Received: 25 May 2023 Revised: 27 July 2023 Accepted: 31 July 2023

DOI: 10.1002/sus2.154

REVIEW

Screening, preparation, and prototyping of metal–organic

frameworks for adsorptive carbon capture under humid
conditions

Long Jiang1 Jinyuan Yong1 Renyu Xie1 Pengfei Xie2 Xuejun Zhang1
Zhijie Chen3 Zongbi Bao2

1 KeyLaboratory of Refrigeration and

Cryogenic Technology of Zhejiang Abstract
Province, Institute of Refrigeration and Adsorption-based carbon capture has been recognized as an attractive method
Cryogenics, Zhejiang University,
for mitigating global warming. Metal–organic frameworks (MOFs) are promis-
Hangzhou, Zhejiang, China
2 College
of Chemical and Biological
ing candidate adsorbents for this purpose due to their high adsorption uptake and
Engineering, Zhejiang University, selectivity for carbon dioxide. However, in real-world applications, such as direct
Hangzhou, Zhejiang, China air capture, the presence of moisture in the feed gas may pose a grand challenge
3 StoddartInstitute of Molecular Science,
for CO2 adsorption in MOFs. This paper aims to address the issue of water–CO2
Department of Chemistry, Zhejiang
University, Hangzhou, Zhejiang, China co-adsorption in MOFs and present screening criteria for selecting MOFs that
preferentially adsorb CO2 under humid conditions. First, we uncover a compre-
Correspondence
hensive overview of CO2 –water co-adsorption characteristics of various MOFs.
Zongbi Bao and Pengfei Xie, College of
Chemical and Biological Engineering, Then, the high-throughput screening methods are summarized. Both computa-
Zhejiang University, Hangzhou, Zhejiang tional and experimental efforts have been dedicated to identify the promising
310027, China.
Email: [email protected];
MOFs for humid CO2 capture. According to the screening results and adsorp-
[email protected] tion mechanism, the optimal preparation strategies are proposed to modulate the
effect of water on CO2 uptake in MOFs. Finally, current MOF-based CO2 capture
Funding information
National Key Research and Development prototypes are presented to evaluate their practical feasibility and performance.
Program of China, Grant/Award This work could offer valuable guidance for the development and application of
Numbers: 2022YFB4101700,
MOFs for CO2 capture in the presence of water and inspire further research in
2022YFE0128600; National Natural
Science Foundation of China, this field.
Grant/Award Numbers: 52276022,
22278365, 22225802; Basic Research Funds KEYWORDS
for the Central Government ‘Innovative carbon capture, CO2 adsorption, high-throughput screening, MOFs
Team of Zhejiang University’,
Grant/Award Number: 2022FZZX01-09

1 INTRODUCTION 400 parts per million, although the COVID-19 pandemic

has caused a drastic reduction in greenhouse gas emis-
Carbon dioxide concentration continuously refreshes the sions in recent years.1,2 The global surface temperature in
highest ever average annual level in the atmosphere, over 2021 increased by 1.12◦ C relative to 1880–1920 average value

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the
original work is properly cited.
© 2023 The Authors. SusMat published by Sichuan University and John Wiley & Sons Australia, Ltd.

SusMat. 2023;1–30. wileyonlinelibrary.com/journal/susmat 1

2 JIANG et al.

in the Goddard Institute for Space Studies analysis and tial of MOFs to adsorb CO2 in the aqueous environment,
constantly reached new records.3 Carbon dioxide removal there is still a lack of full understanding about adsorption
(CDR) method plays an important role in mitigating global analysis, mechanism clarification, and the optimal design
warming. of MOFs in such a hot research field. Therefore, further
Carbon capture and storage is regarded as a promis- research is required to fill these gaps and provide efficient
ing CDR approach to reduce CO2 concentrations from solutions to mitigating CO2 emissions in the presence of
large industrial plants or discrete emission sources of water.
atmosphere. It can be achieved by various methods, for In order to describe the modification strategies for
example, absorption, adsorption, cryogenics, membranes, enhancing CO2 adsorption of MOFs in a moist environ-
and microbial/algal systems.4 Adsorption using solid sor- ment thoroughly and systematically, it is desirable to seek
bents takes the advantage of less regeneration energy over high-throughput screening methods and analyze the char-
absorption via liquid sorbents.5 The energy penalty dra- acteristics of CO2 adsorption in the presence of water. The
matically decreases by reducing evaporation heat of water common traits of the screened MOFs would provide the
and the sensible and latent heat required for the desorption insights for developing MOFs for the target applications.
process. Metal–organic frameworks (MOFs) is an emerg- This review aims to conduct a comprehensive discussion
ing and favorable branch of solid sorbents among these on CO2 adsorption of MOFs under the humid conditions,
porous sorbents. It is a network of solids that are com- and the concerning general schematic is shown in Figure 1.
posed of metal ions or metal cluster vertices and organic First, the high-throughput screening of MOFs for CO2
linkers.6,7 Due to the modular and crystalline charac- adsorption is revealed in Section 2, which includes prelimi-
teristics, pore sizes, pore shapes, and surface chemical nary and further processes. Second, modification strategies
potentials of MOFs can be easily adjusted for CO2 cap- for the improvement of capture performance based on
ture through the modifications of their metal clusters and adsorption mechanisms are summarized in terms of phys-
organic ligands. Due to the bridged/intertwined porous ical and chemical factors in Section 3. Finally, lab-scale
structures with a high surface area, MOFs are well-defined carbon capture prototypes using MOFs are presented and
hosts for gas molecules.8,9 The extraordinary adsorption verified with regard to CO2 –water co-adsorption processes
capacity and selectivity, especially at high pressure, could in Section 4. In real CDR plants, more water-related issues
then be obtained, which greatly correlates with adsorp- are involved. Based on research achievements, future
tion heat.10–14 In addition, some MOFs are desirable under challenges and perspectives of MOFs for CO2 –water co-
low CO2 concentrations and are emerging as the right adsorption are revealed in the end. This work would
adsorbents for direct air capture (DAC).15,16 However, other contribute to better CO2 adsorption performance in wet
component gases in gas source may have an impact on gases by elaborating contemporary exemplary MOFs in
their adsorption characteristics of CO2 , even the stability terms of adsorption properties and mechanisms.
of original structures.
Water is a ubiquitous component in air and most large
industrial CO2 sources and cannot be completely elim- 2 SCREENING OF MOFS FOR CO2
inated through a pretreatment stage. MOFs with open ADSORPTION
metal sites (OMS), also known as coordinatively unsat-
urated sites (CUS), prefer to bind with water over CO2 CO2 adsorption by MOFs is based on van der Waals
due to the permanent dipole of H2 O molecules.17–19 Due forces, chemical bonds, or Coulombic forces from atoms
to these hydrophilic structures with polar bonds, CO2 or ions of sorbents. The driving force is determined by
adsorption capacities have a dramatic drop.20 It has been the surface properties of the adsorbates/adsorbents and
demonstrated in CO2 /H2 O mixtures for Mg-MOF-74 via environmental factors, for example, working temperature
Grand Canonical Monte Carlo (GCMC) and/or density and pressure. However, the presence of other impurities in
functional theory (DFT) simulation.21–23 Some exceptions feed gas may change adsorption sites or adsorption affin-
are described as follows. For MIL-101(Cr),24 ZIF-8,25 ZIF- ity, thereby having synergistic or competitive effects on
Cl,26 SALEM-2,26 SGU-29,27 etc., a given amount of water CO2 adsorption. Table 1 lists the major gas constituents
might only have a negligible impact on CO2 adsorp- from the post-combustion process, the pre-combustion
tion. Moreover, moisture-enhanced CO2 adsorption can process, and the atmosphere. This indicates that CO2 /N2
be observed in HKUST-1,28 MIL-100(Fe),29 InOF-1,30 UiO- adsorption selectivity is generally higher due to the lower
66,31 PCN-250(Fe3 ), and PCN-250(Fe2 Co).32 Therefore, quadrupole moment of N2 , and its kinetic diameter is large
based on the properties of sorbents and particular moisture enough to separate by tailoring pore size. But water has
contents, CO2 adsorption of MOFs performs differently a smaller kinetic diameter and a higher dipole moment
under humid conditions. Despite the remarkable poten- when compared to other two gases. Therefore, pores in the
JIANG et al. 3

F I G U R E 1 The general schematic of the

main parts: high-throughput screening of
metal–organic frameworks (MOFs),
modification strategies to enhance CO2
adsorption in the presence of water and the
lab-scale prototype demonstration.

T A B L E 1 Physical parameters of major gas compositions under the condition of post-combustion, pre-combustion, and the
atmosphere.35–37
Post- Pre- Dipole Quadrupole
combustion combustiona DAC Kinetic moment moment Polarizability
Composition (vol%) (vol%) (vol%) diameter (Å) (D) (1040 cm2 ) (1024 cm3 )
CO2 15–16 35.5 0.038 3.30 0.00 −14.31 ± 0.74 2.91
H2 O 5–7 0.2 0.1–3.0 2.65 1.8546 ∼0.0 1.45
N2 70–75 0.25 78.08 3.64 0.00 −4.65 ± 0.08 1.74
Abbreviation: DAC, direct air capture.
a
After gas shift reaction.

adsorbent have a stronger affinity toward water. Due to the The CO2 uptake of various MOFs, as well as molecu-
smaller kinetic diameter of water, “molecular sieve effect” lar sieve, carbon, and zeolite, is compiled under different
by decreasing pore size is invalid to improve CO2 /H2 O humidities, as shown in Figure 2. The larger symbols
selectivity. Meanwhile, water molecules linked to CUS are not limited to areas with low adsorption amounts.
reduce binding energy between CO2 and framework of Some types of MOFs have the better adsorption perfor-
adsorbents.21 Water adsorption hinders CO2 capacity and mance under high humidity condition. They consist of
selectivity of gas mixtures in most MOFs, especially when both physical adsorbents, such as dptz-CuTiF6 ,38 PCN-
humidity is high.33,34 Yu and Balbuena21 investigated the 250(Fe2 Co),32 and CO2 (dobdc),21 and chemical adsorbents
effect of water on CO2 adsorption of Mg-MOF-74. As con- such as dmpn-Mg2 (dobpdc)39 and mmen-Mg2 (dobpdc).40
firmed by molecular simulation, coordinated water abates In contrast, the hydrophilic nature of zeolite 13X renders
CO2 interaction with framework of adsorbents, and CO2 it essentially incapable of absorbing CO2 in the presence
uptake decreases as water content increases. It is still chal- of water. Mg-MOF-74, MIL-96(Al), and other compounds
lenging to achieve selective CO2 adsorption under humid exhibit similar reductions in CO2 adsorption uptake. More
conditions. attention should be paid to adsorbents that exhibit a slight
4 JIANG et al.

MOFs
5 Co2(dobdc) ndmpn-Mg2(dobpdc)

Amount adsorbed (mol/kg)

0.2 Graph2
1

Amount adsorbed (mol· kg-1)

1%RH dmpn-Mg2(dobpdc) Ni2(dobdc)
100%RH 0.1
4 dptz-CuSiF6 Ni-MOF-74
dptz-CuTiF6 NOTT-400
0.0
0 200 400 600 800
PCN-250(Fe2Co)
3 CO2 concentration (ppm) HKUST-1
InOF-1 PCN-250(Fe 3)
Mg2(dobdc) SIFSIX-3-Ni
2 Mg-MOF-74 SIFSIX-3-Zn
MIL-100(Fe) UiO-66
MIL-101(Cr) ZIF-8
1 Zn2(dobdc)
MIL-96(Al)
mmen-CuBTTri Zn-MOF-74
0 mmen-Mg2(dobpdc) Others
mmen-Ni2(dobpdc) 1°-MCM-41
AX-21 carbon
0.00 0.10 0.12 0.14 0.16 0.18 0.20 0.22 MOF-5
PEI-MCM-41
MOF-808-Gly
CO2 partial pressure (bar) zeolite 13X
MOOFOUR-1-Ni

F I G U R E 2 A summary of CO2 adsorption capacity of metal–organic framework (MOF) materials and several other types of adsorbents
under different humidity conditions. The humidity is shown by the symbol’s size. The adsorption temperature is in the range of
298–313 K.21,24,25,29,32,33,38–48

reduction in CO2 uptake under the conditions of an ele- First, its screening methods and performance indicators
vated humidity or even perform better when exposed to are discussed. The common evaluation indicators are
water molecules. It is also critical to identify the mecha- shown in Table 2. Among these indicators, Henry con-
nism responsible for the abnormal rise in CO2 uptake. stant (KH ), Henry selectivity (SH ), and adsorption enthalpy
CO2 –water co-adsorption characteristics vary greatly (∆Hads ) are used for the preliminary screening. The work-
among different materials. Thus, it is essential to iden- ing capacity (∆N) and regenerability (R%) are the basic
tify preferable MOFs under humid conditions. High- indicators that only focus on CO2 adsorption of materials,
throughput screening via molecular simulation is an regardless of other components. The additional parame-
effective method for screening MOFs with high potential. ters include selectivity (S), percent uptake loss (PUL ), and
The evaluation criteria and material screening results are separation potential (∆Q), which are related to the per-
expounded. formance of multicomponent adsorbates’ adsorption or
separation. The adsorbent performance score (APS) is a
comprehensive evaluation indicator.
2.1 High-throughput screening
indicators
2.1.1 Indicators of preliminary screening
A large number of MOFs with the adjustable pores and
functional groups have been discovered. In the Cam- Although GCMC is an ideal method, the computational
bridge Structural Database, crystallographic information cost of the simulation for adsorption with the participa-
files of over 10 000 non-disordered MOFs are indexed.49 tion of water molecules will significantly increase because
It is convinced that variable pore sizes, functional group acceptance probabilities of insertion and deletion of water
diversities, and structural differences all have a significant molecules are very low. In addition, the accuracy of simu-
influence on CO2 adsorption properties.50 In general, the lation strongly depends on H2 O model. The TIP4P model
experimental conditions are challenging to complete large- represents each water molecule with four atoms, which
scale material screening, especially when water effect is widely used.51 Zhao et al.52 performed 1 × 107 work-
needs to be counted. Thus, with the rapid advancement ing cycles on water adsorption, but others only conducted
of computer technology, more researchers have conducted 5 × 105 cycles. With regard to computer performance and
molecular simulations to screen ideal CO2 adsorption screening efficiency, more convenient indicators should
materials under wet conditions using a high-throughput be adopted for the preliminary screening instead of wet
method that is validated by the experimental results. adsorption isotherms of materials.
High-throughput screening studies on CO2 adsorption The Henry constant is the most commonly used screen-
materials are summarized by considering water effect. ing index for the preliminary high-throughput screening
JIANG et al. 5

TA B L E 2 Evaluation indicators for material screening.

