I approached the challenge of lithium dendrite suppression in solid electrolyte cells from the viewpoint of interfacial transport rather than from the viewpoint of bulk electrolyte replacement or simple anode alloying. In my work, I constructed a mixed ion–electron conductive interphase, which I refer to as MIEC, between a sulfide solid electrolyte and a lithium metal anode. My central hypothesis was that a carefully balanced ionic and electronic conductivity at the anode interface would redistribute the local electric field, smooth the lithium-ion flux, and delay the onset of filamentary lithium growth. I tested this hypothesis by preparing a series of composite interphase layers containing different mass fractions of a carbon-based electronic conductor, and I evaluated them in symmetric solid electrolyte cells and full solid electrolyte cells. The results indicated that a narrow composition window, corresponding to 1% conductive additive, delivered the best combination of high ionic conductivity, low but nonzero electronic conductivity, high critical current density, long symmetric-cell cycling life, and excellent full-cell capacity retention. I describe here the design logic, fabrication, electrochemical data, spectroscopic evidence, morphological observations, and mechanistic interpretation that support this conclusion.
Throughout this article, I use the term solid electrolyte cell to describe an electrochemical cell in which a solid electrolyte separates the anode and cathode and in which the anode is lithium metal. The solid electrolyte cell is not a single material system; it is an integrated architecture in which bulk electrolyte transport, grain-boundary transport, interphase transport, charge-transfer kinetics, and mechanical contact all act together. My experiments were designed to isolate the interphase contribution as much as possible while preserving the realistic operating conditions of a solid electrolyte cell. I therefore combined impedance spectroscopy, distribution of relaxation times analysis, critical current density measurements, galvanostatic cycling, X-ray diffraction, X-ray photoelectron spectroscopy, scanning electron microscopy, and energy-dispersive X-ray spectroscopy.
1. Design hypothesis and research questions
I began with a simple but important observation: in many solid electrolyte cells, failure at the lithium metal interface is not caused by a single catastrophic event but by progressive current constriction. Pores, grain boundaries, surface roughness, and imperfect contact create regions where the local current density is much higher than the nominal applied current density. In those regions, lithium plating becomes uneven, the local potential drops, and dendrites nucleate. A purely insulating interphase can improve chemical stability, but it may also increase interfacial resistance and block lithium transport. A purely electronic interphase can spread current, but if its electronic conductivity is too high, lithium ions are reduced too far from the electrolyte surface, causing lithium deposition inside the interphase and creating dead lithium. Therefore, I asked whether an intermediate condition could be found: an interphase that is ionically conductive enough for fast lithium transport, electronically conductive enough to smooth the potential distribution, and chemically stable enough to survive extended cycling.
I formulated this question in terms of a mixed conduction parameter. If the interphase has ionic conductivity \(\sigma_{\text{ion}}\) and electronic conductivity \(\sigma_e\), then the ratio
$$ \gamma = \frac{\sigma_e}{\sigma_{\text{ion}}} $$
controls how the electrochemical reaction is distributed across the interphase thickness. When \(\gamma\) is extremely small, the reaction is confined to the electrolyte–lithium contact and becomes sensitive to local defects. When \(\gamma\) is extremely large, the reaction extends deep into the interphase and loses selectivity. I reasoned that an optimal \(\gamma\) should exist, producing a uniform reaction front and a uniform lithium deposition morphology. My experimental program was designed to find that optimum for a sulfide solid electrolyte cell based on \(\mathrm{Li_6PS_5Cl}\).
2. Materials, compositions, and cell architecture
I used \(\mathrm{Li_6PS_5Cl}\) as the sulfide solid electrolyte because it offers high ionic conductivity, favorable mechanical deformability, and reasonable compatibility with lithium metal when an appropriate interphase is present. I used a high-surface-area carbon black, Super P, as the electronic conductor. I prepared composite interphase powders by manually grinding the solid electrolyte and Super P in an agate mortar for 30 min. I varied the Super P mass fraction from 0% to 3% and labeled the samples as shown in Table 1. The mass fraction was calculated as
$$ w_{\mathrm{SP}} = \frac{m_{\mathrm{SP}}}{m_{\mathrm{SP}} + m_{\mathrm{Li_6PS_5Cl}}} \times 100\% $$
where \(m_{\mathrm{SP}}\) is the mass of Super P and \(m_{\mathrm{Li_6PS_5Cl}}\) is the mass of the sulfide solid electrolyte. The samples were stored and handled under an inert atmosphere to avoid hydrolysis and oxidation.
