a Upon resonant MIR excitation with the energy of hν_O–H, the O–H stretch amplitude is increased to facilitate proton transfer. b MIR light excites the O–H stretch vibration from the ground state to an excited state. c Schematics of the corresponding EIS spectra with (red) and without (blue) MIR irradiation. Inset: photo of a 0.4-mm thick protonated polycrystalline BZY pellet. In (a) the purple, red and orange spheres represent B-site (Zr or Y), O and H atoms respectively, while the light blue squares denote BO₆ (B = Zr, Y) octahedra.

We demonstrate reversible switching between high and low bulk and grain boundary (GB) resistances controlled by MIR irradiation measured by electrochemical impedance spectroscopy (EIS) while maintaining the sample temperature constant. The enhancement effect in proton conductivity is 2-3 times greater than the estimated IR heating effect. Bandpass filtering was further applied to highlight the enhancement effects of MIR wavelength resonant to the O–H stretch frequency. We rationalize our observation by modelling the excitation of the O–H stretch vibration in the potential energy surface (PES) of the proton.

a Schematics of the electrical conductivity measurement setup. The MIR radiation covers the wavelength range of O–H stretch vibration. b Vibrational profile of O–H stretch vibration band (ν_O–H) characterized by DRIFTS (red). The orange shadow shows the typical emission spectrum of the CW MIR light. c Typical EIS spectra at 160 °C in wet N₂ with (red) and without (blue) MIR i0rradiation at an effective MIR power density (p) of 195.2 mW cm⁻². The insets illustrate the magnification of bulk features near 150 kHz, and the equivalent circuit used for fitting. The reversibility of proton conductivity enhancement is represented by single-frequency Z’ at (d) 150 kHz for bulk (brown) and (e) 1 kHz for GB (turquoise). The yellow shadows highlight the time intervals for MIR irradiation. The dashed red and grey lines show steady-state Z’ averages with and without irradiation respectively; error bands indicate the corresponding standard deviations. f Comparison of bulk and GB proton conductivities deconvoluted from c and corresponding enhancement ratio of conductivity (Δσ/σ). The error bars in (f) incorporate standard deviations from three parallel measurements and fitting errors per sample across three samples prepared using the same method. Source data are provided in the Source Data file.

Enhanced proton conductivity upon MIR irradiation in BZY

Figure 2c shows representative EIS spectra at 160 °C with and without MIR irradiation. The corresponding effective MIR power density (p), defined as the output power density of the MIR light within the wavelength range of O–H stretch vibration, is 195.2 mW cm⁻². Figure 2c and Supplementary Fig. 12, 13 indicate that the real part of impedance (Z’ ) at 150 kHz and 1 kHz reflect bulk and GB proton conduction features in the EIS spectra. The single-frequency impedance at these two frequencies shows that MIR irradiation resulted in an immediate drop in Z’ for both features (Fig. 2d, e). When the MIR light was switched on, a small fluctuation of Z’ was observed within the first 100 seconds, as a result of temperature modulation by the thermostat to compensate for the heat induced by MIR irradiation. After around 100 s, Z’ reached stable values. Upon removal of MIR irradiation, Z’ returned to their original values, confirming the reversibility of the enhancement effect on proton conduction through MIR irradiation. Figure 2f compares the proton conductivities after the MIR light was switched on or off for 100 s.

The EIS spectra in Fig. 2c were then fitted to an equivalent circuit consisting of three serial R(CPE) elements, assuming proton conduction as the origin of the impedance. Each R(CPE) element represents the contribution from the bulk, GB, and electrode process to the overall impedance. From the fitting parameters of the EIS spectra, an obvious decrease in both bulk and GB resistances (R) under MIR irradiation was revealed, as presented in Fig. 2f. The enhancement in proton conductivity under MIR irradiation is then quantified with the enhancement ratio or percent change of conductivity (Δσ/σ):

