DFT study of proton transfer and hydrogen bonding in PVA/citric acid membranes

Scientists have unveiled a detailed quantum-mechanical portrait of how phosphoric acid molecules anchor themselves inside proton-conducting polymer membranes, revealing cooperative binding effects so strong that they more than double the stability of a single acid attachment, and demonstrating that water-soaked, crosslinked environments can nearly halve the energetic barrier that protons must overcome to hop between binding sites. The study, published in Polymer Bulletin, uses density functional theory to dissect, atom by atom, the chemistry underlying poly(vinyl alcohol)/citric acid membranes doped with phosphoric acid — a class of inexpensive, bio-derived proton exchange membranes being explored as alternatives to fluorinated materials such as Nafion in fuel cells and electrochemical devices.

The research team, comprising Shahid Ali of the National University of Science and Technology (NUST) MISIS in Moscow, Faiq Umar of Abdul Wali Khan University Mardan in Pakistan, and Sikandar Azam of the University of West Bohemia in the Czech Republic, carried out twenty-three separate quantum-chemical calculations using the ORCA 6.1.1 software package. All geometries were optimized at the B3LYP-D3BJ/6-31G* level of theory, a hybrid functional augmented with Grimme’s dispersion correction to capture the weak but numerous van der Waals contacts that shape hydrogen-bond networks in polymer membranes. Beyond simple energy minimizations, the workflow included potential energy surface scanning, transition state verification, nudged elastic band analysis to map minimum-energy proton-transfer pathways, natural bond orbital analysis to quantify charge donation and bond orders, and electron localization function mapping to visualize where electrons actually reside during a proton hop. The centerpiece of the computational effort was a 65-atom model of a fully crosslinked, hydrated membrane segment, designed to approximate the real chemical environment of a working device.

The numbers that emerged from these calculations tell a striking story about the energetics of acid retention. The esterification reaction that chemically tethers citric acid crosslinks to the poly(vinyl alcohol) backbone was found to release 12.4 kilojoules per mole — modest but sufficient to lock the network in place. A single phosphoric acid molecule, by contrast, binds to the membrane with an energy of 78.1 kilojoules per mole, a substantially stronger interaction that reflects the three hydroxyl groups of H₃PO₄ acting as a versatile hydrogen-bond donor and acceptor platform. The most remarkable finding, however, concerns cooperativity: when two phosphoric acid molecules bind simultaneously in a mutually reinforcing arrangement, the total binding energy climbs to 203.0 kilojoules per mole. This is far more than twice the single-molecule value, indicating that the first acid molecule reorganizes the local hydrogen-bond landscape in a way that dramatically strengthens the grip of the second — a cooperative stabilization effect that helps explain why acid-doped PVA membranes hold onto their proton carriers even under demanding operating conditions.

Proton mobility is the other half of the puzzle. A fuel cell membrane must not only retain its acid dopant but also allow protons to migrate through it, typically via the Grotthuss mechanism, in which protons relay along chains of hydrogen bonds rather than diffusing as heavy hydrated ions. In a gas-phase model, the barrier for a local proton transfer between binding sites was calculated at 37.8 kilojoules per mole. When the same proton-transfer event was modeled inside the fully crosslinked, water-containing 65-atom membrane environment, the barrier dropped to 19.5 kilojoules per mole — a reduction of nearly half — and the product of the transfer emerged 9.3 kilojoules per mole more stable than the reactant state. The physical interpretation is that surrounding hydrogen-bond donors and water molecules stabilize the charge-separated transition state, effectively lubricating the proton’s passage. The authors emphasize, with appropriate scientific caution, that these results describe a single elementary proton-hopping step and do not by themselves constitute proof of complete long-range Grotthuss conduction across a macroscopic membrane, though NBO bond-order analysis and electron localization function mapping both support the assignment of the event as a genuine, hydrogen-bond-mediated transfer.

The study also offers a quantum-mechanical account of a long-observed experimental phenomenon: partial acid loss during membrane operation. By simulating a twenty-step desorption pathway along the potential energy surface, the team found that pulling a phosphoric acid molecule out of the membrane costs 75.9 kilojoules per mole — an energetic wall high enough to prevent rapid acid leaching, but not so high as to make retention permanent. Hydration analysis added a complementary perspective, showing that water insertion into the acid-membrane complex is strongly favorable at a free energy change of minus 278.0 kilojoules per mole. Together, these results depict a membrane in which acid molecules are thermodynamically anchored yet dynamically exchangeable, rationalizing why experimentally measured acid retention is high but incomplete. The hydration energy also hints at why water plays a dual role in these systems, simultaneously promoting proton transfer and reshaping the binding equilibria that determine acid retention.