| Sample label | \(\mathrm{Li_6PS_5Cl}\) mass parts | Super P mass parts | Super P mass fraction | Intended transport role |
|---|---|---|---|---|
| LPSCl | 100 | 0 | 0% | Baseline solid electrolyte cell interphase |
| 0.5%SP | 100 | 0.5 | 0.5% | Low electronic percolation |
| 1%SP | 100 | 1 | 1% | Optimized mixed conduction |
| 3%SP | 100 | 3 | 3% | Excessive electronic conduction |
For symmetric solid electrolyte cells, I first pressed the sulfide solid electrolyte powder in a 10 mm die at 300 MPa for 1 min. I then placed the composite interphase powder on both sides of the electrolyte pellet and pressed at 500 MPa for 2 min. Finally, I attached lithium foil at a low pressure of about 10 MPa to form a Li|interphase|solid electrolyte|interphase|Li configuration. For full solid electrolyte cells, I prepared a cathode composite by mixing \(\mathrm{LiNi_{0.89}Co_{0.06}Mn_{0.05}O_2}\), sulfide solid electrolyte, and Super P in a mass ratio of 70:27:3. I pressed 115 mg of electrolyte at 300 MPa for 1 min, then applied 5 mg of the interphase composite on one side at 500 MPa for 2 min, then applied the cathode composite on the other side at 300 MPa for 2 min, and finally attached a 20 µm lithium foil at 10 MPa. This architecture allowed me to compare the same solid electrolyte cell chemistry with different anode interphases while keeping the cathode composition and loading constant unless otherwise stated.
3. Fabrication protocol and experimental control
I paid particular attention to pressure, pressing time, and atmosphere because these variables strongly affect the performance of a solid electrolyte cell. The pressing steps are summarized in Table 2. I used a lower pressure for the lithium attachment step to avoid extruding the lithium foil and short-circuiting the cell. I kept all assembly steps in an argon-filled glovebox with oxygen and moisture levels below 1 ppm. After assembly, I rested each solid electrolyte cell for 1 h before electrochemical testing. For galvanostatic cycling, I used a battery tester and performed three formation cycles at 0.1C between 2.5 and 4.3 V versus \(\mathrm{Li^+/Li}\). For symmetric cells, I applied constant current densities and recorded the voltage–time response. For impedance measurements, I used a frequency range from 1 Hz to 7 MHz with a 10 mV alternating-current perturbation. For in-situ impedance measurements, I collected spectra before cycling and after defined charge–discharge intervals, then deconvoluted the impedance into distribution of relaxation times peaks.
| Step | Material | Pressure | Time | Purpose |
|---|---|---|---|---|
| Electrolyte pellet | \(\mathrm{Li_6PS_5Cl}\) | 300 MPa | 1 min | Form dense solid electrolyte layer |
| Interphase lamination | Composite powder | 500 MPa | 2 min | Create intimate mixed conductive contact |
| Cathode lamination | Cathode composite | 300 MPa | 2 min | Form cathode–electrolyte contact |
| Lithium attachment | Li foil | 10 MPa | Short press | Attach lithium without mechanical damage |
| Rest | Full solid electrolyte cell | Ambient | 1 h | Allow contact relaxation |
4. Structural and chemical compatibility
I first asked whether mechanical mixing of Super P with the sulfide solid electrolyte caused chemical decomposition. X-ray diffraction showed that the composite interphase retained the characteristic reflections of the \(\mathrm{Li_6PS_5Cl}\) phase. I did not observe new reflections that could be assigned to crystalline decomposition products or to a separate carbon phase, which is consistent with the amorphous or nanoscale nature of the carbon additive. The positions of the main sulfide electrolyte reflections were preserved in all composite samples. This indicated that the grinding and pressing steps did not destroy the \(\mathrm{PS_4^{3-}}\) structural units or convert the electrolyte into a different crystalline phase. The chemical compatibility was therefore sufficient to justify electrochemical testing.
I also examined the cross-section of the 1%SP interphase using scanning electron microscopy and energy-dispersive X-ray spectroscopy. The carbon signal was distributed uniformly across the interphase, with no large agglomerates or carbon-rich islands visible at the scale of the measurement. The sulfur and phosphorus signals from the sulfide electrolyte were also uniformly distributed. This morphology was important because a nonuniform carbon network would create local variations in electronic conductivity, which in turn would create local hot spots for lithium deposition. My observation of a uniform carbon distribution supported the idea that the 1%SP interphase could act as a distributed mixed conductor rather than as a random collection of conductive particles.