Evaluation indicators Formula Components
1 ⟨𝑊⟩
Henry’s constanta 𝐾H = (a) CO2 , H2 O
RT𝜌f ⟨𝑊 IG ⟩
( ( ) )
ln 𝐾H
Δid 𝐻 = 𝑅𝑇 2 𝜕 (b)
𝜕𝑇 𝑛

𝐾H,CO2
Henry selectivitya 𝑆H = CO2 , H2 O
𝐾H,H2 O
( )
𝜕𝑈
Adsorption enthalpya Δ𝐻ads = − ⟨𝑈g ⟩ − RT H2 O
𝜕⟨𝑁⟩ 𝑉,𝑇

Adsorbent performance scoreb APS = 𝑆 × Δ𝑁 CO2

b 𝑦𝐻2 O
Separation potential Δ𝑄m,H2 O = 𝑁ads,CO2 × − 𝑁ads,H2 O CO2 /N2 , CO2 /CH4 , CO2 /H2 O
𝑦CO2

𝑁dry −𝑁wet
Percent uptake lossb 𝑃UL = × 100% CO2
𝑁dry

𝑛CO2 ∕𝑛𝐻2 O
Selectivityb 𝑆CO2 ,H2 O = CO2 /N2 , CO2 /CH4 , CO2 /H2 O
𝑝CO2 ∕𝑝𝐻2 O

Working capacityb Δ𝑁 = 𝑁ads − 𝑁des CO2

Δ𝑁
Regenerabilityb 𝑅% = × 100% CO2
𝑁ads

a
Evaluation indicator usually used in preliminary screening.
b
Evaluation indicator usually used in further screening.

100 bulk phase concentration), but it is not suitable for high-

throughput screening. Besides, KH can also be calculated
by Widom insertion method53 or deduced from the infinite
80 dilution enthalpy,54 which could correspond to Formulas
CO2 capacity loss (wt%)

(a) and (b) in Table 2.

60 In the actual screening process, SH for CO2 to H2 O
MIL-53(Al) is often used as an indicator of the preliminary screen-
MIL-127(Fe)
ing. Li et al.56 screened out top 15 structures with the
40 largest SH from CoRE MOF database, which has 5109
MOF structures.57 Figure 4A shows that SH is greater than
MIL-96(Al) that calculated by GCMC simulation in most cases. A
20
reasonable explanation is that KH only considers adsorp-
tion properties at low pressure but ignores the changes at
0 high pressure. It is obvious that interaction between adsor-
0 2 4 6 8 10 12 14
bate molecules increases at high pressure, which is almost
lnKH (H2O) ignored at low pressure. Especially for water molecules, a
hydrogen bonding network formed at high pressure greatly
F I G U R E 3 The relationship between CO2 capacity loss and the enhances the interaction between water molecules. Fur-
KH of H2 O. Reproduced with permission.55 Copyright 2017, Wiley.
thermore, their team also investigated the influence of
different charge assignment methods on KH , as shown in
of the materials for CO2 adsorption when considering the Figure 4B. KH for CO2 and N2 are slightly influenced by the
influence of water. The relationship between PUL and KH method of charge calculation. Particularly in the case of a
of water is indicated in Figure 3, where each point rep- large KH of H2 O, the extended charge equilibration (EQeq)
resents a kind of MOF. Except for a few MOFs, the loss method tends to underestimate its value, which also leads
of CO2 capacity of most types shows a positive linear cor- to a larger SH than GCMC selectivity.
relation with the logarithm of 𝐾H,H2O . Several common Besides, Li et al.58 studied the effect of charge assign-
calculation methods can be used to obtain KH of water. One ment methods on KH . Except for EQeq method, the
method is to calculate the slope of the straight line region density-derived electrostatic and chemical methods are
of adsorption isotherm (ratio of gas adsorption capacity to adopted to assign the partial charges. A similar conclusion
6 JIANG et al.

(A) (B)

F I G U R E 4 (A) Comparison between CO2 selectivity calculated by KH and Grand Canonical Monte Carlo (GCMC) simulation. (B)
Comparison between KH calculated by extended charge equilibration (EQeq) method and REPEAT method. Reproduced with permission.56
Copyright 2016, ACS.

can be yield as follows. There is little difference between

KH of CO2 and N2 calculated by different charge meth-
ods, but a great difference between the KH of H2 O can be
found. To explain the difference, van der Waals force and
Coulomb force of three molecules are calculated, as shown
in Figure 5. Compared with CO2 and N2 , energy composi-
tion of water adsorption is significantly different, in which
the Coulomb force has a large proportion. This explains
why charge assignment method has a large influence on
the estimation of KH of H2 O. In addition to KH , ∆Hads can
also be used for the preliminary screening, which partially
reflects the strength of the interaction between adsorbate
and adsorbent. An effective screening strategy is to discard
structures, in which the enthalpy of water adsorption is
greater than evaporation heat.56

2.1.2 Indicators of further screening

After a preliminary screening of desirable adsorbents for

CO2 adsorption under humid conditions based on the
parameters of KH and ∆Hads , more accurate charge assign-
ment methods will be used. Also more detailed GCMC
simulations involving the multicomponent adsorption
isotherm and higher adsorption pressure are considered
to further explore performance differences and identify F I G U R E 5 Adsorption reaction enthalpy of (A) CO2 , (B) H2 O,
the top-performing materials. Various indicators can be and (C) N2 within the selected metal–organic frameworks (MOFs)
computed and analyzed to identify adsorption perfor- from density-derived electrostatic and chemical (DDEC)-based
mance. A comparison of the adsorbate-host interaction is screening. Reproduced with permission.58 Copyright 2017, Wiley.
required to achieve high adsorption capacity and selectiv-
ity. Figure 6A–C shows S, ∆N, APS, and R% of CO2 of the
selected MOFs from the mixed CO2 /N2 and CO2 /N2 /H2 O. four grades: A (less than 25%), B (25%–50%), C (50%–75%),
This indicates that water exerts a non-negligible impact on and D (more than 75%).
various performance indicators.59 Figure 6D compares PUL For real capture process, it is necessary to pursue a
of CO2 in the presence of water, which can be divided into high ΔN𝐶𝑂2 while minimizing adsorption uptake of other
JIANG et al. 7

(A) (D)
MIL-53(Fe) Activated sample
Active carbon Norit
Wet sample
ZIF-93
Active carbon Takeda
ScBDC
MIL-121(Fe)
UiO-66(Zr)-2Et
MIL-69(Al)
ScBDC-NO2
ZIF-8 BASF
Sc-BDC-NH2 A
ZIF-8(2)
UiO-66(Zr)-CF3
MIL-103(La)
ScBDC-Br
MIL-121(Al)
MIL-53(Cr)-Cl
ZIF-94
MIL-140A(Zr)-Br
(B) UiO-66(Zr)-2Me
MIL-163(Zr)
MIL-53(Fe)-Br
MIL-140A(Zr)
UiO-66(Zr)-Me
MIL-53(Cr)
UiO-66(Zr)-2CF3
UiO-66(Zr)
MIL-53(Cr)-Br
MIL-100(Cr)
MIL-140C(Zr)

HKUST-1
MIL-101(Cr)
UiO-66(Zr)-NH2
B
MIL-96(Al)
MIL-53(Fe)-Cl
MIL-103(Eu)
MIL-110(Al)
MIL-127(Fe)
MIL-53(Al)
(C) 100 MIL-100(Cr)-Me

MIL-53(Al)-NH2
C
R% for CO2/CH4 mixture

MIL-130(Al)
in H2O loaded MOFs

MIL-68(Ga)
90 MIL-102(Cr)
MIL-100(Fe)
AFEHUO
CIFCEB
MIL-91(Ti)
D
80 Zeolite chabazite-Cs
FIRMUQ
CPO-27(Ni)
KOSLUB
Zeolite 13X
SUTBIT Mg(H2gal)
WONZIJ
70 0 5 10 15 20 25
70 80 90 100 Amount of CO2 adsorbed (wt%)
R% for CO2/CH4 mixture

F I G U R E 6 Comparison of (A) S𝐶𝑂2 ∕𝑁2 , (B) ΔN𝐶𝑂2 (mol kg−1 ) and APS (mol kg−1 ), and (C) R% of metal–organic frameworks (MOFs)
calculated for separation of the mixed CO2 /N2 and CO2 /N2 /H2 O. Reproduced with permission.59 Copyright 2017, ACS. (D) Summary of the
results obtained for the 43 MOFs and four reference materials tested. Reproduced with permission.55 Copyright 2017, ACS.

gases. Thus, S is an important indicator, for example, optimization algorithm to obtain the optimal adsorption
between CO2 /N2 , CO2 /CH4 , and CO2 /H2 O. Because sim- process setting.60
ulation processes are often simplified to binary or ternary
systems, S and ∆Q are significant for CO2 and another gas
component. However, it is also possible to include more 2.2 High-throughput screening outputs
gas components in the practical application and calculate
S between CO2 and other remaining components, which After considering the indicators for the preliminary and
is similar to the purity calculation. Since the adsorbed further screening, some MOFs appear to be the candidates
CO2 is not completely desorbed, calculating N and R% is for carbon capture under humid conditions. Table 3 sum-
preferable to comparing absolute adsorption capacity. marizes researches on high-throughput screening of MOFs
Besides, the multi-objective optimization arises in terms for CO2 adsorption in the presence of water. It reveals
of various measurement indicators and working con- that the computational screening is more rapid and effi-
ditions. A simple approach is to use a comprehensive cient to select the potential adsorbents, which has recently
evaluation indicator such as APS or a multi-objective become a hot research area. Some insights into the optimal
8 JIANG et al.

