Separation and Purification Technology 320 (2023) 124202

Contents lists available at ScienceDirect

Separation and Purification Technology

journal homepage: www.elsevier.com/locate/seppur

Electrospinning-based nanofiber architectures for outstanding CO2 capture

Sangmo Kang a, Jungho Hwang a, *, Jeong Hoon Byeon b, *
a
School of Mechanical Engineering, Yonsei University, Seoul 03722, Republic of Korea
b
School of Mechanical Engineering, Yeungnam University, Gyeongsan 38541, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: The worsening climate change has made the development of alternative and sustainable technologies to achieve
PEI/CeO2/ACNFs carbon neutrality a pressing issue. Numerous strategies for carbon dioxide (CO2) capture/conversion and rele­
CO2 adsorbent vant material preparation have been dealt with on a case-by-case basis. Therefore, a reconfigurable process to
Electrospinning-based architecting
secure optimal procedures and materials is urgently required to reduce CO2 emissions in a convenient, rapidly
Simple process manipulation
Electrical polarity
implementable manner. On this account, this study attempted to construct polyethylenimine (PEI)/cerium oxide
(CeO2)/activated carbon nanofibers (ACNFs) as a reliable CO2 adsorbent from an electrospinning (ES)-based
architecting as a modular process to accelerate their field applications because adsorption has outstanding ad­
vantages for CO2 capture, such as process simplicity, wide operational range, low cost, and stable performance.
By selecting ES with commercially available materials (PEI (providing CO2 chemisorption), cerium nitrate
hexahydrate (CO2 selectivity), and polyacrylonitrile (base of porous nanofibers)), a composite nanofibrous ar­
chitecture (PEI/CeO2/ACNFs) with outstanding CO2 capture can be achieved by simple process manipulation
(manipulating the electrical polarity applied to the nozzle and concentration of the cerium nitrate for the ES).
This study may provide a reconfigurable manufacture to ensure a reliable adsorbent for CO2 capture and base
materials for CO2 conversion from electro- and photocatalysis.

1. Introduction required to ensure enhanced adsorption capacity considering adsorbate

species, operating temperature and pressure, and space velocity [8,9]. A
Climate change is recognized as a critical global issue by most people trade-off problem between the higher performance and process
around the world [1]. Recently, it has been suggested that the carbon complexity may impede the practical applications of developed mate­
dioxide (CO2) concentration is related directly to the ecosystem and rials for CO2 reduction. A convenient, modular, high-yield platform is
biodiversity of the air and the ocean as an air filter of the world [2]. required to screen architectures systematically for reliable CO2
Advanced adsorption materials are essentially required as an active adsorption.
alternative to reducing CO2 [3]. On the other hand, only limited data or Because of the “mix and go” convenient configuration, electro­
complex approaches for highly capacitive and selective adsorption have spinning (ES) has been used extensively to produce a diverse spectrum of
been provided a case-by-case basis, and there is a lack of sufficient in­ nanofibrous nonwovens for separation and purification purposes in the
formation, even for workable architectures of adsorbents. In that sense, past three decades [10]. Nanofibrous structures can be installed in most
a high-throughput screening platform is required as a priority to accel­ environmental pollutant treatment plants because of their proper ar­
erate the utilization of reliable CO2 adsorbents for practical applications. chitecture, flexibility, and textural property with tolerable pressure
As a primary viable option to capture CO2, adsorption using porous drop. These can be extendable for energy production and storage ap­
materials, such as metal–organic frameworks (MOFs) [4], zeolites [5], plications [11,12]. An electric potential with polarity modulation on the
porous polymer networks [6], and porous carbon [7], has been spinning nozzle can control the arrangement of ionic compounds
considered because of convenience, material abundance, simple facility, (directly added to the polymeric precursor for functionalizing/modi­
and low cost and energy consumption. On the other hand, surface fying purposes) in the discharging of a mixture precursor [13,14] that
modifications based on hydrothermal or vacuum (sputtering for nano­ conveniently generates various composite nanofibers with no post-
roughness or vapor deposition for functional overlayers) reactions are treatments under harsh hydrothermal or vacuum conditions [15].

* Corresponding authors.
E-mail addresses: [email protected] (J. Hwang), [email protected] (J.H. Byeon).

https://doi.org/10.1016/j.seppur.2023.124202
Received 12 March 2023; Received in revised form 14 May 2023; Accepted 24 May 2023
Available online 27 May 2023
1383-5866/© 2023 Elsevier B.V. All rights reserved.
S. Kang et al. Separation and Purification Technology 320 (2023) 124202

