Focused Ion Beam And Scanning Electron Microscopy Research Articles

Investigation of optical polarization characteristics of ultraviolet-C AlGaN multiple quantum wells by angle-resolved cathodoluminescence

AlGaN-based ultraviolet-C (UV-C) light-emitting diodes (LEDs) face challenges related to their extremely low external quantum efficiency, which is predominantly attributed to the remarkably inadequate transverse magnetic (TM) light extraction efficiency (LEE). In this study, we employ angle-resolved cathodoluminescence (ARCL) spectroscopy to assess the optical polarization of (0001)-oriented AlGaN multiple quantum well (MQW) structures in UV-C LEDs, in conjunction with a focused ion beam and scanning electron microscopy (FIB/SEM) system to etch samples with various inclination angles (θ) of sidewall. This technique effectively distinguishes the spatial distribution of TM- and transverse electric (TE)-polarized photons contributing to the luminescence of the MQW structure. CL spectroscopy confirms that UV-C LEDs with a θ of 35° exhibit the highest CL signal compared to samples with other θ. Furthermore, we establish a model using finite difference time domain (FDTD) simulation to validate the mechanism of the outcomes. The complementary contribution of TM and TE photons at different specific angles are distinguished by ARCL and confirmed by simulation. At angles near the sidewall, the CL is dominated by the TM photons, which mainly contribute to the increased LEE and the decreased degree of polarization (DOP) to make the spatial distribution of CL more uniform. Additionally, this method allows us to analyze the polarization of light without the need for polarizers, enabling the differentiation of TE and TM modes. This distinction provides flexibility for selecting different emission mode based on various application requirements. The presented approach not only opens up new opportunities for enhanced UV-C light extraction but also provides valuable insights for future endeavors in device fabrication and epitaxial film growth.

Optical power degradation mechanisms in 271 nm AlGaN-based deep ultraviolet light-emitting diodes.

The degradation of AlGaN-based UVC LEDs under constant temperature and constant current stress for up to 500 hrs was analyzed in this work. During each degradation stage, the two-dimensional (2D) thermal distributions, I-V curves, optical powers, combining with focused ion beam and scanning electron microscope (FIB/SEM), were thoroughly tested and analyzed the properties and failure mechanisms of UVC LEDs. The results show that: 1) the opto-electrical characteristics measured before/during stress indicate that the increased leakage current and the generation of stress-induced defects increase the non-radiative recombination in the early stress stage, resulting in a decrease in optical power; 2) the increase of temperature caused by the deterioration of the Cr/Al layer of p-metal after 48 hrs of stress aggravates the optical power in UVC LEDs. The 2D thermal distribution in conjunction with FIB/SEM provide a fast and visual way to precisely locate and analyze the failure mechanisms of UVC LEDs.

A robust approach for MicroED sample preparation of lipidic cubic phase embedded membrane protein crystals

Crystallizing G protein-coupled receptors (GPCRs) in lipidic cubic phase (LCP) often yields crystals suited for the cryogenic electron microscopy (cryoEM) method microcrystal electron diffraction (MicroED). However, sample preparation is challenging. Embedded crystals cannot be targeted topologically. Here, we use an integrated fluorescence light microscope (iFLM) inside of a focused ion beam and scanning electron microscope (FIB-SEM) to identify fluorescently labeled GPCR crystals. Crystals are targeted using the iFLM and LCP is milled using a plasma focused ion beam (pFIB). The optimal ion source for preparing biological lamellae is identified using standard crystals of proteinase K. Lamellae prepared using either argon or xenon produced the highest quality data and structures. MicroED data are collected from the milled lamellae and the structures are determined. This study outlines a robust approach to identify and mill membrane protein crystals for MicroED and demonstrates plasma ion-beam milling is a powerful tool for preparing biological lamellae.

Quantitative analysis of the structural evolution in Si anode via multi-scale image reconstruction

Despite the high theoretical capacity, silicon (Si) anode suffers from dramatical capacity loss, due to its massive volume swings (up to 300%) during cycling. Hence, thorough understanding of the structural evolution mechanism is necessary and essential for performance optimization of Si anode. Herein, a multi-scale three-dimensional (3D) image reconstruction technique is firstly applied to visualize the structural evolution process of Si anodes. Three key components (Si particles, inactive components, and voids) in the electrode are quantitatively analyzed by the focused ion beam and scanning electron microscope (FIB-SEM) technology. Furthermore, the average sizes of Si particles were run statistics during the cycling. By combining the componential observation within the electrode (macroscopic information) and the 3D models of the particle with solid electrolyte interphase (SEI) layer (microscopic information), the failure mechanism of Si anode is vividly demonstrated. This work establishes a new methodology to quantitatively analyze the structural and compositional evolution of Si anode, which could be further applied for the studies of many other electrode materials with similar issues.