5. Conductivity optimization and percolation behavior
I measured the ionic and electronic transport properties of the composite interphase to determine whether a percolation threshold could be identified. The Nyquist plots of the symmetric solid electrolyte cells showed that the grain-boundary impedance remained nearly unchanged when the Super P content was below 1%. I interpreted this as evidence that the sulfide electrolyte still provided continuous ionic pathways through the composite. At 3%SP, however, a pronounced semicircle appeared in the impedance spectrum, indicating a sharp increase in interfacial or grain-boundary resistance. This suggested that excessive carbon disrupted the ionic contact network between electrolyte particles and increased the tortuosity for lithium-ion transport.
The electronic conductivity increased monotonically with Super P content. I calculated the electronic conductivity from direct-current polarization measurements using
$$ \sigma_e = \frac{L}{R_e A} $$
where \(L\) is the interphase thickness, \(A\) is the active area, and \(R_e\) is the electronic resistance obtained from the linear current–voltage response. I also used the ionic conductivity relation
$$ \sigma_{\text{ion}} = \frac{L}{R_{\text{ion}} A} $$
to compare the ionic transport of the composites. The measured electronic conductivities are summarized in Table 3. The baseline LPSCl had an electronic conductivity of \(9.7 \times 10^{-10}\ \mathrm{S\,cm^{-1}}\). The 0.5%SP composite increased this value to \(8.25 \times 10^{-9}\ \mathrm{S\,cm^{-1}}\). The 1%SP composite reached \(1.8 \times 10^{-8}\ \mathrm{S\,cm^{-1}}\), which I considered the optimum because it was high enough to redistribute current but still many orders of magnitude lower than the ionic conductivity. The 3%SP composite reached \(2.4 \times 10^{-5}\ \mathrm{S\,cm^{-1}}\), which was high enough to allow significant lithium reduction within the interphase and to promote dead lithium formation.
| Interphase | Ionic transport behavior | Electronic conductivity / \(\mathrm{S\,cm^{-1}}\) | Interpretation |
|---|---|---|---|
| LPSCl | Continuous, baseline grain-boundary impedance | \(9.7 \times 10^{-10}\) | Nearly insulating electronically |
| 0.5%SP | Nearly unchanged grain-boundary impedance | \(8.25 \times 10^{-9}\) | Electronic conduction begins to appear |
| 1%SP | Stable ionic pathways, low grain-boundary impedance | \(1.8 \times 10^{-8}\) | Optimal mixed conduction window |
| 3%SP | Increased semicircle, degraded ionic contact | \(2.4 \times 10^{-5}\) | Excessive electronic conduction |
I also considered the percolation behavior of the carbon network. A simple effective-medium expression for the composite conductivity can be written as
$$ \sigma_{\text{eff}} = \phi_{\text{ion}} \sigma_{\text{ion}} + \phi_e \sigma_e $$
where \(\phi_{\text{ion}}\) and \(\phi_e\) are the volume fractions of the ionic and electronic phases. Near the percolation threshold, the electronic conductivity can be described by
$$ \sigma_e = \sigma_0 (\phi_e – \phi_c)^t $$
where \(\phi_c\) is the percolation threshold and \(t\) is a critical exponent. My data indicated that the 1%SP composite was near the onset of a connected carbon network but had not yet crossed into a fully percolating electronic pathway. This intermediate state was exactly what I wanted for a mixed conductive interphase in a solid electrolyte cell.
6. Critical current density and symmetric solid electrolyte cell cycling
The critical current density is one of the most informative metrics for a solid electrolyte cell because it integrates interfacial kinetics, mechanical stability, and dendrite resistance. I defined the critical current density as the maximum current density at which the symmetric solid electrolyte cell could cycle without a sudden voltage collapse or short circuit. In practice, I increased the current density stepwise and monitored the voltage–time response. The criterion was
$$ J_{\mathrm{CCD}} = \max \{ J \mid V(t) \text{ remains stable without a sudden polarization collapse} \} $$
My measurements are summarized in Table 4. The 0.5%SP interphase improved the critical current density to about \(1.1\ \mathrm{mA\,cm^{-2}}\), which was only slightly better than the baseline. The 1%SP interphase increased it to \(1.6\ \mathrm{mA\,cm^{-2}}\). I attributed this improvement to the three-dimensional ion–electron network that smoothed the local current distribution. The 3%SP interphase reduced the critical current density to \(0.9\ \mathrm{mA\,cm^{-2}}\), which confirmed that too much electronic conductivity was detrimental.
