Chemical Capacitor: Nanostructured Charge Storage

• Chemical Capacitor (CC) is a nanoscopic layered device that stores charge nonvolatily by harnessing built-in chemical potential differences in a RED|SEP|OX stack.
• The charge transfer and doping levels are tunable by adjusting geometry, separator thickness, and material chemistry, which can influence superconductivity and other electronic phenomena.
• CC technology bridges the gap between electrostatic and electrochemical energy storage, offering practical insights for designing nonvolatile doping platforms and advanced capacitor materials.

Chemical Capacitor (CC) most directly denotes a nanoscopic stack $RED|SEP|OX$, where $RED$ is a reducing layer, $OX$ is an oxidizing layer, and $SEP$ is an electrically insulating separator. In that usage, charge is stored nonvolatily by aligning chemical potentials across $SEP$, so that electrons flow from $RED$ to $OX$ until electrostatic and band-energy costs balance the chemical-energy gain. Closely related literature uses chemical capacitance for the thermodynamic response $C_\mu = e^2 \partial N_i/\partial \mu_i$ or, in bulk electrolytes, $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$; electrochemical-capacitor studies also use CC as an abbreviation for capacitive devices with non-Debye, correlation-governed, or double-layer behavior (Wolański et al., 28 Jul 2025, Molak, 2013, Usler et al., 20 Jun 2026, Allagui et al., 2022).

1. Definitions and conceptual scope

In the self-biased heterostructure sense, a CC is a layered device in which a strong oxidizer and a reductor are separated by an insulating spacer, optional substrate, and optional top caps. No external voltage is required: the driving force is the built-in chemical potential difference $\Delta \mu$. Charge injection occurs without compositional change, both layers become mutually doped, and the effect is permanent while the stack remains intact and separation is maintained. This distinguishes the CC from a conventional physical capacitor, which stores electrostatic energy through an applied voltage, and from an electrochemical cell, which converts chemical energy to electrical energy by redox reactions at electrodes with ionic transport through an electrolyte. It also differs from ionic-liquid gating and electrostatic capacitor gating, which require external bias and are limited by dielectric breakdown, leakage, mobile ions, or transient near-surface charging (Wolański et al., 28 Jul 2025, Szudlarek et al., 2023, Jezierski et al., 2022).

A second, older usage emphasizes response functions rather than heterostructure architecture. In that thermodynamic formulation, chemical capacitance is the change in carrier number under a change in chemical potential, $RED$0, with additive contributions from free carriers and trapped carriers. In electrolyte theory, the corresponding bulk quantity is $RED$1, which enters the low-frequency capacitance of closed systems when charging requires salt adsorption from a finite reservoir. These usages are not contradictory; they address different levels of description of how chemically driven charge storage is established and limited (Molak, 2013, Usler et al., 20 Jun 2026).

2. Self-biased heterostructures and governing relations

The minimal CC architecture is $RED$2. If the top of $RED$3's valence band lies above the bottom of $RED$4's conduction band, electrons flow from $RED$5 to $RED$6 according to

$RED$7

The transferred charge per $RED$8 unit is then set by the competition between chemical-energy gain and capacitive cost. A minimal energetic picture writes the chemical gain as

$RED$9

while the electrostatic cost grows with separator thickness $OX$0 as $OX$1. This yields the characteristic decrease of $OX$2 with increasing separator thickness $OX$3 and interlayer spacing $OX$4, with $OX$5. The corresponding capacitor relations are the familiar

$OX$6

and, in the nanoscale stack language,

$OX$7

with $OX$8 and $OX$9 acting analogously to self- or quantum-capacitance terms and $SEP$0 the geometrical capacitance (Wolański et al., 28 Jul 2025, Jezierski et al., 2022).

This framework immediately explains several design rules. Large $SEP$1 favors large $SEP$2; high-permittivity and thin separators raise $SEP$3; and the total nanoscale capacitance is limited by both geometry and DOS, $SEP$4. Ferroelectric separators add a further control parameter through polarization $SEP$5, with

$SEP$6

so that switching $SEP$7 shifts the final charge state. In the AgF$SEP$8/MgO prototype, the same logic appears in a chemically specific form: an MgO donor layer transfers electrons across fluoride spacers to a flat AgF$SEP$9 monolayer, and the number $SEP$0 of stoichiometric fluoride layers between donor and acceptor directly controls the effective plate separation and hence the doping level (Wolański et al., 28 Jul 2025, Jezierski et al., 2022).