FIB-SEM tomography reveals the nanoscale 3D morphology of virus removal filters

Virus removal filters are designed to ensure very high levels of removal of viruses and virus-like particles during bioprocessing, but the performance of some virus removal filters can be compromised at high-throughputs, after process disruptions, or during operation at low pressures. Several studies have hypothesized that the different behavior is due to differences in underlying pore morphology, but current techniques are limited to examining the 2D pore structure. Here, we use the combination of a focused ion beam and scanning electron microscopy (FIB-SEM) to obtain 3D reconstructions of the pore structure of the asymmetric Viresolve® Pro virus removal filter. Images were obtained through a 3 μm section into the membrane starting at the size-selective skin. The membrane porosity decreases from 41 to 17% as one approaches the filter exit. Model simulations based on flow through the full 3D pore reconstruction show 100% virus retention, with all virus particles captured at least 400 nm from the filter exit. A pore-network model was developed from the reconstructions and used to evaluate the body and throat size distribution and pore interconnectivity. The number of throats is approximately twice the number of bodies, with an average throat size of 21 nm within the selective skin. The pore interconnectivity remains relatively constant at a value of 4 throats per body. These results provide insights into the underlying pore structure of the Viresolve® Pro virus removal filter as well as a general framework for characterizing the 3D pore space in nanoporous membranes.

Tessellation-based stochastic modelling of 3D coating structures imaged with FIB-SEM tomography

To facilitate printing, coatings are typically applied to paperboard used for packaging to provide a good surface for application. To optimise the performance of the coating, it is important to understand the relationship between the microstructure of the material and its mass transport properties. In this work, three samples of paperboard coating are imaged using combined focused ion beam and scanning electron microscope (FIB-SEM) tomography data appropriately segmented to characterise the internal microstructure. These images are used to inform a parametric, tessellation-based stochastic three-dimensional model intended to mimic the irregular geometry of the particles that can be seen in the coating. Parameters for the model are estimated from the FIB-SEM image data, and we demonstrate good agreement between the real and virtual structures both in terms of geometrical measures and mass transport properties. The development of this model facilitates exploration of the relationship between the structure and its properties.

Convolutional neural networks for segmentation of FIB-SEM nanotomography data from porous polymer films for controlled drug release.

Phase-separated polymer films are commonly used as coatings around pharmaceutical oral dosage forms (tablets or pellets) to facilitate controlled drug release. A typical choice is to use ethyl cellulose and hydroxypropyl cellulose (EC/HPC) polymer blends. When an EC/HPC film is in contact with water, the leaching out of the water-soluble HPC phase produces an EC film with a porous network through which the drug is transported. The drug release can be tailored by controlling the structure of this porous network. Imaging and characterization of such EC porous films facilitates understanding of how to control and tailor film formation and ultimately drug release. Combined focused ion beam and scanning electron microscope (FIB-SEM) tomography is a well-established technique for high-resolution imaging, and suitable for this application. However, for segmenting image data, in this case to correctly identify the porous network, FIB-SEM is a challenging technique to work with. In this work, we implement convolutional neural networks for segmentation of FIB-SEM image data. The data are acquired from three EC porous films where the HPC phases have been leached out. The three data sets have varying porosities in a range of interest for controlled drug release applications. We demonstrate very good agreement with manual segmentations. In particular, we demonstrate an improvement in comparison to previous work on the same data sets that utilized a random forest classifier trained on Gaussian scale-space features. Finally, we facilitate further development of FIB-SEM segmentation methods by making the data and software used openaccess.

Crack-Free Silicon Monoxide as Anodes for Lithium-Ion Batteries.