where σ^on, R^on, and σ^off, R^off denote the conductivity and resistance, with and without MIR irradiation, respectively. The enhancement ratios of bulk and GB conductivities, denoted as (Δσ/σ)_Bulk and (Δσ/σ)_GB, are as high as 36.8% and 53.0% (Fig. 2f). The GB conductivity (σ_GB) is defined as the specific value derived from the brick-layer model (Supplementary Note 6) to account for the microstructure^52,53. Additionally, no significant changes were observed in the EIS spectra measured in wet N₂, ambient air, and dry N₂ atmospheres (Supplementary Fig. 5), indicating a negligible atmosphere effect at 160 °C. The changes in impedance were not significant when changing water partial pressure either with or without MIR irradiation, suggesting that the conductivity is not a surface protonic effect⁵⁴. Nevertheless, the enhancement was prominent when the MIR light was switched on, confirming that the measured enhancement in electrical conductivity originates from bulk transport.

To explore the potential of enhancement in proton conductivity through MIR excitation in BZY, Δσ/σ was evaluated versus various p by adjusting the distance (d) between the sample and MIR light source. p was calibrated as a decreasing function of d (Supplementary Fig. 10b). Δσ/σ for both bulk and GB were found to linearly scale with p (Fig. 3a, b), suggesting that Δσ/σ is linearly correlated with the number of MIR photons incident onto the samples. The intercepts of both lines with the x-axis are >0, which may arise from the absorption of MIR radiation by H₂O molecules in humid atmosphere. To the best of our knowledge, the currently available p for MIR laser working in CW mode can be up to 16 W cm⁻²⁵⁵, and such power density is also achievable with an IR furnace^34,56. Then, σ ^on/σ ^off for both bulk and GB could be further enhanced by 30–50 times. Consequently, the bulk and GB conductivities are expected to be above 10⁻³ S cm⁻¹ and 10⁻⁵S cm⁻¹, respectively. Such enhancement in conductivity is equivalent to heating the sample to 335 °C for the bulk and 275 °C for the GB, while maintaining sample temperature at 160 °C. Therefore, it is possible to achieve an over-ten-fold enhancement in proton conductivity through MIR irradiation.

Fig. 3: Enhancement in proton conductivity versus MIR irradiation intensity.

Δσ/σ for (a) bulk (orange) and (b) GB (turquoise) versus p controlled by adjusting the distance between the sample and MIR light source (d). The sample was maintained at 160 °C. Shadowed areas in (a) and (b) illustrate the 95% confidence band of the fitted curves. Comparison of Δσ/σ for (c) bulk and (d) GB observed in the experiments (closed circles) and estimated with the radiative heating effect (open circles connected with dashed line). The error bars in (a–d) incorporate standard deviations from three parallel measurements and fitting errors per sample across three samples prepared using the same method. Source data are provided in the Source Data file.

Full size image

IR irradiation is known to induce heat, which may also cause enhancement in the ionic conductivity. Since our measurement setup cannot completely exclude the IR heating effect, here, we distinguish the contribution of heat to the observed enhancement in conductivity through an estimation of Δσ/σ by IR heating effect. It is assumed that the samples absorb all spectral components radiated from the MIR light source, and all the incident optical power is directly converted into heat. The consequent increase in temperature is estimated by considering the sample as a thermal resistor for thermal conduction⁴⁴, as elaborated in Supplementary Note 8. Depending on different values of d (corresponding to different power densities for heating), the increase in sample temperature is within 1–5 °C, and the corresponding results of Δσ/σ are plotted in Fig. 3c, d. The observed Δσ/σ for both bulk and GB due to MIR irradiation are about 2 times greater than the estimated results due to the IR heating effect. Therefore, IR heating effect is not a major contributor to the enhancement in proton conductivity. Instead, the excitation of the O–H stretch vibration by absorbing the MIR radiation is considered to dominate the enhancement in proton conductivity. To reach the same proton conductivity, the temperature rise and thermal expansion due to IR heating is still less than the effect caused by thermal heating (Supplementary Notes 8–9).

