction (XRD), and dynamic vapor sorption (DVS) analysis show the pore geometry, morphology, and sorption capacity. The goal is to integrate these performance improvements and propose a more effective, energy-efficient desiccant-based AWH system.

1. Introduction

The atmosphere is a vast source of fresh water, which remains as vapor before precipitation. There is an estimated 12,900 km³ of water in the atmosphere, around six (x6) times more than all combined lakes and rivers. Thus, this vast pool of water in the air can be harnessed to meet future challenges. Cloud seeding is a technique wherein precipitation occurs due to the addition of chemicals such as silver iodide and dry ice in the clouds. This helps the formation of nuclei where the water drop resides in the clouds can be condensed into a water droplet. This practice requires a sufficient moisture level and these external agents in the clouds to trigger the rainfall [1]. Conversely, the atmospheric water harvesting (AWH) process includes capturing the water moisture directly from the air, which is then condensed into liquid water by cooling. The condensed water can be filtered with moderate treatment and can be used in domestic, industrial, and agricultural applications [2]. The water can be collected using passive, active, or semi-active methods. The passive harvesting method consumes less energy compared to active harvesting. The active system uses extra energy sources such as magnetic refrigeration, compression refrigeration, and thermoelectric refrigeration. This enables the system to operate efficiently at any temperature and humidity [3].

When passive water harvesting is considered, various devices can capture atmospheric water vapor from ambient air, including fog capture by dew point condensation (i.e., both natural dewing and with refrigerant) and desiccant-based (adsorber) systems [4]. Fog-based water collection occurs naturally at cooler and higher altitudes, and the process can collect about 65 L (17 gallons) of safe drinking water per square meter for 600 m² of fog collector per day [5]. Natural atmospheric water capture technology (i.e., fog collector) is efficient and scalable. However, it is limited by elevation and geographical locations and may not be effective for all climatic conditions [2]. The dew water is harvested through a passive cooling condenser, solar-regenerated desiccant, and active cooling condenser technologies [6]. The thermal equilibrium equation between sensitive and latent heat fluxes determines the dew yield. In Serbia, the amount of water collected from this technology is 1.4% of their consumption, which is insufficient for large-scale activities [7]. For both, the techniques depend on the prevailing meteorological conditions, the working efficiencies of the fog-collecting mesh, and the surface characteristics of the dew collector [8].

AWHs are currently favored for various reasons, including addressing global water shortages and weather-related challenges, where some geographical areas experience frequent rain while others endure infrequent rainfall and prolonged droughts. These changes contribute to climate-related challenges, resulting in disasters and emergencies (such as flooding, hurricanes, and droughts) in many regions to face water shortages. Consequently, securing safe drinking water faces a significant number of challenges. In addition to these issues, rising energy costs create a need to explore alternative resources and more sustainable methods for securing these precious water supplies, aside from groundwater, surface water, and seawater. This is particularly critical in arid and semi-arid regions with limited freshwater availability. Due to the limitations of the fog and dew collector water harvesters, two other methods of AWH are becoming widely used, refrigeration-based condensation and desiccant-based absorption/desorption systems for capturing atmospheric moisture.

A refrigeration-based AWH system relies on cooling the moisture in the air, using refrigeration to cool to a temperature below the condensation dew point to convert the moisture to liquid water. The system has cold refrigerant in the refrigerant coils. Although this method works well in areas with relatively high (relative) humidity (>40%), it faces severe challenges in regions with lower relative humidity (RH < 30 to 40%) levels and/or low ambient temperatures (<25 °C), where these systems struggle to produce water. Cooling the air sufficiently below the dew point temperature requires significant energy to result in condensation. Airborne particulate matter and volatile organic matter also collect on the refrigerant coils, which in turn enter the storage tank [4]. Moreover, refrigeration-based systems are very dependent on the ambient humidity level and temperature, as mentioned above, making it even more challenging to generate water in drier regions like Arizona, where humidity is lower (RH < 40%).

On the other hand, desiccant-based systems offer a promising alternative to such conditions by overcoming these challenges. They yield lower water conductivity and turbidity compared to that from a refrigerant-based system. This is due to the ambient air filtration and the prevention of dust being trapped in the desiccant before condensation [4]. The key feature of a desiccant-based AWH system is that it uses moisture-absorbing adsorbent materials to capture the water moisture directly from the ambient air through the adsorption process, regardless of temperature or humidity levels in the air [9]. Adsorbents can be divided into physical, chemical, composite, and polymer adsorbents [10]. Depending on the desiccant type, the captured water vapor (as liquid water) can vary from 10 to 50% (0.1 to 0.5 g/g) of the desiccant’s dry weight [11]. Therefore, this type of desiccant-based AWH can capture moisture at low humidity levels (<30%) as well as in higher humidity conditions (>>40%). This versatility makes the system robust and particularly suitable in regions where refrigeration-based AWH offers limited water generation. Additionally, desiccant-based systems can be regenerated (i.e., desorbed) through heating, typically using electrical or gas-derived thermal energies, which require a substantial amount of energy (kWh/L) and act as a drawback for this system. However, integrating alternative low-grade thermal energy sources (e.g., solar, geothermal, thermal, photo-thermal, photovoltaic heating, or waste heat) into the desiccant-based system could reduce overall heating energy needs and enhance sustainability by incorporating renewable energy (RE) into AWH operations.