| Interphase | Critical current density / \(\mathrm{mA\,cm^{-2}}\) | Observed behavior |
|---|---|---|
| LPSCl baseline | Below the first stable comparison point | Early short circuit and unstable polarization |
| 0.5%SP | 1.1 | Modest improvement |
| 1%SP | 1.6 | Best dendrite resistance |
| 3%SP | 0.9 | Excess electronic conduction, dead lithium |
I then tested long-term cycling of symmetric solid electrolyte cells at a constant current density of \(0.5\ \mathrm{mA\,cm^{-2}}\). The baseline LPSCl cell short-circuited after only about 170 h, as indicated by a sudden drop in polarization voltage. In contrast, the cell with the 1%SP interphase cycled for 2800 h with a stable polarization voltage of about 25 mV. This result was striking because it showed that a very small amount of carbon additive could extend the life of a solid electrolyte cell by more than an order of magnitude. The overpotential remained stable throughout the test, which suggested that the interphase did not progressively thicken or become ionically blocked. The cycling data are summarized in Table 5.
| Interphase | Current density / \(\mathrm{mA\,cm^{-2}}\) | Cycle life | Polarization voltage | Failure mode |
|---|---|---|---|---|
| LPSCl | 0.5 | About 170 h | Sudden collapse | Short circuit |
| 1%SP | 0.5 | 2800 h | About 25 mV | No short circuit observed |
The voltage–time behavior can be analyzed using the overpotential relation
$$ \eta(t) = V(t) – V_0 $$
where \(V(t)\) is the measured cell voltage and \(V_0\) is the open-circuit voltage. A stable \(\eta(t)\) indicates that the interfacial resistance and the charge-transfer resistance are not increasing with time. In my 1%SP solid electrolyte cell, the stability of \(\eta(t)\) was a strong indication that the interphase maintained both mechanical contact and ionic percolation during repeated lithium plating and stripping.
7. In-situ electrochemical impedance and distribution of relaxation times
I used in-situ electrochemical impedance spectroscopy to separate the different kinetic processes in the solid electrolyte cell. To interpret the impedance spectra, I applied distribution of relaxation times analysis. The impedance can be written as
$$ Z(\omega) = R_\infty + \int_0^\infty \frac{\gamma(\tau)}{1 + j\omega\tau} \, d\tau $$
where \(R_\infty\) is the high-frequency resistance, \(\gamma(\tau)\) is the distribution function of relaxation times, and \(\tau\) is the relaxation time. In practice, I represented the impedance as a sum of discrete relaxation processes:
$$ Z(\omega) = R_\infty + \sum_k \frac{R_k}{1 + j\omega\tau_k} $$
Each peak in the distribution of relaxation times corresponds to a physical or chemical process. I assigned the peaks as shown in Table 6. The high-frequency peak at \(\tau_1\) corresponds to grain-boundary transport in the solid electrolyte and the composite interphase. The mid-frequency peaks at \(\tau_2\) and \(\tau_3\) correspond to the solid electrolyte interphase and charge-transfer processes at the lithium metal interface. The low-frequency peak at \(\tau_4\) corresponds to diffusion limitations in the solid electrolyte cell.
| Relaxation time range | Assigned process | Effect of 1%SP interphase |
|---|---|---|
| \(10^{-7}\) to \(10^{-5}\) s | Grain-boundary transport | Nearly unchanged; ionic pathways preserved |
| \(10^{-5}\) to \(10^{-3}\) s | Interphase and solid electrolyte interphase resistance | Reduced peak area; improved contact |
| \(10^{-3}\) to \(10^{-1}\) s | Charge-transfer resistance | Reduced peak area; faster kinetics |
| \(10^{-1}\) to \(10\) s | Diffusion resistance | Reduced peak area; better lithium transport |
Before cycling, the 1%SP solid electrolyte cell showed smaller peak areas in the mid-frequency and low-frequency regions than the baseline cell. This indicated that the interphase improved contact and reduced the interfacial resistance. After the first charge–discharge cycle, the baseline cell showed a significant increase in grain-boundary resistance, followed by partial stabilization. In contrast, the 1%SP cell maintained a low grain-boundary resistance after cycling. I interpreted this as evidence that the interphase suppressed the degradation of the electrolyte grain boundaries and preserved the lithium-ion transport network. The in-situ impedance results therefore supported the galvanostatic cycling results and provided a mechanistic explanation for the improved stability.