The volume expansion of Si and SiO particles was investigated using a single-particle battery assembled with a focused ion beam and scanning electron microscopy (FIB-SEM) system. Single Si and SiO particles were galvanostatically charged and discharged as in real batteries. Microstructural changes of the particles were monitored in situ using FIB-SEM from two different angles. The results revealed that the volume expansion of micrometer size particle SiO was not only much smaller than that of Si, but it also kept its original shape with no sign of cracks. This isotropic mechanical property of a SiO particle can be attributed to its microstructure: nanosized Si domains mixed with SiO2 domains. The nanosized Si domains can mitigate the anisotropic swelling caused by the orientation-dependent lithium-ion insertion; the surrounding SiO2 domains can act as a buffer to further constrain the localized anisotropic swelling.

Three-dimensional reconstruction of porous polymer films from FIB-SEM nanotomography data using random forests.

SummaryCombined focused ion beam and scanning electron microscope (FIB‐SEM) tomography is a well‐established technique for high resolution imaging and reconstruction of the microstructure of a wide range of materials. Segmentation of FIB‐SEM data is complicated due to a number of factors; the most prominent is that for porous materials, the scanning electron microscope image slices contain information not only from the planar cross‐section of the material but also from underlying, exposed subsurface pores. In this work, we develop a segmentation method for FIB‐SEM data from ethyl cellulose porous films made from ethyl cellulose and hydroxypropyl cellulose (EC/HPC) polymer blends. These materials are used for coating pharmaceutical oral dosage forms (tablets or pellets) to control drug release. We study three samples of ethyl cellulose and hydroxypropyl cellulose with different volume fractions where the hydroxypropyl cellulose phase has been leached out, resulting in a porous material. The data are segmented using scale‐space features and a random forest classifier. We demonstrate good agreement with manual segmentations. The method enables quantitative characterization and subsequent optimization of material structure for controlled release applications. Although the methodology is demonstrated on porous polymer films, it is applicable to other soft porous materials imaged by FIB‐SEM. We make the data and software used publicly available to facilitate further development of FIB‐SEM segmentation methods.Lay DescriptionFor imaging of very fine structures in materials, the resolution limits of, e.g. X‐ray computed tomography quickly become a bottleneck. Scanning electron microscopy (SEM) provides a way out, but it is essentially a two‐dimensional imaging technique. One manner in which to extend it to three dimensions is to use a focused ion beam (FIB) combined with a scanning electron microscopy and acquire tomography data. In FIB‐SEM tomography, ions are used to perform serial sectioning and the electron beam is used to image the cross section surface. This is a well‐established method for a wide range of materials. However, image analysis of FIB‐SEM data is complicated for a variety of reasons, in particular for porous media. In this work, we analyse FIB‐SEM data from ethyl cellulose porous films made from ethyl cellulose and hydroxypropyl cellulose (EC/HPC) polymer blends. These films are used as coatings for controlled drug release. The aim is to perform image segmentation, i.e. to identify which parts of the image data constitute the pores and the solid, respectively. Manual segmentation, i.e. when a trained operator manually identifies areas constituting pores and solid, is too time‐consuming to do in full for our very large data sets. However, by performing manual segmentation on a set of small, random regions of the data, we can train a machine learning algorithm to perform automatic segmentation on the entire data sets. The method yields good agreement with the manual segmentations and yields porosities of the entire data sets in very good agreement with expected values. The method facilitates understanding and quantitative characterization of the geometrical structure of the materials, and ultimately understanding of how to tailor the drug release.

A nanostructure platform for live-cell manipulation of membrane curvature.

Membrane curvatures are involved in essential cellular processes, such as endocytosis and exocytosis, in which they are believed to act as microdomains for protein interactions and intracellular signaling. These membrane curvatures appear and disappear dynamically, and their locations are difficult or impossible to predict. In addition, the size of these curvatures is usually below the diffraction limit of visible light, making it impossible to resolve their values using live-cell imaging. Therefore, precise manipulation of membrane curvature is important to understanding how membrane curvature is involved in intracellular processes. Recent studies show that membrane curvatures can be induced by surface topography when cells are in direct contact with engineered substrates. Here, we present detailed procedures for using nanoscale structures to manipulate membrane curvatures and probe curvature-induced phenomena in live cells. We first describe detailed procedures for the design of nanoscale structures and their fabrication using electron-beam (E-beam) lithography. The fabrication process takes 2 d, but the resultant chips can be cleaned and reused repeatedly over the course of 2 years. Then we describe how to use these nanostructures to manipulate local membrane curvatures and probe intracellular protein responses, discussing surface coating, cell plating, and fluorescence imaging in detail. Finally, we describe a procedure to characterize the nanostructure-cell membrane interface using focused ion beam and scanning electron microscopy (FIB-SEM). Nanotopography-based methods can induce stable membrane curvatures with well-defined curvature values and locations in live cells, which enables the generation of a library of curvatures for probing curvature-related intracellular processes.