Thus, the primary aim of this research is to explore options for further enhancing the desiccant-based atmospheric water harvester (AWH) as an alternative to refrigerant-based systems. To achieve these objectives, this research begins by comparing the performance of commonly used desiccants, including zeolite (as a baseline (BL) desiccant, which is commercially employed) and three alternative types (activated alumina (AA) and two forms of silica gel), to reduce overall energy requirements for regeneration (i.e., heating). We also examine three types of desiccant materials (i.e., beaded, powdered, and coated forms). Their adsorption capacity (g/g), desorption capacity (g/g), kinetic rates of both adsorption and desorption, and energy requirements for regeneration are evaluated. Unlike previous studies, this work introduces a novel approach by integrating desiccant-coated substrates, which improves adsorption efficiency and mechanical stability while decreasing energy consumption for regeneration. Finally, the research offers recommendations for an alternative desiccant-based AWH design compared to the currently prevalent zeolite-based (baseline) rotary desiccant wheel (RDW) or immobilized desiccant-coated surfaces, addressing specific energy-related challenges and suggesting areas for future research to enhance overall system performance.

2. Materials and Methods

A benchtop scale test rig (Figure 1a) was designed and built to help analyze the behavior of various desiccants and their forms under different test conditions. The desiccants used in the study are zeolite, activated alumina (AA), traditional silica gel, and blue silica gel. The composition of the four desiccants is based on manufacturer-provided data or specification sheets, which state that zeolite 13x is a crystalline aluminosilicate material made of silicon, aluminum, and oxygen. Activated alumina (AA) is a porous solid form of aluminum oxide (Al₂O₃) made from hydrated alumina. Silica gel is a highly porous form of silicon dioxide (SiO₂) units made from sodium silicate. Finally, blue silica gel is a commercially available desiccant, a highly porous form of SiO₂ units made from sodium silica that contains 0.005% cobalt chloride as a color indicator [12,13].

Table 1. Physicochemical properties of the beaded desiccants [14,15,16,17,18].

Characteristics	Zeolite	Activated Alumina	Silica Gel	Blue Silica Gel
Bead diameter (mm)	3 to 5	3.175	3.5	2 to 4
Pore diameter (nm)	1.3	4.8	2 to 3	Not given
Density (kg/m3)	689	769	700	829
BET surface area (m2/g) (measured)	781	298	759	578
Regeneration temperature (°C)	250 to 300	110 to 198	~100	~121

shows the physicochemical properties of the beaded desiccant. The surface area is estimated using the Tristar II Plus instrument (Micromeritics, USA) to measure the materials’ surface area through the Brunauer–Emmett–Teller (BET) model and nitrogen adsorption. All the samples were degassed at 120 °C for 24 h before dosing nitrogen.

The horizontal setup features two chambers equipped with Atlas Scientific sensors connected to a Raspberry Pi, which measures the chambers’ incoming and outgoing dry-bulb temperature (°C), relative humidity (%), and pressure (psi). Openings are precisely drilled/threaded in the chambers to accommodate the sensors. A desiccant holder, a 3D-printed cylinder with a perforated bottom and a removable perforated top cover, is positioned horizontally between the two chambers, and the desiccant inside the cylinder is supported by 3D-printed perforated disks. The 3D-printed cylinder is constructed from polycarbonate or acrylonitrile–butadiene–styrene terpolymer blend (PACBS) material, which can withstand high temperatures. The number of internal supporting disks or plates used in the 3D-printed cylinder or holder varies depending on the desiccant loading (Figure 1b).

