Hydroxide exchange membrane carbon capture using a nickel hydroxide symmetric battery cell

W4406249767 · 수집 2026-09-26 12:33:05

T11690 Advanced battery technologies research 이차전지

CoT 라벨링

라벨링 전

PDF 파일

자동 다운로드한 PDF PDF 보기

기본 정보

DOI 제목 발행년도 저널 URL 페이지 열기 PDF URL PDF 열기 발행기관 피인용수 라이선스

추가 정보 초록 본문

# Hydroxide exchange membrane carbon capture using a nickel hydroxide symmetric battery cell

Yushan Yan

University of Delaware [https://orcid.org/0000-0001-6616-4575](https://orcid.org/0000-0001-6616-4575)

James Buchen

University of Delaware

Teng Wang

University of Delaware

Ben Achrai

RepAir

Jean-Philippe Hiegel

RepAir

Naama Gluz

RepAir

Maurice Artoul

RepAir

Braden Geiger

University of Delaware

Brian Setzler

University of Delaware

Physical Sciences - Article

Keywords:

Posted Date: January 10th, 2025

DOI: [https://doi.org/10.21203/rs.3.rs-5627423/v1](https://doi.org/10.21203/rs.3.rs-5627423/v1)

License:   This work is licensed under a Creative Commons Attribution 4.0 International License.
[Read Full License](https://creativecommons.org/licenses/by/4.0/)

Additional Declarations: Yes there is potential Competing Interest. Y.Y. is the Founder and CEO and P.B.S.
is a Co-Founder for Versogen, the producer of Piperion®. Y.Y. and B.A. are Co-Founders for RepAir the

scale-up company.

Version of Record: A version of this preprint was published at Nature Energy on September 22nd, 2026.
[See the published version at https://doi.org/10.1038/s41560-026-02129-z.](https://doi.org/10.1038/s41560-026-02129-z)

## Abstract

Electrochemical carbon capture devices can be a low energy cost solution for direct air capture (DAC)
using renewable electricity. Historically electrochemical carbon-capture has targeted a range of

concentrations from atmospheric (400 ppmCO2) (DAC), 1 to life-support (5000 ppmCO2), 2 to point-source

capture (10% CO2). 3 The hydroxide exchange membrane nickel hydroxide symmetric battery cell with two

identical electrodes has low voltage requirements making it more suitable for DAC than other

electrochemical approaches. A 25 cm 2 laboratory cell shows an average energy cost of 1.15 MWh·tonCO2
−1 and a CO2 flux of 78 kgCO2·m2·yr− 1 at 2 mA·cm− 2. A manufacturable 25 cm 2 cell is durability tested for

5000 hours and achieves an energy of 0.46 MWh·tonCO2 −1 and a flux of 62 kgCO2·m 2 ·yr − 1 at the end of the

test. A DAC pilot system with a stack of 9 scaled-up 300 cm 2 cells demonstrates an energy of 0.83

MWh·tonCO2 −1 and a flux of 75 kgCO2·m 2 ·yr − 1 and meets the 300 Pa pressure drop required for DAC. 4

Modular design projects improvements in cost from manufacturing and economies of scale, and a

pathway to below 100 $·tonCO2 −1 from learning rates of past energy technologies. 5

## Introduction

Direct air capture (DAC) has been identified as one of the key net negative carbon technologies to achieve

a net zero future. 6,7 DAC is required to offset continued emissions from dilute CO2 sources e.g.,

agriculture and construction. 8 The majority of current DAC technologies at scale (> 1 ktonCO2∙yr − 1 ) are

sorbent based with significant energy cost. 9 The energy cost is the major operating expense and is

primarily driven by the temperature swing required to regenerate the sorbent. 10,11 The United States

Department of Energy (DOE) has set forth a cost target of 100 $·tonCO2 −1 which is not possible to meet

with current industrial electricity costs near 70 $·MWh − 1 and current temperature swing adsorption

devices energy costs above 1.5 MWh·tonCO2 −1 . 12,13 A combination of lower electricity costs and lower

DAC energy costs are required to meet the target DAC cost.

Electrochemical pH gradient devices are a growing research area for low energy cost carbon capture. 1,14–

19 The pH gradient is built by generating hydroxide (OH−) at the cathode and consuming OH− at the anode.
An acid-base equilibrium with CO2 allows for the capture of CO2 at the cathode and release at the anode.

20 This compliments other electrochemical CO2 capture devices based on p _Ka_ gradients of an

electrochemically active species allowing for the capture and release of CO2. 2,21,22

This work explores a hydroxide exchange membrane carbon capture (HEMCC) device with symmetric
Ni(OH)2 electrodes to produce the pH gradient for CO2 capture and release (Fig. 1a). At the cathode NiOOH

is reduced to Ni(OH)2 while at the anode Ni(OH)2 is oxidized to NiOOH. This redox chemistry is known for

its reversibility and cycle durability and has been commercially used in the cathode of alkaline Ni-MH

battery technologies. 23–25 This symmetric cell has a thermodynamic equilibrium voltage of zero. Most of
th lt b d i t d th H di t ith th i d d i i th l i ti f th

electrodes. There is a resistance component as well, but this is small due to the low current densities used

in the device, typically < 5 mA·cm − 2 .

