Search

1,4-Dioxane is a toxic chemical that is used in the chemical industry. The limit concentration of 1,4-dioxane in environment was set at 0.3 mg/kg. Here, for efficient degradation of 1,4-dioxane, combined use of a stripping column and a continuous stirred tank reactor (CSTR) is proposed. The stripping column concentrates 1,4-dioxane, and photo-Fenton reaction in the CSTR decomposes 1,4-dioxane by OH radicals generated by an ultraviolet lamp. In the photo-Fenton reaction, the pseudo-first-order reaction constant was determined to be 1.88 × 10−4 1/s at 318 K. The separation coefficient of 1,4-dioxane, the ratio of the concentrations at the top and bottom of the stripping column, was approximately 500–2700 depending on the heat supplied to the column. The optimum combination of the stripping column and CSTR was investigated on the basis of three proposed models. The order of degradation in the CSTR followed by concentration in the stripping column, along with circulation from the top of the column to the CSTR, gave an overall reaction percentage of 99.9%. From energy calculation, a higher separation coefficient (α) decreased the total heat duty from 4.5 × 104 J/min at α = 640–1.2 × 104 J/min at α = 2400. A stripping column for concentration of 1,4-dioxane combined with a CSTR for photo-Fenton reaction shows potential for efficient degradation of 1,4-dioxane. [Display omitted] • 1,4-Dioxane is toxic requiring the process for removal. • Combined use of a stripping column and a CSTR is proposed for degradation of 1,4-dioxane. • The order of degradation in the CSTR and stripping column with circulation through CSTR gave an overall reaction percentage of 99.9%. • Higher separation coefficient decreased the total heat duty from 4.5 × 104 J/min at α = 640–1.2 × 104 J/min at α = 2400. [ABSTRACT FROM AUTHOR]

In water electrolysis experiments on hydrogen production, boiling effectively decreases the overpotential of the oxygen evolution reaction (OER) [Li et al. Int. J. Hydrog. Energy. 2022, 47(61):25499–510]. This study constructed a theoretical model to quantitatively explore the mechanism by which boiling reduces the OER overpotential. The model considers dissolved oxygen, liquid water, gaseous oxygen, and water vapor flowing through the porous transfer layer (PTL). The results showed that boiling increases the molar flux of gaseous oxygen and reduces the dissolved oxygen concentration. The PTL embedded in the catalyst layer (CL) increased the mass transfer resistance of dissolved oxygen and increased the dissolved oxygen concentration in the CL. Once boiling is superimposed on the OER, it attracts more dissolved oxygen to the water vapor bubbles and significantly decreases the dissolved oxygen concentration at the CL, thereby reducing the Nernst loss of the OER. • A dimensionless OER model with boiling is developed. • The model quantitatively clarifies the OER characteristics at boiling. • Boiling decreases dissolved oxygen concentration at the CL. • Current contribution ratio for the OER under bubbles rises at boiling. [ABSTRACT FROM AUTHOR]

Background: Magnetic resonance imaging is widely used to evaluate the intraplaque components of the cervical carotid artery. The non-gated T1-weighted spin-echo (SE) technique has been reported to have an excellent ability for discriminating stable and unstable plaques. However, the diagnostic performance of various SE-based techniques remains unclear. Hence, we compared plaque signals obtained by 3 kinds of SE-based methods with histological findings.Methods: We prospectively examined 40 patients who underwent carotid endarterectomy by using 1.5-T scanners and obtained 2-dimensional (2D) conventional spin-echo (CSE), 2D fast spin-echo (FSE), and 3-dimensional (3D)-FSE images with identical repetition times. We calculated contrast ratios (CRs) of the plaques against adjacent muscles and compared these values with the pathological classification of the specimens.Results: The CRs of type VII-VIII (calcific/fibrous), IV-V (lipid-rich/necrotic), and VI (complex/hemorrhagic) plaques were significantly different between all the methods (P <.001) and were discriminated from each other at sensitivities of 83%-100% and specificities of 94%-100%. The CRs of type IV-V plaques significantly differed between the methods (low to high, 2D-FSE, 2D-CSE, and 3D-FSE; P <.05); those of the type VI plaques were significantly lower with the 2D-FSE method than with the other methods (P <.01).Conclusions: The SE-based T1-weighted images can readily discriminate plaque characteristics with high sensitivities and specificities, although the signal intensity of unstable plaques was significantly high on the 3D-FSE images and significantly low on the 2D-FSE images. [ABSTRACT FROM AUTHOR]

