Literature DB >> 35350337

Enhanced Performance of Protonic Solid Oxide Steam Electrolysis Cell of Zr-Rich Side BaZr0.6Ce0.2Y0.2O3-δ Electrolyte with an Anode Functional Layer.

Hajime Toriumi1, SeongWoo Jeong1, Sho Kitano2, Hiroki Habazaki2, Yoshitaka Aoki2.   

Abstract

Proton-conducting solid oxide electrolysis cells (H-SOEC) containing a 15-μm-thick BaZr0.6Ce0.2Y0.2O3-δ (BZCY622) electrolyte thin film, porous cathode cermet support, and La0.6Sr0.4Co0.2Fe0.8O3-δ anodes were fabricated using a reactive cofiring process at approximately 1400 °C. Steam electrolysis was conducted by supplying wet air to the anode at a water partial pressure of 20 kPa. The performance was evaluated using electrochemical measurements and gas chromatography. At 600 °C, the cells generated an electrolysis current of 0.47 A cm-2 at a 1.3 V bias while the Faradaic efficiency reached 56% using 400 mA cm-2. The electrolysis performance was efficiently improved by introducing a 40-nm-thick La0.5Sr0.5CoO3-δ (LSC) nanolayer as an anode functional layer (AFL). The cells with LSC AFL produced an electrolysis current of 0.87 A cm-2 at a 1.3 V bias at 600 °C, and the Faradaic efficiency reached 65% under 400 mA cm-2. Impedance analysis showed that the introduction of the AFL decreased the ohmic resistances and improved interfacial proton transfer across the anode/electrolyte interface and polarization resistances related to the anode reaction. These results demonstrate opportunities for future research on AFL to improve the performance of H-SOECs with Zr-rich BaZr x Ce1-x-y Y y O3-δ electrolytes.
© 2022 The Authors. Published by American Chemical Society.

Entities:  

Year:  2022        PMID: 35350337      PMCID: PMC8945173          DOI: 10.1021/acsomega.2c00569

Source DB:  PubMed          Journal:  ACS Omega        ISSN: 2470-1343


Introduction

The development of sustainable energy technologies is driven by the growing interest to address current environmental and energy issues. Solar and wind power are considered primary sources of renewable energy; however, their potency as a reliable energy source is directly affected by several factors, such as climate and seasonality. Energy supply fluctuations can result in grid instability, particularly when the energy demand is higher than the current output. This issue is addressed by using grid-scale energy storage, which enables energy to be stored during peak production and be released during peak demand, thus transforming intermittent energy generators into grid-feasible resources. As such, large-scale green hydrogen production using water electrolysis has gained increasing popularity. The process generates hydrogen fuels that are subsequently stored in high-pressure tanks or converted to liquid fuels such as ammonia and methane. The three standard methods used in the process are (1) alkaline water electrolysis, (2) polymer electrolyte membrane water electrolysis, and (3) solid oxide steam electrolysis. The first two methods are carried out near room temperature using liquid water while the latter is carried out with steam and at high temperatures. It is found that the use of steam electrolysis, in lieu of water electrolysis, lowers the vaporization heat of the water decomposition reaction. The theoretical voltage decreases from roughly 1.6 to 1.3 V as the temperature increases from 25 to 500 °C. Therefore, steam electrolysis by solid oxide electrolysis cells (SOECs) can be considered an alternative and highly efficient process to produce green hydrogen fuels using renewable energy. Studies show that O2– ion conductivity in yttria-stabilized zirconia (YSZ) is directly proportional to the process temperature.[1,2] Thus, conventional SOECs using O2– ion-conducting YSZ (O-SOECs) are operated at temperatures above 800 °C to meet the high activation energy requirement (∼1 eV) to conduct O2– ions in YSZ. However, the need for a high operating temperature leads to higher process costs and thermal degradation of materials.[3−6] Recent studies are now focused on protonic or H+ ion-conducting SOEC (H-SOECs) because of its ability to operate at lower temperatures compared to O-SOECs. The activation energy required for proton conductivity (∼0.5 eV) is also observed to be lower in SOECs using H+-conducting BaZrCe1–YO3−δ (BZCY) perovskites relative to O-SOECs.[7−15] Similarly, the use of Zr-rich-phase BZCY perovskites in steam electrolysis applications is advantageous since the material becomes more tolerant to CO2 and H2O as the Zr content increases.[16,17] However, cells using such a Zr-rich phase exhibits lower performance compared to cells using a Ce-rich phase. This is due to the highly resistive nature of the grain boundary in the Y-doped BaZrO3 moiety,[18] which leads to substantial ohmic loss and interfacial polarization.[19] Considering the process limitations, the development of highly efficient H-SOECs using a Zr-rich BZCY electrolyte with significantly lower interfacial resistance is significantly challenging. Tang et al. recently reported that the performance of H-SOECs using the BaZrCe0.8–Y0.1Yb0.1O3−δ electrolyte in terms of ohmic loss and anode polarization resistance improves significantly by introducing several 10-nm-thick La0.5Sr0.5CoO3−δ (LSC) anode functional layers (AFL) at the anode–electrolyte interface.[20] Although the proton conductivity of LSC is low in ambient atmospheres,[21] AFL was found to promote interfacial proton transfer near the gas–anode–LSC triple phase boundary (TPB). Thus, the resistances related to proton incorporation from the anode to the electrolyte are reduced significantly. In this paper, we demonstrated that the LSC AFL can decrease both the interfacial proton transfer resistances and anode reaction resistances of H-SOEC with BaZr0.6Ce0.2Y0.2O3−δ (BZCY662). Consequently, the thin film cells with AFL gain about twice the electrolysis current as compared with the cells without AFL without lowering the Faradaic efficiency.