Subsequent thermal carbonization and activation formed textures in the 2.2. CO2 adsorption
resulting nanofibers to secure micro- and mesoporosity, whereas uti­
lizing this technique is still immature for CO2 adsorption. A static volumetric analyzer (TriStar II 3020, Micromeritics, USA)
This paper proposes an ES-based platform to build an overlayer/ was used to obtain the capacities of CO2 adsorption for the resulting
interlayer/base fiber (securing compositional hierarchy) architecture samples. Before the test, the samples were degassed for 12 h at 150 ◦ C to
for highly capacitive, reliable CO2 adsorption, where polyethelenimine remove moistures (ensuring the residual pressure of <1 × 10− 3 mbar). A
(PEI; providing CO2 chemisorption), cerium dioxide (CeO2; CO2 selec­ steady flow of CO2 (0.015 L/min) was used to test each sample (0.1 g),
tivity), and activated carbon nanofibers (ACNFs; base for adsorption while different temperatures (0, 25, 50, and 75 ◦ C) were conducted to
capacity) were selected as the individual layer materials by successive examine the adsorption capacity. The nitrogen (N2) adsorption capacity
ES, thermal treatment, and bath immersion (Fig. 1). The polarity of the was also examined under identical conditions to identify the selectivity
electric potential and concentration of CeO2 precursor (cerium nitrate of CO2 adsorption. The CO2 desorption was conducted at 100 ◦ C for 2 h
hexahydrate (Ce(NO3)3), a Ce3+ source) were selected as vital parame­ under pure nitrogen (as desorption gas) atmosphere.
ters to ensure modular property and control the arrangement and den­
sity of Ce3+ for screening optimal interlayer/base fiber (CeO2/ACNFs) 2.3. Characterization
architecture upon a thermal treatment for CO2 adsorption. Subsequent
immersion of the CeO2/ACNFs in the PEI bath for impregnating amine The samples were examined by field-emission scanning electron
groups (constructing a PEI/CeO2/ACNFs hierarchy) was conducted to microscopy (SEM; 7800F, JEOL, Japan) and transmission electron mi­
reinforce the chemisorption of the adjacent CO2. croscopy (TEM; JEM-F200, JEOL, Japan). Elemental maps were ob­
tained using an energy-dispersive X-ray spectroscopy (EDS) included in
2. Experimental the TEM. To prepare the specimens, ethanol was used to disperse each
sample before placing it on a carbon-coated copper grid. The crystal­
2.1. ES-based construction linity of the samples was analyzed by X-ray diffraction (XRD; SmartLab,
Rigaku, Japan) with a Cu Kα radiation source. X-ray photoelectron
Cerium nitrate hexahydrate (CAS#10294–41-4; Ce(NO3)3 • 6H2O) spectroscopy (XPS; K-Alpha, Thermo Scientific, USA) was used to obtain
with different doses (0.05, 0.10, 0.15, and 0.20 g) was dissolved in 10 the Ar+ depth profiles with an etching rate of 0.1 nm/s for quantifying
mL of N,N-dimethylformamide (CAS#68–12-2; DMF) and stirred for 1 h the surface elements of the samples. The surface chemistry and pyrolysis
at 60 ◦ C. The solution was then incorporated with 1 g of poly­ characteristics were examined using Fourier transform infrared (FTIR)
acrylonitrile (PAN; CAS#25014-41-9; Mw = 150,000 g/mol) for one day spectroscopy (VERTEX 70, Bruker, USA) and thermogravimetric anal­
with magnetic stirring at room temperature to form Ce(NO3)3/PAN ysis (TGA; LABSYS evo, Setaram Instrumentation, France), respectively.
composite precursors. The resulting precursor solutions were fed into a The textural properties of the samples were analyzed using a poros­
plastic syringe for ES at a flow rate of 5 μL min− 1, which was regulated imeter (Autosorb-iQ 2ST/MP, Quantachrome, Austria).
using a syringe pump (Fig. S1). The metal nozzle (inner diameter = 0.67
mm) was connected to an electric potential of − 12 kV, where the dis­ 3. Results and discussion
tance between the tip of the nozzle and the revolving drum collector
(200 rpm) was 15 cm. The diameter and length of the drum collector Different doses of the inorganic salt (Ce(NO3)3 • 6H2O) were mixed
were 10 and 20 cm, respectively. As shown in Fig. S2, the as-spun NFs individually with PAN in DMF to produce precursor solutions for con­
were put in an alumina tube and treated thermally through the following structing CeO2/ACNFs-1, CeO2/ACNFs-2, CeO2/ACNFs-3, and CeO2/
steps: (1) stabilization (250 ◦ C), (2) carbonization (250–900 ◦ C), (3) ACNFs-4. An external electric potential with different polarities (+, -)
activation (900 ◦ C), and (4) oxidation (250 ◦ C). The thermally treated was applied to derive the electrohydrodynamic movement of the ionic
NFs were denoted as CeO2/ACNFs-(1–4). For comparison, analogous compounds in the precursor solution during its discharging through the
NFs were synthesized without cerium nitrate (ACNFs) or under the ES nozzle. For the negative potential, Ce3+ and NO–3 ions migrated to­
opposite electric polarity (+12 kV; +CeO2/ACNFs). ward the outer and inner regions of the nozzle, respectively, while the
To prepare PEI-impregnated samples, 2 g of vacuum-dried (at 150 ◦ C ions moved in opposite directions with the positive potential (Fig. 2A
for 12 h) ACNFs was immersed in a PEI (Mw = 800 Da) solution (10 g PEI and 2B). The figures also present representative SEM images of the
dissolved in 200 mL of ethanol for 1 h under magnetic stirring) and electrospun Ce(NO3)3/PAN fibers under negative and positive electric
stirred magnetically at room temperature for 3 h to ensure the PEI potentials. Both fibers showed smooth surfaces with a diameter range
impregnation on the sample surfaces (approximately 0.85 g of PEI/1.00 from 250 to 300 nm.
g of ACNFs). The treated sample was placed in a vacuum oven at 100 ◦ C The concentrations of Ce3+ (Ci) as a function of the distance from the
for 5 h to secure dried PEI/ACNFs. An analogous procedure was carried center of the nozzle (x) were determined analytically using the Boltz­
out for CeO2/ACNFs and +CeO2/ACNFs to provide PEI/CeO2/ACNFs mann distribution and zeta potential profile equations (Eqs. S1 and S2)
and PEI/+CeO2/ACNFs. [16]. Ce3+ was distributed narrowly with distance from the center of the
nozzle, where the concentrations were dependent on the electrical

Fig. 1. Schematic of the ES-based process to

construct hierarchical composite nanofibers
for reliable CO2 adsorption from a combined
physi- and chemisorption. To ensure a proper
distribution of Ce3+ ions within PAN,
different polarities (+ and − ) of electric po­
tential were applied to the ES nozzle, where
the ion pairs (Ce3+ and NO–3 from the disso­
lution of Ce(NO3)3) were oppositely moved
by switching the polarity that generates
different distributions of the ion pairs within
discharging precursor solution (Ce(NO3)3/PAN in DMF). The dose of was Ce(NO3)3 regulated in the four stages to ensure optimal distribution of CeO2 particles and
pore size of ACNFs. The as-spun nanofibers (Ce(NO3)3/PAN NFs) were thermally treated to produce CeO2 particles on porous nanofibers (CeO2/ACNFs), and sub­
sequently immersed in PEI solution to form PEI-impregnated CeO2/ACNFs (PEI/CeO2/ACNFs) for their use in CO2 adsorption tests.