To confirm that the observed effect originates from enhanced proton conductivity, we further conducted the experiments on deuterated samples in D₂O-saturated N₂ (p(D₂O) = 0.02 atm) and dry samples in dry N₂ (p(H₂O) < 10^–4 atm). As shown in Supplementary Table 2, the enhancement ratios of both bulk and GB conductivities for the deuterated samples are 23.1% and 33.6%, approximately 0.6 times those of the protonated samples due to the heavier deuteron. For the dry samples, the enhancement ratios are only 3.91% and 6.54%. These results confirm that the enhancement in conductivity is dominated by contributions specific to protons.

Analysis of proton conductivity parameters with and without MIR irradiation

To illustrate the mechanism of enhanced proton conductivity under MIR irradiation, parameters for temperature-dependent proton conductivity with and without irradiation are quantified and compared. Using classical hopping theory, the prefactor σ₀ and activation energy E_a (also known as activation enthalpy) of Grotthuss-type proton conductivity σ in BZY are evaluated according to the Arrhenius relation^15,57:

where k_B is the Boltzmann constant, T is the absolute temperature in Kelvin, ν₀ is the attempt frequency, and ΔS is the activation entropy arising from the change in vibration frequencies at the initial and transition states during a local hopping process⁵⁸. Eq. (2) shows that σ₀ is strongly dependent on the entropic term exp(ΔS/k_B) and attempt frequency ν₀. Given the thermal activation feature of the proton hopping process, the hopping frequency ν is described by:

In practice, the variation trend of ν upon MIR irradiation can be evaluated from the fitting results of EIS spectra⁵⁹. The equivalent circuit model in Fig. 2c provides the values of resistance (R) and pseudocapacitance of CPE (C). Specifically, C = Y^1/nR^(1−n)/n, where Y and n are capacitance factor and ideality factor, respectively. Subsequently, we can obtain the characteristic frequency for bulk or GB processes, ω, using ω = 1/(RC)^15,59. The concerted effect of attempt frequency and entropy on proton conductivity was estimated similarly as a recent study on solid-state lithium-ion conductors by Li et al.⁵⁷, using a vibrational factor Q = ω₀exp(ΔS/k_B) = ωexp(E_a/k_BT), as derived from Eq. (3). Although ω and ω₀ are obtained from EIS and represents macroscopic measurements, they can reflect the changes in microscopic proton hopping frequency ν and attempt frequency ν₀, respectively^49,60,61.

If considering proton conduction with and without MIR irradiation at a fixed temperature T as two different processes following the Arrhenius relation, Δσ/σ can be expressed by combining Eq. (2) and (3):

where σ0on/σ0off denote the ratios of the prefactors and vibrational factors, and Δ?a=?aon−?aoff refers to the difference in the activation energies with and without MIR irradiation. These parameters indicate the changes in proton conduction properties under different irradiation states. According to Eq. (4), larger Δσ/σ at a given temperature can be achieved by decreasing the activation energy (ΔE_a < 0) or increasing the prefactor (σ0on/σ0off > 1). Δσ/σ for the bulk gradually decreases with increasing T, when the average thermal energy k_BT contributes more to activate proton hopping. However, in this work, we emphasize that the enhancement effect is larger at low temperatures, where k_BT is much lower than the activation energy for proton hopping, and excitations in atomic vibrations contributing to proton hopping would cause a significant enhancement in proton conductivity.