Experimental Methods

Materials Preparation

Jeong et al. reported the fabrication of cermet anode-supported thin film fuel cells with BZCY622 electrolytes using a single-step cofiring process.[22] In this experiment, BZCY622 thin film cells were also fabricated using a similar process. The raw mixture of BZCY622 was prepared by mixing stoichiometric amounts of BaCO3 (High Purity Chemicals, 99.95%), CeO2 (High Purity Chemicals, 99.99%), ZrO2 (High Purity Chemicals, 98%), Y2O3 (High Purity Chemicals, 99.99%), and Zn(NO3)2·6H2O (Wako Chemicals; 99.9%) additives as a sintering aid. The amount of additive was adjusted to a Zn/Ba molar ratio of 0.03. The starting materials were dispersed in ethanol, ball-milled for 24 h, and dried at 80 °C. A BZCY622 raw mixture and NiO were mixed in ethanol at a weight ratio of 40:60. After drying, the resultant mixed powders were uniaxially pressed under 20 MPa and isostatically pressed under a hydrostatic pressure of 100 MPa to obtain green pellets (12 mm ϕ, 1.6–1.8 mm d) for Ni-BZCY622 cermet cathode supports. The precursor layers of the electrolyte films were spin-coated on the green pellets using a MISAKA 1H-D7 spin coater. The slurry was prepared by dispersing the BZCY622 raw mixture in a solution containing a dispersant (20 wt % polyethylenimine (Mw = 28 000) dissolved in α-terpineol) and a binder (5 wt % surfactant dissolved in α-terpineol) at a weight ratio of 10:3:1. The mixture was spin-coated once or twice on the surfaces of the green pellets at 3000 rpm for 30 s. After spin coating, the pellets were dried at room temperature and cofired at 1400 °C for 12 h to form half-cells. The backside of the sintered pellet was polished with SiC paper. The LSC AFL was deposited on the electrolyte surface by radio frequency (RF) magnetron sputtering with an LSC target. RF sputtering was performed at a sputtering power of 50 W under a flow of 4% O2/Ar gas at 50 sccm while the substrate temperature was maintained at 500 °C. Last, the La0.6Sr0.4Co0.2Fe0.8O3−δ (LSCF) anode was screen-printed using an LSCF commercial ink (Fuel Cell Materials). The phase purity was checked by X-ray diffraction (XRD) analysis in the 2θ range between 20° and 80° at a scan rate of 5° min–1 using a Rigaku Ultima IV (Cu Kα radiation). The microstructures of the fabricated cells were examined using a field emission scanning electron microscope (FESEM; SIGMA500, ZEISS) operated at 3 kV.