2
S. Kang et al. Separation and Purification Technology 320 (2023) 124202

Fig. 2. Schematics of the ion pair distribu­

tion for different electrical polarities and the
resulting morphology, microstructure, and
composition of CeO2/ACNFs. (A, B) SEM
images of as-spun fibers from the positive
and negative electric potentials for ES of Ce
(NO3)3/PAN solution. (C–F) SEM images of
CeO2/ACNFs from the four doses of Ce
(NO3)3 to examine the formation of particles
on the fiber surfaces. (G–J) The correspond­
ing TEM images. (K) Representative EDS
maps of CeO2/ACNFs-3. (L) High-
magnification TEM image of the outermost
region of CeO2/ACNFs-3 to identify the mi­
crostructures of the fiber and particle simul­
taneously. (M) XRD profiles CeO2/ACNFs
from the four doses of Ce(NO3)3 to confirm
the crystallinities of the fiber and particle
with those in the TEM image.

polarity and strength, as shown in Fig. S3A and S3B. In particular, the morphologies of ACNFs and +CeO2/ACNFs-3 were observed by SEM and
ions were densely distributed (>450 M) at a distance of 160 to 180 nm TEM, where no dark dots on the fiber surfaces were identified (Fig. S5).
for the negative potential. Longer distances (i.e., closer to the fiber There were no significant differences in appearance between ACNFs
surface) were obtained at higher electric strengths because of the (Fig. S5A) and +CeO2/ACNFs-3 (Fig. S5C). On the other hand, the TEM
stronger electrostatic attractions between the Ce3+ and the negatively image of +CeO2/ACNFs-3 exhibited fine dark dots in the core region,
charged nozzle surface. By contrast, relatively negligible amounts of which may be caused by the inside-concentrated distribution of Ce3+
ions were distributed at these distances because of electrostatic repul­ ions for the positive potential. The dots were smaller than those of CeO2/
sion, showing that modulating the electric polarity was the driver for the ACNFs-3 probably due to limited oxygen exposure to the inner region
concentrated distribution of Ce3+ in the outer or inner region of the fi­ located Ce3+ ions to form CeO2 particles that generate weak intensities
bers. The atomic depth profiles Fig. S3C) from argon ion (Ar+) etching of the characteristic CeO2 bands compared to those of CeO2/ACNFs-3
matched the surface concentrated distribution of Ce3+ (6.6%) for the despite the differences between the ACNFs and +CeO2/ACNFs-3
negative potential (cf., nearly zero % for the positive potential), sup­ (Fig. S6). This represented that the inside-dominant distribution of Ce3+
porting the adjustability of the ES for the site-selective distribution of could retain surface texture of ACNFs likely due to the confined oxida­
Ce3+. FTIR spectroscopy showed a significantly greater surface distri­ tion of Ce3+ ions to form CeO2 particles.
bution of NO–3 (absorption band at 1380–1350 cm− 1) for the positive The thermal decomposition behavior of Ce(NO3)3/PAN NFs (as-spun
potential [17], as shown in Fig. S4. prior to ACNFs, +CeO2/ACNF-3, and CeO2/ACNFs-3) was investigated
With regard to morphology, the thermally treated samples (CeO2/ by TGA in steps 1 and 2 of the thermal treatment (Fig. S2). As shown in
ACNFs-1 to CeO2/ACNFs-4) exhibited a fine roughness from small par­ Fig. S7, the weight losses in the temperature range (25–250 ◦ C) in the air
ticles (brighter dots) on the fiber surfaces (Fig. 2C–2F) unlike the as- atmosphere resulted from the cyclization of nitrile groups and dehy­
spun fibers. The roughness was caused by oxidation of the outermost drogenation of PAN (i.e., the stabilization of PAN) [20]. Until the tem­
Ce3+ ions (forming CeO2 particles) with shrinkage of the PAN fibers perature reached 250 ◦ C, the as-spun fiber samples exhibited weight
during thermal treatment. Different doses of the cerium nitrate induced losses of 1.17%, 2.52%, and 3.95%, respectively. Additional weight
different distributions of CeO2 particles on the fiber surfaces, where the losses (2.42%, 2.96%, and 4.29%) were observed when the temperature
density and size of the particles were proportional to the dose of Ce was maintained at 250 ◦ C for 30 min. The highest weight loss of Ce
(NO3)3, as observed by TEM (Fig. 2G–2 J). EDS maps of the CeO2/ (NO3)3/PAN may be caused by the catalytic oxidation of adjacent PAN
ACNFs-3 revealed the co-existence of Ce and O on a single carbon fiber molecules, which was induced by oxidation of the outermost Ce3+ to
(Fig. 2K), while the characteristic lattice spacing (0.318 nm) for CeO2, CeO2. The higher loss of + Ce(NO3)3/PAN than PAN reflected this hy­
(1 1 1) was identified by high-magnification TEM (Fig. 2L). XRD (Fig. 2M) pothesis. This is consistent with a previous report where a significant
revealed the characteristic 2θ bands at 28.6◦ , 33.1◦ , 47.6◦ , and 56.3◦ , shrinkage of PAN was observed upon oxidation of transition metal to
which were indexed to the (1 1 1), (2 0 0), (2 2 0), and (3 1 1) lattice metal oxide in PAN during thermal treatment [21]. Over 250 ◦ C with a
planes of CeO2, respectively [18]. A broad diffraction band at approxi­ nitrogen atmosphere, the samples exhibited nearly identical weight loss
mately 25◦ was assigned to amorphous carbon formed from the thermal profiles, where the losses were mostly driven by pyrolytic carbonization
treatment of PAN [19]. The intensity of the characteristic CeO2 bands of PAN. The weight differences between the samples were similar during
was also proportional to the Ce(NO3)3 dose, which is consistent with the the temperature increase, suggesting that no more catalytic oxidation of
size increases in TEM images (Fig. 2G–2J) by estimating the average the PAN occurred because of a lack of oxygen molecules. After the
crystal sizes using the Scherrer equation (Table S1). For comparison, the temperature reached 900 ◦ C, the residues of the PAN, +Ce(NO3)3/PAN,