Scaleup is necessary for DAC technologies with several orders of magnitude growth required to meet net

zero targets. 26 This HEMCC device was first developed at a 25 cm 2 laboratory scale able to capture 200

gCO2·yr − 1 (Fig. 1b). It was built with electrodes formed by electrochemical precipitation of nickel hydroxide

on nickel foam. A manufacturable version of the 25 cm 2 electrode was developed and durability tested for

5000 hours. This durable electrode was scaled to 300 cm 2 and integrated into a 9-cell stack for a pilot DAC
system (Fig. 1c). The first commercial modular DAC system is under construction for operation in 2026

with a capacity target of 1 ktonCO2·yr − 1 (Fig. 1d).

## Results and Discussion

Figure 2 shows the data of a typical laboratory 25 cm 2 cell. Figures 2a,b are time series of the operating
cell voltage and cathode outlet CO2 concentration with an initial conditioning cycle followed by ten regular

cycles, all at 2 mA·cm − 2 . Within each regular cycle there are two steps, one for the charge/discharge and
the other for the regeneration. CO2 is captured in both steps, but at different voltages. The cell voltage and

the difference between the 400 ppm cathode inlet and cathode outlet CO2 allow for calculation of key

process parameters of capacity, electron efficiency, energy cost, and flux. Figure 2c-f presents these
parameters excluding the regeneration steps. Capacity tracks long-term changes in the device. Electron
efficiency is defined here as the ratio of CO2 molecules captured per electron passed in the cell with a

maximum stoichiometric ratio of 1 CO2 per electron. Energy cost is the largest operating cost of the device

while the flux determines the size and thus capital cost of the system. An average of 1.15 MWh·tonCO2

energy cost including the regeneration step was achieved in this single cell dataset with a flux of 78

−1

kgCO2·m − 2 ·yr − 1 . Each cycle was integrated using the definitions in SI Discussion 5 and the mean was taken

excluding the conditioning cycles and assuming all CO2 that is captured on the cathode is released in the

anode to a product stream. The HEMCC has a lower energy cost than the incumbent technology presented

by Carbon Engineering of 1.47 MWh·tonCO2

−1 for their device when excluding the blower and compressor.

Figure 2 | Laboratory scale 25 cm 2 HEMCC and its behavior. The HEMCC is comprised of two identical
electrochemically precipitated Ni(OH)2 electrodes and an 80µm Piperion® membrane. The cathode air

flow was 1 L - min -1 with 400 ppm CO2 and 90% RH. The anode received a sweep gas of 90% RH N2 at 1

L - min -1 . a. The cell voltage (V) and b. the cathode outlet CO2 concentration (ppm) are used to calculate

the key process variables c-f during the charge/discharge step

c. capacity, d. electron efficiency, e. energy, and f. flux, all excluding the regeneration step. Regeneration
steps are excluded to highlight the low energy cost of the charge-discharge step. Capture still occurs
during regeneration but at higher energy cost. The first cycle on each side is a 1.5-hour regeneration hold

for break-in followed by 10–75 minute cycles on each electrode. The cycle length is chosen based on an
ex-situ capacity test to minimize regeneration (SI Discussion 7 ). c. Capacity achieved an average of 1.89

mAh·cm -2 which was 85% of the ex-situ capacity tests. d. Electron efficiency achieved 0.25. e. Energy cost

averaged 0.76 MWh∙tonCO2

-1 while f. flux averaged to 72 kgCO2∙m2∙yr-1. g. and h. show a single cycle to

highlight the change in voltage and capture in the charge/discharge and regeneration steps. i., j., and k.
are graphic aids to describe the build ΔpH, maldistributed ΔpH, and rebuild ΔpH sections in g. and h. The
ΔpH between the peak pH at the cathode and the CO2 release pH at the anode drives the CO2 capture rate

in the cell, see SI Discussion 2. The CO2 release pH is the pH that achieves equilibrium between CO2 in the

anode and the concentrated CO2 gas product stream. The dotted red primary reaction coordinate lines are

rough position where OH - is produced in the cathode and consumed in the anode. i. The build ΔpH section

is directly after the gas lines switch in the cell. Redox reactions take place close to the membrane with the
cathode near 100% state of charge and the anode near 0% state of charge. The pH at the back of the
cathode electrode is close to the pH of the anode in the previous cycle. j. The maldistributed ΔpH section
is at the end of a charge/discharge step. This occurs because of the uneven mass loading of
Ni(OH)2/NiOOH when making the electrode and uneven distribution of resistance through the electrode.

Near the end of the charge/discharge step, low loading and low resistance volumes of the cathode will

have fully been consumed, no longer producing OH - . Surrounding higher loading and higher resistance

volumes continue producing OH - but the average pH across the electrode drops lowering the average

driving force for capturing CO2. The anode in this section is still charging and has not reached 100% state

of charge because of competing OER. k. The rebuild ΔpH section is the beginning of the regeneration after
the cathode has reached 0% state of charge. The cathode reaction coordinate is brought to the front
towards the membrane where catalytic ORR reactions are predominant. The full area of the electrode is
used again, and the average pH increases to have a higher rate of CO2 capture. The anode continues to

charge towards 100% for the next cycle.