Background: Magnetic resonance vessel wall imaging (VWI) techniques have been developed to assess atherosclerotic plaques in intracranial arteries, which are a cardinal cause of ischemic stroke. However, the clinical roles of plaque-related vulnerability and inflammation remain unclear. Hence, we evaluated plaque characteristics using VWI of the proximal middle cerebral artery (M1) in patients with acute ischemic stroke.Methods: We prospectively examined 30 consecutive patients with acute noncardioembolic stroke in the M1 territory using pre-/postcontrast T1-weighted (T1W) three-dimensional (3D) VWI with a 3-Tesla scanner. The contrast ratio (CR) and contrast enhancement of the plaques were measured bilaterally at M1.Results: Plaques were identified in the bilateral M1s of all patients, and no substantial stenosis existed. The M1 plaque CRs ipsilateral to the infarct (46.7%-67.9%) were significantly higher than the plaque CRs on the contralateral side (34.3%-69.4%), particularly in patients with lacunar infarcts (P <.01). In contrast, the occurrence of plaque enhancement was not different between the ipsilateral (20.0%) and contralateral (16.7%) sides. Further, the CRs in the nonlacunar group were significantly higher than the CRs in the lacunar group (P <.05), whereas enhanced plaques tended to be more frequent in the nonlacunar group, but this difference was not significant (P = .09).Conclusions: T1W 3D-VWI revealed that the signal intensity of M1 plaques was significantly higher in the affected side and in nonlacunar-type infarcts of patients with acute stroke, suggesting that unstable plaques in the M1 can cause stroke events presumably due to atherothrombotic mechanisms. [ABSTRACT FROM AUTHOR]

When a solid polymer water electrolyzer (SPWE) is operated under high-pressure conditions, a large pressure difference occurs between the anode and cathode. This causes crossover of produced gas, especially hydrogen gas, leading to a decrease in the production efficiency of an SPWE. As a countermeasure against gas crossover, water should be supplied into the cathode channel, as well as into the anode channel, because the water flow will facilitate the drainage of hydrogen gas outside of the cell, resulting in decreased crossover and increased efficiency of the SPWE. This countermeasure is evaluated by observing SPWE operation at a pressure of 2 MPa, with a visualization of hydrogen bubbles in the cathode channel. The evaluation revealed that supplying water into the cathode channel increases the efficiency by several percent at 0.33 A/cm 2 . Further, the visualization of the hydrogen bubbles revealed an enhancement in the separation of hydrogen bubbles from the surface of the current supplier. This suggests that additional water flow can increase the hydrogen production efficiency through promoting bubble detachment. [ABSTRACT FROM AUTHOR]

A fusion enzyme composed of an Aspergillus flavus- derived flavin adenine dinucleotide glucose dehydrogenase (AfGDH) and an electron transfer domain of Phanerochaete chrysosporium- derived cellobiose dehydrogenase (Pcyb) was previously reported to show the direct electron transfer (DET) ability to an electrode. However, its slow intramolecular electron transfer (IET) rate from the FAD to the heme, limited the sensor signals. In this study, fusion FADGDH (Pcyb-AfGDH) enzymes were strategically redesigned by performing docking simulation, following surface-electrostatic potential estimation in the predicted area. Based on these predictions, we selected the amino acid substitution on Glu324, or on Asn408 to Lys to increase the positive charge at the rim of the interdomain region. Pcyb-AfGDH mutants were recombinantly produced using Pichia pastoris as the host microorganism, and their IET was evaluated. Spectroscopic observations showed that the Glu324Lys (E324K) and Asn408Lys (N408K) Pcyb-AfGDH mutants showed approximately 1.70- and 9.0-fold faster IET than that of wildtype Pcyb-AfGDH, respectively. Electrochemical evaluation revealed that the mutant Pcyb-AfGDH-immobilized electrodes showed higher DET current values than that of the wildtype Pcyb-AfGDH-immobilized electrodes at pH 6.5, which was approximately 9-fold higher in the E324K mutant and 15-fold higher in the N408K mutant, than in the wildtype. Glucose enzyme sensors employing N408K mutant was able to measure glucose concentration under physiological condition using artificial interstitial fluid at pH 7.4, whereas the one with wildtype Pcyb-AfGDH was not. These results indicated that the sensor employed the redesigned mutant Pcyb-AfGDH can be used for future continuous glucose monitoring system based on direct electron transfer principle. (247 words). • The docking simulation suggested interaction sites on the interface of domains. • The surface electrostatics suggested mutations were effective to reduce repulsion. • Recombinantly expressed mutants showed original AfGDH enzymatic characters. • The spectroscopic observation of mutants showed significantly higher IET. • The electrochemical investigation showed improvement of DET by modified IET. [ABSTRACT FROM AUTHOR]