New Workflows Broaden Access to S/TEM Analysis and Increase Productivity

Abstract

Characterization and analysis of Coscinodiscus genus frustule based on FIB-SEM

Coscinodiscus genus, a type of diatoms with complex frustule, has been widely studied and reported because of their delicate nanostructures. In this paper, a dual beam system focused ion beam and scanning electron microscope (FIB-SEM) were applied to precisely investigate the microstructure of diatom cell walls, in particular, aiming to reveal the hierarchical pores of the valve. The microstructures of the valve, valve mantle and girdle bands were illustrated in details through a series of high resolution images after performing the cross section milling of the specific region. The 3D morphology of frustule valve was reconstructed based on the 2D image series, which demonstrated the presence of cribellum in valve, the external foramen, hexagonal cavity and the internal foramen. The four dense porous membranes and girdle bands were characterized clearly based on FIB-SEM dual beam system. An integrated model of Coscinodiscus frustule could be simulated based on the 2D and 3D results. This study provided a systematic approach to measure the morphological features of diatoms at a nanoscale, which could be applied to other nanoporous structure in three dimensions.

Formulation of ammonia decomposition rate in Ni-YSZ anode of solid oxide fuel cells

Ammonia decomposition rate on a conventional porous Ni-YSZ anode of solid oxide fuel cells (SOFCs) is experimentally investigated under various temperatures and gas compositions to formulate a reaction rate equation. It is confirmed that the decomposition rate depends not only on the partial pressure of ammonia but also on that of hydrogen at the common operating conditions of SOFCs, with larger hydrogen concentration significantly reducing the decomposition rate. 3D microstructure of the tested Ni-YSZ anode is analyzed using the focused ion beam and scanning electron microscope (FIB-SEM), and microstructural parameters are quantified. By using the quantified Ni–Pore interface area density, a reaction rate model for a unit reaction area in Ni-YSZ anodes is obtained. The proposed model is then applied to a simple computational fluid dynamics (CFD) model for validation. The proposed reaction rate model can be applied to a variety of numerical analysis of ammonia decomposition on Ni-YSZ anodes having different microstructures, and expected to be useful for the analysis of ammonia-fed SOFCs.

Two-scale analysis of a tight gas sandstone

Tight gas sandstones are low porosity media, with a very small permeability (i.e., below 1 mD). Their porosity is below 10%, and it is mainly composed of fine noncemented microcracks, which are present between neighboring quartz grains. While empirical models of permeability are available, their predictions, which do not compare well with macroscopic measurements, are not reliable to assess gas well productivity. The purpose of this work is to compare the permeability measured on centimetric plugs to predictions based on pore structure data. Two macroscopic measurements are performed, namely dry gas permeability and mercury intrusion porosimetry (MIP), together with a series of local measurements including focused ion beam and scanning electron microscopy (FIB-SEM), x-ray computed microtomography (CMT), and standard two-dimensional (2D) SEM. Numerical modeling is performed by combining analyses on two scales, namely the microcrack network scale (given by 2D SEM) and the individual 3D microcrack scale (given by either FIB-SEM or CMT). The network permeability is calculated by means of techniques developed for fracture networks. This permeability is proportional to the microcrack transmissivity, which is determined by solving the Stokes equation in the microcracks measured by FIB-SEM or CMT. Good correlation with experimental permeability values is only found when using transmissivity from 3D CMT data.

Far-reaching geometrical artefacts due to thermal decomposition of polymeric coatings around focused ion beam milled pigment particles.

Focused ion beam and scanning electron microscope (FIB-SEM) instruments are extensively used to characterize nanoscale composition of composite materials, however, their application to analysis of organic corrosion barrier coatings has been limited. The primary concern that arises with use of FIB to mill organic materials is the possibility of severe thermal damage that occurs in close proximity to the ion beam impact. Recent research has shown that such localized artefacts can be mitigated for a number of polymers through cryogenic cooling of the sample as well as low current milling and intelligent ion beam control. Here we report unexpected nonlocalized artefacts that occur during FIB milling of composite organic coatings with pigment particles. Specifically, we show that FIB milling of pigmented polysiloxane coating can lead to formation of multiple microscopic voids within the substrate as far as 5 μm away from the ion beam impact. We use further experimentation and modelling to show that void formation occurs via ion beam heating of the pigment particles that leads to decomposition and vaporization of the surrounding polysiloxane. We also identify FIB milling conditions that mitigate this issue.