Relatively low energy cost is the main benefit of this electrochemical carbon capture approach. Figure 2e

shows several cycles of CO2 capture at an average of 1.15 MWh·tonCO2

−1; when broken down by the steps

the charge/discharge and regeneration steps averaged 0.76 MWh·tonCO2

−1 and 2.15 MWh·tonCO2

respectively. Both steps can be improved by improving the electron efficiency of the device. 14 Electron

efficiency of 1 indicates ionic transport of only HCO3

- in the cell where 0.5 would be predominantly CO32−,

below 0.5 as observed in this cell has either unreacted OH − being transported or back diffusion of HCO3−

or CO3

2− from the anode. Future work can improve electron efficiency by adjusting membrane parameters

such as ionic exchange capacity (IEC), conductivity, and water uptake. Improved electron efficiency at the
same current density directly increases flux of CO2 which decreases energy and capital costs.

The CO2 capture rate changes with time and the state of charge of the electrode. Figure 2h shows the

single half cycle behavior of CO2 capture. At the beginning of the charge/discharge step there is time lag

as the pH gradient is reversed from the previous cycle. A steady increase in capture rate is seen as the pH
gradient is formed, which takes time proportional to the total IEC of the membrane and ionomer. Piperion®

is used for the membrane and ionomer based on PAP-TP-85 with an IEC of 2.37 meq.·g − 1 . 29 The state of
charge determines the CO2 capture behavior for the remainder of the charge/discharge step. The surface

of the electrode near the membrane discharges first due to the difference in conductivity between the

metal, metal hydroxide, and ionomer phases of the electrode. 30,31 The primary reaction coordinate, and
thus the peak pH, moves from the membrane to the back of the electrode as the state of charge changes
in each electrode. At a moderate state of charge, higher capture rates are expected because CO2 diffuses

into and reacts with a larger volume of high pH ionomer. A similar behavior was seen in a hydrogen

HEMCC previously studied by using an interlayer volume. 15 Near the end of the cycle volumes of anode
with lower mass loading of NiOOH/Ni(OH)2 or lower local resistance will run out of NiOOH earlier than the

surrounding volumes and no longer produce OH − ; these volumes no longer contribute to the pH gradient
and lower the average pH at the primary reaction coordinate. This manifests in a dip in CO2 capture near

the end of the charge/discharge step. During the charge/discharge step the flux averages 72 kgCO2·m −−

2·yr−− 1, while the average peak flux is 97 kgCO2·m− 2·yr− 1.

The capture rates during regeneration do not depend on the state of charge of the cathode which has been
depleted. The pH gradient is reestablished by a relatively thin layer of the electrode near the membrane

performing catalytic ORR. 30,31 There is a short time dependent growth of capture at the beginning of
regeneration to reestablish the pH gradient across the area of the electrode after the decrease observed in
the charge/discharge step. After that a steady state constant capture rate is achievable. During the

regeneration step the flux averages 93 kgCO2·m −−2 ·yr −−1, while the average peak flux is 97 kgCO2·m − 2 ·yr − 1 .

Technical pathways to improve the device include improving mass transport (diffusion) from the flow field
channels into and within the porous electrodes, decreasing the effect of time lag for CO2 and decreasing

the requirement for regeneration. Decreasing the effect of time lag can improve both the flux and energy
cost of the cycle. There are two ways to reduce the time lag, one by decreasing the total ion capacity of
membrane and electrode or the other by increasing the electrode capacity. Either option has a trade-off to
optimize. Decreasing the total ion capacity by decreasing membrane thickness and ionomer amount
decreases the number of ions required to displace at the beginning of each cycle. This cannot be
decreased to zero because there needs to be sufficient ionic conductivity and mechanical stability.
Additionally, thinner membranes can lead to increased bicarbonate back diffusion offsetting gains in time

lag. 1,15 Increasing the capacity of an electrode lengthens the time for a full cycle and thus a higher
percentage of time is spent in a high flux region of a cycle. Increasing the electrode capacity can increase
the mass transport resistance for diffusion of CO2 into the electrode by decreasing the porosity of the

electrode or increasing the diffusive path length. The laboratory prepared electrodes nominally have a
porosity of 30–40%. Optimization of this porosity can lead to flux and energy cost gains.

Decreasing the amount of time required for regeneration can also decrease the overall energy cost of the
system. In Ni-MH batteries the amount of OER that occurs depends on the C-rate used and the dopants

present. 27,32 The C-rate is defined as a ratio of the current density in mA to the capacity of the electrode in
mAh. The electrodes in Fig. 2 ran at a C-rate of 0.85. Lower C-rates allow for lower voltage operation of the

battery limiting the catalytic OER which has a high kinetic overpotential. Dopants can either improve the
kinetics of the Ni(OH)2 oxidation or hinder the kinetics of the OER reaction. Dopants such as cobalt have

shown the ability to limit OER and can be added to limit need for regeneration. 33,34 Leaning on nickel

hydride battery research to limit regeneration requirement to 10% of a cycle length could decrease the

average full cycle energy cost to 0.90 MWh·tonCO2

−1 with this laboratory scale system.