Electrochemical Measurements

Prior to conducting electrochemical measurements, the cathode side was exposed to a humidified 50% H2/Ar gas mix at 700 °C for 6 h to convert NiO to metallic Ni and produce a porous Ni-BZCY622 cermet cathode. Steam electrolysis performance of cells was evaluated at 700 and 600 °C with a wet 10% H2/Ar gas mix to the cathode and a 20% H2O/air mix to the anode. The 20% H2O/air mixture was prepared by passing air through a water bath maintained at approximately 60 °C and subsequently supplied to the anode chamber through a stainless-steel tube heated at 150 °C using a ribbon heater to prevent condensation. The I–V characteristics were measured using an electrochemical station (Biologic SP-300). The electrochemical impedance spectra of the SOECs were determined using a frequency response analyzer (Biologic SP-300) with a frequency range of 106 to 0.1 Hz and an AC amplitude of 30 mV. The hydrogen evolution rates (v) in the cathode were quantified by analyzing the cathode exhaust gas using gas chromatography (490 Micro GC, Agilent Technologies). The Faraday efficiency, η, was calculated using the observed and theoretical hydrogen evolution rates (vmeas and vtheo, respectively) using the following equation:where I is the applied current, z is the electron transport number of steam electrolysis, and F is Faraday’s constant (96 485 C/mol).

Results and Discussion

Morphology of Prepared Materials

Figure shows the powder XRD pattern of pulverized thin film cell before screen printing the LSCF anodes (half-cell). Results show that all of the peaks generated during the XRD analysis were from NiO and BZCY622 phases, while the BaY2NiO5 secondary phase[23] was not observed. This confirms that the reaction between BZCY622 and NiO to produce BaY2NiO5 does not occur.
Figure 1

XRD pattern of pulverized half cells.

XRD pattern of pulverized half cells. Figure a,b show the cross-sectional SEM images of H-SOEC with BZCY622 films prepared by spin-coating once and twice, respectively, after H2 reduction of the cermet cathode. The cermet cathode is extremely porous with interconnected macropore networks that act as gas diffusion paths. The films prepared by spin-coating once and twice have a thickness of 15 and 30 μm, respectively, both of which are highly dense and formed uniformly over a porous cermet cathode. The LSCF anode is a porous layer composed of 50 nm diameter particles with a thickness of ∼40 μm. Figure c shows the SEM image of the BZCY622 thin film bare surface on the porous Ni-BZCY662 cermet cathode after H2 reduction at 700 °C. The BZCY622 electrolyte layer maintains dense matrices without any crack formation and pinholes after reduction. This implies that the electrolyte film maintains adequate gas tightness. The BZCY622 grains grew up to 5 μm in diameter, which is larger than the BZCY622 grain size prepared by the sol–gel method.[24]Figure d and e show cross-sectional and surface SEM images of the LSC AFL on the BZCY622 thin film. The 40-nm-thick LSC layer, comprised of the nanoparticles with about 10 nm diameters, fully covers the electrolyte surface.
Figure 2

(a,b) Cross-sectional SEM image of cermet support H-SOEC with BZCY622 electrolyte films prepared by spin-coating (a) once and (b) twice. (c) Surface SEM image of BZCY622 electrolyte film prepared by once spin-coating. (d) Cross-sectional and (e) surface SEM images of LSC-AFL deposited on the BZCY622 electrolyte.

(a,b) Cross-sectional SEM image of cermet support H-SOEC with BZCY622 electrolyte films prepared by spin-coating (a) once and (b) twice. (c) Surface SEM image of BZCY622 electrolyte film prepared by once spin-coating. (d) Cross-sectional and (e) surface SEM images of LSC-AFL deposited on the BZCY622 electrolyte. Figure shows the current–voltage (I–V) curves for the cell without the AFL, i.e., {20%-H2O/air, LSCF | BZCY622 (15 μm) | Ni-BZCY622, wet 10%H2/Ar}, at 700 and 600 °C. Open circuit voltage (OCV) at 700 and 600 °C are measured to be 0.84 and 0.91 V, respectively, which is lower than the ideal value of 0.93 and 0.96 V. The relatively low OCV value for the current cells must be attributed to the partial hole conductivity of BZCY622. On the basis of the theoretical OCV of 0.93 V as calculated by the Nernst equation, the proton transport number of BZCY622 films can be roughly estimated to 0.89, which is in agreement with the proton transfer number of BaZr0.7Ce0.1Y0.2O3 (<0.9)[25] at 700 °C under pO2 ≥ 0.01. This clearly indicates that leakage by hole conduction occurs on the BZCY662 thin film at a relatively high temperature.
Figure 3

Performance (I–V curves) of the H-SOEC using 15-μm-thick BZCY622 film with and without LSC-AFL. Solid and dotted lines indicate the cell with and without LSC, respectively, and red and blue denote 700 and 600 °C, respectively.