Operation at high current density of polymer electrolyte membrane electrolysis cell (PEMEC) can reduce its electrode area and capital expense. The high current density operation is limited by the limiting current density (LCD), where electrolysis voltage abruptly increases. Effects of operating temperature and pressure on the LCD were evaluated in a lab-scale PEMEC within a range from 70 to 90 °C and from 0.1 to 0.3 MPa. Also, a theoretical model was developed to predict the LCD, and to clarify what determines LCD and how the operating conditions influence the LCD. The theoretical analysis suggests that when current density is close to the LCD, water mass transport at anode reaches a limitation, which indicates that water saturation reaches almost zero at anodic catalyst layer. The analysis also finds that lower operating temperature and higher operating pressure can reduce oxygen gas saturation and raise the water saturation in porous transport layer at anode, and finally enlarge the LCD. • Limiting current density (LCD) is decided by mass transport limitation. • Oxygen gas saturation near catalyst layer is close to 1 at LCD. • Bubble overvoltage abruptly rises at LCD. • Lower temperature and higher pressure enlarge LCD. [ABSTRACT FROM AUTHOR]

Novel Ru-core Ir-shell catalyst-integrated porous transport electrodes (PTEs) for polymer electrolyte membrane water electrolysis (PEMWE) cells are prepared, in which Ru-core Ir-shell catalyst nanostructures are directly deposited onto a porous transport layer (PTL) via arc plasma deposition (APD). The PTL has a nanostructured TiO 2 surface prepared via NaOH etching, acting as a catalyst support. The performance and durability of these Ru-core Ir-shell catalysts depend strongly on the ratio of Ir and Ru. The current-voltage (I–V) characteristics of PEMWE cells were improved by applying these core-shell catalysts with a low Ir loading of around 0.1 mg cm−2. The core-shell catalyst-integrated PTEs can operate at current densities of up to 10 A cm−2 without exhibiting limiting current behavior. This unique combination of the core-shell catalyst and the PTE structure enables PEMWE cell operation with low iridium loading and high current density, potentially reducing the cost of green hydrogen. • Ru-core Ir-shell electrocatalysts were prepared on Ti-based porous transport layers. • The activation overvoltage was significantly lower compared to pure Ir electrocatalysts. • The durability of the electrocatalyst was improved by increasing the Ir-to-Ru ratio. • This unique electrode can operate at high current densities up to 10 A cm−2. [ABSTRACT FROM AUTHOR]

The high initial cost of polymer electrolyte membrane water electrolyzers (PEMWEs) has delayed their widespread commercialization. A possible means to reduce cost is by reducing the overvoltage and increasing the current density to reduce the electrode area. This study proposes a novel method in which boiling is superimposed on the oxygen evolution reaction (OER) to decrease electrolysis voltage. The vapor bubbles formed by boiling are expected to decrease the concentration overvoltage. The boiling effect was experimentally analyzed using a three-electrode cell. Although a general catalyst layer (CL) was formed on a working electrode (WE) bar, the structure of the working electrode (WE) bar was special, in which a 10-W heater was embedded and made boiling on the electrode under 1 bar condition. Increasing the electrode temperature under static OER current density slightly decreased the OER potential. However, an abrupt decrease in potential was observed when the temperature was scaled over the boiling temperature. Moreover, this abrupt decrease substantially intensified when, similar to a practical PEMWE, a porous transfer layer (PTL) and flow channel were assembled on the CL. These experimental results suggest that boiling can reduce the concentration overvoltage by reducing the oxygen concentration on the CL, especially when the mass transport resistance caused by the PTL is considerable. Innovatively and simply utilizing boiling, as proposed here, can enhance the oxygen transfer and contribute to reducing the initial cost of PEMWEs. • OER is coupled with boiling. • Three-electrode cell confirms boiling effect on OER. • Theoretical modeling qualitatively predicts the boiling mechanism. • Oxygen activity in catalyst layer is estimated electrochemically. • Boiling reduces concentration overvoltage of OER. [ABSTRACT FROM AUTHOR]