For the beaded desiccant study, the desiccant is placed inside a 3D-printed cylinder (holder) with several vertically arranged perforated (3D-printed honeycomb) disk fillers to adjust the desiccant loading vertically between plates (grams of desiccant; loaded between the top disk and cover). A single perforated 3D-printed rectangular tray is used for the powder desiccant study. It is supported by fine screen mesh (400 US mesh size), which helps to position the powdered desiccants horizontally on the tray. For the coated desiccant study (i.e., coated onto the 3D-printed disks), eight desiccant-coated (drop casting) disks are placed vertically inside the 3D-printed cylinder. Humid air from the pre-conditioned chamber (Chamber 1) travels to the 3D-printed cylinder/holder (placed horizontally in the setup). For the powdered desiccant, the moist air flows through the top half-section of the powdered desiccant and exits the bottom half-section of the 3D-printed cylinder. Thus, air flows over the tray and through the desiccant materials, moving downward to exit the cylinder into the post-chamber. For the coated desiccant study, 3D-printed filler plates are replaced by the eight desiccant-coated disks (also 3D printed). A ball valve is placed before the desiccant cylinder or holder to regulate the humid airflow. To supply humid air to the pre-conditioned chamber (before the desiccant, i.e., Chamber 1), a humidifier (Honeywell, HUL520L) is used, assisted by an external blower (axial fan), where a Variac adjusts the fan speed and, therefore, the airflow rate, controlling the blower speed by regulating voltage. The flow rate is measured by an anemometer (Bonvoisin) (m/s). The pre-conditioned chamber (Chamber 1) also has a side sampling port to adjust the humidity level.

The experimental setup (Figure 1) is designed to assess the desiccant behavior under controlled conditions, where the humidifier’s moisture level (%) can be adjusted as needed. This setup aims to develop profiles of adsorption and desorption behavior (i.e., kinetic rates) for desiccants, along with their water adsorption and desorption capacities (g/g). A specified amount of desiccant (g), ranging from 5 to 10 g (discussed below), is placed inside the 3D-printed cylinder (holder). The system’s humidity and airflow rate can vary according to targeted humidity levels, which range from low to moderate to high, as shown in

Table 2. Relative humidity ranges and flow rates.

Humidity	Range (%)	Flow Rates During Adsorption Process (m/s)
Low humidity	~20 to 40	~1
Moderate humidity	~60 to 65	~1
High humidity	~90 to 95	~1

.

The operation is divided into three stages to perform the adsorption experiment:

(a) Stage 1 sets a baseline (BL) condition before desiccant loading (i.e., no desiccant is placed inside the 3D-printed holder or cylinder). For this study, the desiccant cylinder is placed between the chambers (first without desiccant), and the ball valve is kept open. Stage 1 is made to run for 30 min to check that both chambers reach equilibrium (pre- and post-conditioned chambers) as per the applied condition to maintain targeted humidity (%), temperature (°C), pressure (psi), and flow rate (m/s).

(b) After completing Stage 1, Stage 2 begins. Oven-dried desiccant (maintained at its regeneration temperature for 24 h to remove moisture; available in beads, powder, or coated disks) is loaded. For beaded desiccant, 10 g is placed inside the holder/cylinder between the filler disks. Alternatively, for powdered desiccant, 5 g of dry powder is spread uniformly on the desiccant tray (mounted on a 3D-printed tray within the cylinder, which restricts how the powder can be loaded onto the tray). In the case of the coated desiccant study, eight coated disks are utilized (each disk coated with 0.6 to 0.8 g of desiccant, with the total desiccant loading for eight disks ranging from 4.8 to 6.4 g). The desiccant cylinder or holder is positioned horizontally between the chambers; the ball valve is closed (preventing any moisture or airflow to the desiccant) while the side port is partially opened. With the ball valve closed, the humidity in the post-conditioned chamber (Chamber 2) is significantly reduced (any residual moisture from Stage 1 is purged out by dry air). Only the pre-conditioned chamber (Chamber 1) retains the targeted humidity level before engaging Chamber 2 (the post-condition chamber before opening the valve). Stage 2 operates for 30 min to complete.

(c) Soon after Stage 2 is completed, the ball valve is opened (Stage 3), and the humid air is allowed to flow from the pre-condition (Chamber 1) to the post-condition chamber (Chamber 2) through the desiccant holder. This enables the adsorption process to take place on the desiccant. To determine the adsorption behavior, temperature, humidity, and pressure drop are recorded over time (i.e., time-stamped by sensors) and on the log sheets. Once the humidity ranges in both the chambers (pre- and post-condition chambers) come closer to saturation, the adsorption profile over time is completed, and the desiccant weight is measured. The measured desiccant weight is used to determine the adsorption capacity (Equation (1)), as the weight of moisture in g divided by the weight of the desiccant in g) of the dry desiccant (g/g), and the profile is analyzed for the rate of reaction and kinetic rates:

$$\mathrm{Adsorption}\mathrm{capacity}({\displaystyle \frac{gofwater}{gofdesiccant}})={\displaystyle \frac{weightofdesiccantafteradsorption-weightofdrieddesiccant}{weightofdrieddesiccant}}$$