TA B L E 3 High-throughput screening studies of metal–organic frameworks (MOFs) for CO2 capture from the wet gases.
Screening
Database method Screening result Adsorbent feature Ref.
45 MOFs Experimental 4 zones with different effects ∙ Pores with different sizes or chemistry with 55
of adsorbed H2 O on CO2 “bifunctionality.”
uptake ∙ Adsorbed H2 O attract CO2 via
diploe–quadrupole interaction.
CoRE MOF Computational 15 MOF structures ∙ High van der Waals interaction between CO2 56
and MOFs.
∙ Small pore size provides strong interaction
with CO2 and prevents water cluster
formation.
CoRE MOF Computational 15 MOF structures ∙ Low surface area (<1000 m2 g−1 ). 58
∙ Small pore size (∼5 Å).
CoRE MOF Computational GaPO and AlPO zeolites ∙ The adsorption heat of CO2 is higher than 61
and IZAa H2 O at zero loading.
∙ The inter-adsorbates interaction of H2 O is
stronger than CO2 .
325 000 Computational MOFs with adsorbaphore of ∙ The parallel aromatic rings provide the 62
hypothetical type A1 favorable interaction with CO2 .
MOFs ∙ This adsorbaphore hinders hydrogen bonding
formation.
CoRE MOF Computational 13 MOF structures ∙ Generally, contain Zn ions metallic centers. 63
∙ Generally, contain alkaline nitrogen groups.
MOF-74-Mb Computational Noble metals in MOF-74 ∙ The redox characteristics of noble metals 64
enhance the interaction with CO2 and
repulse H2 O adsorption.
a
IZA: International Zeolite Association.
b
M: Be, Mg, Al, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Sr, Zr, Nb, Ru, Rh, Pd, La, W, Os, Ir, Pt.

selection or design of adsorbents are revealed by the ture. Similarly, Rogacka et al.63 revealed 13 promising MOF
existing screening results. The features of MOFs can be structures with high CO2 /CH4 selectivity under ambient
classified into two types according to their top-performing conditions. Nine of the screened MOFs have ligands with
adsorption characteristics: (i) high adsorption interaction alkaline nitrogen groups that have a stronger chemical
with CO2 provided by adsorbents with superb chemical interaction with CO2 . According to Canepa’s research,64
structure or co-adsorption adsorbates and (ii) the con- MOF-74 with the noble metals Rh, Pd, Os, Ir, and Pt
finement effect arising from appropriate pore sizes of competes with water due to the redox nature of noble
adsorbents. metals. Boyd et al.62 identified three types of effective
Interaction strength between adsorbates and adsorbents CO2 adsorption sites in the screening process of MOFs
can be reflected by reaction heat. Findley and Sholl61 used for wet CO2 capture, inspired by traditional drug molec-
adsorption heat to screen MOFs and zeolites using Widom ular design strategies.65,66 They are parallel aromatic rings
insertion Monte Carlo simulations. It is demonstrated that (A1), metal–oxygen bridges (A2), and OMS (A3). As shown
almost all MOFs from the 2014 CoRE MOF database are in Figure 7A,B, those materials with parallel aromatic
deemed to unsuitable for DAC in the presence of water due rings have a high 𝐾H,CO2 and a low 𝐾H,H2 O . Later, they
to the much higher adsorption heat of H2 O under humid further studied MOFs with frz topology, which was charac-
atmospheric conditions. This situation can be improved terized by tetra-carboxylated organic ligands coordinated
in the application of higher CO2 concentrations such as to one-dimensional metal–oxygen rods and contained
post-combustion carbon capture. Li et al.56 screened 15 the preferred adsorption sites.62 By choosing appropri-
MOFs from the CoRE MOF databases with high CO2 /H2 O ate metal element and parallel aromatic ring space, one
and CO2 /N2 selectivity, which are primarily due to strong highly promising MOF material, Al2 (OH)2 (H2 TCPP) (Al-
van der Waals interactions between CO2 and adsorbent PMOF) H2 TCPP, meso-tetra(4-carboxyphenyl)porphyrin),
at 80% relative humidity (RH) for post-combustion cap- was designed and experimentally validated for its good
JIANG et al. 9

F I G U R E 7 (A) KH for H2 O against those of CO2 . The colors represent three different adsorbaphores in the top-performing structures:
A1, parallel aromatic rings; A2, metal–oxygen bridges; A3, open metal sites. Some materials have both A1 and A2 sites. (B) Adsorbaphore
containing parallel aromatic rings (A1) (adapted from Ref. 62). (C) CO2 /H2 O selectivity based on the ratio of Henry’s constants against the
largest cavity diameter (LCD). Data points are colored based on accessible surface area (ASA) and (D) void fraction (VF). Reproduced with
permission.58 Copyright 2017, Wiley.

hydrophobicity, water stability, and high CO2 adsorp- screening process, researchers investigated common char-
tion uptake. Other studies have reported that MOFs with acteristics of adsorbents such as pore size, surface area,
different proprietary adsorption sites can achieve better pore volume, and chemical functionality, and designed
CO2 –water co-adsorption, which is conducive to solve materials based on these findings.67 However, few studies
competitive adsorption issues.55 of carbon capture have focused on the effect of water based
The majority of small pores have a confinement effect on on the high-throughput screening. The related researches
CO2 adsorption and impede water cluster formation.56,58,62 have mainly concentrated on MOFs, while other materi-
Li et al.58 summarized characteristics of materials suit- als, for example, zeolites and molecular sieves are rarely
able for CO2 adsorption under wet conditions through reported.68–70 Thus, relevant databases of other sorbents
screening methods. The results show that the Henry selec- need to be extend.
tivity of CO2 /H2 O decreases as the largest cavity diameter The results of a high-throughput screening for CO2
(LCD) increases. Materials with the lower accessible sur- adsorption in the presence of water show that prefer-
face area and void fraction usually have a higher selectivity, ential CO2 adsorption is related to adsorbent structures.
as shown in Figure 7C,D. Therefore, porous materials with CO2 /H2 O adsorption selectivity is not just a simple func-
smaller pore sizes may be better suited for CO2 adsorption tion of MOF structure properties, for example, pore size,
under wet working conditions, and these small pores can surface area, etc.63 The operation conditions for adsorp-
prevent the formation of hydrogen bond networks among tion processes, such as humidity, temperature, and partial
water molecules. pressure should also be taken into consideration. To
In summary, high-throughput screening performance understand the mechanism of MOFs with high CO2 /H2 O
indicators that are simple to calculate, such as KH , are fre- selectivity based on existing researches, a comprehen-
quently evaluated for the preliminary screening to save the sive review of optimization strategies for increasing CO2
computational resources, and then further simulations are adsorption in wet gases should be conducted, which could
performed to obtain more effective indicators. During the provide some insights for material development.
10 JIANG et al.

FIGURE 8 Explanation of adsorption mechanism by using a flow chart.

3 STRATEGIES FOR THE IMPROVED with the quadrupole moment of CO2 .55 The interaction
CO2 ADSORPTION changes depending on the magnitude and orientation
of the intermolecular force, resulting in different effects.
In order to obtain the materials with good adsorption The key to promoting CO2 adsorption is to erode water’s
performance through screening, the internal adsorption competitive advantages or increase synergistic effects. To
mechanisms should be further explored to guide material improve CO2 uptake in wet gases, some approaches based
synthesis. The mechanism is explained by a flow chart, on the affinity of CO2 and H2 O to adsorbents have been
as shown in Figure 8. Physisorption produced by inter- proposed: (i) increasing the hydrophobicity within MOFs
molecular interactions is favored by surface affinity and to decrease the polarity of adsorption sites; (ii) building the
confinement effect. However, chemisorption occurs only specific adsorption sites; and (iii) adsorbing an appropriate
through surface affinity due to the action of chemical amount of H2 O to append CO2 adsorption sites.
bonds. The confinement effect arises from the overlap of
the host–guest potential energy surface in cavities of adsor-
bent structure, which leads to CO2 adsorption in narrow 3.1.1 Hydrophobic modification
pores.71 Surface affinity is determined by the chemical
composition of the solid, which has no bearing on frame- There are two common methods for MOF modifica-
work topology. According to the existing research, the tion treatment, that is, pre-functionalization, also called
promoting effect on CO2 adsorption could occur in some direct synthesis and in situ synthesis and post-synthetic
sorbents via thermodynamic (surface affinity) or kinetic modification. To reduce water affinity in MOFs by pre-
(confinement effect) factors. These characteristics that functionalization, some hydrophobic groups, such as alkyl
promote CO2 adsorption under humid conditions can be groups, aryl groups, fluorinated alkyl groups, and aryl
classified by action mechanism as follows: (i) CO2 adsorp- groups could be introduced to protect weak coordination
tion sites with the suitable polarization interactions72–75 ; bonds from hydration.86–88 Yu et al.72 investigated a series
(ii) adsorption sites with strong chemical interactions with of UiO-66 derivatives decorated by alkyl or perfluoroalkyl
CO2 76–79 ; and (iii) confinement effect of channel structure groups bound to ligands of different lengths (terephthalate
or the coordinated H2 O molecules.28,38,80 Table 4 lists some or 4,4′-azobenzenedicarboxylate). This demonstrates that
advanced MOFs with high CO2 uptake under humid con- increasing the number of substituents increases hydropho-
ditions. Their excellent performance is evaluated by using bicity. As a result of fluorination, perfluoro-alkylated
adsorption mechanism. groups among the derivatives of UiO-66 exhibit more
hydrophobicity than alkyl groups. Although the polarity of
pore walls decreases, introduction of functional groups can
3.1 Polarization interaction also increase the confinement effect to promote adsorp-
tion. This effect is more apparent in UiO-66D, which
The binding forces for CO2 adsorption include van der possesses the large pores, as shown in Figure 9.
Waals forces, chemical bonds, or Coulombic forces. For Fluorine-functionalized MOFs have a low surface
physical adsorption, there is no need to consider chemical energy, which could be attributed to fluorine’s high elec-
bond formation. Adsorption enthalpy is only proportional tronegativity and relatively low electric polarizability. The
to intermolecular interactions, for example, van der Waals treatment of fluorinated building blocks can generate the
forces and Coulombic forces.85 Physical adsorption forces fabulous hydrophobic porous structures.73 MOFs contain-
are related to the polarizability of adsorbents and adsor- ing inorganic fluorinated monoanions achieve preferential
bates, which is an important fundamental fact for the CO2 adsorption over the interaction of F atoms and CO2 .
following content. The polarity of major gas compositions Noro et al.75 synthesized a flexible and hydrophobic cop-
can then be determined. Water with a higher polarizabil- per porous coordination polymer, which indicates a high
ity competes with CO2 for adsorption sites. Furthermore, selectivity for CO2 over CH4 and the high stability under
the dipole moment of water has an undeniable interaction humid conditions. The whole structure is composed of low
TA B L E 4 The structure, CO2 uptake, and features of advanced metal–organic frameworks (MOFs).
JIANG et al.

CO2 Adsorption
pressure temperature CO2 uptake
MOFs Metal Ligands Components (kPa) (K) (mol kg−1 ) Features Ref.
Mg-MOF-74-COOH Mg DHTP-COOH CO2 /H2 O (50/50 vol%) 50.5 298 51.1 (N u.c.−1 ) The carboxyl groups have weakly interacted with 23
water molecules.
MIL-101 (Cr) Cr BDC CO2 /N2 /H2 O 10 298 0.509 The increase electrostatic interactions between 24
(10/90 vol%/10% RH) coordinated water and quadrupole moment of
CO2 .
MIL-100 Fe BTC CO2 /N2 /H2 O 20 298 2.32 Water adsorption on the CUS leads to stronger 29
(20/80 vol%/40% RH) adsorption sites for CO2 and it would seem that
the CO2 displaces some water.
UiO-66 Zr BDC CO2 /pre-adsorbed H2 O 100 298 3.2 Pre-adsorbing water favors dipole–quadrupole 31
(1.8 wt%) interaction and molecule–framework
interaction toward CO2 .
PCN-250 (Fe3 ) Fe ABTC CO2 /N2 /H2 O 15 298 1.82 The plier effect of coordinated H2 O. 32
(15/85 vol%/50% RH)
PCN-250 (Fe2 Co) Fe, Co ABTC CO2 /N2 /H2 O 15 298 2.23 The plier effect of coordinated H2 O and the higher 32
(15/85 vol%/50% RH) ionic character of the Co–O bond forms
stronger interaction with CO2 .
Mg-CUK-1 Mg 2,4-PCD CO2 /H2 O (18% RH) 100 303 3.3 Favorable intermolecular interactions between 81
CO2 and H2 O.
MIL-53 Al BDC CO2 /H2 O (20% RH) 100 303 1.18 Confinement effect induced by H2 O and directing 82
effect of the hydroxo functional groups.
InOF-1 In BPTC CO2 /H2 O (20% RH) 100 303 2.5 Confinement effect induced by H2 O and directing 83
effect of the hydroxo functional groups.
NOTT400 Sc BPTC CO2 /H2 O (20% RH) 100 303 2.32 Confinement effect induced by H2 O. 45
[Cu(PF6)2 (BPP)2 ] Cu PF6 − , BPP CO2 /CH4 /H2 O 300 298 1.4 Low electric polarizability of fluorine. 75
(39.98/59.66/0.06 mol)
een– Mg2 (dobpdc)/Al–Si Mg een, H4dobpdc CO2 /He/H2 O 15 353 2.5 Diamine grafting and silane coating. 84
(15/81.25/3.7 vol%)
NOTT401 Sc TDA CO2 /H2 O (5% RH) 100 303 0.89 Confinement effect of H2 O and hydrogen bond 46
formation between H2 O and CO2 .
Note: ABTC: 3,3′,5,5′-azobenzenetetracarboxylate; BDC: 1,4-benzenedicarboxylate; BPTC: 3,3′,5,5′-biphenyltetracarboxylate; BPP: 1,3-bis(4-pyridyl)propane; BTC: benzene-1,3,5-tricarboxylate; DHTP:
2,5-dihydroxyterephthaliate; een: N-ethylethylenediamine; PCD: pyridinecarboxylate; TDA: thiophene-2,5-dicarboxylate.
Abbreviations: CUS, coordinatively unsaturated sites; RH, relative humidity.
11
12 JIANG et al.