Performance (I–V curves) of the H-SOEC using 15-μm-thick BZCY622 film with and without LSC-AFL. Solid and dotted lines indicate the cell with and without LSC, respectively, and red and blue denote 700 and 600 °C, respectively. The cells without AFL exhibited electrolysis currents of 0.92 and 0.47 A cm–2 at 700 and 600 °C, respectively, at a 1.3 V bias. The OCV of the cell with AFL, i.e., {20%-H2O/air, LSCF | LSC/BZCY622 (15 μm) | Ni-BZCY622, wet 10%H2/Ar}, is comparable to the results observed in cells without the AFL–OCV values are 0.82 and 0.92 V at 700 and 600 °C, respectively. However, at 1.3 V, the electrolysis currents at 700 and 600 °C are 1.43 and 0.87 A cm–2, respectively, which are nearly twice higher than the corresponding values of the cells without AFL. The electrolysis current of the cell with the LSC-AFL at 600 °C is similar to or higher than that of the H-SOECs using a Ce-rich side BaCe1–ZrMO3 (M = Yb, Y, Co; x ∼ 0.1) thin film electrolyte at 1.3 V under similar conditions.[26−29] Figure a,b show the AC impedance spectra for the cells of 15-μm-thick electrolyte with and without the AFL at 600 °C subjected to different DC conditions. In the H-SOECs spectra, the high-frequency x intercept is normally attributed to the electrolyte’s ohmic resistances (RO), while the succeeding arcs denote the interfacial polarization mainly because of the anode reactions.[30,31] The RO of the cells without and with the AFL at 600 °C under OCV conditions are 0.81 Ω cm2 and 0.30 Ω cm2, respectively. This feature indicates that the RO related to proton conduction efficiently decreases with the use of the AFL. Moreover, the RO of H-SOEC was found to include the protonic resistances near the anode/electrolyte interfaces together with the bulk resistance of the electrolyte film. To separate interfacial resistance from bulk resistance of the electrolyte, the RO is evaluated for the cells with different thicknesses of electrolyte films (Figure c). The RO at 600 °C of the cell without AFL increases from 0.81 to 0.92 Ω cm2 when the thickness of the electrolyte increases from 15 to 30 μm. Assuming the linear dependence of RO on electrolyte thickness reveals the existence of an RO of about 0.7 Ω cm2 at zero thickness. Recent studies reported that proton-conducting solid oxide fuel cells (H-SOFCs) possess large impedances (∼105 Hz) due to slow proton transfer at the electrolyte–anode–gas TPB.[32−36] On the basis of these, the BZCY622 cells involve relatively large resistances related to the interfacial proton transfer from the electrolyte to the anode. The value estimated for the zero thickness electrolyte cell without AFL, i.e., 0.7 Ω cm2, is comparable to 80% of RO for the 15-μm-thick electrolyte cell without AFL (0.81 Ω cm2), which confirms that the interfacial resistances are a dominant component in the RO of the BZCY622 base cells. The RO for the 15-μm-thick electrolyte cells with AFL is only 0.30 Ω cm2, which is much smaller than the value of the corresponding cell without AFL. These results unambiguously demonstrate that interfacial resistance is significantly reduced with the use of LSC–CFL.
Figure 4

AC impedance spectra for H-SOEC of 15-μm-thick electrolyte (a) without and (b) with LSC-AFL, measured at 600 °C under various DC conditions. (c) AC impedance spectra of H-SOEC of 15-μm-thick electrolyte without (red, ○) and with LSC-AFL (blue, □) and of 30-μm-thick electrolyte without LSC-AFL (green, △). In a–c, digits show the frequencies at each blackened point in the spectra.