Figure 4a presents the Arrhenius plot of the bulk and GB proton conductivities with and without MIR irradiation at 130–200 °C. The corresponding ω and Q are plotted in Fig. 4b, c, respectively. Q were calculated at different temperatures as presented in Fig. 4c for further analysis. The comparison for the trends in E_a, σ₀, and resulting Q are presented in Fig. 4d, e. Quantitative comparison results based on Eq. (4) are summarized in Supplementary Table 5. Upon MIR irradiation, the E_a for bulk proton conduction (E_a,Bulk) decreased from 0.450 ± 0.006 eV to 0.409 ± 0.007 eV, with a statistically significant difference of ~0.04 eV. Notably, the change in E_a,Bulk for the deuterated BZY is 0.02 eV (Supplementary Fig. 7b). The different enhancement ratio (Supplementary Table 3) and activation barrier originate from the greater mass of deuteron and lower energy of O–D stretch vibration. Conversely, the E_a for GB proton conduction (E_a,GB) remains unchanged within the error (0.716 ± 0.016 eV and 0.698 ± 0.011 eV with and without MIR irradiation, respectively). The E_a data measured without MIR irradiation agree with previous literature^20,49,51. Despite a reduction in E_a,Bulk upon MIR irradiation, the photon energy and optical power density of the MIR radiation are much lower than those required to induce direct flow of protons or lattice atoms⁶², again indicating that the absorbed energy of MIR radiation excites the O–H stretch vibration. Notably, both σ₀ and Q exhibit opposite trends for bulk and GB upon MIR irradiation—for bulk proton conduction, the values decrease (σ0,Bulkon < σ0,Bulkoff and ?Bulkon < ?Bulkoff); but for GB proton conduction, the values increase (σ0,GBon > σ0,GBoff and ?0,GBon > ?0,GBoff) (Fig. 4d, e). σ0on/σ0on and Q^on/Q^off are roughly consistent for both bulk and GB (Supplementary Table 5). It can be concluded that the enhancement in bulk proton conductivity of BZY is mainly attributed to the reduction in E_a, accompanied by a decrease in σ₀ and Q. On the other hand, the increase in σ₀ and Q is the major reason for the enhancement in GB proton conductivity.

Fig. 4: Temperature dependence of the proton conductivity and corresponding parameters with and without MIR irradiation.

Arrhenius plots of (a) proton conductivity (σT) and (b) characteristic frequency of the bulk and GB processes (ω). c Vibrational factor (Q = ω₀exp(ΔS/k_B)) of bulk (circle) and GB (triangle) measured with (red) and without (blue) MIR irradiation at 130–200 °C. The calculated Q and their errors are represented by dashed lines and shadowed areas. Comparison of the corresponding E_a (orange), σ₀ (blue), and Q (green) of d bulk and (e) GB. Data in (a–e) are presented as mean values from three parallel measurements per sample across three samples prepared using the same method. The error bars in (a–e) incorporate the corresponding standard deviations and fitting errors. Source data are provided in the Source Data file.

We further discuss the origin of the distinct behaviours of σ₀ and Q for bulk and GB proton conductivities, by considering the possible effects of MIR irradiation on the trends for ν₀ and ΔS. For bulk proton conduction, if we take a constant ω₀ (since ν₀ is usually taken as a constant)^63,64 the entropic term exp(ΔS/k_B) changes similarly as σ₀ and Q, corresponding to a decrease of ΔS upon MIR irradiation. The decrease in both E_a and ΔS obeys the Meyer-Neldel rule, which states that the decrease in E_a is compensated by a contribution to decrease ΔS for bulk proton conduction^65,66. Such contribution to ΔS is associated with the excitations providing the energy for protons to overcome E_a, underpinning the role of exciting the O–H stretch vibration in enhancing bulk proton conductivity. In fact, ν₀ may vary upon the perturbation of the local chemical environment of protons⁶⁴, and it is necessary to consider the variation of ν₀ with and without MIR irradiation. The decrease in E_a, σ₀, and Q has been reported as a consequence of increased lattice anharmonicity and lattice softening, leading to a decrease in ν₀⁶⁴. Lattice softening also causes a decrease in ΔS. The increased lattice anharmonicity and lattice softening also indicate the correlation between atomic vibrations and the enhancement in proton conductivity. An origin of these effects has been proposed to be the excitation of atomic vibrations altering O–O separation, such as O–H wag and lattice vibrations, thus changing the shape of the potential energy surface (PES)^14,24,64. It has also been demonstrated that variation of PES caused by different vibration amplitudes accounts for changes in E_a values, where a large amplitude is accompanied by flat PES to facilitate solid-state ionic conduction²⁵. Given this, the excitation of atomic vibrations that can modulate the PES of protons can assist proton hopping. Furthermore, exciting O–H stretch vibration with infrared radiation has been shown to influence such atomic vibrations in some proton-conducting oxides, through the coupling of O–H stretch with low-frequency atomic vibrations, as observed by Spahr et al.^35,36 and Sakurai et al.⁶⁷. On the other hand, for GB proton conduction, Fig. 4e shows a slightly increased σ₀ upon MIR irradiation, indicating that ΔS increases regardless of constant or decreased ν₀. This trend can be attributed to the significant configurational entropy on the highly distorted GB region⁶⁵. The change in configurational entropy during proton conduction may be amplified by the excitation of lattice vibrations coupled with O–H stretch vibration, thus could explain the increase in ΔS for GB proton conduction.