3
S. Kang et al. Separation and Purification Technology 320 (2023) 124202

and Ce(NO3)3/PAN were 32.65%, 31.19%, and 29.53%, respectively, reflecting that the PEI was loaded on the nearly entire the pore walls
which were probably due to pyrolysis-induced shrinkage in the absence through the simple immersion (Fig. S9A). The lower N2 adsorption levels
of oxygen. at low relative pressures represented the PEI loading on most of the
Textural properties of the thermally treated CeO2/ACNF samples micropore region upon impregnation. The CeO2-included samples still
(1–4), including ACNFs and +CeO2/ACNF-3, were obtained from N2 exhibited type-IV isotherms even after PEI incorporation, suggesting
adsorption–desorption isotherms using the Brunauer–Emmett–Teller that the loading was performed uniformly along the pore walls. The
(BET) method. As shown in Fig. 3A, all samples exhibited a substantial stronger hysteresis intensity at >0.5 P/P0 for PEI/CeO2/ACNFs-3 than
increase in N2 adsorption at a low relative pressure (<0.05), repre­ PEI/+CeO2/ACNFs-3 was related to the higher amount of CeO2 particles
senting a significant portion of micropores. In particular, ACNFs on the fiber surfaces. On the other hand, the loading changed the type of
demonstrated type-I isotherms, whereas the others (CeO2-containing the hysteresis loop from H4 to H3, which followed the existence of non-
configurations) exhibited type-IV isotherms with H4 hysteresis loops in rigid aggregates of CeO2 particles, according to a previous report [25].
the range of 0.5–1.0 (P/P0) due to the mesoporosity of slit-shaped pore Although CeO2/ACNFs retained the CO2 selective microporosity after
channels owing to CeO2 formation [22]. The N2 adsorption capacity of the loading, the micropore volumes in both BJH (Fig. S9B) and HK
the samples was observed in the following order: CeO2/ACNFs-4 < (Fig. S9C) models were reduced significantly. PEI/+CeO2/ACNFs
ACNFs < CeO2/ACNFs-1 < CeO2/ACNFs-2 < CeO2/ACNFs-3. Even showed a similar pore size distribution to PEI/ACNFs than PEI/CeO2/
though increasing Ce(NO3)3 content in PAN led to improvement of the ACNFs, suggesting that the insufficiently developed CeO2 on the fiber
adsorption performance, an excessive dose of Ce(NO3)3 caused the for­ surface cannot generate favorable textures for CO2 adsorption. The PEI
mation of unfavorable architecture (forming larger CeO2 particles and loading was analyzed further by FTIR (Fig. S10), where the C=O and
their blocking the pore channels) for transport within the microstructure C–N stretching vibrations of the amide were observed at 1620 and 1575
and interaction with CeO2 and ACNF surface of CO2 molecules. The cm− 1, respectively, as an amine-grafted architecture [26]. Unlike the Ce
lowest capacity of CeO2/ACNFs-4 may be due to the blockage of slit- (NO3)3/PAN NFs, the strong absorption bands at 1581 cm− 1 and 1251
shaped pores by the densest distribution of CeO2 particles on the fi­ cm− 1 were assigned to C=C and C–C bonds instead alkyl, nitrile,
bers, which could be confirmed by the modulation of Ce(NO3)3 doses. carbonyl, and ether groups, also proving successful carbonization of
The pore size distributions of the samples were determined using the PAN NFs through the thermal treatment [27].
Barrett–Joyner–Halenda (BJH) and Horvath–Kawazoe (HK) models. As Fig. 4A and S11 show the CO2 and N2 adsorption capacities and the
shown in Fig. S8, drastic increases in pore volume were observed for the equilibrium isotherms of ACNFs, CeO2/ACNFs-3, PEI/ACNFs, PEI/
diameters smaller than 3 nm in all samples, showing that the thermal +CeO2/ACNFs-3, and PEI/CeO2/ACNFs-3 at different temperatures for
treatment induced a significant portion of micropores. On the other pressure up to 850 mmHg (1.13 bar), respectively. The adsorption ca­
hand, CeO2/ACNFs-2, 3, and 4 included relatively larger portions of the pacities of all samples were proportional to pressure, which were due to
pores with diameters >3 nm because of the coexistence of the meso­ the forced deep diffusion of CO2 into the pores regardless of PEI incor­
porous CeO2 particles (Table S1). The highest volume of CeO2/ACNFs-3 poration. In the absence of PEI (ACNFs and CeO2/ACNFs-3), CO2
matched the N2 adsorption capacity, suggesting that modular coordi­ adsorption was inversely proportional to the operating temperature,
nation of Ce3+ is required to ensure optimal textural properties. The pore showing exothermic physisorption to be the main driver of the CO2
size distributions for diameters <1 nm were also analyzed to identify the uptake (Fig. S11A and S11B). The higher uptake for CeO2/ACNFs-3 than
portion of pores with diameters of 0.5 to 0.6 nm (Fig. 3B) because these ACNFs proved that the formation of CeO2 particles on the surface of
diameters (close to the kinetic diameters (<1 nm) of CO2) are favorable activating NFs could provide a favorable pore structure and selectivity
for physical adsorption of CO2 (increasing adsorption energy of CO2 due for enhanced CO2 adsorption. The microstructure of the resulting CeO2
to van der Waals potential field formed by the pore walls) [23,24]. particles may also have oxygen basic and vacant sites as active surfaces
Compared to the others, CeO2/ACNFs-3 exhibited a prominent portion that enable the storage of extra oxygen. These oxygen vacancies may act
of the pores for CO2 selectivity, showing the requirement of the modular as active sites for the chemisorption and activation of CO2, leading to the
approach to derive a favorable effect of the simultaneous oxidation of reduction to CO or other products from intermediate carbonates [28].
Ce3+ ions and PAN molecules for constructing optimal textural Owing to the presence of the active sites with adjacent hydroxyl groups,
properties. adsorbed CO2 onto the active sites was reacted with surface oxygen ions
The textural properties of PEI-impregnated (for reinforcing CO2 or hydroxide species to be formed as carbonates [29,30]. This suggests
chemisorption) CeO2/ACNFs-3 (PEI/CeO2/ACNFs-3), including PEI/ that CeO2/ACNFs-3 may provide greater selectivity to CO2 than N2 (<5
ACNFs and PEI/+CeO2/ACNFs-3 were also investigated. All samples cm3/g at 0 ◦ C, 850 mmHg) owing to its large quadrupole moments and
exhibited significantly reduced N2 adsorption after impregnation, polarizability to CO2 molecules [31]. In the presence of PEI

Fig. 3. Textural properties of the resulting CeO2/ACNFs from the four doses of Ce(NO3)3, including ACNFs and +CeO2/ACNFs-3 for comparison. (A, B) N2
adsorption–desorption isotherms and pore size distributions (from the HK model) of the samples.

4
S. Kang et al. Separation and Purification Technology 320 (2023) 124202

Fig. 4. CO2 adsorption performance and

stability at different operating temperatures.
(A) CO2 adsorption capacities of the samples
in the absence (ACNFs, CeO2/ACNFs-3) and
presence (PEI/ACNFs, PEI/+CeO2/ACNFs-3,
and PEI/CeO2/ACNFs-3) of PEI at 850
mmHg. The figure includes N2 adsorption
results for comparison. (B) Comparison in
the adsorption capacity between the present
and previous studies with and without PEI.
(C) Cyclability of PEI/CeO2/ACNFs-3 for five
consecutive CO2 adsorption tests at 25 ◦ C
and 75 ◦ C. After each cycle, the sample was
degassed at 100 ◦ C for 1 h in dry air.