F I G U R E 9 Water sorption isotherms measured at 298 K. (A) UiO-66 derivatives; (B) UiO-66D derivatives. UiO-66D represents a longer
UiO-66 analog, in which ligand is exchanged by 4,4′-azobenzenedicarboxylate linkers (adapted from Refs. 72, 89).

polar ligands, fluorinated PF6 − anions, BPP containing Besides, some MOF-derived porous materials are stable
two pyridine moieties and one cross-linking −(CH2 )3 − in wet gas. Kim et al.94 studied CO2 adsorption of porous
moiety without any strong Lewis acid and base sites. carbon materials zinc-containing MOFs. Due to the intrin-
Another viable method for post-synthetic exchange that sic hydrophobic feature of porous carbon and lack of metal
allows for simple modification of existing MOFs is coating sites, the breakthrough curves of CO2 and N2 are identical
the crystals with hydrophobic polymers. Qian et al.90 under dry and wet conditions. Also, some zeolitic imida-
fabricated an organosilicon coating on surface of various zolate frameworks (ZIFs) have been reported to have the
MOFs via facile solution-immersion process and obtained high thermal and chemical stability in refluxing organic
water-stable MOFs with a high contact angle. Choe et al.84 and aqueous media.95 Nguyen et al.96 synthesized three
synthesized Mg2 (dobpdc)–alumina composite microbeads hydrophobic ZIFs by incorporating two distinct imidazo-
(MOF/Al) through spray dry method, and then further late links. These materials exhibit remarkable hydrophobic
functionalized MOF/Al with een and coated with long properties and CO2 uptakes. Additionally, the series ZIF
alkyl chain silanes to obtain splendid MOFs with the exhibits significant CO2 separation performance over three
high carbon capture capacity under humid conditions. working cycles from a ternary mixture of N2 and water
As shown in Figure 10, een–MOF/Al–Si exhibits the without degradation of ZIF structure or performance. Hu
best stability and the lowest affinity toward water when et al.26 also reported a series of ZIFs with the same sodalite
compared to MOFs without a hydrophobic silane coating. (SOD) topology with different functional groups. The pres-
Nanochannels of MOFs can also be used as the desir- ence of water is found to have a negligible effect on
able spaces for polymerization, which belongs to in situ CO2 adsorption in terms of ZIF-8, ZIF-Cl, and SALEM-
synthetic modification.91 Ding et al.92 developed a new 2, which contain no or weakly polar groups. Therefore,
strategy to partition the channels of MOF-5 into the con- ZIFs are the potential candidates for the hydrophobic
fined compartments by in situ polymerization of aromatic treatment.
acetylene to improve the stability of the framework to
moisture, as shown in Figure 11A,B, respectively. Recently,
Yu’s group proposed a novel in situ microencapsulated syn- 3.1.2 Pores with the specific adsorption sites
thesis method to synthesize MOF-based hybrid sorbents.93
As shown in Figure 11C,D, three fluids flow through glass In most cases, water has an influence on CO2 adsorption
double-capillary devices: precursors of MOFs (inner fluid), due to the higher affinity toward adsorption sites. Recently,
hydrophobic ultraviolet photopolymerizable silicone (shell some researchers discovered MOFs with CO2 -specific
fluid), and an aqueous surfactant-containing carrier fluid, sites, such as amine groups, which perform exceptionally
which are used to form MOF-based hybrid adsorbents with well under humid conditions.74 There is less interaction
excellent hydrophobicity (outer fluid). It is demonstrated between the adsorption of gases in MOFs with distinct
that CO2 adsorption capacity of the encapsulated HKUST-1 adsorption sites due to the varied pore sizes or bifunc-
was slightly lower than that of HKUST-1, but the encapsu- tionality. Benoit et al.33 also found a similar principle
lated HKUST-1 showed the good recyclability and stability through the experiments and GCMC simulations for MIL-
after 10 working cycles. 96(Al) to investigate its adsorption behavior. Although CO2
JIANG et al. 13

(A) 5 (B) 25
140°C een-MOF
120°C een-MOF/Al
4 100°C 20 een-MOF/Al-Si

CO2 adsorbed (mmol· g-1)

H2O adsorbed (mmol· g-1)

80°C
60°C
3 15
40°C

2 10

1 5

0 0
0.01 0.1 1 10 100 1000 0.0 0.2 0.4 0.6 0.8
Pressure (mbar) P·P-1
0

(C) (D)
16
16
CO2 adsorbed (wt%)

12

CO2 adsorbed (wt%)

12

8
8

een-MOF 4 een-MOF
4
een-MOF/Al een-MOF/Al
een-MOF/Al-Si een-MOF/Al-Si
0 0
0 12 24 36 48 60 72 0 3 6 9 12 15
Exposure time (day) Exposure time (day)

F I G U R E 1 0 (A) Adsorption CO2 isotherms for een–MOF/Al–Si at various temperatures. (B) Water vapor isotherms of een–MOF,
een–MOF/Al, and een–MOF/Al–Si. (C) CO2 adsorption capacity of een–MOF, een–MOF/Al, and een–MOF/Al–Si after exposure to ambient
air. (D) CO2 adsorption capacity of een–MOF, een–MOF/Al, and een–MOF/Al–Si after exposure to 10% H2 O and 90% CO2 at 140◦ C. The flow
rate is 50 cm3 min−1 . Reproduced with permission.84 Copyright 2021, Springer Nature. MOF, metal–organic framework.

uptake significantly decreases in moisture, CO2 adsorp- Moreover, the mechanism of specific sites requires more
tion enthalpy displays a maximum value at 10% RH with research.
regard to the higher confinement effect and interaction
of pre-adsorbed molecules. Moreover, MIL-96(Al) con-
tains more hydrophilic groups, that is, Al(1)–H2 Oterminal , 3.1.3 New sites generated by pre-adsorbed
Al(3′)–OHterminal , and Al(3)–H2 Oterminal , which leads to water
the adsorption of CO2 molecules at the remaining sites
with the lower polarity, as shown in Figure 12. The adsorbed water molecules exhibit the antagonistic
Similarly, Datta et al.27 reported a highly stable micro- effects to some degree, which block intrinsic CO2 adsorp-
porous coppersilicate that has the square bipyramidal tion sites while also having the potential to become
crystal morphology (SGU-29). From Fourier transform new CO2 adsorption sites. Several recent studies have
infrared spectra of SGU-29 in Figure 13, the growth rates, shown that pre-adsorption of a spot of water prefer-
peak shapes, and intensities of adsorbed CO2 peaks in wet entially interacts with hydroxyl groups within the pore
gases could be comparable to those observed in the absence wall of MOFs, resulting in a significant increase in CO2
of H2 O, which only shows a broad peak of the adsorbed adsorption enthalpy due to an increase in co-adsorbate
water. The results verify that there are CO2 -specific and interactions.33,81 As shown in Figure 14, it is verified by
H2 O-specific sites but no sharing sites; thus, moisture does the molecular simulation that CO2 can interact with H2 O,
not influence CO2 capture ability of SGU-29. This kind but the interaction is relatively weak and has only a tiny
of sorbent can be seriously considered as a promising promoting effect.
candidate for binary adsorption. The exploration of CO2 - In 2009, Yazaydın et al.98 initially discovered that water
specific sites is the key to filling existing research gaps. molecules can promote CO2 uptake due to the increased
14 JIANG et al.

F I G U R E 1 1 (A) Illustration of the competitive adsorption of CO2 against H2 O at the surface and edge of polynaphthylene (PN). (B)
Polymerization of 1,2-diethynylbenzene (DEB) in metal–organic frameworks (MOFs). Reproduced with permission.92 Copyright 2016, ACS.
(C) Schematic of the microfluidic system for in situ encapsulated synthesis of MOFs. (D) The overall CO2 sorption and desorption steps by the
encapsulated MOFs. Reproduced with permission.93 Copyright 2021, KeAi.

F I G U R E 1 2 Maps of the occupied positions of (A) CO2 (orange), (B) N2 (blue), and (C) H2 O (cyan) in 1000 equilibrated frames extracted
from the Grand Canonical Monte Carlo (GCMC) simulations for co-adsorption of CO2 /N2 (20/80) in MIL-96(Al) at 303 K and 1.0 bar in the
presence of humidity, relative humidity (RH) = 8.5% (adapted from Ref. 33).

interactions between the quadruple moment of CO2 and coordinated water blocks the pores for nonpolar CH4 /N2 ,
the electric field of the hydrated framework. In Figure 15A, and a breathing effect takes place, which could open
both non-Coulombic energy and Coulomb energy increase the pores in favor of CO2 . As a result, CO2 selectivity
by introducing the coordinated water molecules to MOF rises, as depicted in Figure 15B. Jajko et al.31 revealed
structure, and the latter plays a leading role in energy that the CO2 loading is enhanced by pre-adsorption of
rise. The simulation results support a greater interaction water in UiO-66. To understand adsorption mechanism of
between CO2 and adsorbent, in which water interacts performance enhancement in CO2 adsorption, the inter-
with the coordinatively unsaturated Cu sites. Besides, the molecular adsorption energy is calculated, which includes
JIANG et al. 15

(A) 3
min
40
2 20
10
5
1 3

Absorbance (a.u.)
1
0
(B) 3
40
2 20
10
5
1 3
1
0
0
4000 3600 3200 2800 2400 2000 2450 2400 2350 2300 2250
Wavenumber (cm-1)

F I G U R E 1 3 The effect of adsorbed water on the adsorbed CO2 . Progressive change in the Fourier transform infrared spectra of dried
SGU-29 with time (as indicated) under the flow of (A) the dry flue gas and (B) the humid flue gas. Dry flue gas consists of CO2 , O2 , and N2
(100/190/723 mbar, respectively), and humid flue gas consists of H2 O, CO2 , O2 , and N2 (29/97/185/702 mbar, respectively) at 298 K. The flow
rate is equal to 5 mL min−1 . Reproduced with permission.27 Copyright 2015, AAAS.

F I G U R E 1 4 (A) The radial distribution functions calculated by Grand Canonical Monte Carlo (GCMC) simulations at 1 bar and 303 K
for the pair OCO2 –Hw of hydrated Mg-CUK-1. (B) The representative snapshot showing the interactions between CO2 and the atoms of
metal–organic framework (MOF) pore wall. The dashed lines in the snapshots denote interactions between CO2 molecules and the
surrounding atoms at distances below 3 Å (adapted from Ref. 81).

CO2 –CO2 , CO2 –H2 O, and CO2 –UiO-66. It is indicated that that of the dehydrated MIL-101 at low pressure because
adsorption energy of CO2 –H2 O is greater than that of CO2 – the terminal water molecules act as extra CO2 adsorption
CO2 at pressures of 0–1 bar. This demonstrates that the sites, which increases the interaction between quadrupole
interaction of pre-adsorbed water increases CO2 adsorp- moment of CO2 and dipole moment of adsorbents.97
tion capacity. The average occupational density profiles are It is worth noting that pre-adsorption may not accurately
also calculated, and the results show that water molecules reflect the real situation of carbon capture from humid
adsorb in the corners of octahedrons, which is explained streams. After water saturation, the moisture content does
by entropy impact of the size and shape of water molecules. not remain stable during adsorption process, which will
CO2 molecules begin to fill the octahedral cages after water cause a slight underestimate of CO2 uptake, particularly
adsorption, which is different from pure gas adsorption. It when only using gravimetric method to monitor CO2 –
is concluded that water transforms unfavorable sites into water co-adsorption. Other methods for measurement, for
favorable sites for CO2 adsorption. Thus, water promotes example, monitoring outlet gas concentration by using gas
the interaction of CO2 adsorption. Similarly, the hydrated chromatography, are more reliable to obtain a reasonable
MIL-101 exhibits a higher CO2 adsorption capacity than CO2 adsorption capacity.
16 JIANG et al.

(A) 25 (B) 35
CO2/N2 (hydrated 4 wt%)
30 CO2/N2 (dry)
20
CO2/CH4 (hydrated 4 wt%)

Energy (kJ· mol-1)

25
CO2/CH4 (dry)
15

Selectivity
Lennard-Jones (hydrated) 20
Lennard-Jones (dry)
10 Coulombic (hydrated) 15
Coulombic (dry)
10
5
5
0 0
0 5 10 15 20 25 30 35 40 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
CO2 loading (molecules·unit cell-1)
Partial pressure (bar)

F I G U R E 1 5 (A) Interaction energies between CO2 and dry Cu-BTC and hydrated Cu-BTC (4 wt%) from Grand Canonical Monte Carlo
(GCMC) simulations at 298 K. To convert molecules per unit cell to milligrams of gas adsorbed per gram of sorbent, multiply by 4.3557. (B)
CO2 selectivity from the simulations of the equimolar mixtures of CO2 /CH4 and CO2 /N2 at 298 K. Reproduced with permission.98 Copyright
2009, ACS. BTC, benzene-1,3,5-tricarboxylate.