(1)

After the adsorption capacity (g/g) is measured, desorption is performed at 95 to 110 °C (air temperature) at regeneration temperature. This temperature is selected to mimic solar-based thermal energy from evacuated tubes and flat-plate collectors that can reach a maximum of ~100 °C [19,20]. A Variac (voltage regulator) controls the heater output to maintain the target regeneration temperature. The airflow rate from the heater is 2.4 m/s. Once the temperature reaches 95 to 110 °C, the humidity of the post-conditioned chamber (Chamber 2) is measured. The measurements are taken until the moisture (removed by heating) reaches the ambient level (the desiccant is completely dry). The desorption capacity (g/g) (Equation (2)) is calculated once the ambient humidity is reached (i.e., all water vapor is desorbed) in the post-conditioned chamber (Chamber 2). This is performed to determine the percentage of water lost through the desorption process:

$$\begin{gathered} \mathrm{Desorption}\mathrm{capacity}({\displaystyle \frac{gofwater}{gofdesiccant}}) ={\displaystyle \frac{weightofsaturateddesiccant-weightofdrieddesiccant}{weightofdrieddesiccant}}= \\ {\displaystyle \frac{weightofdesiccantbeforedesorption-weightofdesiccantafterdesorption}{weightofdesiccantafterdesorption}} \end{gathered}$$

(2)

These adsorption and desorption experiments (Equations (1) and (2)) assess the overall adsorption and desorption behavior, provide kinetic profiles, and assess the reaction order and rates of adsorption and desorption capacities (g/g) as well as the adsorption and desorption cycling times achievable for various desiccants. These profiles and capacities are critical in comparing the effectiveness among the various desiccant types and selecting effective desiccant types and forms. An effective desiccant will have rapid adsorption and desorption kinetics and higher water adsorption capacity (g/g) at lower regeneration energy and temperature (heating).

Experimental Uncertainty

Every experiment, along with the use of equipment and analytical devices, carries potential errors and variations.

Table 3. Parameters and uncertainty.

Parameters	Uncertainty
Weighing pan (Weight) (g)	±1%
Sensors: temperature, humidity, and pressure (°C,% and psi)	±0.4%, ±2%, ±0.1%
Thermocouple integrated into the heater (temperature) (°C)	±0.1%
Air flow rate (m/s)	±3%

illustrates the uncertainty of the experimental setup. This experimental uncertainty aids in evaluating the reliability and accuracy of the measurements (i.e., weight, temperature, humidity, pressure, airflow rate, etc.) taken throughout this study. Rather than displaying error bars in the figures, Table 3 enumerates the uncertainty based on the error parameters of the instruments or sensors. The capacity involves the weight of the desiccant, and since the uncertainty is only ±1%, the error bars are excluded.

3. Results and Discussion

The results of this study provide a comparison of the performance of three desiccant forms—beaded, powdered, and coated (drop casting)—among four desiccant types, which include zeolite, activated alumina (AA), traditional silica gel, and blue silica gel [21]. Key performance indicators include adsorption capacity, regeneration energy, and kinetics. The findings from this study highlight significant differences in performance based on desiccant type and form. This section presents a detailed evaluation of the metrics that guide the selection of optimal desiccant configurations for energy-efficient AWH systems.

Table 4. Differences between beaded, powdered, and coated desiccants.

Characteristics	Beaded Desiccants	Powdered Desiccants	Coated Desiccants
Shape and Size	Spherical beads (2 to 5 mm)	Finely crushed particles (<500 microns)	Powder coated on the perforated disks
Porosity	Mesopores (2 to 50 nm)	Micropores (<2 nm)	Coated onto a highly porous substrate
Mechanical Stability	High durability, easy to handle	Hard to handle due to being a powder, clumping	Easy to handle
Airflow	Better airflow	Airflow might be affected due to clomping	Excellent airflow

shows the physical and structural differences among the beaded, powdered, and coated desiccants. Beaded desiccants (larger) are more suited for various applications, such as air dryers for compressed air use and special drying agents. In contrast, this study shows that powdered desiccant provides faster adsorption and desorption kinetics and has a much higher surface area than beaded desiccant. However, it needs to be handled carefully due to the dust-prone nature of the powder, which can cause respiratory issues. At the same time, this study also shows that the coated desiccant on a perforated disk structure or substrate has an excellent airflow rate through the coated porous system. They are also easy to handle and use, which enables them to last longer with extended operability.