(Fig. S11C–S11E), PEI/CeO2/ACNFs-3 exhibited the highest CO2 portion of chemisorption. The inversely proportional adsorption ca­
adsorption capacity (reached 185.79 cm3/g at 75 ◦ C, 850 mmHg; pacity of ACNFs and CeO2/ACNFs-3 to the operating temperature re­
derived the highest Qmax in the Langmuir–Freundlich model (Table S2 flected the increases in ΔG◦ with increasing temperature, which matched
and Eqs. S3 and S4)), even better than those in most recent relevant the CO2 uptake results (Fig. S11A and S11B). In the presence of PEI, the
studies [32]. This highest performance was probably due to a combi­ positive ΔH◦ values of the samples represented an endothermic adsorp­
natorial effect of chemical and physical CO2 affinities of PEI and CeO2 tion process for CO2, indicating better performance at high tempera­
from the optimal CeO2 particle and pore size distributions that confers tures. On the other hand, the momentum to offset the CO2 adsorption
enhanced bindings to adjacent CO2 molecules, even under lower oper­ will be strengthened due to the exothermic nature of CO2 adsorption
ating pressures (<100 mmHg) [33]. Other bindings from electrostatic that accelerates the desorption of CO2 as the temperature increases. This
(between the positively charged groups in PEI and negatively charged induced relatively lower R2 levels (dependence on the van’t Hoff
CeO2), Lewis acid (CeO2)–base (PEI), and π–π stacking (between the equation in the CO2 adsorption) for PEI/ACNFs (R2 = 0.8051) and PEI/
aromatic groups on PEI and ACNF) interactions might also derive the +CeO2/ACNFs-3 (R2 = 0.4883) that that of PEI/CeO2/ACNF-3 (R2 =
enhancement in CO2 adsorption [34–36]. In the case of PEI/ACNFs, the 0.9432), proving the robust performance of PEI/CeO2/ACNF-3 in a
higher capacities than those of ACNFs alone at higher temperatures were temperature-dependent manner.
related to the strong interactions between the CO2 and amine groups of Fig. 4B shows a comparison in CO2 adsorption capacity between the
PEI on the microporous channels of ACNFs. The temperature-dependent present and previous studies in the absence and presence of PEI [42–48].
uptake manner was attributed to the increased molecular flexibility and The developed CeO2/ACNFs-3 and PEI/CeO2/ACNFs-3 in this study
intermolecular distance of the branched chain of PEI molecules, pro­ demonstrated the highest specific surface area and CO2 adsorption ca­
moting the reactivity with CO2 [37–39]. The PEI-impregnated CeO2/ pacity within each category. The consecutive CO2 adsorp­
ACNFs-3 also exhibited the temperature-dependent manner, suggesting tion–desorption tests were also carried out at different temperatures
that the PEI/CeO2/ACNFs hierarchical architectures may be a viable (25 ◦ C and 75 ◦ C) to investigate the cycling behavior and stability of
option even for post-combustion CO2 capture [7,32]. As shown in PEI/CeO2/ACNFs-3 (Fig. 4C). The adsorption capacity of PEI/CeO2/
Fig. S11A–S11E, the Freundlich and Langmuir–Freundlich models were ACNFs-3 exhibited relatively constant capacities at 25 ◦ C (130–134
well fitted (R2 > 0.99) to the experimental results of CO2 adsorption for cm3/g) and 75 ◦ C (185–187 cm3/g) after the fives adsorp­
all tested samples (Table S2). In particular, the operating temperatures tion–desorption cycles. This stability was probably due to PEI overlayers
were relevant (inversely or directly proportional) to the value of n as protective coatings to suppress the deactivation of CeO2 particles on
(determined by the slope of the plot in the Freundlich model) [40], while the fiber surface, further suggesting that the overlayer (PEI)/interlayer
all of the n values in the Freundlich model were >1, representing the (CeO2)/base fiber (ACNFs) hierarchical architecture warrants further
favorable multilayer adsorption of CO2 [41]. investigation for its practical applications.
Fig. S11F and Table S3 (with Eqs. S5 and S6) present the van’t Hoff In the light of practical implementation, the necessary expense for
plot and thermodynamic parameters of ΔHo, ΔSo, and Go of the samples, the proposed method was estimated based on a previous study [49],
respectively. The negative values of ΔHo in the absence of PEI (ACNFs which was compared to that from chemical surface etching of Ce(NO3)3/
and CeO2/ACNFs-3) confirmed the exothermic nature of the adsorption PAN NFs for surface distribution of Ce components. Even though the cost
process, driven by a net decrease in entropy. The higher ΔHo level of reduction by adopting the proposed method was marginal for the PEI/
CeO2/ACNFs-3 than ACNFs indicated that CeO2/ACNFs-3 has a greater CeO2/ACNFs production, wastewater generation and chemicals to be