3.2 Chemical interaction multipoint interactions of framework toward CO2 , which

includes π-quadrupole system of aromatic rings and CO2 ,
To overcome the drawbacks of physisorption, amine potentially exposed nitrogen atoms and C atom of CO2 ,
functional solid adsorbents have attracted considerable and partial C–H bonds and O atoms of CO2 . Besides,
interest due to their exceptional chemisorption prop- post-synthesis can prepare amine-functionalized MOFs.
erties, especially under humid conditions, which are Bahamon et al.78 investigated six amine models of dif-
also considered in MOF functionalization.99,100 Amine- ferent chain lengths and substitutions grafted onto the
functionalized MOFs are prepared in three general routes, unsaturated metal sites of M2 (dobpdc) by using molecu-
that is, in situ synthesis, post-modification, and physical lar simulation. The results show that mmen/Mg/dobpdc
mixing of unfunctionalized MOFs and polyamines.101 A has a higher CO2 adsorption capacity and cycling perfor-
common method for in situ synthesis is to use ligands mance than MOFs with other structures. However, due to
with amine. There are myriad pathways to decorate MOFs an increase in steric hindrances, additional amine func-
according to the type of amine functional groups, the tional groups or full functionalization of metal centers do
position, the possibility of coordination to node, and the not show the improvement in CO2 separation capabili-
environmental conditions.102 ties. Hence, functionalizing MOFs by amino groups could
The stability, polarity, alkalinity, and pore structure, occupy the partial pore sites and result in the decrease in
can be influenced by introducing functional groups to void space and surface area, which is not conducive to
pores. Free electrons located on amino nitrogen result CO2 adsorption. This is the reason why no enhancement
in a remarkably high affinity toward CO2 due to Lewis of adsorption capacity is found in MIL-101(Cr) –NH2 .103
acid–base interactions and non-covalent interactions.102 However, MIL-101(Fe)–NH2 displays 1.4-fold increase in
Ethiraj et al.76 synthesized a series of the mixed-ligand CO2 adsorption capacity than MIL-101(Fe) at 4 MPa and
UiO-66 through two different methods. The volumetric is more stable against water and ethanol.104 Mahdipoor
CO2 isotherms show a higher CO2 capacity in terms of low- et al.104 also deduced the mechanism of CO2 adsorption
temperature synthesis since the larger surface area could by amino functional groups based on three factors: elec-
be obtained. The results also demonstrate that the affin- tric field generated by different charges, hydrogen bonding
ity toward CO2 increases with the increase in NH2 content between CO2 and NH2 functional groups, and dipolar ion
in samples. The heteroaromatic amines that have nitro- (zwitterion) formation by the acidic CO2 and basic NH2
gen atoms in aromatic rings also exhibit the high CO2 groups.
adsorption capacity. Cui et al.77 synthesized a zeolite-like Regarding adsorption performance in the pres-
microporous tetrazole-based MOF, which is composed of ence of water, many researches have observed an
flexible tetrazole derivative ligands and Zn2+ ions. It has enhanced CO2 capture capacity under various humid
a high CO2 adsorption capacity of up to 8.09 mmol g−1 conditions.42,79,105–107 Lee et al.106 prepared a diamine-
at 273 K and ambient pressure. This can be attributed to functionalized MOF by grafting the primary amine to
JIANG et al. 17

(A) 1.0
0.9 MOF-808-Gly adsorption (dry)
0.8 MOF-808-Gly adsorption (RH 10%)

CO2 uptake (mmol· g-1)

0.7
0.6
0.5
0.4 (C)
0.3
0.2
0.1 15%
4%
0.0
0 5 10 15
Absolute pressure (kPa)

(B)
10
MOF-808 (RH 0%)
MOF-808 (RH 15%)
CO2 effluent/Influent (%)

8
MOF-808-FR (RH 0%)
MOF-808-FR (RH 15%)
6 MOF-808-Gly (RH 0%)
MOF-808-Gly (RH 15%)
4

0
-10 0 10 20 30 40 50 60 70
Normalized Time (min·g-1)

F I G U R E 1 6 (A) Overlay of CO2 /H2 O binary adsorption isotherm and dry CO2 isotherm at 25◦ C. (B) The dynamic CO2 breakthrough
curve comparison among MOF-808, MOF-808-FR, and MOF-808-Gly, in dry and humid (relative humidity [RH] 15%) conditions at 25◦ C. (C)
The proposed chemisorption mechanism of CO2 uptake in MOF-808-AAs with (humid) and without (dry) the presence of water. Reproduced
with permission.42 Copyright 2022, ACS. MOF, metal–organic framework.

open coordination sites of Mg2 (dobpdc). The material has formation can be removed by vacuum process, which is
a significant CO2 uptake (2.83 mmol g−1 ) at ambient CO2 desirable for vacuum swing adsorption (VSA).
pressure. Notably, it can retain CO2 adsorption capacity Siegelman et al.79 reported a diamine-appended
after exposure to humidity. However, this structure is not Mg2 (dobpdc) MOF with a high CO2 capacity under humid
stable enough in which H2 O and CO2 usually compete conditions. DFT calculations show that the pre-adsorbed
with metal sites and therefore it loses original grafted water increases CO2 binding energy by 18 kJ mol−1 , as
amine. To avoid this, Lyu et al.42 used a robust MOF-808 shown in Figure 17. This is because ammonium carbamate
system to append amino acids through carboxylates to chains can form easily due to the hydron bonding inter-
Zr centers, which forms a strong MOF backbone and action provided by water molecules and the carbamate
prevents water or CO2 from breaking the bonds between nitrogen atoms of nearby ammonium carbamate chains.
metal atoms and the carboxylates with amino acids. Both Introduction of the basic amine functional groups onto
CO2 adsorption isotherm and dynamic CO2 breakthrough the surface of a solid material can increase the interac-
curve reveal a significant increase in CO2 uptake in tion between CO2 and solid surface by physisorption and
MOF-808-Gly upon humidification of stream, as shown chemisorption. For CO2 –water co-adsorption, the increase
in Figure 16A,B. Later, they used solid-state nuclear in CO2 adsorption occurs mainly with the increase in
magnetic resonance to explore the mechanism behind the amine efficiency. The presence of water also changes
binary adsorption. As demonstrated in Figure 16C, bicar- adsorption heat, which affects the affinity toward CO2 .
bonate formation under humid conditions only needs an Furthermore, grafting amine functional groups reduces
amine to adsorb each CO2 molecule from natural gas flue pore sites and surface area. Water molecules as impurities
emissions, which is less than what is needed when the can clog some adsorption sites and decrease CO2 adsorp-
conditions are dry. Moreover, it is noted that bicarbonate tion capacity. A balance is considered by tuning up the
18 JIANG et al.

F I G U R E 1 7 Projections along pore axis and first coordination spheres of Mg(II) sites for van der Waal (vdW)-corrected, density
functional theory (DFT) calculated structures of evacuated L-2-ampd-R-Mg2 (dobpdc) (top left) and the framework following CO2 insertion
(bottom left), water adsorption (top right), and co-adsorption of CO2 and H2 O (bottom right). The vdW-corrected, DFT-calculated binding
energies (ΔE) are provided for each adsorption process, and the available experimental differential binding enthalpies (ΔHads ) are included in
parentheses. Green, blue, gray, red, and white spheres represent Mg, N, C, O, and H atoms, respectively. Reproduced with permission.79
Copyright 2019, ACS.

amount of amine grafted. Besides, the long-term stability et al.71 discovered new adsorption regions around inor-
of amine-functionalized adsorbents is necessary to avoid ganic building units, resulting in the large H2 uptake in
the loss of amine. PCN-12, NOTT-103, NOTT-112, and HKUST-1 due to the
confinement effect caused by small pore size (D ≤ 8.5 Å)
or narrow windows in cavities. For MOF-5 and MOF177
3.3 Confinement effect with larger pore sizes (10.8 and 11.2 Å), no adsorption site
is ascribed to confinement effect, which is confirmed by
Adsorption sites are classified into two types: binding sites first-principle calculations. It can be inferred that confine-
via electrostatic interaction that dominates CO2 adsorption ment effect could happen in MOFs with the appropriate
and pore sites through confinement effect.108 The excel- pore sizes. Apparently, the pore size must be larger than
lent adsorption capacity of pores with confinement effect dynamic diameter of adsorbate. Based on existing stud-
can promote CO2 uptake in some MOFs with optimal pore ies on CO2 adsorption of MOFs, it is concluded that the
sizes. The confinement effect, termed as nest effect, was rapid and strong CO2 adsorption could be attributed to
initially proposed by Derouane,109 which represents that the combination of high charge density and the optimal
a molecule and its environment tend to reciprocally opti- pore size.110–114 With an appropriate channel size, zinc-
mize their intermolecular force. This effect is related to based Calgary Framework 20 (CALF-20) may selectively
crystal topology and the auxiliary effects of atom dispersive physisorb CO2 .80 The CO2 binding sites are in the middle
interactions on cavity walls.71 The confinement impact of of the pores, which may prevent the creation of networks
the channel construction and the adsorbed H2 O molecules of hydrogen bonds between water molecules. Liang et al.38
should be considered in the binary adsorption of CO2 and used an interpenetration approach to develop a fluorinated
H2 O. MOF with ideal pore aperture and channel sizes (4.5–6
Å). It allows for favorable CO2 adsorption versus N2 and
reduces adsorption heat because of the strong interaction
3.3.1 Confinement effect of framework via confinement effect. Besides, it indicates the extraordi-
topology nary resistance to moisture, as shown in Figure 18, which
is presumably due to the nucleophilic nature of TiF6 2−
Some researchers have dedicated their efforts to deter- pillars. The sieving effect is a useful technique for gas sep-
mining the ideal pore size for gas separation. Gomez aration that involves changing the pore window and size
JIANG et al. 19

(A) 1.6 (B) (C)

He N2
CO2 (298 K, dry) CO2 (298 K, 74% RH)
CO2 (328 K, dry) CO2 (328 K, 74% RH)
1.2
CA /C0

0.8

65.6 cm 3·g-1
64.0 cm3·g -1
m ·g -1

31.8 cm3·g -1
27.6 c 3
0.4

0.0
0 20 40 60 80 100 120
Time (min·g-1)

F I G U R E 1 8 (A) Dynamic column breakthrough tests for dptz-CuTiF6 with a gas mixture of 10% CO2 and 90% N2 (8 cm3 min−1 flow
rate) at 298 or 328 K and 1 bar under dry and humid conditions. Direct visualization of CO2 molecules inside the crystal structure of
dptz-CuTiF6 , as viewed along the (B) c-axis and (C) b-axis. This highlights the optimal arrangement of the CO2 molecules within the
square-shaped channels. Reproduced with permission.38 Copyright 2019, Elsevier.

F I G U R E 1 9 (A) Synthesis procedure of three isoreticular metal–organic frameworks (MOFs). (B) Views of site I and (C) site II in
activated NbOFFIVE-dps-Cu, GeFSIX-dps-Cu, SIFSIX-dps-Cu, and UTSA-300a with varied pore aperture sizes. Color code: Cu, green; F, light
green; S, bright yellow; N, light blue; C, gray; Si, orange; Ge, navy blue; Nb, wine red; solvent molecules are omitted for the clarity.
Reproduced with permission.117 Copyright 2022, Springer Nature.

to better match the size of the gas molecules.115 Pore topol- specific binding sites for C2 H2 /CO2 separation, as shown
ogy plays a crucial role in the capacity and selectivity of in Figure 19. This reveals that [Cu(dps)2 (SiF6 )] has the
adsorbents for multicomponent adsorption. highest C2 H2 uptake and selectivity for C2 H2 /CO2 . Simi-
Tuning the pore size can be achieved by controlling the larly, Cui et al.118 reported several “SIFSIX” materials with
size of metal ions or clusters and adjusting the lengths different pore structures, which were tuned by changing
of ligands.116 Wang et al.117 displayed a fine pore engi- organic linkers, metal nodes, or framework interpenetra-
neering by altering anionic linkers and metal nodes in tion. The best sorbent could be found with the ideal hybrid
[Zn(dps)2 (SiF6 )] to explore appropriate pore sizes and the pore chemistry and the optimal pore size to form gas
20 JIANG et al.