3.1. Beaded Desiccants

The beaded desiccants used for the study varied in diameter from 3 to 5 mm. Zeolite, activated alumina (AA), and silica gel were purchased from Delta Adsorbents, and blue silica gel was purchased from Dry-n-Dry. In the adsorption study, three relative humidity ranges were investigated in the test rig (Figure 1a) in adsorbing water moisture at levels of 16 to 21% (low), 60 to 65% (moderate), and 90 to 95% (high). For the low humidity range (16 to 21%), the humidifier was turned off entirely while the fan (axial blower) was engaged, and the ambient humidity (~20 to 30%) in the lab was used. Ambient air was blown into the pre-condition chamber (Chamber 1) (Figure 1a). To achieve a moderate humidity level, the humid air (from the humidifier) was combined with dry ambient air (from the laboratory) to meet the targeted humidity level. To meet the higher humidity requirements (90 to 95%), the humidifier was set to its highest setting in the pre-condition chamber (Chamber 1). These experiments were run at ambient temperature (22 to 23 °C) and a moderate fan speed of 1.1 m/s.

The desiccant adsorption and desorption behaviors and their capacities under different humidity levels (low, moderate, and high) are further analyzed and compared. It took 7 h for adsorption and less than an hour for desorption. It is observed (Figure 2a) that the humidity adsorption (water uptake) (g/g) increases as the humidity increases. Meanwhile, desiccants like zeolite seem less sensitive to humidity levels (i.e., flatter line; less sensitive to changing humidity conditions). However, among the other desiccants, silica gel shows the highest adsorption capacity (g/g) compared to the other desiccant types. While Figure 2b shows the desorption behavior (i.e., regeneration) for all desiccants, zeolite tends to have a significantly lower desorption capacity (g/g) than other desiccants. This is due to its lower efficiency in releasing the adsorbed water from the desiccant (i.e., zeolite) under the applied regeneration temperature (~95 to 110 °C), which is significantly lower than the recommended regeneration temperature (~250 to 300 °C) for zeolite [15]. As the ambient relative humidity rises, the desorption capacity increases due to the greater water vapor concentration.

Figure 3a,b show adsorption and desorption kinetics rates for all tested desiccants. It can be seen that, unlike other desiccants, zeolite’s kinetic rate constant decreases as the humidity level increases. Meanwhile, other desiccants follow upward trends, suggesting that their adsorption rate constants increase as the humidity increases. Moreover, Figure 3b shows that silica gel and blue silica gel have a higher desorption rate at lower humidity than in moderate to high humidity ranges. The kinetic rate constant is higher for desorption, possibly due to a higher air flow rate (~2.4 m/s, whereas it is ~1.0 m/s for all adsorption studies; the effect of a higher air flow rate beyond 1.0 m/s was shown to have a lower impact on the overall water adsorption). Adsorption uses a lower flow rate for better contact time and mass transfer, while desorption requires a higher flow rate to remove desorbed species and quickly prevent re-adsorption. The fan speed has no impact on the adsorption or desorption capacity, but the kinetics and speed of the reaction drive the difference.

The above results suggest some inherent differences in adsorption and desorption behaviors among the desiccant types. Zeolite tends to be less sensitive to moisture levels during adsorption and desorption, while the behaviors of other desiccants, such as silica gels, tend to be sensitive to moisture levels. The water uptake and release under the targeted regeneration temperature of ~95 to 110 °C are adequate for releasing the water vapor for silica-based desiccants, while this regeneration temperature is inadequate for zeolite [15]. After multiple regeneration cycles, the material retained its adsorption and desorption efficiency, showing uniform performance.

3.2. Powdered Desiccants

The beaded desiccants were crushed using ball milling to assess the behavior of the powdered desiccants on the test rig (Figure 1a). The mesh size of the powdered desiccants varied from 100 to 60 mesh, corresponding to particle sizes of 150 to 250 $\mathsf{\mu }$m. All beaded desiccants were purchased from vendors and later ball-milled. Thus, the desiccant powders are the same as beaded desiccants but ball-milled in the powdered form.

The data from the adsorption and desorption experiments were further analyzed. The capacities of adsorption and desorption are plotted, and their behavior across the humidity ranges is illustrated. Figure 4a shows that the capacity of powdered desiccants increases with rising humidity, similar to beaded desiccants. Among the desiccants, zeolite in powdered form exhibits a higher adsorption capacity across all humidity ranges. Activated alumina (AA) demonstrates insufficient adsorption capacity compared to other powdered desiccants. In contrast to zeolite’s performance in adsorption, it has the lowest water desorption capacity (Figure 4b). Similar to the discussion of the beaded desiccants above, this is due to the relatively low regeneration temperature (95 to 110 °C) used to release the adsorbed water from the zeolite’s surface [4,15]. Blue silica gel and silica gel show greater desorption capacity compared to zeolite for the applied regeneration temperature.