This study evaluated the mechanical reliability of glass seal in planar type solid oxide fuel cells and clarified the relationship between glass wettability and the mechanical reliability of glass seal. A visualization experiment was developed to observe a specific model seal in a furnace subjected to temperature variation and the relationship between the glass wettability and the seal shape was measured. The seal model consists of glass (G018-311, Schott, Germany) and interconnectors (Crofer 22 APU, VDM Metals GmbH, Germany). The visualization experiment and analysis were based on the sessile drop method to quantify glass surface tension and contact angle between glass and interconnector. The equation balance of forces with these surface properties succeeded to reproduce the seal shape observed in the visualization. Subsequently, the seal stress field was numerically solved, and the mechanical strength of the seal was quantitatively evaluated using the stress intensity factor. The analyses concluded that the stress intensity factor is 1.5 times higher than its minimum value as the contact angle raised from 90 to 130° and glass wettability decreased. • The correlation between glass wettability and seal strength in SOFCs was clarified. • The FEM stress analysis based on the glass wettability can predict the seal strength. • Medium wettability is most effective to prevent the crack propagation inside the glass. • High wettability can prevent peeling off the glass from the interconnector. [ABSTRACT FROM AUTHOR]

• Critical current density (CCD) of alkaline water electrolysis was improved by honeycomb porous plate (HPP). • Capillary suction and path separation of gas-liquid effects are necessary for the CCD improvement. • CCD and boiling critical heat flux (CHF) showed similar trends with the thickness of HPP. • Honeycomb cooling method to improve the boiling CHF was confirmed to be applied to enhance the water electrolysis CCD. Since water electrolysis is a pollution-free and economical method to produce hydrogen which has a great potential as an energy carrier, it is imperative to improve the performance of water electrolysis for efficient hydrogen production. Supposing that the analogy between boiling and water electrolysis holds, the next interesting question is whether the water electrolysis performance can also be significantly improved by using a method that improves the heat transfer coefficient and the critical heat flux (CHF). The critical current density (CCD), which is the upper operation limit of water electrolysis, is considered to be one of the keys for effective hydrogen generation. However, the study on the enhancement of CCD has rarely been conducted. It is known that the boiling CHF can be improved by a honeycomb porous plate (HPP) due to two effects, capillary force and path separation of gas-liquid. In this work, we applied a cooling method using an HPP which has succeeded in improving the boiling CHF to alkaline water electrolysis. For the first time, we have succeeded in improving the CCD to approximately 1.3 times (CCD: 6.6 A/cm2) compared to without capillary force case (CCD: 5.1 A/cm2). As regards the mechanism, the effect of the liquid supply by capillary action and the improvement of the bubble discharge by separating the gas–liquid flow path using the honeycomb structure of the HPP play an important role in increasing the CCD. [ABSTRACT FROM AUTHOR]

Abstract The overpotentials generated in an electrochemical hydrogen pump (EHP) are experimentally separated using a hydrogen reference electrode. An intentionally thick polymer electrolyte membrane and a low-Pt-catalyst-loaded layer are embedded in the EHP cell for high accuracy in separation. The results show that the non-ohmic overpotential of the cathode is larger than that of the anode. Electrochemical impedance spectroscopy (EIS) is employed to analyze the non-ohmic overpotential in detail, and different features are observed in the Nyquist plots for the anode and the cathode, where the cathode spectra exhibits charge transfer rate-limiting features and the anode spectra displays mass transfer rate-limiting features. These different characteristics are theoretically clarified by the classical Volmer–Heyrovsky–Tafel mechanism. The hydrogen evolution reaction (HER) in the cathode is dominated by the Volmer–Heyrovsky route, and the hydrogen oxidation reaction (HOR) in the anode is dominated by the Volmer–Tafel route. Due to the slow reaction rate imposed by the Volmer–Heyrovsky route, the cathode HER displays high non-ohmic overpotential. With increasing cathode pressure, the rate of HER is increased due to the increasing coverage of the hydrogen adsorbed on the surface of the catalyst. [ABSTRACT FROM AUTHOR]

