Automated Operando Multimodal In-Liquid SDM

• The technique quantifies the coupled ionic, electronic, and mechanical dynamics in OMIECs operating under live bias.
• It integrates AFM and inverted optical microscopy with automated voltage sweeps for synchronized, voltage-resolved multimodal mapping.
• Insights from SDM provide guidance for optimizing electrolyte composition and polymer design in OECTs.

Automated operando multimodal in-liquid scanning dielectric microscopy (SDM) is an advanced quantitative technique for spatially and temporally resolving the coupled ionic, electronic, and mechanical dynamics within mixed ionic-electronic conductors (OMIECs) operating under live bias in aqueous environments. This methodology is central for revealing localized electrochemical transitions and mapping multimodal responses—electrical, mechanical, and morphological—across complex device architectures such as organic electrochemical transistors (OECTs). SDM provides direct access to internal states via real-time multimodal maps, transforming phenomenological device readouts into voltage-resolved fingerprints of underlying physical processes (Tanwar et al., 8 Jan 2026).

1. Instrumentation and Automated Operando Operation

SDM employs an atomic force microscopy (AFM) platform (JPK NanoWizard 4 BioAFM) combined with an inverted optical microscope, utilizing a gold-coated silicon AFM tip (MikroMasch SCM-PIT Au, tip radius ≲ 35 nm, spring constant ∼1 N/m, resonance ∼36 kHz). The tip concurrently acts as a gate electrode and as the probe for local electrostatic, mechanical, and topographic measurements. Experiments are conducted within an aqueous electrolyte (e.g., 10 mM NaCl or 10 mM NH₄Cl), fully immersing the sample. OMIEC-based OECTs feature a channel of poly(benzimidazobenzophenanthroline) (BBL), swelling from ∼25 nm to ∼80 nm under bias (W = 20–100 μm, L = 10 μm), bounded by Au source/drain contacts patterned and passivated except at the channel.

Automated operando mapping synchronizes voltage sweeps (V_GS: −0.4 V to +0.8 V in 20 mV increments; V_DS held at 0–0.5 V, varied across experiments) with AFM acquisition. Force-volume mode executes approach-retract cycles per pixel (extend ∼75 ms, retract ∼4 ms, overhead ∼5 ms, setpoint ∼50 nN, Z-range ∼3 μm) over scan fields (∼30 μm × 5.6 μm, 64 × 12 pixels) that span source, channel, and drain. Custom Python scripts automate synchronization for high-throughput, bias-resolved multimodal imaging.

2. Physical Principles and Signal Transduction in SDM

SDM exploits tip-sample electromechanical coupling under combined DC+AC voltage excitation (V_DC + V_AC sin 2π f t) to actuate and read out electrostatic forces sensitive to the local dielectric permittivity ($\varepsilon$) and conductivity ($\sigma$). Tip-sample capacitance $C(z, \epsilon, \sigma)$ modulates the electrostatic force:

$$F_\text{elect}(z) = \frac{1}{2}\frac{\partial C(z, \epsilon, \sigma)}{\partial z} [V_\text{DC}^2 + \frac{1}{2}V_\text{AC}^2]$$

High-frequency AC drive (f ≈ 105–115 MHz, V_AC ≈ 1–2 V_pp) is amplitude-modulated at f_mod ≈ 10 kHz; cantilever oscillations at f_mod yield signals proportional to $\partial C/\partial z \cdot V_\text{AC}^2$, discretely demodulated via lock-in amplifiers. Calibration-free normalization leverages

$$\Delta F/F = \frac{F_{\text{elect}}(z_{\text{lift}}) - F_{\text{elect}}(z_{\text{ref}})}{F_{\text{elect}}(z_{\text{ref}})}$$

Mechanical stiffness is extracted from indentation slope $s = dD/dZ_\text{piezo}$, correlating with effective sample modulus $k_\text{sample}$. Morphological topography is measured via contact height and root-mean-square roughness $R_q$.

The capacitance model treats the tip apex region as a parallel-plate capacitor:

$$C_\text{apex}(z) \approx \epsilon_0\,\epsilon_\text{sol}/A_\text{apex}\cdot 1/z$$

with in-plane and out-of-plane conduction parameterized for distinct device regions by an equivalent circuit framework.

3. Multimodal Spectral Mapping and Local Threshold Extraction

SDM simultaneously acquires spatially co-registered maps of electrical, mechanical, and morphological signals. Data analysis segments the topographic image into source, channel, and drain, averaging spectral signals per region as functions of gate bias V_G. Line profiles of the dielectric signal $\Delta F/F(x)$ at each V_GS are concatenated over 61 gate voltages, forming 2D electrostatic spectral maps $\Delta F/F(x, V_G)$; mechanical and morphological spectral maps are analogue constructed.