While the aforementioned mechanisms may increase proton mobility, higher proton concentration represents another potential origin of the enhanced proton conductivity. Computational evidence has demonstrated that phonons can influence hydration entropy of acceptor-doped BaZrO₃^68,69, thus changing the proton concentration. However, these investigations focused specifically on lattice phonons rather than the O–H stretching vibration central to this work. Furthermore, as shown in Supplementary Fig. 5, the variation in conductivity of the protonated sample was not significant with changing water partial pressure at 160 °C, indicating kinetically limited hydration/dehydration processes at this temperature. On the contrary, the observed MIR-induced conductivity enhancement occurs instantaneously below 200 °C. Given this rapid response, we believe that our observation is not dominated by the light-induced effects on hydration enthalpy and the subsequently increased proton concentration.

Overall, the enhancement in proton conductivity by MIR irradiation may originate from the excitation of O–H stretch vibration. The mechanisms can be that the O–H vibration is excited to a higher energy state with a lower proton hopping barrier and larger vibration amplitude. Such excitation may further reinforce atomic vibrations related to oxygen ions, thus assisting proton hopping. In the following section, we will elaborate on the O–H vibration excitation mechanism.

Wavelength effect and suggested mechanism

To confirm that the enhanced proton conductivity originates from the excitation of O–H stretch vibration, we further analyse the wavelength effect of MIR radiation. Narrowband MIR radiations of two specific wavelengths are obtained by two different bandpass filters (25 mm diameter, 1 mm thick). As shown in Fig. 5a, the passing wavelength range of filter A is 2.95 ± 0.055 μm, which is close to the resonant frequency of O–H stretch vibration. For filter B, the passing wavelength range is 2.70 ± 0.060 μm, where only slight absorption was observed. The filtered MIR radiation power incident to the sample was maximized by fixing the filters between the MIR light source and the sample stage, since the radiation power through the filters is small and decreases with increasing d (Supplementary Fig. 10b). To identify the heating effect of contact heat transfer between the light source, filter, and sample stage, a control group was set up by replacing the filters with a stainless-steel plate (25 mm × 25 mm size, 1 mm thick). Δσ/σ for bulk and GB proton conductivities versus filters A and B and the control group are compared in Fig. 5b. The enhancement effect for filter B is insignificant compared to that of the control group; whereas for filter A, with passing wavelength matching with O–H stretch vibration, the enhancement effect is ~5 times and 3 times in bulk and GB conductivities of the control group, respectively. Such measured MIR light resonant enhancement effect is even greater compared to the estimated results in Fig. 3 (about 2 times of IR heating effect), since the IR heating effect is weaker with narrowband bandpass filtering. We hypothesize that the O–H stretch vibration absorbs the photons matching its frequency in MIR radiation. The resonant photon energy excites the vibrational energy to a higher vibrational state (Fig. 5c and Supplementary Fig. 20a), and promotes proton transfer in the lattice.