The electrodes shown in Fig. 2 were electrochemically precipitated; this method was chosen for tunability
of the Ni(OH)2 loading early in research. There are several other methods to produce Ni(OH)2 including

electrodeposition and bulk precipitation. 24 Improvements were made to the laboratory scale electrode to

produce large manufacturable electrodes for pilot scale operation. Such 25 cm 2 electrodes were tested for
5000 hours with the cell supplied with ambient air held at 32°C and 82–87% RH. Three segments of the
corresponding data are shown in Fig. 3. The end of the 5000-hour run highlighted a low energy cost near

0.46 MWh·tonCO2−1 with a flux of 62 kgCO2·m− 2·yr− 1.

The more manufacturable 25 cm 2 cell shows an increase in cycle length or capacity over time (Fig. 3). This

is likely the result of a self-limiting corrosion process in the electrode. The nickel foam used in the
electrode as a scaffold for Ni(OH)2 can be oxidized to Ni(OH)2 at potentials above 0.11 V vs reversible

hydrogen electrode (RHE). 23 Any exposed bare nickel can go through the oxidation process. The corrosion
stops when accessible nickel metal is fully converted. An increase in capacity is seen between the first
and second segments, but not between the second and third segments suggesting the complete
conversion of the available nickel sites. This phenomenon was also observed at the laboratory scale and is
shown in SI Discussion 4. An increase of the average flux is also seen, contributing to a lower energy cost.
This corroborates the concept that reducing the effect of time lag with a larger capacity electrode will lead
to higher average flux and lower average energy cost.

This manufacturable electrode was scaled up from 25 cm 2 to 300 cm 2 and used in a 9 x 300-cm 2 stack

(Fig. 3d,f). This stack averaged an energy cost of 0.83 MWh·tonCO2

−1 with a flux of 75 kgCO2·m− 2·yr− 1. At

the system level a blower is required to push air through the DAC device and 300 Pa is considered an

upper limit for DAC. 4 The laboratory scale pressure drop suffers from triple serpentine flow fields adapted

from bench top fuel cells with a pressure drop of 2800 Pa for a 1 L·min − 1 flow rate as discussed in SI

Discussion 3. The scaled system exhibits a total pressure drop of less than 300 Pa (10x lower in
comparison to the lab cells) while maintaining similar performance metrics. Such low pressure drops
range is obtained by an innovative design of the stack hardware (inlets/outlets and flow field morphology)

- needed to allow energy efficient Air Streaming Unit (ASU, e.g., axial fan) and an overall energy-efficient
DAC process.

Figure 4 shows a levelized cost reduction pathway to 92 $·tonCO2 −1 in line with the United States DOE

targets. Durability at the 25 cm 2 size and performance at the 9-cell stack size lays the groundwork for the

1000 tonCO2·yr − 1 Gen-0 pilot scale. The Gen-1 system will follow research pathways discussed to

decrease time lag and regeneration as well as increasing CO2 flux. Gen-1 will additionally focus on the full

system requirements including a blower and compressor. Improved airflow design can lower pressure

drop to decrease blower costs. 4,35 Electrochemical compression is to be explored as shown in proton
exchange membrane electrolyzers (PEMEL) for hydrogen production with a cost benefit compared with

conventional compression alone. 36–38 Module standardization for assembly line production and expected

improvement in energy consumption are the main drivers for the Gen-2 system. 39 Further improvement of

the CO2 flux and economies of scale will enable Gen-3 First-Of-A-Kind (FOAK) megaton systems. 40–42 At

this scale pricing improvements can be achieved for membrane polymers and peripheral components. 43

Finally, centralized control rooms, automation, digital twin machine learning, and AI-driven maintenance
strategies are expected to drive final costs towards the Nth-Of-A-Kind (NOAK) Gen-3 systems.

## Conclusion

The nickel hydroxide HEMCC technology shown here has a low energy cost compared to conventional
temperature swing adsorption devices. Producing the pH gradient is the primary energy cost for the
device, followed by the polarization of the electrodes, and resistance.

At the laboratory scale the concept was developed to produce a device that has an average energy cost of

1.15 MWh·tonCO2−1 at an average flux of 78 kgCO2·m− 2·yr− 1. The energy cost and flux can be improved by

decreasing the time lag for pH gradient formation and minimizing parasitic OER reactions. The time lag
can be decreased by increasing electrode capacity which aids in increasing the average flux while
decreasing the average energy cost. Anodic OER reactions can be limited by employing traditional Ni-MH
battery dopants. By limiting OER, the energy cost can be reduced by reducing the required amount of
regeneration.

Scale up of the nickel hydroxide HEMCC has started where the original concept was expanded from a 25

cm 2 cell to a 9-cell stack of 300 cm 2 cells. A pilot module containing these stacks is in operation capturing
CO2 from ambient air. A 5000-hour durability test has shown stability of the electrode design with an end

of-test energy cost of 0.46 MWh·tonCO2

−1 at a flux of 62 kgCO2·m− 2·yr− 1. This electrode design was shown

to be able to scale into a stack with a low pressure drop of 300 Pa meeting technical requirements for
DAC.

The prototype is being iterated to drive down cost along learning curves like those seen in solar photo
voltaic cells and lithium-ion batteries. Device level research is expected to achieve many of the early-stage
improvements, while manufacturability and economies of scale will drive costs down below the 100
$·tonCO2 target set forth for DAC.

## Methods Electrode preparation

Lab scale electrode preparation was done by electrochemical precipitation of nickel hydroxide onto nickel
foam.The nickel foam was from MTI corporation (product name EQ-bcnf-16m) with 99.9% purity, 95%

porosity, 80–110 pores per inch and thickness of 1.6mm. Three 25cm 2 nickel foam substrates were cut

retaining a 1 cm 2 tab for connecting with an alligator clip. The three substrates were compressed in a

hydraulic press to 512 µm by using a 512 µm thick steel plate as a guide barrier. These were weighed and
then placed in prepared 1 M sulfuric acid solution for 1 hour to remove the surface oxide layer. The oxide
layer removed is assumed to be negligible in mass for purposes of measuring amount of mass added to
the working electrode.

