Faster Atmospheric Water Harvesting Without Salt: Why a Near-Neutral Hydrogel Surface Matters

Original Authors: Wenjuan Ma, Wei Zhang, Renke Wei, Shiyu Wen, Xiaoyang Duan, Zhenao Gu, Chenghai Lu, Jiaxing Xu, Chengzhi Hu, Jiuhui Qu

Original paper is accessible at: https://doi.org/10.1016/j.watres.2026.125330

The kinetics problem in sorption-based AWH

Sorption-based atmospheric water harvesting (SAWH) works across a wider humidity range than condensation or fog collection. That is why it has attracted so much attention. Its daily water output, however, depends on more than how much water a sorbent can hold. It depends just as much on how quickly the sorbent can fill up and release water, because a fast sorbent can complete more cycles per day.

Hydrogels are attractive sorbents. They are flexible, tunable, and can hold a lot of water. Their sorption and desorption are mediated by hydrogen bonds, though, and they tend to be slow. The most common fix is to load them with hygroscopic salts such as LiCl or CaCl₂. That improves performance but brings familiar problems: salt leakage, corrosion, and performance loss over time.

This paper takes a different route. Instead of adding salt, the authors ask a more basic question:

What happens to a water molecule before it is captured, and can the polymer surface be designed to make that step faster?

The overlooked step: getting close enough

Most discussions of sorption focus on the binding event, meaning the hydrogen bond that forms between water and a functional group on the sorbent. A hydrogen bond is only about 2.7 Å long, though. Before it can form, a water molecule moving randomly in the gas phase must first get close enough to the surface, and in the right orientation.

Water is a polar molecule. Its oxygen end is slightly negative and its hydrogen end slightly positive. When it approaches a charged surface, it rotates to align its dipole with that charge. The authors argue that this approach-and-reorientation step strongly affects sorption kinetics and has been largely overlooked.

If the surface charge controls how water approaches, then tuning that charge should tune the speed of water harvesting.

A hydrogel with adjustable surface charge

To test this, the authors built a polyelectrolyte complex hydrogel, which they call the water harvesting gel (WHG), from two chitosan derivatives with opposite charges:

• Carboxymethyl chitosan (CMCS) carries negatively charged carboxyl groups (zeta potential about −31.9 mV).

• Quaternary chitosan ammonium salt (QCS) carries positively charged quaternary ammonium groups (about +59.6 mV).

Mixing the two in different ratios lets the charges partially neutralize each other. The leftover charge sets the net surface charge of the gel. Across the series, called WHG-10 to WHG-90 (the number is the QCS percentage), surface charge ranged from about −22.5 mV to +51.3 mV.

The gels were directionally frozen in liquid nitrogen and freeze-dried. This produced vertically aligned channels that help moisture move from the surface into the gel. Conventional freezing in a refrigerator produced only random, tortuous pores.

The material contains no hygroscopic salt. It is chitosan-based and intrinsically hygroscopic.

The result: speed peaks near zero charge

All gels were tested for two hours at 90% RH and about 25 °C, with weight recorded continuously. The relationship between surface charge and water uptake was not linear:

• Starting from a negative surface, uptake rose as the charge moved toward neutral.

• It peaked at 1.08 g/g in 2 hours near electrical neutrality (about −10 mV).

• It then dropped sharply to a minimum of 0.75 g/g at about +33 mV.

• At even higher positive charge, uptake partially recovered but stayed below the near-neutral value.

The near-neutral gel was therefore about 44% faster than the slowest one. A sample with surface charge even closer to zero (+0.6 mV) reached 1.12 g/g, which supports the trend.

Several checks confirm that this is a surface kinetics effect rather than a change in capacity or internal transport:

• Equilibrium capacity barely changed. Sorption isotherms of the near-neutral and negatively charged gels were similar in shape and final capacity. The charge mainly affects how fast water arrives, not how much the gel can ultimately hold.

• Bulk diffusion was comparable. Fick diffusion modeling of negative, neutral, and positive samples showed similar internal diffusion behavior.

• Thinner gels amplified the effect. Reducing thickness from 5.0 mm to 1.0 mm increased performance, with little further gain at 0.5 mm. That is consistent with the surface, not the interior, controlling the rate.

• QCM-D measurements on thin films showed onset sorption rates on the near-neutral surface more than double those on a strongly positive surface (about 1.02 × 10⁻⁴ vs 0.49 × 10⁻⁴ g g⁻¹ s⁻¹).