Fig. 5: Wavelength effect and suggested mechanism of MIR light enhanced proton conductivity.

a Representation of passing wavelength ranges of the MIR narrowband bandpass filters A (brown shadow) and B (green shadow). The range for filter A is closest to the resonant frequency of O–H stretch vibration. b Wavelength effect of MIR light enhanced Δσ/σ of bulk and GB for filters A and B, and the control group. Schematic representations for the lattice vibration-assisted proton hopping process showing (c) excitation and relaxation of the O–H stretch vibration, (d) modulation of the potential energy surface (PES) of the proton due to subsequent excitation of the coupled lattice vibrations, and (e) proton transfer from the donor (O_I) to the acceptor (O_II). For reference, the dashed curve in (d) depicts the shape of the PES prior to resonant excitation, corresponding to configuration shown in c. The error bars in (b) incorporate standard deviations from three parallel measurements and fitting errors per sample across three samples prepared using the same method. Source data are provided in the Source Data file.

We model the effective PES of the proton for a detailed discussion of the above hypothesis. Several studies have elaborated that a proton incorporated into metal oxides vibrates locally around the proton donor (O_I in Fig. 5c–e)^14,36,64. The photon energy derived from the absorption frequency in the vibrational spectra represents the transition energy of the vibration, normally from the ground state to its first excited state. Following previous works^29,36,67, we model the motion of the proton in the vicinity of O_I by a Morse potential, a typical anharmonic potential (Supplementary Fig. 20a):

where r is the distance between O_I and H atoms, r₀ is the equilibrium O_I–H distance, V₀ is the activation barrier, α defines the curvature of the potential and can be solved from ν_e elaborated below. The energy of the i-th vibrational level in the Morse potential is given by E_i = (i + 1/2) hν_e [1 – χ (i + 1/2)], leading to the transition energy between the i-th and i + 1-th levels E_i,i + 1 = hν_i,i + 1 = hν_e [1 – 2χ (i + 1)]. ν_e is a theoretical quantity related to the measured frequency of O–H stretch vibration ν_0,1 via ν_e = 1/(1 – 2χ), and χ is the anharmonic coefficient given by χ = hν_e/4V₀^29,70. Here, ?0=?0+?a,Bulkeff means the depth of the potential, where E₀ = 1/2hν_e (1 – 1/2χ) refers to the ground-state energy, and ?a,Bulkeff is the effective activation energy defined as the difference between the top of the activation barrier and E₀.

To find E_i, we take ?a,Bulkeff = 0.45 eV (the measured activation energy for bulk proton conduction). This value reasonably reflects the activation barrier for proton conduction of most protons at 130–200 °C, since the majority of protons are reported to remain in a trapped state with an apparent activation barrier of 0.44–0.47 eV at low temperatures⁴⁹. By applying ν_0,1 = 1 × 10¹⁴Hz (O–H stretch frequency from the vibrational spectra) and r₀ = 1.003 Å (estimated by an empirical correlation between ν_O–H and O–H bond length)⁷¹, we obtain χ = 0.249. Therefore, the O–H stretch vibration behaves as an anharmonic oscillator, and the energy gap between its ground state (E₀) and the first excited state (E₁) is E_0,1 = 0.41 eV. By absorbing an MIR photon matching its vibration frequency, the oscillator is brought to its excited state⁷⁰. This process is analogous to the photoexcitation of electrons or holes in semiconductors⁴⁴. High excitation rate can be achieved by increasing p, which promotes the enhancement in proton conductivity. Subsequently, the effective barrier for proton hopping would reduce from 0.45 eV at the ground state to 0.04 eV at the first excited state. comparable with the average thermal energy k_BT in 130–200 °C (0.03–0.04 eV). However, if proton transfer occurs via direct breaking of the O–H bond when the proton is at the first excited state, it would lead to a significant decrease in the activation energy, deviating from our observation that the change in E_a,Bulk upon MIR irradiation is only 0.04 eV. Therefore, we propose that the enhanced proton conduction originates from relaxation of excited O–H stretch vibration into lower-frequency oxygen lattice vibrations coupling to O–H stretch vibration (Fig. 5c). These coupled vibrations alter the O–O distance, thereby tuning the height and width of the PES of the proton (Fig. 5d), and facilitating proton transfer (Fig. 5e)^14,72,73. For proton-conducting metal oxides, experimental evidence has confirmed that the temperature-dependent excited-state lifetimes of O–H stretch mode of KTaO₃ is coupled to the O–Ta–O bending motion, thus reducing the activation barrier height^35,67,74. Our previous works have also shown that the temperature-dependent proton jump times of BZY and Y-doped BaCeO₃ measured by quasielastic neutron scattering follow Samgin’s proton polaron model^75,76,77, suggesting that proton hopping is strongly coupled to lattice phonons, in particular B–O stretch modes^28,46. The above evidence indicates that the lattice vibrations coupled to O–H stretch vibration are able to tune the PES of the proton, and, in turn, assist proton transfer.