3. Materials realizations and tunable charge transfer

The most extensively quantified CC realizations span magnetic fluoride monolayers, hydride monolayers, cuprate-related oxides, charge-ordered halides, and highly oxidizing molecular or extended $SEP$1 layers. In all of them, the transferred charge is tuned by chemistry, separator thickness, and, in some cases, ferroelectric polarization.

Representative values reported for AgF$SEP$2, hydrides, and broader $SEP$3 stacks are summarized below (Jezierski et al., 2022, Szudlarek et al., 2023, Wolański et al., 28 Jul 2025).

System	Reported charge transfer or doping	Reported consequence
AgF$SEP$4 monolayer with MgO donor and fluoride separator	$SEP$5 at $SEP$6, $SEP$7 at $SEP$8, $SEP$9 at $RED$0; fine tuning to $RED$1–$RED$2 with additional MgO layers	Access from underdoped to overdoped electron regime
$RED$3	$RED$4 per Ru	Largest computed charge transfer per $RED$5 unit
$RED$6	$RED$7	$RED$8
$RED$9	$OX$0	Metallization without structure collapse up to that level
CaCuO$OX$1 monolayer in CC geometry	up to $OX$2 with Li cap; up to $OX$3 with $OX$4 cap	Fine tuning toward the cuprate “optimal” $OX$5 carriers/Cu
CsAuCl$OX$6 with Li cap on Na$OX$7MgF$OX$8 separator	$OX$9	Metallization of the charge-density-wave solid

Within this set, the AgF$C_\mu = e^2 \partial N_i/\partial \mu_i$0 case is unusually explicit about geometric control. Flat AgF$C_\mu = e^2 \partial N_i/\partial \mu_i$1 layers on tetragonal RbMgF$C_\mu = e^2 \partial N_i/\partial \mu_i$2 use $C_\mu = e^2 \partial N_i/\partial \mu_i$3 Å, and the electron doping decreases with increasing spacer thickness exactly as the $C_\mu = e^2 \partial N_i/\partial \mu_i$4 picture suggests. Direct contact to MgO overdopes the layer, whereas $C_\mu = e^2 \partial N_i/\partial \mu_i$5 places the system near $C_\mu = e^2 \partial N_i/\partial \mu_i$6–$C_\mu = e^2 \partial N_i/\partial \mu_i$7, and additional MgO layers farther from the AgF$C_\mu = e^2 \partial N_i/\partial \mu_i$8 plane provide quasi-continuous tuning around that value. By contrast, hole doping of AgF$C_\mu = e^2 \partial N_i/\partial \mu_i$9 is difficult because generating Ag(III) sits at the verge of fluoride stability and the resulting holes avoid the in-plane $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$0 manifold (Jezierski et al., 2022).

The broader CC survey extends the accessible range well beyond the AgF$$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$1 prototype. It reports $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$2 per F in $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$3, $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$4 per PtF$$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$5 in $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$6, and $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$7 per PbO$$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$8 in $$C_\mathrm{chem} = e^2 V_\mathrm{bulk}(\partial c_0/\partial \mu)$$9. Switching from non-ferroelectric Na$\Delta \mu$0MgF$\Delta \mu$1 to ferroelectric Sr$\Delta \mu$2Ti$\Delta \mu$3O$\Delta \mu$4 increases $\Delta \mu$5 by about $\Delta \mu$6 for the same oxidizer chemistry, for example $\Delta \mu$7 for RuO$\Delta \mu$8. This suggests that CC performance is not governed by chemistry alone, but by the full series combination of redox mismatch, separator response, and nanoscale geometry (Wolański et al., 28 Jul 2025).