8. Full solid electrolyte cell performance
I assembled full solid electrolyte cells with a high-nickel layered oxide cathode, a sulfide solid electrolyte separator, the composite interphase, and a lithium metal anode. The first-cycle charge and discharge capacities at 0.1C are summarized in Table 7. The baseline LPSCl cell delivered a first-cycle discharge capacity of about 176 mAh g\(^{-1}\) with a Coulombic efficiency of 81.69%. The 0.5%SP cell delivered about 185 mAh g\(^{-1}\) with an efficiency of 82.22%. The 1%SP cell delivered about 203 mAh g\(^{-1}\) with an efficiency of 83.45%. The 3%SP cell delivered about 176 mAh g\(^{-1}\) with an efficiency of 82.35%. These data showed that the 1%SP interphase provided the best initial capacity and the best first-cycle efficiency.
| Interphase | Charge specific capacity / mAh g\(^{-1}\) | Discharge specific capacity / mAh g\(^{-1}\) | First-cycle Coulombic efficiency / % |
|---|---|---|---|
| LPSCl | 216 | 176 | 81.69 |
| 0.5%SP | 225 | 185 | 82.22 |
| 1%SP | 243 | 203 | 83.45 |
| 3%SP | 213 | 176 | 82.35 |
The Coulombic efficiency can be written as
$$ \eta_C = \frac{Q_{\text{discharge}}}{Q_{\text{charge}}} \times 100\% $$
and the capacity retention can be written as
$$ R_c = \frac{Q_N}{Q_1} \times 100\% $$
where \(Q_1\) is the initial discharge capacity and \(Q_N\) is the discharge capacity after \(N\) cycles. I used these metrics to compare the long-term stability of the solid electrolyte cells. The 1%SP solid electrolyte cell showed an initial discharge capacity of about 150 mAh g\(^{-1}\) at 1C. After 1000 cycles, the capacity retention was 79.7%. After 2000 cycles, the capacity retention was still 68.3%, and no internal short circuit occurred. When I increased the cathode active material loading from about 5.18 mg to about 20.32 mg, the first-cycle discharge capacity remained nearly unchanged, and the cell cycled for 250 cycles at 0.3C with almost no capacity fade. These results demonstrated that the mixed conductive interphase was effective not only in symmetric solid electrolyte cells but also in practical full solid electrolyte cells with high areal loading.
| Condition | Rate | Initial discharge capacity | Cycling result |
|---|---|---|---|
| 1%SP, standard loading | 0.1C | 203 mAh g\(^{-1}\) | High first-cycle efficiency |
| 1%SP, standard loading | 1C | 150 mAh g\(^{-1}\) | 79.7% retention after 1000 cycles |
| 1%SP, standard loading | 1C | 150 mAh g\(^{-1}\) | 68.3% retention after 2000 cycles |
| 1%SP, high loading | 0.3C | Nearly unchanged | Stable for 250 cycles |
I also examined the differential capacity curves. The 1%SP solid electrolyte cell showed sharper and more intense redox peaks than the baseline cell, which indicated lower interfacial resistance and improved electrochemical reversibility. The 3%SP cell showed a lower discharge capacity, which I attributed to excessive electronic conduction and the formation of dead lithium inside the interphase. This observation reinforced the idea that the electronic conductivity must be tuned within a narrow window; it is not a case in which more electronic conduction is always better.
9. Interfacial morphology and lithium deposition
I examined the interface between the interphase and the lithium metal after cycling using scanning electron microscopy. In the baseline solid electrolyte cell, lithium deposition was uneven, and dendritic features were observed near the interface. In the 1%SP solid electrolyte cell, lithium deposition was much more uniform. The lithium grew at the interphase–lithium interface rather than penetrating into the solid electrolyte. I did not observe dendrites extending through the interphase into the bulk electrolyte. The cross-sectional images showed that the interphase and the lithium metal remained in good contact, with no large voids or delamination. This was important because void formation at the anode interface is a common failure mode in solid electrolyte cells.
I also fabricated a Li|interphase|solid electrolyte|interphase|Cu half-cell and charged it at \(0.5\ \mathrm{mA\,cm^{-2}}\) to visualize lithium deposition. The optical images showed that lithium deposited as a flat and dense layer on the interphase surface. This supported the conclusion that the interphase had good lithiophilicity and that its uniform composition promoted uniform lithium nucleation. The morphological evidence was consistent with the electrochemical data and with the proposed mechanism of current redistribution.