It should be noted that the potential model derived here only depicts the time-averaged effective PES of the proton, since the actual PES shape is dependent on time, position, and temperature due to lattice vibrations and various oxygen chemical environments^14,24. On the other side, the activation energy for GB proton conduction has a strong contribution from the high Schottky barrier (up to 0.6 V) due to the depletion of protons in the space-charge region^44,51. As presented in Supplementary Note 6, upon MIR irradiation, the Schottky barrier height decreases from 0.287 ± 0.011 eV to 0.275 ± 0.009 eV, resulting in the higher GB enhancement ratio. Although the enhancement effect presented in this work is only up to about 50%, it is worth noting that the transmitted MIR light intensity of the sample (Supplementary Fig. 17b) decreases exponentially against the sample thickness. This thickness dependence suggests that higher enhancement ratios could be achieved in thinner samples.

In conclusion, we enhanced the proton conductivity in protonated polycrystalline BZY proton conductor using CW MIR irradiation with dominant wavelengths of 2–6 μm. Vibrational spectroscopy showed that the radiation between 2.7–4.0 μm can be absorbed by the O–H stretch vibration in BZY. The enhancement in bulk and GB proton conductivities at 160 °C were 36.8% and 53.0% under an effective irradiation power density of 195.2 mW cm⁻². The enhancement effect was reversible when the MIR irradiation was switched on and off for multiple cycles. The results implicate the possibility of an over-ten-fold enhancement of both bulk and grain boundary proton conductivities to over 10⁻³ S cm⁻¹ if irradiated with a high-power CW MIR laser. Conductivity measurements under narrowband MIR irradiation demonstrated that the MIR radiation strongly absorbed by the O–H stretch vibration (2.95 μm) exhibits stronger enhancement in conductivity than that with weak absorption (2.70 μm). The evidence indicates that MIR radiation enhances proton conductivity in BZY by exciting the O–H stretch vibration, which subsequently relax into coupled lattice vibrations that modulate the PES of the proton to lower the energy barrier for proton transfer.

Our findings demonstrate MIR irradiation as a novel strategy to enhance ionic conductivity in fast ionic conductors in addition to existing approaches. This effect holds particular promise for tubular^78,79 and thin-film^80,81 protonic ceramic electrochemical cells (PCEC), with the thickness of electrolyte and air electrode within 400 μm, using IR furnaces^34,56. This strategy offers advantages including rapid start-up, reduced operation temperature and thermal expansion, power consumption, and costs. Notably, the potential to enhance the ionic conductivity with a low-power light source would inspire the exploration towards low- or even room-temperature PCECs. Furthermore, our findings go beyond conventional chemical approaches to improve the ionic conductivity, and will inspire further exploration of tuning ionic conductivity as well as chemical reaction rates based on selective modulation of vibrational properties of solids.