The kinetic rate constants (1/min) for four desiccants in the powdered forms—zeolite, activated alumina (AA), silica gel, and blue silica gel—under varying humidity ranges (20 to 25%, 60 to 65%, and 90 to 95%) are shown in Figure 5. Adsorption rates are shown in Figure 5a, where blue silica gel exhibits the maximum adsorption rate at medium humidity (60 to 65%). On the other hand, silica gel exhibits consistently reduced adsorption rates, while activated alumina (AA) and zeolite maintain consistent performance across all humidity ranges in the powdered form. The desiccants’ rates of desorption are shown in Figure 5b. While blue silica gel, activated alumina, and silica gel show reasonable desorption rates across all humidity levels, with minor differences, zeolite shows the maximum desorption rate at medium humidity (60 to 65%). Activated alumina (AA) and blue silica gel exhibit more stable and consistent performance under all humidity levels, whereas the silica gel exhibits superior adsorption and desorption behavior at mid-level humidity. This indicates that silica gel, with its enhanced performance in mid-range humidity conditions, offers a more targeted and efficient solution than other desiccants, particularly in environments where mid-level humidity is predominant. In low-humidity conditions, blue silica gel and zeolite show suitable behavior.

3.3. Coated Desiccants

This research also focused on assessing the desiccant’s coated form (i.e., as 3D-printed coated disks) to determine the characteristics of an effective desiccant. The beaded desiccants were first ball-milled and crushed into powdered desiccant (60 to 100 mesh) form. However, based on the above results, only the blue silica gel was applied to coat the 3D-printed disks. A similar coating could have been used for the other desiccants; however, the blue silica gel was chosen as both silica gel and blue silica gel performed relatively equally. Research was also performed by integrating the supporting substrate and adsorbent material by developing a coated desiccant heat exchanger [22].

Coated desiccant is prepared using the drop cast method. Ethanol and desiccant powder are mixed and used in this method. Initially, 5 g of desiccant powder is mixed with 45 mL of acetone and 5 mL of DI water. Two batches of these are made. The solution is then sonicated to ensure proper mixing of the particles. Once the solution is mixed thoroughly, the 3D-printed porous disks (with honeycomb-structured pores) are drop casted [23] to ensure that the coating is applied correctly. The coating thickness is measured using a TOMLOV optical microscope. The 3D-printed coated disks are then dried overnight in the oven at 100 °C so the desiccant dries up and drives off the moisture, and the pore openings (honeycomb), if covered after coating, can be opened (unclogged) using a sharp needle. The coated disks are kept in the desiccator to ensure dryness. The number of dip-coats can vary based on the desired desiccant loading (g) on the disks. This method ensures that the coating is durable and ideal for subsequent experiments.

The coated desiccant contains 4.8 to 6.4 g of blue silica gel, while a comparable commercial system contains 20 g of zeolite. The commercial system features a rotating wheel, allowing more incoming air to enter the dehumidifier. The adsorption capacity of the coated desiccants ranges from 0.04 to 0.18 g/g at low and high humidity, respectively. In contrast, the commercial desiccant dehumidifier has a 0.25 g/g capacity across a humidity range of 20 to 100% [4].

Figure 6 illustrates the cycling process of adsorption and desorption for the coated desiccants and their use over three-month periods, where the same disks were utilized after reactivation (or regeneration). The relative humidity (RH) of the air leaving Chamber 2 is directly linked to the adsorption/desorption behavior of the material inside. During adsorption, the material adsorbs moisture, lowering the RH of the exiting air. Conversely, during desorption, moisture is released from the material, increasing the RH of the exiting air. The RH trend reflects the material’s ability to adsorb or release water, which depends on the operating conditions (e.g., temperature, airflow rate, and desiccant properties). Experiments were performed on 23 July, 9 September, and 11 September, 2024. Different experiments were conducted using the same coated disks over this period. These were performed under variable conditions (i.e., humidity levels and airflow rates) to illustrate that these coated desiccants are robust and can be used multiple times without much deterioration in their performance. The 23 July and 6 August experiments were conducted at moderate air flow rates of 1.1 m/s (i.e., fan speed) under moderate (60 to 65%) and high humidity (90 to 95%) conditions, respectively. Later, an additional experiment was conducted at high fan speed (2.9 m/s) and lower humidity (40 to 45%) on 11 September. It showed (results not included here) that as the fan speed increases, the kinetics of moisture uptakes (adsorption) improve. However, it did not improve the overall adsorption capacity (g/g), which remained the same. The cycles repeating in Figure 6 indicate the performance and stability of coated disks over three months. The adsorption phase shows rapid uptake followed by desorption, which means that the coated desiccant remains stable and can be used for multiple cycles without much deterioration. The amount of desiccant loaded remains constant for all the runs (6.1 to 6.6 g). These cycling data provide the basis for the durability and reusability of the coated desiccants over time.