10. X-ray photoelectron spectroscopy and chemical stability
I used X-ray photoelectron spectroscopy to examine the chemical state of the interphase before and after cycling. The sulfur 2p spectrum showed peaks at about 161.37 and 162.47 eV, which I assigned to the \(\mathrm{PS_4^{3-}}\) units in the sulfide electrolyte. Additional sulfur 2p peaks at about 162.87 and 162.97 eV were assigned to P–S–P bonding. The phosphorus 2p spectrum showed peaks at about 132.87 and 133.97 eV, which were also consistent with the \(\mathrm{PS_4^{3-}}\) structural unit. After cycling, these peaks remained at nearly the same binding energies and with similar intensities. I did not observe new peaks corresponding to sulfur oxides, phosphorus oxides, or other decomposition products. The chlorine 2p spectrum was more difficult to resolve because the binding energies of \(\mathrm{Cl^-}\) in the sulfide electrolyte and in lithium chloride overlap. However, the overall spectra indicated that the interphase remained chemically stable during cycling.
| Core level | Species | Binding energy before cycling / eV | Binding energy after cycling / eV | Interpretation |
|---|---|---|---|---|
| S 2p | \(\mathrm{PS_4^{3-}}\) | 161.37, 162.47 | Retained | Sulfide framework preserved |
| S 2p | P–S–P | 162.87, 162.97 | Retained | No major decomposition |
| P 2p | \(\mathrm{PS_4^{3-}}\) | 132.87, 133.97 | Retained | Phosphorus environment stable |
| Cl 2p | \(\mathrm{Cl^-}\) and possible LiCl | Overlapping | Stable envelope | Limited interfacial decomposition |
The X-ray photoelectron spectroscopy results provided atomic-level evidence that the mixed conductive interphase did not sacrifice chemical stability. If the carbon additive had catalyzed decomposition of the sulfide electrolyte, I would have expected new sulfur–oxygen or phosphorus–oxygen peaks. I did not observe such peaks. This was important because a chemically unstable interphase would eventually increase resistance and consume lithium, even if it initially improved current distribution.
11. Mechanistic model of dendrite suppression
I propose that the mixed conductive interphase suppresses lithium dendrites through three coupled effects: current redistribution, potential flattening, and reaction-front broadening. The first effect can be described by the current continuity equation. In a mixed conductor, the total current is the sum of ionic and electronic contributions:
$$ \mathbf{i} = \mathbf{i}_{\text{ion}} + \mathbf{i}_e $$
with
$$ \mathbf{i}_{\text{ion}} = -\sigma_{\text{ion}} \nabla \phi_{\text{ion}} $$
and
$$ \mathbf{i}_e = -\sigma_e \nabla \phi_e $$
At steady state, charge conservation requires
$$ \nabla \cdot \mathbf{i} = 0 $$
If the electronic conductivity is zero, the ionic current must pass through the entire interphase without any electronic shortcut, and any local variation in contact resistance produces a local variation in ionic current density. If the electronic conductivity is finite but moderate, the electronic current can redistribute around high-resistance regions and reduce the local ionic current crowding. This is the current-redistribution effect. It smooths the lithium-ion flux at the interface and reduces the probability of dendrite nucleation.
The second effect is potential flattening. The electrochemical potential difference across the interface can be written as
$$ \Delta \tilde{\mu}_{\mathrm{Li^+}} = F(\phi_{\text{Li}} – \phi_{\text{SE}}) + RT \ln \frac{a_{\mathrm{Li^+}}^{\text{Li}}}{a_{\mathrm{Li^+}}^{\text{SE}}} $$
A uniform electronic potential in the interphase reduces lateral variations in \(\phi_{\text{Li}} – \phi_{\text{SE}}\), which in turn reduces lateral variations in the driving force for lithium deposition. When the driving force is uniform, lithium nucleates uniformly rather than at a few high-field points.
The third effect is reaction-front broadening. In a mixed conductor, the electrochemical reaction is not confined to a mathematical plane. The characteristic reaction depth can be approximated by
$$ L_r \sim \sqrt{\frac{\sigma_{\text{ion}} \sigma_e}{\sigma_{\text{ion}} + \sigma_e} \frac{RT}{F j_0}} $$
where \(j_0\) is the exchange current density. When \(L_r\) is very small, lithium deposition is interfacial and sensitive to defects. When \(L_r\) is very large, deposition occurs deep inside the interphase and creates dead lithium. The 1%SP interphase appears to produce an intermediate \(L_r\) that broadens the reaction zone just enough to smooth the current but not enough to bury lithium irreversibly.
I also considered the Sand’s time concept as a qualitative analogy for dendrite onset. In a simplified form,
$$ \tau_s = \frac{\pi D}{4} \left( \frac{e C_0}{J} \right)^2 $$
where \(D\) is the lithium-ion diffusion coefficient, \(C_0\) is the initial lithium-ion concentration, \(e\) is the elementary charge, and \(J\) is the current density. Although this expression was developed for liquid electrolytes, it captures the idea that dendrite onset is delayed when the effective current density is reduced and the lithium-ion concentration is uniform. The mixed conductive interphase effectively reduces the local current density at the most vulnerable points and thus extends the time before dendrite nucleation.