clusters. Li et al.119 designated this method as “single- energies of Ni/dobdc for hydration.122 Soubeyrand-Lenoir
molecule trap.” This group also successfully synthesized et al.29 investigated adsorption uptake and enthalpy of
a cage-type structure with the desired size and strong three topical porous MOFs. The results show that water is
electrostatic interactions with CO2 molecules. Similar not conducive to CO2 adsorption capacity of UiO-66(Zr)
techniques may likely be applied to separate additional under different RHs, which is inconsistent with the find-
gases. Nevertheless, studies on tuning pore size for bet- ing from the research by Jajko et al.31 MIL-100(Fe) exhibits
ter separation of CO2 and H2 O are rarely reported. It can a striking fivefold improvement in CO2 uptake with the
be explained by the fact that H2 O has a smaller kinetic increase in RH. Given the poor MIL-100(Fe) test results,
diameter (2.65 Å) than CO2 (3.30 Å), and consequently, this conclusion is dubious. Álvarez et al.45 demonstrated
the sieving effect for CO2 capture in wet conditions is NOTT-400 that pre-adsorbed water could enhance CO2
invalid. However, most high-performing MOFs with good uptake via confinement effect. The CO2 uptake presents
CO2 /H2 O selectivity have narrow pores, which result in an approximate 2.5-fold increase after exposure to 20%
the confinement effect and prevent the formation of water RH conditions. Adsorbed water molecules form strong
clusters via hydrogen binding networks. According to the hydrogen bonds with the bridging hydroxo functional
study from Li’s group,58 pore size of ∼5 Å can attain the groups (μ2 -OH) in NOTT-400, which leads to the aug-
highest CO2 /H2 O selectivity in most MOFs based on 1627 mented CO2 sequestration. Likewise, PCN-250(Fe3 ) and
MOFs in CoRE MOF database. Thus, further research on PCN-250(Fe2 Co) represent moisture-enhanced features.32
pore topology is required to focus on this range of pore Water can have strong interactions with the hydroxo func-
diameters. tional groups (μ3 -O) in their framework. The coordinated
H2 O molecule acts as a plier that clamps CO2 on accessible
adsorption site. Furthermore, it is demonstrated that CO2
3.3.2 Confinement effect of the coordinated moves closer to OMS when hydroxo functional groups
H2 O molecules (μ3 -O) are coordinated with H2 O by GCMC. González-
Zamora and Ibarra44 concluded that water molecules
The confinement of small amounts of polar molecules can strongly interact with hydroxo functional groups via
such as water and alcohols or non-polar molecules such as hydrogen bonding, which leads to a confinement effect
hydrocarbons within MOFs can also increase CO2 uptake, allowing for more efficient CO2 packing.
which proposes a novel method to address the issues of Water loading amount is critical to the effectiveness of
CO2 –water co-adsorption.43 It occurs inside the pores, the confinement effect. The directing effect of hydroxo
which leads to the diffusional limitations and structural functional groups is greatly reduced at high RH due to an
constraints on the transition state.108 Also, pre-adsorbed increase in water disorder caused by thermal agitation.44
molecules can interact with CO2 .120 Numerous investi- It is essential to determine the optimal water loading
gations have found a trend of increased CO2 uptake in of confinement effect on CO2 . Sánchez-González et al.41
co-adsorption when compared to that under dry condi- explored the structural transformation of Mg-CUK-1 at dif-
tions. ferent H2 O loadings by using powder X-ray diffraction and
It is noted that CO2 adsorption capacities of some refined it by the Le Bail strategy in Figure 20. With eight
MOFs increase first and then decrease with the increase hydrogen-bonded H2 O molecules per unit cell, the window
in the humidity, such as HKUST-1,28 MIL-101(Cr),97 MIL- dimension was reduced to 2.56 Å inducing non-accessible
100(Fe),29 Mg-CUK-1,81 InOF-1,30 and UiO-66.31 Liu et al.28 for CO2 . A window reduction is also resulted by four or
studied the equilibrium of CO2 /H2 O binary adsorption two H2 O molecules per unit cell and thereby the diffu-
of two types of MOFs: HKUST-1 (Cu-BTC) and Ni/dobdc sion of CO2 is restrained. CO2 has enough free space to
(CPO-27-Ni or Ni/MOF-74). Water adsorption progress on interact with adsorbent and adsorbate only when there is
HKUST-1 is reversible. A small amount of the adsorbed one H2 O molecule per unit cell, which results in “bottle-
water does not inhibit CO2 adsorption but may promote neck effect.” Indeed, Mg-CUK-1 exhibits a 1.8-fold increase
adsorption capacity of HKUST-1, which contributes to in CO2 uptake at 18% RH (one hydrogen-bonded H2 O
the addition of Coulombic interactions between CO2 and molecule per unit cell) due to pore windows partially con-
H2 O molecules. However, there is no enhancement of strained by water. Soubeyrand-Lenoir et al.29 elucidated
CO2 adsorption capacity in Ni/dobdc after introducing the mechanism of the confinement effect in a mesoporous
H2 O molecules into feed gases. Zhao et al.121 reached a architecture. Pre-equilibrated water initially fills microp-
similar conclusion. HKUST-1 slightly degrades, and orous pockets in MIL-100(Fe), as shown in Figure 21A.
Ni/dobdc exhibits a better stability after several working CO2 molecules are then adsorbed at the accessible sites
cycles, which could be ascribed to higher transition state (Figure 21B) and even displace some water (Figure 21C),
JIANG et al. 21

3.4 Hydrolytic stability

Along with the aforementioned techniques to improve

CO2 adsorption selectivity, it is also important for MOFs
to have the high stability and resistance to irreversible
reactions with water in moisture gases.123 If the MOF struc-
ture collapses under humid conditions, its durability and
application will be limited. The hydrolytic reactivity and
initial decomposition pathway of MOFs depend on the
specific type of ligand and second building unit (SBU).124
MOFs with various metal ions and ligands have been
investigated, which involves water-labile and water-stable
MOFs.122,125–127 For water-stable MOFs, the stronger bonds
between metal clusters and organic linkers can be found to
possess a structure with greater stability in moisture.28
Water can react with MOFs via ligand displacement
and/or hydrolysis.122 As shown in Equation (1), ligand dis-
placement inserts an H2 O molecule into a metal–ligand
F I G U R E 2 0 Crystal structures of Mg-CUK-1 with different bond. The hydrolysis reaction involves water dissocia-
H2 O molecule loadings, from top to bottom, 0, 1, 2, 4, and 8 water tion into a hydroxide ion and a hydrion. Bond formation
molecules per unit cell. (A) View through the a-axis showing the between a metal with hydroxide ion and a ligand with
hydrogen-bonded H2 O molecules to the hydroxo functional groups hydrion is shown in Equation (2).
(H2 O⋅⋅⋅OH-μ3), and (B) side view of the channel (through the
c-axis) and accessible surface, yellow circles represent CO2 kinetic M − L + H2 O → M − (OH2 ) ⋯ L (1)
diameter. Reproduced with permission.41 Copyright 2020, PMC.

M − L + H2 O → M − (OH) + LH (2)

Many hydrolytic stability factors, such as metal compo-

which would not happen under anhydrous condition. sition and coordination, chemical functionality of organic
Nevertheless, some MOFs have no accessible microporos- ligands, framework dimensionality, and interpenetration,
ity after water absorption (Figure 21D). As a result, there is have an influence MOF stability from an aspect of ther-
no increase in CO2 adsorption uptake. modynamic or kinetics.127 To explore the stabilities of
High-throughput screening results indicate that a small various MOFs, Low et al.122 conducted an experimental
pore size (LCD of around 5 Å) is better for CO2 /H2 O sepa- test together with molecular modeling. A steam stabil-
ration, which satisfies CO2 diffusion and obstructs water ity map is shown in Figure 22. It is found that MOFs
cluster formation. Moreover, the pre-installing water of bridged by carboxylates need less activation and reac-
MOFs can also cause a confinement effect, thus changing tion energy since the metal–ligand bonds are weaker than
the interaction between CO2 and sorbent. The confine- those of imidazolates, which is linked to imidazolates’
ment effect is definitely a novel approach to develop MOFs greater basicity.128 Hence, MOFs bridged by imidazolates
with a suitable framework for carbon capture in practical such as ZIF-8 tend to be more hydrolytically stable than
humid working conditions. those bridged by carboxylates. Besides, it is significant to
It is noted that not all water loadings give rise to more regard the stability of SBUs in the MOF structure. The
efficient CO2 molecule accommodation by using the above higher oxidation of metals or metal ions, the better stabil-
methods. For the methods of changing the polarization ity toward hydrolysis becomes. In general, metals with the
interaction of surface or creating confinement effects can higher metal valence are discovered to be more resistant
only be used in low humidity ranges because adsorption to hydration. For instance, the relative positive charge per
type is still physisorption. The presence of extra water can Zn of Zn4 O6+ cluster in MOF-5 is 1.5, which is less than
occupy partial adsorption sites, which leads to a reduction metal valence in HKUST-1 and MIL-101. It is suggested
in adsorption amount. Relatively, the chemical method that HKUST-1 and MIL-101 could form stronger metal–
can be employed for larger water contents because the linker bonds than MOF-5, which can prevent the collapse
chemisorption force is stronger. However, as Figure 2 structures.122 Although a small amount of water (0.5 mole
shows, there are also some exceptions. Thus, we herein equivalent) has no impact on the network structure sta-
cannot offer a quantitative description. bility of HKUST-1, higher water adsorption may cause the
22 JIANG et al.

F I G U R E 2 1 (A–D) Schematic diagram of the possible mechanisms of CO2 adsorption in the presence of water. Reproduced with
permission.29 Copyright 2012, ACS.

F I G U R E 2 2 Map of steam stability for four metal–organic frameworks (MOFs). Position of the structure for an MOF represents its
maximum structural stability by X-ray diffraction (XRD) measurement, while activation energy for ligand displacement by a water molecule
determined by molecular modeling is represented by a magenta number (in kcal mol−1 ). Reproduced with permission.122 Copyright 2009,
ACS.

network structure to be partially broken at Cu–O coor- bicity or steric hindrance from a kinetic perspective.
dination sites.129 Besides, chemical stability to water can From a thermodynamic aspect, the enhanced strength
be increased with the increased inertness of central metal of metal–ligand bonds can be achieved by replacing
ions.130 For example, the chemical stability of isotypic metal ions with the higher valence cations or more inert
MOFs such as Al-BDC (Al-benzenedicarboxylate called metal ions since cation–ligand bonds are inclined to
MIL-53-Al), Cr-BDC (MIL-53-Cr), and V-BDC (MIL-47-V) hydrolyze for low-valent metals.72 Alternatively, chang-
decreases in the order, Cr-BDC > Al-BDC > V-BDC.130 ing the ligand can improve the bond energy. Colombo
Thus, it is possible to enhance water stability of et al.131 applied polyazolate-bridging ligands in MOFs,
MOFs by strengthening metal–ligand interactions from which leads to strong metal–nitrogen bonds when com-
a thermodynamic perspective, or enhancing hydropho- pared to carboxylate-based counterparts. This indicates
JIANG et al. 23

that strength of M–N bonds is related to pressure required be evaluated when considering real heat and mass trans-
for the deprotonation of N–H bonds. Thus, frameworks fer processes. In order to evaluate adsorbent features
generated from organic ligands functionalized with 1,2,3- that more closely resemble actual use, this section aims
triazole (pKa = 13.9) have greater thermal stability than to present simplified prototype testbeds using MOFs for
analogs based upon tetrazole (pKa = 4.9).132 carbon capture.
For the kinetic aspect, it is effective to increase water sta- Adsorption CO2 removal can be achieved by tempera-
bility by synthesizing highly hydrophobic surfaces or coat- ture swing adsorption (TSA), pressure swing adsorption
ing MOFs with appropriate polymers to increase structure (PSA), VSA, or other hybrid working processes. For CO2
steric hindrance and hydrophobicity.133,134 For instance, and H2 O binary adsorption, VSA and TSA are more com-
some MOFs with water-repellent groups such as methyl monly used to achieve higher CO2 purities than PSA.145
groups incorporated into organic ligands exhibit excellent Under the circumstances of simulated coal flue gas, Lyu
stability under humid conditions. Wu et al.135 reported an et al.42 investigated MOF-808-Gly’s performance of car-
MOF by strategically with non-polar functional groups, bon capture with a specially designed VSA apparatus,
trifluoromethoxy, in the frameworks. Trifluoromethoxy as shown in Figure 23. The inlet of sorption reactor is
groups in MOFs prevent water intrusion into framework connected to flue gases, and the outlet is used to expel CO2 -
structures, which improves the stability in the humid lean stream. The inlet transforms into an outlet CO2 -rich
air. Linkers modified by either polar or non-polar groups stream during the regeneration process. To avoid water
enhance the overall stability of the lattice by drawing water condensation on the gas sensor and to increase the purity
toward new adsorption sites, thus hindering water from of product gas, a membrane drier is installed along the
reaching the metal center.136 Besides, steric hindrance is regeneration branch. The simulated flue gases are bal-
an undeniable factor in reducing hydrolysis. MOFs with anced in N2 at 20◦ C–23◦ C, which contain 15% CO2 at
high coordination numbers, such as UiO-66(Zr)137 and an RH of 20%. The regeneration process of the system
MIL-125(Ti),138 have high connectivity and provide steric starts once CO2 concentration reaches more than 2%. A
shielding against water for SBUs. This factor can also be glycine-filled MOF-808-Gly-functionalized sorption reac-
confirmed in research by Tang et al.139 Ca-based MOFs tor displays good cycle stability at 80 cycles with an average
are stable up to 98% RH due to ligand distortion, which CO2 uptake of 0.42 mmol g−1 h−1 . Likewise, Hossain
causes more steric hindrance effect of benzene rings to et al.146 investigated similar volumetric system to test mul-
protect Ca2+ ions from water attack. Moreover, contract- ticomponent isotherms of UiO-66 with three distinct water
ing or rigidifying the ligands can also enhance the rigidity loadings (low, pore filling, and saturation). The results
of framework and consequently water stability.140,141 Lv indicate that CO2 uptake slightly increases at low water
et al.142 constructed 13 Zr-MOFs consisting of three classes loading (1.5 mol kg−1 ), which is consistent with the results
of ligands. The results imply that rigid linkers with lower from Jajko et al.’s study,31 and then gradually decreases
degree of rotational freedom have a self-repair mechanism as water loading increases. The cooperative effect of low
after water substitution, and this effect is more effective adsorbed water can be attributed to the confinement effect
with the higher connectivity. Moreover, shorter linkers can and affinity toward CO2 . When water loading approaches
have greater deformation energy barrier due to the larger saturation, CO2 uptake will decrease by up to 60% from 0 to
angle during framework destruction.143 5 kPa. However, the flue gases will be nearly saturated with
water when dropping to ambient temperature since water
content in flue gases is almost 7%, which has been ignored
4 PROTOTYPE VERIFICATION in most researches. Thus, the simulated flue gases with sat-
uration water loadings at a moderate temperature require
To address the large energy penalty of common aque- additional attention. In this work, we are just concerned
ous monoethanolamine solution scrubbing technology for with the adsorption properties of capacity and selectivity.
post-combustion carbon capture, adsorbents and process Other aspects such as adsorption kinetics, thermal conduc-
design need to be more effective and energy intensive. tivity, and adsorption heat should also be considered for
Porous solid sorbents offer the advantage of lower regen- process metrics such as productivity, specific energy, and
eration heat when compared to aqueous amine solvents. purity.147
When desorption heat is assumed to be consistent, a higher The zero length column (ZLC) technique, which is effec-
sorbent operating capacity can lead to a lower thermal tive in obtaining transient desorption curves, has gained
energy input.144 Several promising adsorbents have been application in the determination of the effect of water.
proposed for carbon capture, which have high CO2 selec- In a ZLC experiment, the exterior mass and heat trans-
tive adsorption capacity under wet conditions. Besides, fer resistance of sorption bed can be ignored. Hu’s team148
the capture process of sorbent in the reactor should also assessed the behavior of M/dobdc (M = Co, Ni, Mg) and
24 JIANG et al.