Nanoscale chemical mapping of Li-ion battery cathode material by FIB-SEM and TOF-SIMS multi-modal microscopy

Lithium ion batteries are a leading energy storage technology for electronic portable devices and hybrid electric vehicles. A long-standing challenge to understand Li ion transport is the fact that this light atom escapes direct detection by X-ray techniques (EDX, XRD and XAS) due to weak scattering and emission properties. Simultaneous characterisation of the structure, chemical composition (including Li) and elemental distribution in Li-ion battery materials can reveal the relationship between Li ion transport and structural effects (phase transformation, internal stress), and battery performance and degradation. Nanoscale mapping is achieved by multi-modal correlative microscopy combining focused ion beam and scanning electron microscopy (FIB-SEM) with Time-of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS). Mapping Li, Mn and Co nanoscale distributions reveals the micro-structural consequences of the electrochemical reaction, and allows identifying Li “trapping” sites within the structure that control materials properties, and open the way towards designing better Li-ion cathode materials with superior performance.

Evaluation of Exchange Current Density for LSM Porous Cathode Based on Measurement of Three-Phase Boundary Length

The microstructure of porous electrodes has a strong impact on a power generation performance of solid oxide fuel cells. Sufficient gas diffusion, electron transport and oxide-ion transport to the reaction sites is necessary for active electrochemical reactions. Recent advancement of 3-D microstructure observation by a focused ion beam and scanning electron microscope (FIB-SEM) or micro X-ray computed tomography enabled quantification of electrode microstructure. The three dimensional datasets of porous electrodes are also applied to numerical simulations of the electrode's polarization characteristics. Understanding the correlation between the electrode performance and microstructure is important to obtain guidelines for optimal microstructure. The numerical simulation requires a rate equation for the electrochemical reaction occurring at the reaction sites. It is often expressed by Butler-Volmer like form that includes the density of three-phase or two-phase boundary and the exchange current density per unit reaction site, i.e. unit triple phase boundary length for three phase boundary (TPB) reactions and unit area for two phase boundary reactions. Generally, the exchange current density per unit TPB length was derived from experiments using flat and dense patterned electrodes since there is an advantage that TPB length can be determined from its geometric shapes. However, we often observe a discrepancy between the polarization characteristics obtained from the simulation and the experimental measurements. One of the possible reasons for the discrepancy is the rate equation evaluated from the patterned electrodes which structure is very different from the real porous electrodes. This study focuses on LSM (Lanthanum Strontium Manganite) porous cathode. Although the number is still limited, we can find reports on the exchange current density per unit TPB length obtained from patterned LSM electrodes typically with a thickness of several hundred nanometers. It is also reported in literature that the thickness of patterned LSM electrodes affects activation overpotential. Since LSM has small oxide ion conductivity, oxide ion can be transferred through LSM phase, and the charge transfer at interface of LSM phase and pore phase (two phase boundary) can occur. This phenomenon may bring overestimation of the exchange current density per unit TPB length in the experiment with patterned LSM. In this study, we derive a model of exchange current density per unit TPB length considering its temperature and oxygen partial pressure dependency from experiments using LSM porous cathode with a particle diameter of approximately 3 micrometers. We assume that the charge transfer occurs only at TPB in this large particle. We first conducted power generation experiments using LSM/YSZ/Ni-YSZ button cells. For the reference electrode, a platinum wire was attached so as to surround the side edge of the electrolyte disk. Therefore the potential difference between the cathode and the reference electrode can be obtained. Activation overpotentials were obtained by impedance spectroscopy varying temperature (1073 - 1223 K) and oxygen partial pressure (5 - 21 %). After the power generation experiment, the cathode microstructure was observed by FIB-SEM and three-dimensional microstructure dataset with dimensions of 18 μm / 25 μm / 12 μm was obtained. From the geometry of its microstructure, TPB length was extracted and quantified. The exchange current densities per unit TPB length were calculated by substituting current density, activation overpotential and TPB density into Butler-Volmer like form. Eventually, obtained exchange current densities per unit TPB length are fitted into a power-law equation that has temperature and oxygen partial pressure term. As a result, obtained exchange current densities per unit TPB length in this study is found to be smaller than the ones estimated from patterned electrode at any condition. The discrepancy between them increased with decreasing temperature and/or with increasing oxygen partial pressure. For instance, the exchange current density derived in this study was less than half of that derived by Radhakrishnan et al. from patterned electrode at 1023 K for 21 % of oxygen. For actual LSM or LSM/YSZ composite cathodes using relatively large LSM particles, the equation of exchange current density per unit TPB length derived in this study is appropriate. Figure 1