4. Electronic structure, superconductivity, and electrochemical implementations

The materials interest of CCs derives from the fact that the stored charge is not merely electrostatic; it reorganizes low-energy manifolds, magnetic exchange, transport gaps, and electron-phonon or spin-fluctuation spectra. In flat AgF$\Delta \mu$9, the relevant low-energy manifold is the Ag $RED$00 orbital on a square lattice. Electron doping fills the upper Hubbard band, while attempted hole-doping routes populate $RED$01 or apical-F states instead. The undoped flat monolayer on RbMgF$RED$02 has $RED$03 meV; $RED$04 decreases roughly linearly with electron doping and falls to $RED$05 meV at $RED$06. Around $RED$07–$RED$08, $RED$09 to $RED$10 meV. Weak-coupling RPA/FLEX analysis finds leading singlet $RED$11 pairing with a maximum at $RED$12 holes, and prior estimates for flat epitaxial AgF$RED$13 place the optimal $RED$14 near $RED$15–$RED$16 K under a magnetic-glue scenario (Jezierski et al., 2022).

Hole-doped ionic hydrides provide a distinct, BCS-like branch of CC-enabled superconductivity. In truncated CC calculations, LiH, MgH$RED$17, and NaH monolayers are stabilized by inert support and separator layers while hole density is varied until imaginary phonons appear. LiH reaches $RED$18 in several fluoride-perovskite sandwiches, with $RED$19 of order $RED$20–$RED$21. For $RED$22, the reported values are $RED$23 and $RED$24 K; the double-sandwich $RED$25 yields $RED$26 K at the same hole density. NaH reaches $RED$27 in $RED$28, while MgH$RED$29 reaches $RED$30 with $RED$31 K (Szudlarek et al., 2023).

A broader CC program extends this to other electronically active layers. CaCuO$RED$32 monolayers can be doped to $RED$33 or $RED$34, CsAuCl$RED$35 is metallized at $RED$36, and truncated CC calculations report $RED$37 K in graphene near $RED$38, $RED$39 K in NaCl monolayers at $RED$40, and $RED$41 between $RED$42 and $RED$43 K in Mg$RED$44Cu$RED$45H$RED$46 over stable doping ranges (Wolański et al., 28 Jul 2025).

Electrochemical and interfacial implementations use the same capacitive logic in a different regime. A permeable biased gate inserted between Zn and Pt electrodes shifts the local electrolyte potential and modulates ionic current with a coupling factor $RED$47, transconductance $RED$48, and $RED$49 over a $RED$50 V gate swing (Grebel et al., 2015). A “water-only” membrane-electrode assembly using activated carbon, nanodiamond, and pure water identifies an interfacial water layer of thickness $RED$51 nm and an optimal pore radius $RED$52 nm; with a $RED$53 mm separator it reaches $RED$54 and $RED$55 (Melnik et al., 2022). A microemulsion EDLC that is mostly water by mass shows an electrochemical stability window of approximately $RED$56 V on hydrophobic glassy carbon, while symmetric activated-carbon devices operate stably to $RED$57 V with $RED$58, $RED$59, and $RED$60 Coulombic efficiency for over $RED$61 cycles (Hughson et al., 2020). Optically controlled supercapacitors with semiconductor-embedded active carbon add a further degree of control: SiC-containing electrodes show a relative capacitance increase as large as $RED$62, or $RED$63 when the illuminated area is taken into account, and the model attributes the optical contribution to an optically induced dipole (Grebel, 2021).

5. Chemical capacitance, ionic correlations, and impedance formalisms

In the thermodynamic formulation, chemical capacitance is a bulk or local response function rather than a specific stack geometry. For conduction-band electrons in the non-degenerate limit,

$RED$64

and localized states add

$RED$65

This framework was used for manganite-based ceramics, where $RED$66 rises from $RED$67–$RED$68 at $RED$69–$RED$70 K to values reaching $RED$71 at $RED$72–$RED$73 K, with Arrhenius segments interpreted through thermally activated carriers. The same study emphasizes a geometrical distinction: purely chemical capacitance scales with volume $RED$74, while pure electrostatic capacitance scales with $RED$75 (Molak, 2013).