Rotameter
CO2 Volumetric
RH T CO2
Rotameter Flow Meter Simulated
Vacuum Gauge
N2 Flue Gas
Rotameter Stream
Solenoid Solenoid Pressure
Valve Valve Relief Valve
S S

Regeneration
Branch
Saturator Cell
Isothermal Bath RH T RH T CO2
MFC CO2-Rich
Legends Stream
Diaphragm
Gas Line
Pump
Sorption In-Line
Stream Type Bed Dryer Membrane
CO2
Dryer
Adsorption
Branch
Branch Name Compressed Humid
Air Air
RH Sensor
In-Line Adsorption RH T CO2
Filter S Branch MFC CO2-Lean
Stream
Solenoid Piston
Valve Pump

F I G U R E 2 3 Schematic diagram of the practical vacuum swing adsorption (VSA) testing apparatus. Reproduced with permission.42
Copyright 2022, ACS.

commercial 13X zeolite pellets in ZLC with 1% (v/v) water. Figure 24. For example, after the nitrogen oxides, partic-
The water content was purposefully adjusted to a low ulates, and sulfur oxides have been removed, the flue gas
value, presuming that flue gas had already been dried for will be cooled using cooling water from power plant’s cool-
physical sorbents. Co/dobdc and Mg/dobdc lack stability ing tower. The temperature of the flue gas will fall below
to moisture, but Ni/dobdc is less sensitive to water. How- 40◦ C, and its water content will be reduced by at least
ever, when exposed to wet flue gas conditions, adsorption 75% until it reaches saturation. Therefore, we should con-
capacity of Ni/dobdc declines as the duration of the expo- sider CO2 –water co-adsorption and energy consumption of
sure increases. Co/dobdc appears to require a longer time regeneration. Meanwhile, a water zone may be formed in
for the regeneration process; that is, it only has a slight the adsorption bed and alter thermal profiles.151 More heat
loss in CO2 uptake after exposure to wet flue gas. This or vacuum degrees are required to desorb the water zone.
is in line with Kizzie’s work, which finds that Co/dobdc The water that comes along with the CO2 -rich stream is
had the highest CO2 capacity (85% of that of pristine collected in the vacuum pump or other drying devices. This
samples) at 70% RH among M/dobdc.149 Also, Mangano system is only appropriate for MOFs on which the satu-
et al.150 employed ZLC to investigate water stability. For rated water has a favorable or insignificant effect on CO2
water stability of Mg-CPO-27 and Ni-CPO-27, Mg-CPO-27 adsorption ability. For other MOFs with a dramatic decline
is deactivated in the presence of water. However, Ni-CPO- in CO2 adsorption capacity under humid conditions, mul-
27 passes water-stable test and remains at a reduced and tiple stages might be a viable option to pre-adsorbed water
steady CO2 capacity in the moist flue gas. vapor in flue gases.
The performance of MOFs in the volumetric system
is basically comparable to that of gravimetric apparatus.
One issue that needs to be pointed out in the volumet- 5 CONCLUSIONS AND FUTURE
ric experiment is the transformation of the morphology PERSPECTIVES
of MOF. MOFs in pellets have the less uptake and slower
adsorption kinetics when compared to pure powder. This Nanoporous MOFs have great potential for carbon capture
issue cannot be avoided if real carbon capture progress because of their customizable and adjustable pore sur-
proceeds. More CO2 –water co-adsorption factors should face, high porosity, structure diversity, etc. To find MOFs
be considered in a practical CDR facility, as shown in with excellent adsorption performance under humid
JIANG et al. 25

F I G U R E 2 4 Schematic diagram of Cooled CO2 to

metal–organic framework (MOF)-based CO2 Flue Gas Storage
capture system process (adapted from Ref.
152). CO2 and
Feed Sorption Water CO2 compression
Flue Gas Cooler bed and Drying
from FGD

Feed Water Raffinate Evacuation Water

Compressor Compressor
(for PSA) (for VSA)

conditions, it is feasible and convenient to perform numer- a thermodynamic aspect, or it can be refined by hydropho-
ical screenings of MOFs based on their structures for selec- bic treatment or/and enhancing the steric hindrance with
tive adsorption. The Henry constant, Henry selectivity, and respect to adsorption kinetics.
adsorption enthalpy are three more practical indicators For the simplified prototype testbeds, working capac-
that can be used in the initial screening. Then, depending ity of MOFs is also related to heat and mass transfer
on the specific capture objectives, more targeted indica- rate and actual operating conditions. It is not involved
tors such as separation potential, percent uptake loss, and in this review, which is only concerned with humidity
selectivity should be analyzed for top-performing MOFs. impact. It should be noted that the change in humid-
The screening results of CO2 adsorption in the presence ity level depends on the application circumstance. If flue
of water show that high surface affinity toward CO2 pro- gases are pretreated to remove moisture, sorbents with
vided by adsorbents with superb chemical structures or confinement effect are suitable under the condition of
co-adsorption adsorbates and suitable pore sizes results in low humidity since water vapor cannot be completely
good CO2 /H2 O selectivity. removed. Without pretreatment or in DAC, the porous
CO2 uptake or selectivity can be improved through amine-functionalized MOFs are emphasized as having a
these methods. The higher polarization interaction for higher binding strength to collect CO2 in conditions of
CO2 than H2 O can come from hydrophobic modification, high RH or/and the extremely low CO2 partial pressure.
pores with specific- CO2 adsorption sites, or new adsorp- After water–CO2 co-adsorption, the regeneration energy of
tion sites generated by pre-adsorbed water. However, the water and drying of product gas should both be considered.
extremely strong hydrophobicity leads to poor CO2 adsorp- In summary, various modification methods have been
tion capacity. In addition, amine functional sorbents with applied for carbon capture in the presence of water. The
chemisorption ability have drawn wide attention due to intrinsic properties of adsorbents are very important to
their remarkable CO2 capacity, especially in humid con- enhance the affinity of CO2 , which can be tuned by alter-
ditions, which could be attributed to the increase in amine ing the constituent elements and topologies of MOFs. It is
efficiency and promotion of ammonium carbamate chain also crucial to strike a balance between a low affinity for
formation via hydrogen bonding interactions. However, H2 O and a strong affinity for CO2 . This work undoubtedly
the desorption process also faces larger energy barriers. offers practical and useful guidelines to choose and design
The confinement effect is another condition that can MOFs in the near future.
increase CO2 uptake in wet gases. This is caused by water
pre-adsorption and leads to diffusional restriction and AC K N OW L E D G M E N T S
structural constraints. Due to physisorption, they also have This research was supported by the National Key
a low desorption heat, but their CO2 adsorption capacity Research and Development Program of China (Nos.
at relatively low partial pressure is not very impressive. 2022YFB4101700 and 2022YFE0128600), the National
Researches on changing the pore window and size to Natural Science Foundation of China (Nos. 52276022,
improve the confinement effect of cavities in MOFs for 22278365, and 22225802), and the Basic Research Funds
CO2 adsorption and prevent water cluster formation are for the Central Government ‘Innovative Team of Zhejiang
still needed for further exploration. Moreover, the struc- University’ under contract number 2022FZZX01-09. Long
ture of MOFs should be robust enough to avoid structure Jiang and Jinyuan Yong contributed equally to this work.
collapse in the presence of water. The water stability can
be improved by increasing the strength of the metal–linker C O N F L I C T O F I N T E R E S T S TAT E M E N T
bond by replacing more stable metal ions or ligands from The authors declare no conflicts of interest.
26 JIANG et al.