Finally, I consider the mechanical effect. The interphase is a composite of soft sulfide electrolyte particles and hard carbon particles. This composite can deform plastically under stack pressure and maintain contact with both the electrolyte and the lithium metal. The mechanical stability helps prevent void formation during lithium stripping, which is a major cause of current constriction and dendrite growth in solid electrolyte cells. The electrochemical and mechanical effects are therefore coupled: better contact leads to more uniform current, which leads to more uniform lithium deposition, which leads to better contact retention.
12. Design rules for mixed conductive interphases in solid electrolyte cells
From my experiments, I extracted several design rules that I believe are useful for engineering mixed conductive interphases in solid electrolyte cells. First, the electronic conductivity must be nonzero but low. The optimum I found was on the order of \(1.8 \times 10^{-8}\ \mathrm{S\,cm^{-1}}\) for a sulfide solid electrolyte cell with a Super P additive. Second, the ionic conductivity must remain high. The interphase must not block lithium-ion transport or introduce a large grain-boundary resistance. Third, the carbon additive must be uniformly dispersed. Agglomerates create local electronic hot spots and local ionic dead zones. Fourth, the interphase must be chemically stable against both the solid electrolyte and lithium metal. Fifth, the interphase must be mechanically compliant enough to maintain contact during cycling. These rules are summarized in Table 8.
| Design parameter | Too low | Optimum | Too high |
|---|---|---|---|
| Super P mass fraction | 0 to 0.5% | 1% | 3% |
| Electronic conductivity | Below \(10^{-9}\ \mathrm{S\,cm^{-1}}\) | About \(1.8 \times 10^{-8}\ \mathrm{S\,cm^{-1}}\) | Above \(10^{-5}\ \mathrm{S\,cm^{-1}}\) |
| Critical current density | Below 1.1 \(\mathrm{mA\,cm^{-2}}\) | 1.6 \(\mathrm{mA\,cm^{-2}}\) | 0.9 \(\mathrm{mA\,cm^{-2}}\) |
| Lithium deposition | Uneven, filamentary | Uniform, dense | Dead lithium inside interphase |
| Interfacial resistance | High and unstable | Low and stable | High due to ionic blockage |
I also formulated a dimensionless mixed conduction number to guide future optimization:
$$ M = \frac{\sigma_e}{\sigma_{\text{ion}}} \frac{L^2}{L_r^2} $$
where \(L\) is the interphase thickness and \(L_r\) is the reaction depth. When \(M\) is too small, the interphase behaves as an insulator and current crowding occurs. When \(M\) is too large, the interphase behaves as an electronic conductor and lithium deposition becomes spatially uncontrolled. The optimum condition corresponds to \(M \sim 1\), where the electronic and ionic transport lengths are comparable. In my experiments, the 1%SP composite appeared to satisfy this condition.
13. Comparison with other interfacial strategies
I compared my mixed conductive interphase strategy with other approaches that have been used to stabilize lithium metal in solid electrolyte cells. Alloy interphases can improve lithiophilicity but may be consumed over time. Oxide coatings can improve chemical stability but often increase impedance. Fluoride-rich interphases can suppress dendrites but require precise control of fluorination. Three-dimensional current collectors can redistribute current but add mass and volume. My mixed conductive interphase is attractive because it is simple, scalable, and compatible with roll-to-roll processing. It does not require expensive precursors or complex synthesis. It uses a small amount of a common conductive additive and a simple grinding and pressing procedure. Table 9 summarizes the qualitative comparison.
| Strategy | Main benefit | Main limitation | Scalability |
|---|---|---|---|
| Alloy interphase | Improved lithiophilicity | Consumption and volume change | Moderate |
| Oxide coating | Chemical stability | Increased impedance | Moderate |
| Fluoride-rich interphase | Dendrite suppression | Process complexity | Low to moderate |
| Three-dimensional current collector | Current redistribution | Added mass and volume | Moderate |
| Mixed conductive interphase | Current redistribution and kinetics | Narrow composition window | High |
I do not claim that the mixed conductive interphase is a universal solution. Its performance depends on the solid electrolyte chemistry, the stack pressure, the cathode loading, and the operating temperature. However, my results show that it is a powerful and practical tool for improving the stability of solid electrolyte cells. The key is to avoid the extremes: an interphase that is too insulating or too electronically conductive will both fail, but for different reasons.
14. Limitations and future work
I acknowledge several limitations in my study. First, my experiments were performed at room temperature and at a fixed stack pressure. The optimal composition may shift at higher temperatures or different pressures. Second, I used a single sulfide solid electrolyte composition. Other sulfide, oxide, or halide solid electrolytes may require different electronic conductivity targets. Third, I used a single conductive additive. Other carbon morphologies, such as carbon nanotubes, graphene, or carbon nanofibers, may create different percolation networks and different mechanical properties. Fourth, my characterization focused on the anode interface. The cathode interface also contributes to the overall impedance and cycling stability of a solid electrolyte cell, and future work should examine both interfaces together.