AC impedance spectra for H-SOEC of 15-μm-thick electrolyte (a) without and (b) with LSC-AFL, measured at 600 °C under various DC conditions. (c) AC impedance spectra of H-SOEC of 15-μm-thick electrolyte without (red, ○) and with LSC-AFL (blue, □) and of 30-μm-thick electrolyte without LSC-AFL (green, △). In a–c, digits show the frequencies at each blackened point in the spectra. The polarization resistances (Rp) are approximated using the total diameter of the impedance arcs. The calculated values for the cells without and with the AFL under OCV are 0.29 Ω cm2 and 0.95 Ω cm2, respectively. The Rp reduction caused by the AFL is more evident with the DC outputs. The Rp values with the AFL are 0.26 Ω cm2 and 0.23 Ω cm2 under a DC current of 50 and 100 mA cm–2, respectively. These values are observed to be smaller by a factor of 0.33 and 0.53 than the values without the AFL. The results indicate that the LSC AFL can facilitate both interfacial proton transfer and other cathode reaction steps, such as charge transfer to adsorbed oxygen and/or oxygen desorption at the TPB. The hydrogen production rate was calibrated using gas chromatography under galvanostatic electrolysis at 200 mA cm–2 and 400 mA cm–2. Figure illustrates the cell voltage transients, hydrogen evolution rate (vmeas), and Faradaic efficiency (η) at 600 and 700 °C. The first electrolysis was conducted at 200 mA cm–2 for 30 min and at 400 mA cm–2 for 30 min subsequently. The voltages for cells without AFL are observed to be stable at 0.91 to 0.98 V during the electrolysis at 700 °C using 200 mA cm–2 and 400 mA cm–2, respectively. Higher cell voltage is observed at 600 °C compared to the cell voltage at 700 °C. This is primarily due to the increase in total cell resistance (RO + Rp) with decreasing temperature. The cell voltage is 1.15 and 1.26 V at 200 mA cm–2 and 400 mA cm–2, respectively. The nominal overpotentials, defined by the gap between the cell bias and the OCV, are equal to 0.24 and 0.36 for electrolysis at 600 °C in 200 mA cm–2 and 400 mA cm–2, respectively.
Figure 5

Faradaic efficiency (η; blue, □), hydrogen evolution rate (vmeas; green, ○) and cell voltage (red line) of (a,c) LSCF | BZCY622 (15 μm) | Ni-BZCY622 and (b,d) LSCF | LSC | BZCY622 (15 μm) | Ni-BZCY622 electrolysis cells at 600 and 700 °C under galvanostatic conditions at 200 and 400 mA cm–2.

Faradaic efficiency (η; blue, □), hydrogen evolution rate (vmeas; green, ○) and cell voltage (red line) of (a,c) LSCF | BZCY622 (15 μm) | Ni-BZCY622 and (b,d) LSCF | LSC | BZCY622 (15 μm) | Ni-BZCY622 electrolysis cells at 600 and 700 °C under galvanostatic conditions at 200 and 400 mA cm–2. vmeas achieves equilibrium within 20 min upon applying a constant current during electrolysis at 700 and 600 °C. vmeas without the AFL is measured to be 3.6 × 10–5 mol cm–2 min–1 and 6.9 × 10–5 mol cm–2 min–1, and η values of 58% and 56% were axquired for electrolysis conducted at 600 °C using 200 mA cm–2 and 400 mA cm–2, respectively. Similar results for vmeas and η at 700 °C are observed under equilibrium conditions. These η results are consistent with the values reported in previous studies for H-SOECs with an acceptor-doped Ba(Zr,Ce)O3 electrolyte.[11−14,20,32,37−40] Vøllestad et al. determined the theoretical Faradaic efficiency of H-SOEC with BaZr0.1Ce0.7Y0.1Yb0.1O3−δ under different hydrogen partial pressures (pH) and water partial pressures (pH).[40] The approach employed numerical and experimental techniques and concluded that the efficiency is less than 70% when the anode’s pH2O and cathode’s pH2 are below 100 kPa. Hole carrier concentration in the electrolyte is also found above zero under similar conditions. These results confirm that the cells in this study conduct steam electrolysis and exhibited moderate Faradaic efficiency. The cells with the LSC AFL exhibit stable voltages of 0.91 and 0.98 V at 700 °C for the galvanostatic electrolysis in 200 mA cm–2 and 400 mA cm–2, respectively. In contrast, 1.04 and 1.14 V are observed at 600 °C under similar conditions. The cell overpotentials with the AFL at 600 °C are calculated to 0.12 and 0.21 for electrolysis under 200 mA cm–2 and 400 mA cm–2, respectively. This is approximately 49 and 59%, respectively, of the corresponding values for cells without the AFL. vmeas with the AFL reaches 4.1 × 10–5 mol cm–2 min–1 and 8.1 × 10–5 mol cm–2 min–1 for electrolysis at 600 °C using 200 mA cm–2 and 400 mA cm–2, respectively. The values are comparable with the results for cells without the AFL. In addition, η values are 67 and 65% for electrolysis at 600 °C using 200 mA cm–2 and 400 mA cm–2, respectively. These results demonstrate that LSC AFL lowers the overpotentials of H-SOEC efficiently by promoting interfacial proton transfer and anode reactions without increasing the hole leakage current. It was demonstrated in this study that the introduction of the AFL is an efficient way to improve the performance of H-SOEC-based Zr-rich side BZCY electrolytes. The results of this research demonstrate opportunities for future research on optimal AFL materials with excellent durability and high efficiency for H-SOECs.