These substrates were set parallel in a 0.05 M nickel chloride solution using 98% anhydrous nickel chloride
salt from Millipore Sigma (CAS No: 7718-54-9). NiCl2 hexahydrate (Sigma-Aldrich, ReagentPlus) was also

tested without any noticeable performance difference. The substrates were suspended in place with
copper electrode clips as shown in SI Fig. 9 . The outer electrodes were connected via wire to be at the
same voltage as the counter electrode, the middle electrode was the working electrode for Ni(OH)2 to be

precipitated onto. Two outer electrodes were used to evenly grow Ni(OH)2 on both sides of the central

working electrode. The distance between electrodes were approximately controlled to 1 cm.

A reference Ag/AgCl electrode from Pine Research (RREF0021) tracked voltage in the cell during

electrochemical precipitation. A current density of -5 mA·cm 2 for 4 hours was applied to the working

electrode. Afterwards, the electrodes were rinsed with DI water and set out to dry overnight. The next day
the weight of the working electrode was measured to determine the mass of deposition. As a control, the
counter electrodes were also weighed to measure their loss of Ni. The loss of Ni of the counter electrodes

closely aligns with the oxidation to Ni 2+ in the solution as expected this balances the loss of Ni 2+ during

precipitation and maintains a relatively stable bulk Ni 2+ concentration. The pair of counter electrodes lost
0.57 g Ni (27% of initial mass) during the process while the working electrode gained 0.83 g during the
process.

After electrochemical precipitation the electrodes were spray coated with Piperion® ionomer. A solution of
1.3333g of 5% ionomer in ethanol solution was added to 6.2222g IPA and sonicated for 30 min. This
solution was sprayed on one side of the electrode and then set out to dry for 1 day. Weighing the next days
showed that 25% of the initial ionomer mass was transferred to the electrode during the spraying.

## Ex-situ three electrode cycling

The electrode with ionomer was then placed in a 1 M KOH bath in the same set up as the corrosion test in

SI Fig. 7,8 . Three series of current densities were used to cycle each electrode, 1-cycle at 2 mA·cm 2, 30
cycles at 5 mA·cm 2, and 5-cycles at 2 mA·cm 2 . A break in is conducted at 2 mA·cm − 2 . First the electrode
is charged from Ni(OH)2 to NiOOH for three hours finishing with a plateau at OER voltage. Then the

electrode is discharged back to Ni(OH)2 to completion with a cutoff voltage of 1.1 V vs RHE. This break in

charge is longer than the final 5-cycles at 2 mA·cm − 2 to allow for impurities in the deposited Ni(OH)2
structure to escape without physically damaging the deposited layer.

Next, 30 cycles at 5 mA·cm 2 are conducted to stabilized the Ni(OH)2 layer. It was found that 30 cycles

were required to reach a plateau in capacity for an electrode; the current density 5 mA·cm 2 was chosen to

speed up the process. The final 5 cycles were conducted at 2 mA·cm 2 to compare with the 2 mA·cm 2 used

during the direct air capture tests.

## Membrane Electrode Assembly

The final assembly of the laboratory scale 25cm 2 cell uses a Scribner 25 cm 2 cell with triple serpentine
flow fields. (see SI Fig. 11 ) A 2.5”x3” piece of 80µm Piperion® membrane was used between two
electrodes. Two electrochemically precipitated nickel hydroxide electrodes were cut to fit precisely to their
respective gaskets. A total of 336µm in FEP gaskets were used on each side for the 512µm thick
electrodes. This leads to a 0.39 compression ratio once the cell bolts are torqued to 100 PSI.

## Declarations

Acknowledgements

This work is supported by the University of Delaware, U.S. Department of Energy National Energy
Technology Laboratory (DE-FE0031955), and U.S. Department of Defense Army Research Laboratory
(W911NF-23-2-0028). Versogen provided Piperion® membrane and ionomer for this research.

Authors and Affiliations

These authors contributed equally: R. James Buchen & Teng Wang

Department of Chemical and Biomolecular Engineering at University of Delaware

R. James Buchen, Teng Wang, K. Braden Geiger, P. Brian Setzler, & Yushan Yan

RepAir

Ben Achrai, Y. Naama Gluz, Maurice Artoul, & Jean-Phillippe Hiegel

Contributions

Y.Y. and P.B.S. supervised the project. P.B.S. provided the initial concept and guidance on fundamentals of
electrochemical pH driven CO2 capture. R.J.B. designed the test station and laboratory scale electrodes.

R.J.B. and T.W. developed testing protocol and evaluation of the direct air capture system. R.B.G.
conducted electrochemical tests and related data processing. T.W. carried out materials characterization.
B.A. led the development and scale up efforts of the main technological aspects including durable

electrodes, stack hardware design and the accommodating system for operating the DAC process. Y.N.G
led the MEA optimization efforts and M.A conducted the DAC testing J-P.H. developed the TEA analysis.
R.J.B. wrote the manuscript, with input from all authors.

Corresponding author

Correspondence to Yushan Yan yanys@udel.edu

Competing Interests

Y.Y. is the Founder and CEO and P.B.S. is a Co-Founder for Versogen, the producer of Piperion®. Y.Y. and
B.A. are Co-Founders for RepAir the scale-up company.