I plan to extend this work in several directions. I will test the mixed conductive interphase with different solid electrolyte chemistries and different lithium metal alloys. I will vary the interphase thickness and the carbon morphology to determine whether the optimum electronic conductivity is universal or material-specific. I will use operando imaging and operando impedance to observe dendrite nucleation in real time. I will also scale up the cell fabrication to pouch cells and evaluate the process compatibility of the interphase with high-speed manufacturing. Finally, I will develop a multiphysics model that couples ionic transport, electronic transport, electrochemical kinetics, and mechanical deformation to predict the optimal interphase properties for a given solid electrolyte cell design.
15. Conclusions
I have shown that a mixed ion–electron conductive interphase can significantly improve the performance and stability of solid electrolyte cells. By tuning the Super P content to 1%, I achieved an electronic conductivity of \(1.8 \times 10^{-8}\ \mathrm{S\,cm^{-1}}\) while preserving high ionic conductivity and low grain-boundary resistance. The 1%SP interphase increased the critical current density of a symmetric solid electrolyte cell to \(1.6\ \mathrm{mA\,cm^{-2}}\) and enabled stable cycling for 2800 h at \(0.5\ \mathrm{mA\,cm^{-2}}\). In full solid electrolyte cells, the 1%SP interphase delivered a first-cycle discharge capacity of 203 mAh g\(^{-1}\) at 0.1C, and it retained 79.7% of its capacity after 1000 cycles at 1C and 68.3% after 2000 cycles. The interphase also maintained stable performance at high cathode loading. Morphological and spectroscopic analyses showed that the interphase remained chemically stable and promoted uniform lithium deposition without dendrite penetration. I attribute these improvements to three coupled mechanisms: current redistribution, potential flattening, and reaction-front broadening. My results establish a practical design rule for solid electrolyte cells: the anode interphase should be ionically conductive, moderately electronically conductive, chemically stable, and mechanically compliant. When these conditions are met, the solid electrolyte cell can tolerate higher current densities, cycle longer, and operate more safely with a lithium metal anode.
I believe that the mixed conductive interphase strategy is broadly applicable to other solid electrolyte cell chemistries and that it can be combined with other interfacial engineering approaches, such as alloy formation, surface coating, and three-dimensional current collection. The key insight is that the interphase is not a passive barrier; it is an active transport medium that shapes the electrochemical reaction distribution. By designing this transport medium carefully, I can control where and how lithium deposits, and thereby suppress dendrites at their source. This is the central lesson of my work, and I expect it to guide future advances in solid electrolyte cell technology.
For clarity, I summarize the main quantitative findings in Table 10. These values represent the best performance I observed within the scope of this study and provide a benchmark for future optimization.
| Metric | Baseline LPSCl | 1%SP interphase | Improvement |
|---|---|---|---|
| Electronic conductivity / \(\mathrm{S\,cm^{-1}}\) | \(9.7 \times 10^{-10}\) | \(1.8 \times 10^{-8}\) | About 18.6 times higher |
| Critical current density / \(\mathrm{mA\,cm^{-2}}\) | Below stable comparison | 1.6 | Substantially enhanced |
| Symmetric cell life at \(0.5\ \mathrm{mA\,cm^{-2}}\) | About 170 h | 2800 h | About 16.5 times longer |
| First-cycle discharge at 0.1C / mAh g\(^{-1}\) | 176 | 203 | 15.3% higher |
| First-cycle Coulombic efficiency / % | 81.69 | 83.45 | 1.76 percentage points higher |
| Capacity retention after 1000 cycles at 1C / % | Not stable enough to compare | 79.7 | Excellent retention |
| Capacity retention after 2000 cycles at 1C / % | Not stable enough to compare | 68.3 | Long cycle life |
In my view, the most important conclusion is that the electronic conductivity of the anode interphase must be treated as a design variable, not as an unavoidable impurity. A small, controlled amount of electronic conduction can transform the interface from a weak point into a stabilizing element. The solid electrolyte cell is a complex system, but its failure modes are often local. By homogenizing the local current and potential, the mixed conductive interphase removes the local driving force for dendrite formation. This principle is simple, but its consequences for solid electrolyte cell performance are profound. I therefore recommend that future solid electrolyte cell designs explicitly include a mixed conductive interphase layer with a carefully optimized composition and morphology.