• Temperature measurements showed the gel warming slightly above the surrounding air during uptake. That indicates heat release from adsorption rather than simple condensation.

What the simulations show

Molecular dynamics simulations help explain why near-neutral surfaces perform better.

More water arrives. Within 200 picoseconds, about 654 water molecules per nm² reached the near-neutral surface, compared with 555 on the negative surface and 482 on the positive surface.

Charged surfaces “grab” water early. On strongly positive or negative surfaces, water molecules start interacting with the surface from about 17–17.5 Å away. They rotate their dipoles early, with oxygen pointing toward a positive surface and hydrogen toward a negative one. Fewer of them eventually settle on the surface.

Near-neutral surfaces let water come closer. On the near-neutral surface, interaction begins only within about 10 Å. Water molecules move freely closer to the surface before being stabilized, and the system energy drops by 42.7%, indicating stronger stabilization.

Hot spots between opposite charges. Mapping water positions on the surface showed that about 69% of adsorbed molecules sit in “near-neutral” regions between positive and negative groups, compared with 12% near the positive groups and 19% near the negative ones. Neighboring opposite charges seem to work together: they repel water less on approach and then stabilize it more effectively once it arrives.

In situ Raman spectroscopy added a further detail. Water’s hydrogen-bonding configuration at the interface shifts as the surface charge moves from negative to positive, consistent with the flip in molecular orientation seen in the simulations.

Performance under realistic conditions

Using the near-neutral WHG-25, the authors measured 2-hour uptake across a range of humidities at about 25 °C:

RH	2-hour uptake
33%	0.12 g/g
43%	0.16 g/g
61%	0.52 g/g
80%	0.82 g/g
90%	1.08 g/g
95%	1.18 g/g

The material performs best at moderate to high humidity. Its 2-hour performance from 60–95% RH compared favorably with several pure hydrogels and some gel–salt composites reported in the literature.

Other practical findings:

• Easy regeneration. Lowering the drying temperature from 90 °C to 70 °C barely changed performance. At 35 °C, capacity remained at 91.7%. Dropping RH from about 90% to about 20% released roughly 80% of the water within one hour.

• Cycling stability. Five cycles of 2-hour uptake and 1-hour solar desorption showed stable performance.

• Salt aerosol tolerance. In a sea-salt aerosol atmosphere, performance stayed comparable over five cycles. Surface chemistry and zeta potential were essentially unchanged afterwards.

• Purification test. When vapor was harvested above synthetic raw water containing NaCl and methyl orange, the collected water was free of the salt and dye.

Outdoor test

Two samples were left outdoors overnight in Beijing (22:00 to 07:00) at 14.7–18.0 °C and 74.4–84.0% RH. Over nine hours, the gel captured 0.67 g/g. A carbon-black-coated aluminum box under one sun then released more than 62% of that water within 30 minutes.

The collected water had Ca²⁺, Mg²⁺, Cl⁻, TOC, and total nitrogen concentrations one to two orders of magnitude below WHO drinking water guidelines.

Open questions

• Scale. The tests used small samples (centimeter-scale, 1 mm thick). Thin films are where this surface effect matters most, so the design challenge will be maintaining it in larger, thicker devices or engineered high-surface-area architectures.

• Low humidity. Two-hour uptake drops sharply below about 60% RH (0.12–0.16 g/g at 33–43%). This material is better suited to humid or night-time harvesting than to arid daytime conditions.

• Durability. Five cycles is a short-term demonstration. Long-term stability of the chitosan network and its surface charge balance under repeated field cycling remains to be shown.

• Water quality scope. The outdoor analysis covered major ions, TOC, and total nitrogen. A fuller panel, including microbiology and any leaching from the polymer itself, would strengthen claims of drinking water suitability.

Why this paper matters

The main contribution of this work is a design principle, not a record number. It shows that surface charge controls how quickly water molecules approach and settle on a sorbent, and that a near-neutral surface with closely mixed positive and negative groups gives the fastest uptake.

That matters for the field in two ways. First, it offers a route to faster sorbents without hygroscopic salts, which avoids leakage and corrosion. Second, because daily water output in SAWH is often limited by kinetics rather than capacity, a principle that speeds up uptake without sacrificing capacity directly supports more cycles per day.

The approach is not limited to chitosan. Balancing surface charge could become a general tool for designing the next generation of fast, salt-free AWH sorbents.