4. Comparison Among Desiccant Forms

Adsorption and desorption capacities of desiccants with respect to their forms (i.e., beaded versus powdered versus coated) and by the types (zeolite, activated alumina, and two silica gels), along with their ease of handling, stability, reuse, etc., are also compared. The experimental setup (Figure 1a) was designed to evaluate the performance of desiccants in various forms (shapes), ensuring a reliable comparison between them. While beads offer easy handling, the powdered form requires additional safety measures before testing due to its powder or particulate form. On the other hand, the 3D-printed coated desiccant provided an easier approach by making it easier to handle due to its coating on the perforated surface and good airflow exchanges through the coated desiccant. The following sections detail the additional insights gained during the testing process for each desiccant form and type. Figure 7 shows the difference in adsorption capacity (g/g) between the beaded and powdered desiccants for zeolite, activated alumina (AA), and two types of silica gels. The water-holding capacity of the beads is found to be much higher than that of the powder desiccant at three humidity ranges: low (20 to 40%), moderate (60%), and high (>90%). This is due to some intrinsic characteristic or property changes when the bead is transformed into a powdered form (by ball milling). The mesopores were apparently reduced during the ball milling process. This is concluded from the BET measurements described next in Section Characterization of the Desiccant Powders. The powder desiccant was found to have 50–80% less water absorption capacity than that of the bead form. As the humidity increases from low to high, the beaded form had higher water uptake than the powdered form, as shown in Figure 7.

Characterization of the Desiccant Powders

The BET (Brunauer–Emmett–Teller) surface area analysis provides insights into various desiccants’ adsorption and desorption behavior by examining their BET profiles. These analyses help predict the desiccant behavior based on the pores’ qualitative and quantitative nature. These analyses also provide the surface area, total pore volume, and average pore diameter (

Table 5. BET surface area, total pore volume, and average pore diameter of beaded desiccants.

Desiccants	BET Surface Area (m2/g)	Total Pore Volume (cm3/g)	Average Pore Diameter (nm)
Zeolite	781	0.052	5.04
Activated Alumina	298	0.380	4.03
Silica Gel	759	0.354	2.57
Blue Silica Gel	578	0.350	2.61

and

Table 6. BET surface area, total pore volume, and average pore diameter of powdered desiccants.

Desiccants	BET Surface Area (m2/g)	Total Pore Volume (cm3/g)	Average Pore Diameter (nm)
Zeolite	912	0.045	4.39
Activated Alumina	276	0.293	3.64
Silica Gel	359	0.137	2.89
Blue Silica Gel	578	0.176	2.86

). Figure 8a,b show the difference between the powdered and beaded desiccants regarding their isotherm type. Isotherms are calculated using the Langmuir theory, which denotes the surface area calculation. The beaded form of desiccant has a type IV isotherm, whereas the powdered form has a type I isotherm. The type IV isotherm is formed due to increased partial pressure, and capillary condensation happens in the pores [24]. The hysteresis is due to capillary condensation. Zeolite does not show the presence of hysteresis, indicating the absence of mesopores. While analyzing the adsorption isotherm of the powdered desiccants in Figure 8b, the absence of hysteresis is observed for all desiccants, unlike the beaded form (Figure 8a), except for zeolite, which indicates that the powdered form of desiccant shows a lack of mesopores but the presence of micropores.

XRD (X-ray diffraction) analysis was also conducted to provide more significant insights and determine the powdered desiccants’ crystallinity. The peak of the graph suggests the desiccant’s crystalline or amorphous structure. Figure 9 presents the tested desiccant materials’ X-ray diffraction (XRD) pattern. The XRD pattern highlights the difference between the desiccant’s crystalline and amorphous structures, which helps compare the adsorption capacity to surface area. The zeolite exhibits sharp peaks, indicating that it is a crystalline structure, which is expected due to its ordered pore framework. In contrast, activated alumina (AA) shows a more amorphous pattern with less defined peaks, which indicates a lower degree of crystallinity. The two forms of silica gel display broad peaks, indicating their amorphous nature. The presence of cobalt chloride in the blue silica gel did not impact the peaks.

To analyze the sorption behavior of the desiccant powders, 35 to 40 mg of each powder was loaded into a dynamic vapor sorption (DVS) analyzer (IGASorp system) [9]. The instrument was run at ambient temperature (25 °C) to resemble the ambient lab temperature. Once the temperature reaches ambient, the system operates under different humidity levels as programmed. The powder undergoes adsorption for 10 min at every humidity condition. The total run takes approximately 60 to 80 min.