## References

1. Shi, L. _et al._ A shorted membrane electrochemical cell powered by hydrogen to remove CO2 from the
air feed of hydroxide exchange membrane fuel cells. Nature Energy 7, 238–247 (2022).
https://doi.org/10.1038/s41560-021-00969-5

2. Bell, W. L. Synthesis and Evaluation of Electroactive CO₂ Carriers. SAE International 97, 544–552
(1988).

3. Rheinhardt, J. H., Singh, P., Tarakeshwar, P. & Buttry, D. A. Electrochemical Capture and Release of
Carbon Dioxide. ACS Energy Letters 2, 454–461 (2017).
https://doi.org/10.1021/acsenergylett.6b00608

4. Luukkonen, A., Elfving, J. & Inkeri, E. Improving adsorption-based direct air capture performance
through operating parameter optimization. Chemical Engineering Journal 471 (2023).
https://doi.org/10.1016/j.cej.2023.144525

5. McDonald, A. & Schrattenholzer, L. Learning rates for energy technologies. Energy Policy 29, 255–261
(2001). https://doi.org/10.1016/s0301-4215(00)00122-1

6. IEA. Net Zero by 2050. (Paris, 2021).

7. Fuhrman, J. _et al._ Diverse carbon dioxide removal approaches could reduce impacts on the energy–
water–land system. Nature Climate Change 13, 341–350 (2023). https://doi.org/10.1038/s41558023-01604-9

8. UNEP (2017). The Emissions Gap Report 2017. United Nations Environment Programme (UNEP),
Nairobi. (2017).

9. IEA (2022), Direct Air Capture 2022, IEA, Paris https://www.iea.org/reports/direct-air-capture-2022,
Licence: CC BY 4.0. (2022).

10. Keith, D. W., Holmes, G., St. Angelo, D. & Heidel, K. A Process for Capturing CO2 from the Atmosphere.
Joule 2, 1573–1594 (2018). https://doi.org/10.1016/j.joule.2018.05.006

11. Fasihi, M., Efimova, O. & Breyer, C. Techno-economic assessment of CO2 direct air capture plants.
Journal of Cleaner Production 224, 957–980 (2019). https://doi.org/10.1016/j.jclepro.2019.03.086

12. Monthly Energy Review August 2024 (U. S. Energy Information Administration, 2024).

13. Carbon _Negative Shot_, <https://www.energy.gov/fecm/carbon-negative-shot> (2022).

14. Matz, S. _et al._ Hydrogen-powered Electrochemically-driven CO2 Removal from Air Containing 400 to
5000 ppm CO2. Journal of The Electrochemical Society 169 (2022). https://doi.org/10.1149/19457111/ac7adf

15. Matz, S. _et al._ Demonstration of Electrochemically-Driven CO2 Separation Using Hydroxide Exchange
Membranes. Journal of The Electrochemical Society 168 (2021). https://doi.org/10.1149/19457111/abd5fe

16. Zhu, P. _et al._ Continuous carbon capture in an electrochemical solid-electrolyte reactor. Nature 618,

959–966 (2023). https://doi.org/10.1038/s41586-023-06060-1

17. Gao, X., Omosebi, A., Perrone, R. & Liu, K. Promoting CO2 Release from Carbonate Containing Solvents
during Water Electrolysis for Direct Air Capture. Journal of The Electrochemical Society 169 (2022).

https://doi.org/10.1149/1945-7111/ac5fef

18. Gao, X., Omosebi, A., Patrick, A. & Liu, K. Optimizing Direct Air Capture Solvents to Minimize Energy
Consumption of CO2 Release in a Carbonate Electrolyzer. ECS Advances 3 (2024).

https://doi.org/10.1149/2754-2734/ad3928

19. Zheng, Y. _et al._ Editors’ Choice—Power-Generating Electrochemical CO2Scrubbing from Air Enabling

Practical AEMFC Application. _Journal of The Electrochemical Society_ 168 (2021).
https://doi.org/10.1149/1945-7111/abe08a

20. Wang, X. Comprehensive Study of the Hydration and Dehydration Reactions of Carbon Dioxide in

Aqueous Solution. J. Phys. Chem. A 114, 1734–1740 (2010).

21. Diederichsen, K. M., Liu, Y., Ozbek, N., Seo, H. & Hatton, T. A. Toward solvent-free continuous-flow
electrochemically mediated carbon capture with high-concentration liquid quinone chemistry. Joule 6,

221–239 (2022). https://doi.org/10.1016/j.joule.2021.12.001

22. Li, X. _et al._ Redox-tunable isoindigos for electrochemically mediated carbon capture. Nat Commun 15,
1175 (2024). https://doi.org/10.1038/s41467-024-45410-z

23. Huggins, R. A. _Advanced Batteries Materials Science Aspects_ . (Springer, 2009).

24. Hall, D. S., Lockwood, D. J., Bock, C. & MacDougall, B. R. Nickel hydroxides and related materials: a

review of their structures, synthesis and properties. Proc Math Phys Eng Sci 471, 20140792 (2015).
https://doi.org/10.1098/rspa.2014.0792

25. Ash, B., Nalajala, V. S., Popuri, A. K., Subbaiah, T. & Minakshi, M. Perspectives on Nickel Hydroxide
Electrodes Suitable for Rechargeable Batteries: Electrolytic vs. Chemical Synthesis Routes.