Conclusions

The introduction of 40-nm-thick LSC AFL is a promising technique to decrease the ohmic and polarization resistances of H-SOECs with Zr-rich side BZCY622 electrolytes. The cell with the AFL can generate an electrolysis current of 0.87 A cm–2 at 600 °C at 1.3 V. This is nearly twice higher than that of the cell without the AFL. The cells without the AFL at 600 °C in OCV have ohmic resistance and polarization resistances of 0.81 Ω cm2 and 0.95 Ω cm2, respectively. Conversely, cells with the LSC AFL exhibited lower ohmic and polarization resistances of 0.30 Ω cm2 and 0.29 Ω cm2, respectively. The addition of AFL efficiently promoted interfacial proton transfer at the electrolyte/anode interface and corresponding anode reactions. The current results offer a design concept for an efficient H-SOEC.
  6 in total

1.  New, Efficient, and Reliable Air Electrode Material for Proton-Conducting Reversible Solid Oxide Cells.

Authors:  Daoming Huan; Nai Shi; Lu Zhang; Wenzhou Tan; Yun Xie; Wanhua Wang; Changrong Xia; Ranran Peng; Yalin Lu
Journal:  ACS Appl Mater Interfaces       Date:  2018-01-05       Impact factor: 9.229

2.  Mixed proton and electron conducting double perovskite anodes for stable and efficient tubular proton ceramic electrolysers.

Authors:  Einar Vøllestad; Ragnar Strandbakke; Mateusz Tarach; David Catalán-Martínez; Marie-Laure Fontaine; Dustin Beeaff; Daniel R Clark; Jose M Serra; Truls Norby
Journal:  Nat Mater       Date:  2019-06-03       Impact factor: 43.841

3.  The relevance of interfaces for oxide ion transport in yttria stabilized zirconia (YSZ) thin films.

Authors:  Matthias Gerstl; Gernot Friedbacher; Frank Kubel; Herbert Hutter; Jürgen Fleig
Journal:  Phys Chem Chem Phys       Date:  2013-01-28       Impact factor: 3.676

4.  Ionic conductivity and activation energy for oxygen ion transport in superlattices--the semicoherent multilayer system YSZ (ZrO2 + 9.5 mol% Y2O3)/Y2O3.

Authors:  C Korte; A Peters; J Janek; D Hesse; N Zakharov
Journal:  Phys Chem Chem Phys       Date:  2008-06-20       Impact factor: 3.676

5.  3D Self-Architectured Steam Electrode Enabled Efficient and Durable Hydrogen Production in a Proton-Conducting Solid Oxide Electrolysis Cell at Temperatures Lower Than 600 °C.

Authors:  Wei Wu; Hanping Ding; Yunya Zhang; Yong Ding; Prashant Katiyar; Prasun K Majumdar; Ting He; Dong Ding
Journal:  Adv Sci (Weinh)       Date:  2018-08-31       Impact factor: 16.806

6.  Solid-Solid Interfaces in Protonic Ceramic Devices: A Critical Review.

Authors:  Alessandro Chiara; Francesco Giannici; Candida Pipitone; Alessandro Longo; Chiara Aliotta; Marianna Gambino; Antonino Martorana
Journal:  ACS Appl Mater Interfaces       Date:  2020-12-02       Impact factor: 9.229

  6 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.