**2. Methods**

### *2.1. Sample Preparation*

Each sponge of same size was pre-dried and injected with 0.2 mL of 5 M CaCl2 solution. It was ensured that the solution was evenly distributed. The sponge was then dried in the oven at 60 ◦C for 12 h on a Petri dish with a lid. The mass was measured to quantify the salt content in the sponge. The sponge salt samples were then ready for water absorption measurement. The salt control was created by evenly spreading out an equal mass of salt as the salt content in the sponge on a Petri dish.

### *2.2. Moisture Absorption Time Dependence Measurement*

The materials were placed on a microbalance inside a temperature- and humidityregulated chamber. Weight changes of the samples were monitored and logged over time using a custom program written in in LabVIEW. Measurements were carried out at a constant temperature of 27 ◦C and a humidity level of 80% RH. After the absorption measurement, the sponge was put in the oven for 12 h at 60 ◦C to run the next trial.

### **3. Results and Discussion**

### *3.1. Preliminary Experiment*

Four types of sponges were tested to decipher the most effective sponge: cellulose, soundproof, magic, and Dacron (Figure 2a). The results were normalized to the salt's mass in the respective sponges. Hence, it is evident that the cellulose sponge was the most effective, as seen in the water uptake vs. time graphs in Figure 2b. The water uptake percentage was defined by *massmoisture absorbed masssalt* × 100%. Control measurements of each of the sponges without salts did not show any water absorption effects. Thus, in the next sections, we focus our investigation on the salt–cellulose sponge systems only.

**Figure 2.** (**a**) Sponge types tested. Each container has a respective sponge and control; change in mass was measured every 15 min for 6–7 h. (**b**) Sponge type curves. Graph shows water uptake by various commercial sponges. They were soaked in 5M solutions, dried, and left to absorb. The vertical dotted lines indicate a new trial. The weight changes during the drying phase were not tracked, except for the beginning and final weight.

#### *3.2. Salt-Cellulose Sponge Water Uptake Time Dependent Measurements*

The efficacy of the cellulose sponge was examined by testing its maximum absorption capacity by exposing it to a controlled atmosphere until the salts stopped absorbing moisture, as can be seen in Figure 3. There are several key observations. First, we found that the water uptake dynamics fit very well to a double exponential model (Equation (1)), as listed below:

$$\mathcal{W}(t) = \mathcal{W}\_{\text{saturated}} - A\_1 e^{\frac{-t}{\overline{\tau}\_1}} - A\_2 e^{\frac{-t}{\overline{\tau}\_2}}$$

where *W*(*t*) is the water uptake percentage at a given time, *Wsaturated* is the maximum water uptake percentage, and *A*1, *A*2, *τ*1, and *τ*2 are fitting constants. We confirm that a single exponential model cannot explain such behavior. This suggests that there are two types of water uptake mechanisms in our system. Secondly, it is evident that the sponge salt sample exhibit faster water uptake rate. Finally, the sponge salt sample was found to exhibit a higher maximum water uptake, *Wsaturated*, of ~305% while for the control salt sample, it was found to be ~272%. The salt sponge sample performed better than most of the previously reported moisture absorbers in the literature (Table 1).

### *3.3. Proposed Model*

Based on the enhancement observed from the water uptake dynamics, we propose a physical model as illustrated in Figure 4. In this model, salts are spread around the pores of the sponge. When the salts absorb moisture, the absorption is homogenous throughout the sponges. This, in effect, causes even salt precipitate distribution upon drying. As a result, the optimal amount of surface area of the salt species is maintained; for example, it does not suffer from an agglomeration problem. The optimal surface area would cause an improved water absorption rate. The infusion of the salts into the cellulose sponge results in an increase in both maximum water uptake and absorption rate. More analysis is needed to explain the improved maximum water uptake and the two absorption mechanisms that were observed in our system. Nevertheless, it is clear that the simple infusion of salts into the cellulose sponge provided significant improvements that are valuable in the context of atmospheric moisture clean water generation.

**Figure 4.** Proposed physical model of the salt-cellulose sponge system. Cellulose sponge with minimal agglomeration.

### *3.4. Peltier Device Prototype*

To recover the absorbed moisture as clean water, a Peltier-based distillation unit was prototyped (Figure 5). The unit was equipped with three water recovery channels to maximize the transfer of water droplet from the evaporation to the condensation chambers. The primary channel was set directly under the acute end of the roof due where most of the vapor was accumulating. The secondary channel directly transferred moist air to the collection chamber using a small DC motor pump. Finally, a tertiary channel was set up across from the acute end of the slanted roof to capture the flow of the larger water droplets

**Figure 5.** Prototype of the Peltier-based clean water generator. (**a**) Side view showing the three water channels to transfer water from the top evaporation to bottom condensation/collection chambers. (**b**) front view. (**c**) View of the Peltier module.

The device uses a single piece of Peltier module (9 V, 2 A) in a small chamber with a diameter of 8 cm and generates ~5 mL of clean water/hour from atmospheric moisture. Although further optimization is necessary, we note that this system does not waste any sacrificial clean water, as in most reverse osmosis or water distillation systems. Further optimization is also possible, including the incorporation of solar-powered energy sources, development of a multi-module systems, and better material choice to encourage formation and flow of water droplets on the walls of the evaporation chamber.