Nanomaterials (Basel) 10 (2020). https://doi.org/10.3390/nano10091878

26. McQueen, N. _et al._ A review of direct air capture (DAC): scaling up commercial technologies and
innovating for the future. Progress in Energy 3 (2021). https://doi.org/10.1088/2516-1083/abf1ce

27. Becker, M. D. C., D. J.; Salva, H.R.; Rodriguez Nieto, F.; Ghilarducci, A. A.; Visintin A. Oxygen Evolution

Nickel Hydroxide Electrodes in a Ni-MH Battery Prototype. Journal of New Materials for
Electrochemical Systems 12, 277–282 (2012).

28. Hall, D. S., Lockwood, D. J., Poirier, S., Bock, C. & MacDougall, B. R. Raman and infrared spectroscopy

of alpha and beta phases of thin nickel hydroxide films electrochemically formed on nickel. J Phys
Chem A 116, 6771–6784 (2012). https://doi.org/10.1021/jp303546r

29. Wang, J. _et al._ Poly(aryl piperidinium) membranes and ionomers for hydroxide exchange membrane

fuel cells. Nature Energy 4, 392–398 (2019). https://doi.org/10.1038/s41560-019-0372-8

30. Tobias, J. S. N. C. W. Theoretical Analysis of Current Distribution in Porous Electrodes. Journal of The
Electrochemical Society (1962).

31. Newman, J. & Tiedemann, W. Porous-electrode theory with battery applications. AIChE Journal 21,

25–41 (1973). https://doi.org/10.1002/aic.690210103

32. Ayeb, A. & Notten, P. H. L. The oxygen evolution kinetics in sealed rechargeable NiMH batteries.
Electrochimica Acta 53, 5836–5847 (2008). https://doi.org/10.1016/j.electacta.2008.03.023

33. Shinyama, K., Nakamura, H., Nohma, T. & Yonezu, I. Improvement of high- and low-temperature

characteristics of nickel-metal hydride secondary batteries using rare-earth compounds. Journal of
Alloys and Compounds 408–412, 288–293 (2006). https://doi.org/10.1016/j.jallcom.2005.04.111

34. Watanabe, K.-i., Koseki, M. & Kumagai, N. Effect of cobalt addition to nickel hydroxide as a positive

material for rechargeable alkaline batteries. Journal of Power Sources 58, 23–28 (1996).
https://doi.org/10.1016/0378-7753(95)02272-4

35. Bui, M. _et al._ Carbon capture and storage (CCS): the way forward. Energy & Environmental Science 11,
1062–1176 (2018). https://doi.org/10.1039/c7ee02342a

36. Hancke, R., Holm, T. & Ulleberg, Ø. The case for high-pressure PEM water electrolysis. Energy

Conversion and Management 261 (2022). https://doi.org/10.1016/j.enconman.2022.115642

37. Hancke, R., Bujlo, P., Holm, T. & Ulleberg, Ø. High-pressure PEM water electrolyser performance up to
180 bar differential pressure. Journal of Power Sources 601 (2024).

https://doi.org/10.1016/j.jpowsour.2024.234271

38. Abbasi, R. _et al._ A Roadmap to Low-Cost Hydrogen with Hydroxide Exchange Membrane Electrolyzers.
Adv Mater 31, e1805876 (2019). https://doi.org/10.1002/adma.201805876

39. Rubin, E. S., Azevedo, I. M. L., Jaramillo, P. & Yeh, S. A review of learning rates for electricity supply

technologies. Energy Policy 86, 198–218 (2015). https://doi.org/10.1016/j.enpol.2015.06.011

40. Gerke, B. F. N., Allison T.; Fisseha, Kibret S.. Recent price trends and learning curves for household LED
lamps from a regression analysis of internet retail data. (Lawrence Berkeley National Laboratory,

2015).

41. Schmidt, O., Hawkes, A., Gambhir, A. & Staffell, I. The future cost of electrical energy storage based on

experience rates. Nature Energy 2 (2017). https://doi.org/10.1038/nenergy.2017.110

42. Orangi, S. _et al._ Historical and prospective lithium-ion battery cost trajectories from a bottom-up
production modeling perspective. Journal of Energy Storage 76 (2024).

https://doi.org/10.1016/j.est.2023.109800

43. Mathias, M. F. _et al._ Two Fuel Cell Cars In Every Garage? The Electrochemical Society Interface 14, 24–
35 (2005). https://doi.org/10.1149/2.F05053if

## Figures

Figure 1

HEMCC device schematic and scales from 200 gCO2·yr -1 to 200 tonCO2·yr -1 a. A laboratory scale MEA

schematic shown with two porous Ni(OH)2 electrodes separated by a hydroxide exchange membrane.

Flow fields provide a path for air and a purified CO2 outlet stream. During cycling the cathode produces

OH - and the anode consumes OH - generating a pH gradient. A cycle stops when the cathode NiOOH is
consumed. During the cycling step a competing oxygen evolution reaction (OER) occurs at the anode

which prevents fully charging the anode. A regeneration step is performed after the cathode NiOOH is fully
consumed, where the cathode performs oxygen reduction reaction (ORR) to fully charge the anode. The

capture and release of CO2 is based on an acid-base equilibrium driven by the pH gradient. b. A laboratory

scale 25 cm 2 membrane electrode assembly (MEA). c. A stack of 9-300 cm 2 cells in pilot operation. d.
Commercial RepAir® module under construction.