Water boiling in a polymer electrolyte membrane water electrolyzer (PEMWE) has been experimentally demonstrated to reduce the electrolysis voltage [Ito K et al. ECS Transactions. 2017; 80(8):1117]. The boiling phenomenon and formation mechanism are studied comprehensively by introducing boiling into a PEMWE model. The gas saturation at the channel/porous transfer layer (PTL) is first determined as the boundary condition for the gas transfer in the PTL. Then, the gas transfer equation can be used to determine the oxygen and hydrogen pressures at the catalyst layer (CL). The dissolved oxygen and hydrogen activities are considered in the activation overvoltages through a variable (activation fraction). Finally, the Nernst loss, activation, and ohmic overvoltage of the PEMWE over a wide temperature range (encompassing non-boiling, boiling, and gas-phase-only conditions) are determined from the oxygen and hydrogen pressures and gas saturation at the CL. The results show that the PEMWE under boiling conditions has the lowest electrolysis voltage among the non-boiling, boiling, and gas-phase conditions. Boiling can accelerate the detachment of gas bubbles composed of gaseous oxygen and water vapor in the channel (CH). Furthermore, vigorous boiling significantly reduces the Nernst loss by reducing the oxygen and hydrogen partial pressures at the CL. • Boiling is coupled in a PEMWE model. • Model quantitatively clarifies the boiling mechanism on the electrolysis voltage. • Boiling lowers the oxygen and hydrogen activities at the CL. • Anodic and cathodic Nernst losses decrease during boiling. • Boiling activates the anodic reaction under bubbles. [ABSTRACT FROM AUTHOR]

It was recently clarified on a microtubular Solid Oxide Fuel Cell (SOFC) that the range of mass transport limitation might commence from the inlet periphery (inlet opening and inlet pipe), i.e., the concentration gradient of reactants may extend inward the inlet periphery. For demonstrating that this phenomenon occurs regardless of the form and type of the fuel cell operating at high reactant utilization rate, herein we investigate the mass transport in the anode side of a one-cell stack of a planar SOFC. The investigation leans upon experimental and numerical data analyzed from both conventional (non spatial) and spatial perspectives. The experimental data were spatially obtained in the lateral direction by applying the segmentation method. Regarding analyses let us to confirm that mass transport limitation occurs in the inlet periphery of the planar stack. Besides, the critical ratio of the consumed/supplied mass fluxes of hydrogen is valid for assessing whether the concentration gradient of hydrogen extends inward the inlet periphery. Furthermore, the virtual inlet opening is useful for accurately calculating the mass transport within the active field of the stack via hypothetically preventing the mass transport limitation in the inlet periphery. These findings are expected to help researchers and engineers for accurately designing and characterizing fuel cell systems at varying scales from cells to stacks. [ABSTRACT FROM AUTHOR]

This study aimed to optimize the flow-field pattern and flow configuration of a polymer-electrolyte-membrane water electrolyzer, with a particular focus on high-temperature operation up to 120 °C. Three types of flow-field pattern (serpentine, parallel, and cascade) were tested in both the anode and cathode sides of a water electrolyzer cell, and the current-voltage characteristics and high-frequency resistance were measured to examine which overpotential components are impacted by the flow-field pattern. The experimental results revealed that the cathode flow-field pattern only affects the ohmic overpotential, while the anode flow-field pattern significantly affects the overpotential related to liquid water shortage at catalyst layer, and the flow configuration (counter- and co-flow) does not affect the electrolysis performance. Finally, under operating conditions of 120 °C and 0.3 MPa, we found that the optimized cell configuration consisted of cascade and serpentine flow-field patterns in the anode and cathode, respectively; this configuration produced the minimum electrolysis voltage of 1.69 V at 2 A/cm 2 . [ABSTRACT FROM AUTHOR]

The durability of a water electrolysis cell against wind power voltage fluctuations was evaluated by developing an accelerated potential fluctuation test protocol based on wind power voltage fluctuations over an actual 24 h period of operation. Accelerating potential protocol tests corresponding to 160 d of actual operation were performed. The current loss during the test is mostly recovered if rest time was included during the operation. Reversible loss is most likely related to gas stagnation, suppression of oxygen reduction reaction activity owing to the formation less active Ir(V) species, and suppression of gas evolution owing to the increased hydrophilic surface. This irreversible loss can be attributed to the dissolution of IrO 2 and the loss of Nafion® ionomer. Even though the potential fluctuation itself was not a problem if the lowest potential was 1.0 V, the upper limit of the potential was rather problematic. When the upper potential is increased to 2.3 V, the oxidation of IrO 2 is enhanced, leading to the partial dissolution of IrO 2 , followed by the formation of a dense structure, a decrease in the oxygen evolution reaction activity, and an increase in the surface hydrophilicity. • The durability of a water electrolyzer against wind power fluctuation was evaluated. • A protocol evaluating the long-term durability was investigated. • Most of the current density loss was recovered after the rest time. • Suppression of gas evolution due to the increased hydrophilic surface was suggested. • A small portion of the irreversible loss was related to the loss of Ir and ionomer. [ABSTRACT FROM AUTHOR]