5
S. Kang et al. Separation and Purification Technology 320 (2023) 124202

consumed for the chemical etching can be reduced by switching the org/10.1016/j.seppur.2023.124202.
electrical polarity for electrospinning. Nevertheless, the cost issue of Ce
(NO3)3 needs to be compromised through mass production and References
improving supply chain stability to secure an actual competitive
advantage in practical implementation [50]. [1] K. Rennert, F. Errickson, B.C. Prest, L. Rennels, R.G. Newell, W. Pizer, C. Kingdon,
J. Wingenroth, R. Cooke, B. Parthum, D. Smith, K. Cromar, D. Diaz, F.C. Moore, U.
K. Müller, R.J. Plevin, A.E. Raftery, H. Ševčíková, H. Sheets, J.H. Stock, T. Tan,
4. Conclusion M. Watson, T.E. Wong, D. Anthoff, Comprehensive evidence implies a higher social
cost of CO2, Nature 610 (2022) 687–692, https://doi.org/10.1038/s41586-022-
05224-9.
To ensure an optimal compositional hierarchy of nanofibrous ad­ [2] J. Uetake, T.C.J. Hill, K.A. Moore, P.J. DeMott, A. Protat, S.M. Kreidenweis,
sorbents for CO2 emissions, an ES-based assembly with polarity modu­ Airborne bacteria confirm the pristine nature of the Southern Ocean boundary
lation of electric potential was developed and utilized to build composite layer, Proc. Natl. Acad. Sci. U.S.A. 117 (2020) 13274–13282, https://doi.org/
10.1073/pnas.2000134117.
architectures. In particular, the location of Ce3+ in the Ce(NO3)3/PAN [3] J.Y. Lai, L.H. Ngu, S.S. Hashim, A review of CO2 adsorbents performance for
precursor (inner distribution at positive potential while outer at a different carbon capture technology processes conditions, Greenh. Gases Sci.
negative potential) was conveniently manipulated by switching elec­ Technol. 11 (2021) 1076–1117, https://doi.org/10.1002/ghg.2112.
[4] M.-H. Yu, P. Zhang, R. Feng, Z.-Q. Yao, Y.-C. Yu, T.-L. Hu, X.-H. Bu, Construction of
trical polarity when the precursor was discharged through the nozzle
a multi-cage-based MOF with a unique network for efficient CO2 capture, ACS
with electric potential. At a negative potential, in particular, CeO2 Appl. Mater. Interfaces 9 (2017) 26177–26183, https://doi.org/10.1021/
particles were densely distributed on the surface of the nanofibers as acsami.7b06491.
CeO2/ACNFs after the thermal treatment. Their distribution was opti­ [5] M. Debost, P.B. Klar, N. Barrier, E.B. Clatworthy, J. Grand, F. Laine, P. Brázda,
L. Palatinus, N. Nesterenko, P. Boullay, S. Mintova, Synthesis of discrete CHA
mized by regulating the deployed Ce3+ concentration to enhance CO2 zeolite nanocrystals without organic templates for selective CO2 capture, Angew.
adsorption superior to previous studies. Anchoring PEI molecules onto Chem. Int. Ed. 59 (2020) 23491–23495, https://doi.org/10.1002/anie.202009397.
CeO2/ACNFs further enhanced CO2 adsorption even at increased tem­ [6] Y. Ji, M. Zhang, K. Guan, J. Zhao, G. Liu, W. Jin, High-performance CO2 capture
through polymer-based ultrathin membranes, Adv. Funct. Mater. 29 (2019)
perature probably because of the capacity of CO2 chemisorption to 1900735, https://doi.org/10.1002/adfm.201900735.
ACNFs. Switching polarity of the electrical potential for electrospinning [7] A.E. Creamer, B. Gao, Carbon-based adsorbents for postcombustion CO2 capture: a
was effective to distribute conveniently CeO2 particles on ACNFs to critical review, Environ. Sci. Technol. 50 (2016) 7276–7289, https://doi.org/
10.1021/acs.est.6b00627.
maximize CO2 adsorption capacity while only an immersion of CeO2/ [8] H. Sun, J. Wang, J. Zhao, B. Shen, J. Shi, J. Huang, C. Wu, Dual functional catalytic
ACNFs in PEI dissolved solution was required to anchor PEI molecules materials of Ni over Ce-modified CaO sorbents for integrated CO2 capture and
for improving chemisorption of CO2 in spite of the porosity reduction. A conversion, Appl. Catal. B 244 (2019) 63–75, https://doi.org/10.1016/j.
apcatb.2018.11.040.
repetitive adsorption–desorption of CO2 proved the stability in high [9] G. Zhang, P. Zhao, Y. Xu, Z. Yang, H. Cheng, Y. Zhang, Structure property - CO2
temperature and long-term operation. This work may provide a reliable capture performance relations of amine-functionalized porous silica composite
architecture of highly capacitive composite nanofibers for CO2 capture adsorbents, ACS Appl. Mater. Interfaces 10 (2018) 34340–34354, https://doi.org/
10.1021/acsami.8b13069.
and modular preparation platform to screen optimal hierarchies of base
[10] J. Yoon, H.-S. Yang, B.-S. Lee, W.-R. Yu, Recent progress in coaxial electrospinning:
adsorbents for various types of CO2 capture and conversion applications. new parameters, various structures, and wide applications, Adv. Mater. 30 (2018)
1704765, https://doi.org/10.1002/adma.201704765.
[11] H.-Y. Mi, X. Jing, Q. Zheng, L. Fang, H.-X. Huang, L.-S. Turng, S. Gong, High-
CRediT authorship contribution statement
performance flexible triboelectric nanogenerator based on porous aerogels and
electrospun nanofibers for energy harvesting and sensitive self-powered sensing,
Sangmo Kang: Methodology, Validation, Formal analysis, Investi­ Nano Energy 48 (2018) 327–336, https://doi.org/10.1016/j.nanoen.2018.03.050.
gation, Data curation, Writing – original draft. Jungho Hwang: Meth­ [12] Y. Zhang, B.-T. Zhang, Y. Teng, J. Zhao, X. Sun, Heterogeneous activation of
persulfate by carbon nanofiber supported Fe3O4@carbon composites for efficient
odology, Validation, Funding acquisition, Writing – review & editing, ibuprofen degradation, J. Hazard. Mater. 401 (2021), 123428, https://doi.org/
Project administration, Supervision. Jeong Hoon Byeon: Methodology, 10.1016/j.jhazmat.2020.123428.
Validation, Data curation, Conceptualization, Writing – review & edit­ [13] S. Kang, D.H. Park, J. Hwang, Hierarchical ZnO nano-spines grown on a carbon
fiber seed layer for efficient VOC removal and airborne virus and bacteria
ing, Project administration, Supervision. inactivation, J. Hazard. Mater. 424 (2022), 127262, https://doi.org/10.1016/j.
jhazmat.2021.127262.
Declaration of Competing Interest [14] L. Zhang, T. Wei, Z. Jiang, C. Liu, H. Jiang, J. Chang, L. Sheng, Q. Zhou, L. Yuan,
Z. Fan, Electrostatic interaction in electrospun nanofibers: Double-layer carbon
protection of CoFe2O4 nanosheets enabling ultralong-life and ultrahigh-rate
The authors declare that they have no known competing financial lithium ion storage, Nano Energy 48 (2018) 238–247, https://doi.org/10.1016/j.
interests or personal relationships that could have appeared to influence nanoen.2018.03.053.
[15] X. Li, W. Chen, Q. Qian, H. Huang, Y. Chen, Z. Wang, Q. Chen, J. Yang, J. Li, Y.
the work reported in this paper. W. Mai, Electrospinning-based strategies for battery materials, Adv. Energy Mater.
11 (2021) 2000845, https://doi.org/10.1002/aenm.202000845.
Data availability [16] S. Kang, J. Choi, G.Y. Park, H.R. Kim, J. Hwang, A novel and facile synthesis of Ag-
doped TiO2 nanofiber for airborne virus/bacteria inactivation and VOC elimination
under visible light, Appl. Surf. Sci. 599 (2022), 153930, https://doi.org/10.1016/j.
Data will be made available on request. apsusc.2022.153930.
[17] C. Yang, B. Wang, Y. Zhang, H. Wang, Preparation and properties of
Acknowledgements polyacrylonitrile fibers with guanidine groups, Fibers Polym. 16 (2015)
1611–1617, https://doi.org/10.1007/s12221-015-4480-1.
[18] Z. Liu, Q. Wang, J. Wu, H. Zhang, Y. Liu, T. Zhang, H. Tian, S. Zeng, Active sites
This work was supported by Korea Environment Industry & Tech­ and interfacial reducibility of CuxO/CeO2 catalysts induced by reducing media and
nology Institute (KEITI) through Technology Development Project for O2/H2 activation, ACS Appl. Mater. Interfaces 13 (2021) 35804–35817, https://
doi.org/10.1021/acsami.1c09332.
Biological Hazards Management in Indoor Air Program, funded by Korea [19] W. Li, L. Jin, F. Gao, H. Wan, Y. Pu, X. Wei, C. Chen, W. Zou, C. Zhu, L. Dong,
Ministry of Environment (MOE) (2021003370005).This work was also Advantageous roles of phosphate decorated octahedral CeO2 {1 1 1}/g-C3N4 in
supported by Korea Environment Industry & Technology Institute boosting photocatalytic CO2 reduction: charge transfer bridge and Lewis basic site,
Appl. Catal. B 294 (2021), 120257, https://doi.org/10.1016/j.
(KEITI) through Core Technology Development Project for Environ­ apcatb.2021.120257.
mental Diseases Prevention and Management, funded by Korea Ministry [20] M.S.A. Rahaman, A.F. Ismail, A. Mustafa, A review of heat treatment on
of Environment (MOE) (ARQ202201527001). polyacrylonitrile fiber, Polym. Degrad. Stab. 92 (2007) 1421–1432, https://doi.
org/10.1016/j.polymdegradstab.2007.03.023.
[21] S. Kang, J. Hwang, Fabrication of hollow activated carbon nanofibers (HACNFs)
Appendix A. Supplementary data containing manganese oxide catalyst for toluene removal via two-step process of
electrospinning and thermal treatment, Chem. Eng. J. 379 (2020), 122315,
https://doi.org/10.1016/j.cej.2019.122315.
Supplementary data to this article can be found online at https://doi.