Within each region $T \in \{\text{S}, \text{C}, \text{D}\}$, the $\Delta F/F$–$V_G$ response is sigmoidal, enabling identification of two key local electrochemical thresholds: $V_G^\text{low,T}$ (insulator-to-semiconductor) and $V_G^\text{high,T}$ (semiconductor-to-conductor transition). Global OECT transfer curves [$I_{DS}(V_G)$, $g_m = dI_{DS}/dV_G$, $g' = d^2I_{DS}/dV_G^2$] are reconstructed and cross-referenced; eight global curve features uniquely correspond to these local threshold voltages, providing a “voltage-resolved fingerprint” of internal device state.

4. Equivalent Circuit Modeling and Analytical Framework

The physical modeling employs an equivalent circuit with two perpendicular branches under the tip (apex and cone capacitances $C_\text{apex,sol}$, $C_\text{cone,sol}$) in series with the tip/electrolyte capacitance $C_\text{tip}$ and interfacial polymer/electrolyte capacitance $C_\text{int}$. Lateral and out-of-plane conductivity ($\sigma_\text{sem,⊥}$, $\sigma_\text{sem,∥}$), respective permittivities ($\varepsilon_\text{sem,⊥}$, $\varepsilon_\text{sem,∥}$), and thickness $h_\text{sem}$ are explicitly treated.

Electrostatic apex force is:

$$F_\text{apex} = \frac{1}{2}\frac{\partial C_\text{total}}{\partial z} V_\text{AC}^2$$

Analytic fits for force versus conductivity are composed of two overlapping sigmoids:

$$F_{S/D}(\sigma) \propto \frac{1}{2} \left[ \frac{\sigma}{\sigma + \sigma_\text{low,S}} \right] - \frac{1}{2} \left[ \frac{\sigma}{\sigma + \sigma_\text{high,S}} \right]$$

$$F_{ch}(\sigma) \propto \frac{1}{2} \left[ \frac{\sigma}{\sigma + \sigma_\text{low,C}} \right] - \frac{1}{2} \left[ \frac{\sigma}{\sigma + \sigma_\text{high,C}} \right]$$

with conductivities:

$$\sigma_\text{low,S} = 2\pi\,f\,\epsilon_0\,\epsilon_\text{sem,⊥}/h_\text{sem}$$

$\sigma_\text{high,S} = 2\pi\,f\,C_\text{int}$

$$\sigma_\text{low,C} \propto 2\pi\,f\,\epsilon_0\,\epsilon_\text{sem,∥}/L_\text{sem}$$

$$\sigma_\text{high,C} \propto 2\pi\,f\,C_\text{int}$$

Geometric factors (apex/cone areas, channel width/length, substrate thickness) contribute explicitly to the full analytic solution.

$C_\text{int}$ encodes the volumetric (Stern) capacitance of ion-penetrated polymer regions, reflecting the static coupling of ionic and electronic charges. Material parameters ($h_\text{sem}$, $\varepsilon_\text{sem}$, $\sigma_\text{sem}$) determine force response thresholds and amplitude, thus directly linking SDM observables to physical properties.

5. Correlation with OECT Device Physics and Operation

SDM local transition voltages correspond systematically to recognizable features in OECT electrical response:

• $V_G^\text{low,C}$ (channel insulator$\rightarrow$semi): coincident with the onset of $g_m$ (device turn-on).
• $V_G^\text{high,C}$ (full channel conduction): matches the $g_m$ peak.
• $V_G^\text{low,S}$ (source transversal conduction onset): aligned with the $I_{DS}$ minimum.
• $V_G^\text{high,S}$ (source over-doping collapse): reflected in negative curvature (anti-ambipolar drop) at high gate. Interplay between source/drain transitions governs inflection points in $g'$ ($d^2I_{DS}/dV_G^2$), mapping region-switching in channel control.

Ion chemistry modulates these transitions: comparing 10 mM NH₄Cl to Milli-Q water reveals unchanged $V_G^\text{low,C}$, but a narrowed $V_G^\text{high,C} - V_G^\text{low,C}$ window and halved $I_{DS,\max}$. SDM identifies reduced dielectric response plateau at high $V_G$ ($\Rightarrow$ lower $C_\text{int}$), while mechanical mapping registers suppressed swelling/softening, evidencing ion-specific polymer binding that curtails volumetric expansion.

6. Implications for Mixed-Conductor Device Design and Sensing

Automated operando multimodal in-liquid SDM provides a framework to infer local material/interfacial parameters from macroscopic electrical measurements. The method enables co-optimization of electrolyte composition (ion size, binding affinity) and polymer design (swelling, ionic channel architecture) to engineer threshold voltages, transconductance, and anti-ambipolar switching behavior. A plausible implication is the universal extensibility of the approach to other OMIECs (including p-type and ambipolar) and to in situ sensing of biological analytes via OECT transduction.

In summary, automated operando multimodal in-liquid SDM rigorously quantifies ion-electron-mechanical coupling in mixed conductors, underpinning predictive modeling and functional design of next-generation bioelectronic, memory, and neuromorphic devices (Tanwar et al., 8 Jan 2026).