One of the study’s most practical contributions is its predictive spectroscopy. The DFT-predicted infrared vibrational frequencies reproduced every major experimental peak of the membrane system within 0 to 110 wavenumbers — a level of agreement that validates both the molecular models and the computational protocol. Because hydrogen bonding shifts characteristic vibrational bands in well-understood ways, this agreement means researchers can now use the calculated spectra as a fingerprint to interpret real membrane samples, distinguishing between ester crosslinks, free hydroxyl groups, and various acid-bound configurations. For a field in which membrane characterization often relies on indirect inference from broad infrared envelopes, having a validated quantum-chemical reference map is a meaningful methodological advance.

The broader context makes the work timely. Proton exchange membranes are the beating heart of hydrogen fuel cells, electrolyzers, and a growing range of electrochemical technologies central to decarbonization strategies. The incumbent material, Nafion, is a perfluorosulfonic acid polymer with excellent conductivity but significant drawbacks: high cost, fluorine-related environmental concerns, and declining performance at elevated temperatures or low humidity. Polybenzimidazole membranes doped with phosphoric acid have emerged as a high-temperature alternative, and bio-derived systems based on poly(vinyl alcohol) crosslinked with citric acid offer a cheaper, greener route with tunable chemistry. Understanding, at the molecular level, exactly how phosphoric acid binds, cooperates, retains, and conducts in such matrices is a prerequisite for rational design rather than trial-and-error formulation. The present study supplies precisely that understanding for the PVA/citric acid/H₃PO₄ system.

The technical apparatus behind the findings reflects the state of the art in computational chemistry. The B3LYP functional, originally parameterized by Becke and by Lee, Yang, and Parr, remains one of the most widely validated approaches for hydrogen-bonded organic systems, while the D3BJ dispersion correction of Grimme and coworkers is essential for correctly capturing the stacking and contact distances in dense hydrogen-bond networks. The 6-31G* basis set, introduced by Hehre, Ditchfield, and Pople, provides polarization functions necessary for describing the directional character of hydrogen bonds. Harmonic vibrational frequencies were evaluated with standard scaling factors of the kind benchmarked by Scott and Radom, and the climbing-image nudged elastic band method of Henkelman, Uberuaga, and Jónsson enabled rigorous location of the proton-transfer saddle point without biasing the pathway. Counterpoise-style interaction energy corrections in the spirit of Boys and Bernardi, natural bond orbital analysis following Glendening, Landis, and Weinhold, and electron localization function topology in the tradition of Silvi and Savin round out a methodology designed to cross-check every energetic claim with independent electronic-structure evidence.

The implications extend in several directions. For membrane engineers, the demonstration that cooperative acid binding dominates the energetics suggests that maximizing the density of favorable multi-acid binding motifs — through crosslink density, acid loading, or backbone functionalization — could enhance retention without sacrificing the loosely held, mobile acid fraction needed for conduction. For computational chemists, the finding that a realistic hydrated, crosslinked environment halves the proton-transfer barrier underscores the dangers of interpreting gas-phase cluster calculations as proxies for condensed-phase behavior. And for the hydrogen economy at large, the work contributes to a growing portfolio of computational tools that can screen membrane chemistries in silico before expensive synthesis campaigns begin. The authors note that the data supporting the study are available from the corresponding author upon reasonable request, and that the project received no external funding, with ORCA access provided for academic use through the development team led by Frank Neese at the Max Planck Institute Mülheim.

As hydrogen fuel cells and electrolyzers move from niche applications toward infrastructure-scale deployment, the molecular details of proton conduction in next-generation membranes will matter more, not less. This study’s atomically resolved picture — cooperative binding energies of 203 kilojoules per mole, hydration-softened proton-transfer barriers of under 20 kilojoules per mole, and spectroscopic predictions accurate to within about a hundred wavenumbers — gives membrane designers quantitative targets to aim for and mechanisms to exploit. It is a reminder that some of the most consequential questions in clean energy technology are decided at the scale of a single proton crossing a single hydrogen bond.