Figure 2

Laboratory scale 25 cm 2 HEMCC and its behavior. The HEMCC is comprised of two identical
electrochemically precipitated Ni(OH)2 electrodes and an 80μm Piperion® membrane. The cathode air

flow was 1 L∙min -1 with 400 ppm CO2 and 90% RH. The anode received a sweep gas of 90% RH N2 at 1

L∙min -1 . a. The cell voltage (V) and b. the cathode outlet CO2 concentration (ppm) are used to calculate

the key process variables c-f during the charge/discharge step: c. capacity, d. electron efficiency, e. energy,
and f. flux, all excluding the regeneration step. Regeneration steps are excluded to highlight the low energy

cost of the charge-discharge step. Capture still occurs during regeneration but at higher energy cost. The
first cycle on each side is a 1.5-hour regeneration hold for break-in followed by 10-75 minute cycles on

each electrode. The cycle length is chosen based on an ex-situ capacity test to minimize regeneration (SI

Discussion 7 ) . c. Capacity achieved an average of 1.89 mAh·cm -2 which was 85% of the ex-situ capacity

tests. d. Electron efficiency achieved 0.25. e. Energy cost averaged 0.76 MWh∙tonCO2

-1 while f. flux

averaged to 72 kgCO2∙m 2 ∙yr -1 . g. and h. show a single cycle to highlight the change in voltage and capture

in the charge/discharge and regeneration steps. i., j., and k. are graphic aids to describe the build ΔpH,
maldistributed ΔpH, and rebuild ΔpH sections in g. and h. The ΔpH between the peak pH at the cathode

and the CO2 release pH at the anode drives the CO2 capture rate in the cell, see SI Discussion 2. The CO2
release pH is the pH that achieves equilibrium between CO2 in the anode and the concentrated CO2 gas

product stream. The dotted red primary reaction coordinate lines are rough position where OH - is produced
in the cathode and consumed in the anode. i. The build ΔpH section is directly after the gas lines switch in
the cell. Redox reactions take place close to the membrane with the cathode near 100% state of charge

and the anode near 0% state of charge. The pH at the back of the cathode electrode is close to the pH of
the anode in the previous cycle. j. The maldistributed ΔpH section is at the end of a charge/discharge

step. This occurs because of the uneven mass loading of Ni(OH)2/NiOOH when making the electrode and

uneven distribution of resistance through the electrode. Near the end of the charge/discharge step, low
loading and low resistance volumes of the cathode will have fully been consumed, no longer producing

OH - . Surrounding higher loading and higher resistance volumes continue producing OH - but the average pH
across the electrode drops lowering the average driving force for capturing CO2. The anode in this section

is still charging and has not reached 100% state of charge because of competing OER. k. The rebuild ΔpH

section is the beginning of the regeneration after the cathode has reached 0% state of charge. The
cathode reaction coordinate is brought to the front towards the membrane where catalytic ORR reactions

are predominant. The full area of the electrode is used again, and the average pH increases to have a
higher rate of CO2 capture. The anode continues to charge towards 100% for the next cycle.

Figure 3

Durability on manufacturable 25cm 2 electrodes and performance on pilot stack. a-c. Three segments of a

>5000 hour durability test using a manufacturable 25cm 2 cell, operated using ambient air held at 32 °C and
82-87 % RH. Missing cycles in segment 2 and 3 contained software errors and were omitted. a. Energy

cost across three segments averaged 0.67, 0.53, 0.46 MWh·tonCO2

-1 for the respective segments. b. Flux

averaged 48, 60, 62 kgCO2·m -2 ·yr -1 for the respective segments. c. Cycle duration averaged 1.4, 2.3, 2.4

hours for the respective segments. Increased capacity is associated with corrosion process between the
first and second segments. Flux and energy cost improved over time as the capacity increased. d and e.

Performance data of a 9 x 300-cm 2 cell stack operated in ambient conditions between 20-25 °C and 75
90% RH with d. average energy cost of 0.83 MWh·tonCO2-1 and e. average flux of 75 kgCO2·m-2·yr-1.

Figure 4

Levelized cost projection for Ni(OH)2 based HEMCC The Gen-0 pilot scale is in production with a

calculated cost of 566 $·tonCO2

-1. To get from Gen-0 to Gen-1, research developments in the HEMCC

device are expected to follow learning curves like solar and battery industries to reduce cost.

Advancements from Gen-1 to Gen-2 are expected to mostly derive from improvements in
manufacturability and energy consumption. From Gen-2 to Gen-3 First-Of-A-Kind (FOAK) economies of

scale will provide the greatest benefit. To achieve Gen-3 Nth-Of-A-Kind systems, centralized control rooms,
digital twins, and AI-assisted maintenance are expected to drive the final costs.

## Supplementary Files

This is a list of supplementary files associated with this preprint. Click to download.

[SIHydroxideExchangeMembraneCarbnCaptureHEMCCusingsymmetricalnickelhydroxidebatteries.docx](https://assets-eu.researchsquare.com/files/rs-5627423/v1/5c17ba661108d6d5b91d68a6.docx)

본문 출처 PDF: PDF 보기 · 다른 PDF는 위 [PDF 업로드]로 올릴 수 있습니다.

활용가능성

이력 0건

아직 편집 이력이 없습니다. 저장할 때마다 이전 내용이 여기에 백업됩니다.