This study elucidates the effects of current density on membrane degradation under combined mechanical and chemical stress tests. Relative humidity (RH) cycling tests using hydrogen gas and air are conducted on a polymer electrolyte membrane fuel cell based membrane NRE211 at the open circuit voltage (OCV), 0.05 and 0.3 Acm−2 conditions. The different current density conditions result in different in-plane membrane stresses and H 2 O 2 formation rates during the test. After every 200 RH cycles, membrane integrity is assessed via the hydrogen crossover rate and OCV. Furthermore, catalytic combustion is analyzed during OCV measurement using a thermal imaging method employing high-transmittance glass at the cathode side. The membrane failed after 1600, 1800, and 2200 RH cycles under the OCV condition, 0.05 of 0.3 Acm−2, respectively. The vigorous membrane degradation under OCV conditions can be attributed to higher mechanical stress and H 2 O 2 formation rate. Hotspots created owing to the combustion between the crossover hydrogen and air were successfully captured, with a maximum temperature rise ranging from 15 to 16 °C compared with a given cell temperature of 80 °C. Moreover, a post-mortem analysis (SEM imaging) revealed the presence of pinholes, through-membrane cracks, and membrane thinning at the hotspot locations. • RH cycling tests are conducted under OCV, 0.05 Acm−2 and 0.3 Acm−2 conditions. • Different conditions cause different mechanical stress and H 2 O 2 formation rates. • Visualization of in-situ catalytic combustion using a high transmittance glass (ZnS). • RH cycling test at lower current density leads to higher stresses on the membrane. • Hotspots emerge at the locations where fatal cracks and pinholes exist. [ABSTRACT FROM AUTHOR]

Electrochemical hydrogen pump (EHP) with an internal humidifier is reported. The internal humidifier contains liquid water stored in a vessel built in the cathode end plate, and the water directly wets a polymer electrolyte membrane (PEM). By using the internal humidifier, conductivity of the PEM is higher than that using a conventional vapor humidifier. A modeling of water transport is built to discuss the water balance in the membrane. Electrochemical impedance spectroscopy (EIS) is employed for analyzing the resistance components included in the EHP. In particular, the effects of compression ratio on resistance components are investigated. [ABSTRACT FROM AUTHOR]

Solid Polymer Water Electrolysis (SPWE) is a method to efficiently produce high-purity hydrogen gas using a polymer electrolyte membrane-based system. SPWE systems that utilize natural water circulation (resulting from a difference in buoyancy) are a promising technology, which need no additional circulation pump for water supply to the electrolysis cells, and generate no pressure difference between the hydrogen generation and oxygen generation chambers. However, despite not needing an accurate pressure control, gas bubbles formed and trapped within the cell stacks can inhibit heat convection, leading to hot-spot formation and consequent destructive degradation. Improving the reliability is therefore one of the most important technological issues in natural circulation SPWEs. In this study, hot-spot formation is studied both by numerical heat and flow analysis, and by experimental in-situ visualization. This leads to insights into the degradation mechanisms of SPWE stacks, and their possible solutions. An improved design for an SPWE cell stack is successfully fabricated, and reliable operation up to 5000 h is demonstrated. [ABSTRACT FROM AUTHOR]

Abstract: The water produced in a proton exchange fuel cell can exist in supercooled states during cold start operation. Visualization studies of unit cells under cold start conditions have confirmed that supercooled states exist in the cells and that they are eventually released. However, these supercooled states have not been quantitatively characterized, and it is difficult to predict them. Moreover, it has not been clarified what determines the supercooling degree and the release of supercooled states in each part of the cells, such as the gas flow channels, gas diffusion layers, and catalyst layer. In this work, a theoretical model was developed to predict the release of supercooled states on the basis of heterogeneous nucleation theory and by considering the surface wettability of the porous media in the cells. This model was evaluated through comparison to our in situ visualization study of a cell during a cold start. The developed model reproduced the supercooled state in the cell well, specifically its release time, and quantitatively clarified the impact of the pore diameter and wettability on the supercooled states. [Copyright &y& Elsevier]