6
S. Kang et al. Separation and Purification Technology 320 (2023) 124202

[22] D. Ye, R. Wang, X. Wang, X. Liu, H. Liu, H. Wang, Effects of synthesis methods on [36] F. Gu, K. Chen, Y. Du, Y. Song, L. Wang, CeO2-NiO/N, O-rich porous carbon
the physicochemical properties and Hg0 capture capability of MnO2-CeO2 mixed derived from covalent-organic framework for enhanced Li-storage, Chem. Eng. J.
oxides, Appl. Surf. Sci. 578 (2022), 151998, https://doi.org/10.1016/j. 442 (2022), 136298, https://doi.org/10.1016/j.cej.2022.136298.
apsusc.2021.151998. [37] R. Sanz, G. Calleja, A. Arencibia, E.S. Sanz-Pérez, CO2 adsorption on branched
[23] A.C. Dassanayake, M. Jaroniec, Dual optimization of microporosity in carbon polyethylenimine-impregnated mesoporous silica SBA-15, Appl. Surf. Sci. 256
spheres for CO2 adsorption by using pyrrole as the carbon precursor and potassium (2010) 5323–5328, https://doi.org/10.1016/j.apsusc.2009.12.070.
salt as the activator, J. Mater. Chem. A 5 (2017) 19456–19466, https://doi.org/ [38] R. Kishor, A.K. Ghoshal, High molecular weight polyethylenimine functionalized
10.1039/C7TA05523D. three dimensional mesoporous silica for regenerable CO2 separation, Chem. Eng. J.
[24] X. Zhao, X. Zeng, Y. Qin, X. Li, T. Zhu, X. Tang, An experimental and theoretical 300 (2016) 236–244, https://doi.org/10.1016/j.cej.2016.04.055.
study of the adsorption removal of toluene and chlorobenzene on coconut shell [39] F. Liu, S. Chen, Y. Gao, Synthesis of porous polymer based solid amine adsorbent:
derived carbon, Chemosphere 206 (2018) 285–292, https://doi.org/10.1016/j. effect of pore size and amine loading on CO2 adsorption, J. Colloid Interface Sci.
chemosphere.2018.04.126. 506 (2017) 236–244, https://doi.org/10.1016/j.jcis.2017.07.049.
[25] X. Yao, R. Zhao, L. Chen, J. Du, C. Tao, F. Yang, L. Dong, Selective catalytic [40] V.K. Singh, E.A. Kumar, Comparative studies on CO2 adsorption isotherms by solid
reduction of NOx by NH3 over CeO2 supported on TiO2: Comparison of anatase, adsorbents, Mater. Today Proc. 5 (2018) 23033–23042, https://doi.org/10.1016/j.
brookite, and rutile, Appl. Catal. B 208 (2017) 82–93, https://doi.org/10.1016/j. matpr.2018.11.032.
apcatb.2017.02.060. [41] A. Sarwar, M. Ali, A.H. Khoja, A. Nawar, A. Waqas, R. Liaquat, S.R. Naqvi, M. Asjid,
[26] J. Kong, S. Yu, Fourier transform infrared spectroscopic analysis of protein Synthesis and characterization of biomass-derived surface-modified activated
secondary structures, Acta Biochim. Biophys. Sinica 39 (2007) 549–559, https:// carbon for enhanced CO2 adsorption, J. CO2 Util. 46 (2021) 101476, 10.1016/j.
doi.org/10.1111/j.1745-7270.2007.00320.x. jcou.2021.101476.
[27] W. Xu, B. Xin, X. Yang, Carbonization of electrospun polyacrylonitrile (PAN)/ [42] C. Slostowski, S. Marre, P. Dagault, O. Babot, T. Toupance, C. Aymonier, CeO2
cellulose nanofibril (CNF) hybrid membranes and its mechanism, Cellulose 27 nanopowders as solid sorbents for efficient CO2 capture/release processes, J. CO2
(2020) 3789–3804, https://doi.org/10.1007/s10570-020-03006-y. Util. 20 (2017) 52–58, https://doi.org/10.1016/j.jcou.2017.03.023.
[28] X. Zheng, Y. Li, L. Zhang, L. Shen, Y. Xiao, Y. Zhang, C. Au, L. Jiang, Insight into the [43] Q. Li, J. Guo, D. Xu, J. Guo, X. Ou, Y. Hu, H. Qi, F. Yan, Electrospun N-doped
effect of morphology on catalytic performance of porous CeO2 nanocrystals for H2S porous carbon nanofibers incorporated with NiO nanoparticles as free-standing
selective oxidation, Appl. Catal. B 252 (2019) 98–110, https://doi.org/10.1016/j. film electrodes for high-performance supercapacitors and CO2 capture, Small 14
apcatb.2019.04.014. (2018) 1704203, https://doi.org/10.1002/smll.201704203.
[29] L. Lin, S. Yao, Z. Liu, F. Zhang, N. Li, D. Vovchok, A. Martínez-Arias, R. Cataneda, [44] F.E.C. Othman, N. Yusof, S. Samitsu, N. Abdullah, M.F. Hamid, K. Nagai, M.N.Z.
J. Lin, S.D. Senanayake, D. Su, D. Ma, J.A. Rodriguez, In situ characterization of Abidin, M.A. Azali, A.F. Ismail, J. Jaafar, F. Aziz, W.N.W. Salleh, Activated carbon
Cu/CeO2 nanocatalysts for CO2 hydrogenation: Morphological effects of nanofibers incorporated metal oxides for CO2 adsorption: effects of different type
nanostructured ceria on the catalytic activity, J. Phys. Chem. C 122 (2018) of metal oxides, J. CO2 Util. 45 (2021) 101434, 10.1016/j.jcou.2021.101434.
12934–12943, https://doi.org/10.1021/acs.jpcc.8b03596. [45] M. Niu, H. Yang, X. Zhang, Y. Wang, A. Tang, Amine-impregnated mesoporous
[30] W.-F. Kuan, W.-Y. Yu, F.-Y. Tu, C.-H. Chung, Y.-C. Chang, M.M. Lin, T.-H. Yu, L.- silica nanotube as an emerging nanocomposite for CO2 capture, ACS Appl. Mater.
J. Chen, Facile reflux preparation of defective mesoporous ceria nanorod with Interfaces 8 (2016) 17312–17320, https://doi.org/10.1021/acsami.6b05044.
superior catalytic activity for direct carbon dioxide conversion into dimethyl [46] V. Irani, A.G. Khosh, A. Tavasoli, Polyethyleneimine (PEI) functionalized metal
carbonate, Chem. Eng. J. 430 (2022), 132941, https://doi.org/10.1016/j. oxide nanoparticles recovered from the catalytic converters of spent automotive
cej.2021.132941. exhaust systems and application for CO2 adsorption, Front. Energy Res. 8 (2020)
[31] Y. Liu, J. Wilcox, Molecular simulation studies of CO2 adsorption by carbon model 196, https://doi.org/10.3389/fenrg.2020.00196.
compounds for carbon capture and sequestration applications, Environ. Sci. [47] B. Wu, F.Q. Liu, S.W. Luo, L.Q. Zhang, F. Zou, Carbonaceous materials-supported
Technol. 47 (2013) 95–101, https://doi.org/10.1021/es3012029. polyethylenimine with high thermal conductivity: a promising adsorbent for CO2
[32] F.E.C. Othman, N. Yusof, M. Petrů, N.A.H.M. Nordin, M.F. Hamid, A.F. Ismail, A. capture, Compos. Sci. Technol. 208 (2021), 108781, https://doi.org/10.1016/j.
I. Rushdan, S.A. Hassan, Polyethyleneimine-impregnated activated carbon compscitech.2021.108781.
nanofiber composited graphene-derived rice husk char for efficient post- [48] K. Li, J. Jiang, F. Yan, S. Tian, X. Chen, The influence of polyethyleneimine type
combustion CO2 capture, Nanotechnol. Rev. 11 (2022) 926–944, https://doi.org/ and molecular weight on the CO2 capture performance of PEI-nano silica
10.1515/ntrev-2022-0055. adsorbents, Appl. Energy 136 (2014) 750–755, https://doi.org/10.1016/j.
[33] X. Yan, L. Zhang, Y. Zhang, K. Qiao, Z. Yan, S. Komarneni, Amine-modified apenergy.2014.09.057.
mesocellular silica foams for CO2 capture, Chem. Eng. J. 168 (2011) 918–924, [49] B. Stieberova, M. Zilka, M. Ticha, F. Freiberg, P. Caramazana-González,
https://doi.org/10.1016/j.cej.2011.01.066. J. McKechnie, E. Lester, Sustainability assessment of continuous-flow
[34] P. Hu, J. An, M.M. Faulkner, H. Wu, Z. Li, X. Tian, J.P. Giraldo, Nanoparticle hydrothermal synthesis of nanomaterials in the context of other production
charge and size control foliar delivery efficiency to plant cells and organelles, ACS technologies, J. Clean. Prod. 241 (2019), 118325, https://doi.org/10.1016/j.
Nano 14 (2020) 7970–7986, https://doi.org/10.1021/acsnano.9b09178. jclepro.2019.118325.
[35] J. Li, Z. Xia, Q. Xue, M. Zhang, S. Zhang, S. Zhang, H. Xiao, Y. Ma, Y. Qu, Insights [50] Z.C. Sims, M.S. Kesler, H.B. Henderson, E. Castillo, T. Fishman, D. Weiss,
into the interfacial Lewis acid-base pairs in CeO2-loaded CoS2 electrocatalysts for P. Singleton, R. Eggert, S.K. McCall, O. Rios, How cerium and lanthanum as
alkalaine hydrogen evolution, Small 17 (2021) 2103018, https://doi.org/10.1002/ coproducts promote stable rare earth production and new alloys, J. Sustain. Metall.
smll.202103018. 8 (2022) 1225–1234, https://doi.org/10.1007/s40831-022-00562-4.