Figure 10a,b show the water uptake by the desiccant powders under five humidity ranges (20%, 40%, 60%, 80%, and 90%). Figure 10a shows 10-min runs, whereas Figure 10b shows 60-min runs (extended intervals) at several humidity ranges. All the desiccants exhibit similar water uptake behavior at shorter (10 min) and longer (60 min) equilibrium conditions. Zeolite shows low capacity holding for both time intervals. The silica and blue silica gel show good adsorption capacity at lower humidity conditions and favorable behavior in moderate and higher humidity conditions. Additionally, when kinetics (uptakes) are compared (Figure 10), the silica gels offered a faster moisture uptake compared to the zeolite and activated alumina (AA). Based on these and previous results, silica gels (and blue silica gel) would be more effective than zeolite due to faster and higher (adsorption and/or desorption) capacities.

5. Conclusions

The primary objective of this research is to select an effective desiccant and its form that can be utilized in effectively designing an alternative desiccant-based AWH that would capture more atmospheric moisture faster and under different humidity and/or operational conditions and, more importantly, under lower regeneration temperatures (at lower energy). The relative humidity conditions tested in this study include low (ambient, ranges of 21 to 40%), moderate (60 to 65%), and high (90 to 95%). Among the tested desiccant forms, zeolite and blue silica gel in beaded form demonstrate effective water adsorption under low humidity conditions, making them suitable for environments with limited atmospheric moisture. The coating enhances the material’s performance by optimizing surface interactions and mass transfer rates, making it a promising candidate for efficient and scalable atmospheric water harvesting applications. All these experiments showed that silica and blue silica gels have better adsorption capacities under lower regeneration temperatures (more sustainable for renewable heat sources) compared to widely used zeolite or activated alumina (AA), requiring a higher regeneration temperature. The water adsorption and desorption capacity are shown to be impacted by desiccant forms (beads, powdered versus coated). Under the conditions tested (i.e., targeted regeneration temperature), silica gel or blue silica gel seem effective, especially in the 3D-printed coated form with the highest adsorption and desorption kinetics and reliable reuse with rapid cycling.

The experimental results showed the following:

• All beaded desiccants were found to be effective. However, the effectiveness of water adsorption, desorption, and kinetics varied among the different types of desiccants. The analyses of the adsorption and desorption profiles, along with the kinetics and water adsorption and desorption capacities (g/g), reveal that beaded desiccants may require a considerably longer time (hours) to reach their saturation level, exhibiting slower water uptake followed by a slower release of water during regeneration. Additionally, they demonstrate long-term stability with no surface degradation observed over repeated tests and cycles. The results indicate that the adsorption and desorption capacities (g/g) increase as humidity rises.
• The powdered desiccants showed much faster moisture uptakes than beads, but the operation and handling of the powdered form as desiccants are more challenging than the beaded form. As powdered material, these are highly ground with fines (60–100 mesh sizes), can easily aerosolize, especially when dried, and can be harmful if inhaled; precautions are necessary during handling.
• The coated desiccant avoided the above issues of the powdered form and was observed to have the highest adsorption and desorption kinetics. Varying the air flow rate (0.6, 1.1, and 2.9 m/s) also impacts the kinetic speed but not the adsorption capacity. This indicated that it would be the best form (as opposed to bead and powder) for AWH and longer-term repeated use with effective cyclability options.
• A few experiments were repeated to ensure consistency; however, a thorough statistical analysis was not conducted. Future research could integrate rigorous statistical tests to improve the reliability and generalizability of the findings.
• The lifecycle considerations of the desiccant can be implemented for future recommendations.
• The test rig did not have a condenser in the module. The water quality can be assessed using different desiccants and types. Water quality also depends on the air. It varies from residences to labs, offices, and industries [9]. The combination of the parameters can be inculcated as a practical implementation to understand more about the water yield and quality.
• Material characterization, such as Brunauer–Emmett–Teller (BET), has been performed to understand the pore geometry and structure. It has been shown that the mesopores for most desiccants (zeolite, activated alumina, and silica gels) are lost during the conversion of beaded to powder form (by ball-milling). In contrast, the silica gels in the beaded form showed the highest surface area (759 m²/g).
• Dynamic vapor sorption (DVS) data show that silica gel offers the highest water adsorption capacity, suggesting this material as the best water adsorber compared to the other tested materials. Based on the crystallinity analysis results obtained from XRD, no crystalline peaks were detected for the silica gel material, showing its amorphous nature. Collectively, these data show that the amorphous nature of silica gel leads to its higher water adsorption capacity due to its disordered structure and higher surface area, which provide more adsorption sites. The same behavior has been observed in the literature for amorphous adsorbents and adsorptive removal of contaminants from water [25].