For closed electrolytes, the corresponding quantity is

$RED$76

When EDL charging requires salt adsorption from the bulk, the effective series contribution becomes

$RED$77

and, for sufficiently large $RED$78,

$RED$79

In the impedance of a charged flat-plate EDL capacitor, the intermediate-frequency slanted line is attributed to ambipolar salt diffusion, and the correct equivalent element is a Warburg short for antisymmetric salt perturbations or a Warburg open for symmetric perturbations. The resulting framework links the Warburg prefactors directly to the ambipolar diffusion coefficient, bulk chemical capacitance, and differential charge efficiency of the EDLs (Usler et al., 20 Jun 2026).

Exact many-body lattice models show that this ionic problem is intrinsically correlation-sensitive. In the one-dimensional Coulomb lattice fluid capacitor, the Hamiltonian contains only one-dimensional Coulomb interactions and steric exclusion, yet the exact transfer-operator solution exhibits overscreening, layered charge-density oscillations, and strong oscillations in differential capacitance as a function of voltage. The local charge profile decays with correlation length

$RED$80

and the exact solution can exceed the Helmholtz benchmark $RED$81; mean-field theory reproduces only envelope trends and misses layering, overscreening, and plateau structures (Demery et al., 2012).

Frequency-domain modeling of electrochemical capacitors adds a complementary response-theory layer. Fractional-order Cole-Cole, Davidson-Cole, and Havriliak-Negami impedances,

$RED$82

and q-deformed local evolution models were benchmarked on a $RED$83 V, $RED$84 F commercial supercapacitor. In the reported fits, fractional-order models achieved RMSE $RED$85 in frequency-domain data, while the Debye fit gave RMSE $RED$86. The same work stresses that for non-ideal devices $RED$87 with a constant $RED$88; complex $RED$89 or model-based relaxation functions are required (Allagui et al., 2022).

6. Limits, controversies, and experimental tests

The principal limitation of the self-biased CC is not the existence of a chemical potential difference, but maintaining it without shorting it chemically or ionically. Large internal fields can drive ion migration and neutralization; in $RED$90, the computed energy gain for converting the stack into an equivalent $RED$91 slab is $RED$92 eV per LiF unit, which signals a strong thermodynamic drive toward chemical neutralization if defects permit ion transport. More generally, large $RED$93-fields in ultrathin separators can cause leakage or breakdown, symmetry breaking and Peierls-like distortions can open gaps, and $RED$94 for macroscopic separations, so the CC is intrinsically nanoscale (Wolański et al., 28 Jul 2025).

AgF$RED$95 provides a concrete illustration of chemically specific limits. Electron doping is favored by the very large work function of AgF$RED$96, but hole doping that would place holes into the $RED$97 manifold is difficult because Ag(III) formation pushes fluoride anions to the verge of instability. In every explored hole-doping setup, the holes avoid the $RED$98 band. Computation itself introduces additional caution: non-symmetric slabs used for economy in some LiF/MgO scans overestimate electron doping by about $RED$99–$OX$00 relative to symmetric sandwich slabs (Jezierski et al., 2022).

Impedance-based interpretations of chemical capacitance also contain specific pitfalls. In biased flat-plate EDL capacitors, the width $OX$01 of the slanted-line region is not an EDL resistance, and the slope $OX$02 of that line is not a universal measure of $OX$03. The slanted line is instead governed by ambipolar salt diffusion, with Warburg-short or Warburg-open symmetry depending on whether the salt perturbation is antisymmetric or symmetric. This matters because $OX$04 saturates at large bias, vanishes at large packing fraction, and is constrained by finite salt inventory in closed systems (Usler et al., 20 Jun 2026).

The experimental program implied by the CC literature is correspondingly direct. For AgF$OX$05, the proposed route is to build LiF–MgO heterostructures with a single AgF$OX$06 monolayer, vary $OX$07 from $OX$08 to $OX$09, and measure doping and Fermi-surface evolution by spectroscopies such as ARPES, XPS, and EELS while correlating them with Ag–F–Ag angles, Ag–F bond lengths, and AFM signatures. For the broader class of CCs, defect-free ultrathin separators, robust ferroelectric control, and chemically compatible $OX$10 pairs remain the central materials constraints. A plausible implication is that the field will continue to bifurcate into two technically connected directions: chemically self-biased heterostructures for nonvolatile doping of quantum materials, and chemically defined capacitance formalisms for interpreting electrolyte-limited, correlation-limited, or defect-limited charge storage.