Towards the Design-Led Optimization of Solid Oxide Fuel Cell Electrodes

Electrode microstructure plays an important role in determining the performance and durability of solid oxide fuel cells (SOFCs). It needs to be tailored towards a variety of requirement for the electrodes, such as transport properties and electrochemical activity. Long-term stability of the microstructure is also important to guarantee the durability of SOFC systems. In order to satisfy the growing demand for their performance and durability, an improved understanding of the microstructure-performance relationships is desired, along with optimization of the microstructure, through a design-led approach. The focus of this study is on electrodes fabricated with the nano-particle infiltration techniques. Porous framework structures (scaffolds) were first fabricated with gadolinium-doped ceria (GDC), and then nickel metal particles were introduced into the structures in the form of nano particles. 3D microstructural analysis using focused ion beam and scanning electron microscope (FIB-SEM) was carried out and revealed that the infiltration technique has a greater potential, compared with the conventional approach based on powder mixing and sintering, to control the electrode microstructure, helping satisfy multiple requirements for the electrodes. Most notably, the triple-phase boundary (TPB) density in the infiltrated electrodes was found to be one order of magnitude larger than that typically found in the conventional electrodes. A 1D numerical model for Ni-GDC electrodes was also developed assuming that the GDC phase has mixed ionic and electronic conductivity and hence that the electrochemical reaction takes place on the GDC-pore double-phase boundary (DPB), and successfully reproduced the overpotential characteristics obtained from the experiment. Sensitivity analysis was also conducted with the developed model at 700 °C to investigate the effect of electrode microstructure on electrode performance. This revealed that the electrochemical reaction on the DPB is the rate-determining process within the electrodes; therefore increasing the DPB density is recommended as the most effective route to improving performance of ceria-based electrodes, rather than improving species transport rate. Using the insights from the experiment, microstructural analysis and numerical simulation, guidelines for further optimizing the electrode microstructure are proposed Figure 1

Local Evolution of Three-Dimensional Microstructure Morphology of Planar Anode-Supported SOFC

In this research a planar anode-supported cell was tested over the period of 3700 hours. The investigated cell was tested as a part of six cells stack with standard power output 100 [W]. The aging test was conducted at 800oC and under constant current 19.4 A to provide 90 W of initial power output. After the operation subsequent post-test analysis of the selected cell microstructure has been conducted using combination of Focused Ion Beam and Scanning Electron Microscopy (FIB-SEM). Obtained data was reconstructed into 3D images, based on which the microstructure parameters were obtained. The microstructure parameters were quantified at three different location in the selected cell; inlet (downstream), middle and outlet (upstream). The obtained results indicate strong non-homogeneous microstructure morphology change after long-term operation.

Towards the 3D Modelling of the Effective Conductivity of Solid Oxide Fuel Cell Electrodes – Validation against experimental measurements and prediction of electrochemical performance

The effective conductivity of thick-film solid oxide fuel cell (SOFC) electrodes plays a key role in their performance. It determines the ability of the electrode to transport charge to/from reaction sites to the current collector and electrolyte. In this paper, the validity of the recently proposed 3D resistor network model for the prediction of effective conductivity, the ResNet model, is investigated by comparison to experimental data. The 3D microstructures of Ni/10ScSZ anodes are reconstructed using tomography through the focused ion beam and scanning electron microscopy (FIB-SEM) technique. This is used as geometric input to the ResNet model to predict the effective conductivities, which are then compared against the experimentally measured values on the same electrodes. Good agreement is observed, supporting the validity of the ResNet model for predicting the effective conductivity of SOFC electrodes. The ResNet model is then combined with the volume-of-fluid (VOF) method to integrate the description of the local conductivity (electronic and ionic) in the prediction of electrochemical performance. The results show that the electrochemical performance is in particular sensitive to the ionic conductivity of the electrode microstructure, highlighting the importance of an accurate description of the local ionic conductivity.