ORCID 18. Kökçam-Demir Ü, Goldman A, Esrafili L, et al. Coordina-

Zongbi Bao https://orcid.org/0000-0003-4327-3028 tively unsaturated metal sites (open metal sites) in metal–
organic frameworks: design and applications. Chem Soc Rev.
2020;49(9):2751-2798.
REFERENCES 19. Collins DJ, Zhou H-C. Hydrogen storage in metal–organic
1. Rasmussen CE. Atmospheric Carbon Dioxide Growth Rate. frameworks. J Mater Chem. 2007;17(30):3154-3160.
2023. Accessed April 14, 2023. https://mlg.eng.cam.ac.uk/carl/ 20. Li Y, Wang H-T, Zhao Y-L, et al. Regulation of hydrophobic-
words/carbon.html ity and water adsorption of MIL-101(Cr) through post-synthetic
2. Tohjima Y, Patra PK, Niwa Y, Mukai H, Sasakawa M, modification. Inorg Chem Commun. 2021;130:108741.
Machida T. Detection of fossil-fuel CO2 plummet in China 21. Yu J, Balbuena PB. Water effects on postcombustion CO2
due to COVID-19 by observation at Hateruma. Sci Rep. capture in Mg-MOF-74. J Phys Chem C. 2013;117(7):3383-3388.
2020;10(1):18688. 22. Fetisov EO, Shah MS, Long JR, Tsapatsis M, Siepmann JI.
3. James Hansen MSaRR. Global Temperature in 2021. 2022. First principles Monte Carlo simulations of unary and binary
Accessed April 14, 2023. http://www.columbia.edu/∼jeh1/ adsorption: CO2 , N2 , and H2 O in Mg-MOF-74. Chem Commun.
mailings/2022/Temperature2021.13January2022.pdf 2018;54(77):10816-10819.
4. Song C, Liu Q, Qi Y, et al. Absorption-microalgae hybrid 23. Yang Y, Jin L, Zhou L, Du X. A molecular study of humid
CO2 capture and biotransformation strategy—a review. Int J CO2 adsorption capacity by Mg-MOF-74 surfaces with ligand
Greenhouse Gas Control. 2019;88:109-117. functionalization. Comput Mater Sci. 2022;209:111407.
5. Veneman R, Frigka N, Zhao W, Li Z, Kersten S, Brilman W. 24. Liu Q, Ning L, Zheng S, Tao M, Shi Y, He Y. Adsorption of
Adsorption of H2 O and CO2 on supported amine sorbents. Int carbon dioxide by MIL-101(Cr): regeneration conditions and
J Greenhouse Gas Control. 2015;41:268-275. influence of flue gas contaminants. Sci Rep. 2013;3(1):2916.
6. Samanta A, Zhao A, Shimizu GKH, Sarkar P, Gupta R. Post- 25. Yang F, Ge T, Zhu X, Wu J, Wang R. Study on CO2 capture in
combustion CO2 capture using solid sorbents: a review. Ind Eng humid flue gas using amine-modified ZIF-8. Sep Purif Technol.
Chem Res. 2012;51:1438-1463. 2022;287:120535.
7. Li X-M, Gao J. Recent advances of metal–organic frameworks- 26. Hu J, Liu Y, Liu J, Gu C. Effects of water vapor and trace
based proton exchange membranes in fuel cell applications. gas impurities in flue gas on CO2 capture in zeolitic imidazo-
SusMat. 2022;2(5):504-534. late frameworks: the significant role of functional groups. Fuel.
8. Pardakhti M, Jafari T, Tobin Z, et al. Trends in solid adsor- 2017;200:244-251.
bent materials development for CO2 capture. ACS Appl Mater 27. Datta SJ, Khumnoon C, Lee ZH, et al. CO2 capture from
Interfaces. 2019;11:34533-34559. humid flue gases and humid atmosphere using a microporous
9. Wu Y, Li Y, Gao J, Zhang Q. Recent advances in vacancy engi- coppersilicate. Science. 2015;350(6258):302-306.
neering of metal–organic frameworks and their derivatives for 28. Liu J, Wang Y, Benin AI, Jakubczak P, Willis RR, LeVan
electrocatalysis. SusMat. 2021;1(1):66-87. MD. CO2 /H2 O adsorption equilibrium and rates on
10. Furukawa H, Ko N, Go YB, et al. Ultrahigh porosity in metal– metal−organic frameworks: HKUST-1 and Ni/DOBDC.
organic frameworks. Science. 2010;329(5990):424-428. Langmuir. 2010;26(17):14301-14307.
11. Millward AR, Yaghi OM. Metal−organic frameworks with 29. Soubeyrand-Lenoir E, Vagner C, Yoon JW, et al. How water
exceptionally high capacity for storage of carbon dioxide at fosters a remarkable 5-fold increase in low-pressure CO2
room temperature. J Am Chem Soc. 2005;127(51):17998-17999. uptake within mesoporous MIL-100(Fe). J Am Chem Soc.
12. Liu J, Thallapally PK, McGrail B, Brown D, Liu W. Progress in 2012;134(24):10174-10181.
adsorption-based CO2 capture by metal–organic frameworks. 30. Garrido-Olvera LP, Sanchez-Bautista JE, Alvarado-Alvarado D,
Chem Soc Rev. 2011;41(6):2308-2322. et al. Confined toluene within InOF-1: CO2 capture enhance-
13. Yazaydin AO, Snurr RQ, Park TH, et al. Screening of metal– ment. RSC Adv. 2019;9(56):32864-32872.
organic frameworks for carbon dioxide capture from flue gas 31. Jajko G, Kozyra P, Gutiérrez-Sevillano JJ, Makowski W, Calero
using a combined experimental and modeling approach. J Am S. Carbon dioxide capture enhanced by pre-adsorption of water
Chem Soc. 2009;131(51):18198. and methanol in UiO-66. Chem Eur J. 2021;27(59):14653-14659.
14. Lee S-Y, Park S-J. A review on solid adsorbents for carbon 32. Chen Y, Qiao Z, Huang J, et al. Unusual moisture-enhanced
dioxide capture. J Ind Eng Chem. 2015;23:1-11. CO2 capture within microporous PCN-250 frameworks. ACS
15. McDonald TM, Lee WR, Mason JA, Wiers BM, Hong CS, Appl Mater Interfaces. 2018;10:38638-38647.
Long JR. Capture of carbon dioxide from air and flue gas 33. Benoit V, Chanut N, Pillai RS, et al. A promising metal–
in the alkylamine-appended metal–organic framework mmen- organic framework (MOF), MIL-96(Al), for CO2 separation
Mg2(dobpdc). J Am Chem Soc. 2012;134(16):7056-7065. under humid conditions. J Mater Chem A. 2018;6(5):2081-2090.
16. Liu F, Wang T, Dong H, Liu W. Modified metal–organic 34. Liu B, Smit B. Molecular simulation studies of separation
framework by a novel coordinatively unsaturated amine graft- of CO2 /N2, CO2 /CH4 , and CH4 /N2 by ZIFs. J Phys Chem C.
ing mechanism for direct air capture of CO2 . Chem Eng J. 2010;114(18):8515-8522.
2023;454:140431. 35. D’Alessandro DM, Smit B, Long JR. Carbon dioxide capture:
17. DeCoste JB, Peterson GW, Schindler BJ, Killops KL, Browe MA, prospects for new materials. Angew Chem. 2010;49(35):6058-
Mahle JJ. The effect of water adsorption on the structure of 6082.
the carboxylate containing metal–organic frameworks Cu-BTC, 36. Purdue MJ, Qiao Z. Molecular simulation study of wet flue
Mg-MOF-74, and UiO-66. J Mater Chem A. 2013;1(38):11922- gas adsorption on zeolite 13X. Microporous Mesoporous Mater.
11932. 2018;261:181-197.
JIANG et al. 27

37. Lowry RK, Kullberg RC, Rossiter DJ. Harsh environments and 55. Chanut N, Bourrelly S, Kuchta B, et al. Screening the effect
volatiles in sealed enclosures. Surface Mount Technology Asso- of water vapour on gas adsorption performance: application
ciation International Technical Conference. Published online to CO2 capture from flue gas in metal–organic frameworks.
October, 2010. https://orslabs.com/pdf/SMTAI10HE31_Lowry_ ChemSusChem. 2017;10(7):1543-1553.
RobertREV.pdf 56. Li S, Chung YG, Snurr RQ. High-throughput screening of
38. Liang W, Bhatt PM, Shkurenko A, et al. A tailor-made inter- metal–organic frameworks for CO2 capture in the presence of
penetrated MOF with exceptional carbon-capture performance water. Langmuir. 2016;32(40):10368-10376.
from flue gas. Chem. 2019;5(4):950-963. 57. Chung YG, Camp J, Haranczyk M, et al. Computation-ready,
39. Milner PJ, Siegelman RL, Forse AC, et al. A diaminopropane- experimental metal–organic frameworks: a tool to enable high-
appended metal–organic framework enabling efficient CO2 throughput screening of nanoporous crystals. Chem Mater.
capture from coal flue gas via a mixed adsorption mechanism. 2014;26(21):6185-6192.
J Am Chem Soc. 2017;139(38):13541-13553. 58. Li W, Rao Z, Chung YG, Li S. The role of partial atomic charge
40. Mason JA, McDonald TM, Bae T-H, et al. Application of a assignment methods on the computational screening of metal–
high-throughput analyzer in evaluating solid adsorbents for organic frameworks for CO2 capture under humid conditions.
post-combustion carbon capture via multicomponent adsorp- ChemistrySelect. 2017;2(29):9458-9465.
tion of CO2 , N2 , and H2 O. J Am Chem Soc. 2015;137(14):4787- 59. Erucar I, Keskin S. Unlocking the effect of H2 O on CO2 separa-
4803. tion performance of promising MOFs using atomically detailed
41. Sánchez-González E, Flores JG, Flores-Reyes JC, et al. Bot- simulations. Ind Eng Chem Res. 2020;59:3141-3152.
tleneck effect explained by Le Bail refinements: structure 60. Wang J, Deng S, Zhao R, et al. Performance comparison of three
transformation of Mg-CUK-1 by confining H2 O molecules. adsorption cycles for CF4 recovery from waste gas using 13X
Materials. 2020;13(8):1840. zeolite. J Cleaner Prod. 2022;337:130546.
42. Lyu H, Chen OI-F, Hanikel N, et al. Carbon dioxide capture 61. Findley JM, Sholl DS. Computational screening of MOFs and
chemistry of amino acid functionalized metal–organic frame- zeolites for direct air capture of carbon dioxide under humid
works in humid flue gas. J Am Chem Soc. 2022;144(5):2387-2396. conditions. J Phys Chem C. 2021;125(44):24630-24639.
43. Cotlame-Salinas VDC, López-Olvera A, Islas-Jácome A, 62. Boyd PG, Chidambaram A, García-Díez E, et al. Data-driven
González-Zamora E, Ibarra IA. CO2 capture enhancement in design of metal–organic frameworks for wet flue gas CO2
MOFs via the confinement of molecules. React Chem Eng. capture. Nature. 2019;576(7786):253-256.
2021;6:441-453. 63. Rogacka J, Seremak A, Luna-Triguero A, et al. High-throughput
44. González-Zamora E, Ibarra IA. CO2 capture under humid screening of metal–organic frameworks for CO2 and CH4 sepa-
conditions in metal–organic frameworks. Mater Chem Front. ration in the presence of water. Chem Eng J. 2021;403:126392.
2017;1(8):1471-1484. 64. Canepa P, Arter CA, Conwill EM, et al. High-throughput
45. Álvarez JR, Peralta RA, Balmaseda J, González-Zamora E, screening of small-molecule adsorption in MOF. J Mater Chem
Ibarra IA. Water adsorption properties of a Sc(iii) porous coordi- A. 2013;1(43):13597-13604.
nation polymer for CO2 capture applications. Inorg Chem Front. 65. Wolber G, Seidel T, Bendix F, Langer T. Molecule-
2015;2(12):1080-1084. pharmacophore superpositioning and pattern matching
46. Sánchez-González E, Álvarez JR, Peralta RA, et al. Water in computational drug design. Drug Discov Today. 2008;13(1-2):
adsorption properties of NOTT-401 and CO2 capture under 23-29.
humid conditions. ACS Omega. 2016;1(2):305-310. 66. Sliwoski G, Kothiwale S, Meiler J, Lowe EW. Computational
47. Kolle JM, Fayaz M, Sayari A. Understanding the effect of water methods in drug discovery. Pharmacol Rev. 2014;66(1):334-395.
on CO2 adsorption. Chem Rev. 2021;121(13):7280-7345. 67. Wilmer CE, Farha OK, Bae Y-S, Hupp JT, Snurr RQ. Structure–
48. Shi X, Xiao H, Azarabadi H, et al. Sorbents for the direct capture property relationships of porous materials for carbon dioxide
of CO2 from ambient air. Angew Chem Int Ed. 2020;59(18):6984- separation and capture. Energy Environ Sci. 2012;5:9849-9856.
7006. 68. Moghadam PZ, Fairen-Jimenez D, Snurr RQ. Efficient identifi-
49. Groom CR, Bruno IJ, Lightfoot MP, Ward SC. The Cambridge cation of hydrophobic MOFs: application in the capture of toxic
structural database. Acta Cryst. 2016;B72(2):171-179. industrial chemicals. J Mater Chem A. 2016;4(2):529-536.
50. Liu L, Tan S, Horikawa T, Do DD, Nicholson D, Liu J. Water 69. Watanabe T, Sholl DS. Accelerating applications of metal–
adsorption on carbon—a review. Adv Colloid Interface Sci. organic frameworks for gas adsorption and separation by com-
2017;250:64-78. putational screening of materials. Langmuir. 2012;28(40):14114-
51. Daglar H, Keskin S. Recent advances, opportunities, and chal- 14128.
lenges in high-throughput computational screening of MOFs 70. Qiao Z, Zhang K, Jiang J. In silico screening of 4764
for gas separations. Coord Chem Rev. 2020;422:213470. computation-ready, experimental metal–organic frameworks
52. Zhao J, Deng S, Zhao L, et al. Synergistic and competitive for CO2 separation. J Mater Chem A. 2016;4(6):2105-2114.
effect of H2 O on CO2 adsorption capture: mechanism expla- 71. Gomez DA, Combariza AF, Sastre G. Confinement effects in
nations based on molecular dynamic simulation. J CO2 Util. the hydrogen adsorption on paddle wheel containing metal–
2021;52:101662. organic frameworks. Phys Chem Chem Phys. 2012;14(7):2508-
53. Widom B. Some topics in the theory of fluids. J Chem Phys. 2517.
1963;39(11):2808-2812. 72. Yu C, Bourrelly S, Martineau C, et al. Functionalization of Zr-
54. Dubbeldam D, Calero S, Ellis DE, Snurr RQ. RASPA: molecu- based MOFs with alkyl and perfluoroalkyl groups: the effect on
lar simulation software for adsorption and diffusion in flexible the water sorption behavior. Dalton Trans. 2015;44(45):19687-
nanoporous materials. Mol Simul. 2016;42(2):81-101. 19692.
28 JIANG et al.

73. Noro S-I, Nakamura T. Fluorine-functionalized metal–organic 91. Uemura T, Yanai N, Kitagawa S. Polymerization reactions in
frameworks and porous coordination polymers. NPG Asia porous coordination polymers. Chem Soc Rev. 2009;38(5):1228-
Mater. 2017;9(9):e433. 1236.
74. Darunte LA, Walton KS, Sholl DS, Jones CW. CO2 capture via 92. Ding N, Li H, Feng X, et al. Partitioning MOF-5 into con-
adsorption in amine-functionalized sorbents. Curr Opin Chem fined and hydrophobic compartments for carbon capture under
Eng. 2016;12:82-90. humid conditions. J Am Chem Soc. 2016;138(32):10100-10103.
75. Noro S-I, Matsuda R, Hijikata Y, et al. High CO2 /CH4 selectiv- 93. Yu W, Gao M, Rim G, et al. Novel in-capsule synthesis
ity of a flexible copper(II) porous coordination polymer under of metal–organic framework for innovative carbon dioxide
humid conditions. ChemPlusChem. 2015;80(10):1517-1524. capture system. Green Energy Environ. 2023;8(3):767-774.
76. Ethiraj J, Albanese E, Civalleri B, et al. Carbon dioxide 94. Kim HR, Yoon T-U, Kim S-I, An J, Bae Y-S, Lee CY. Beyond
adsorption in amine-functionalized mixed-ligand metal– pristine MOFs: carbon dioxide capture by metal–organic frame-
organic frameworks of UiO-66 topology. ChemSusChem. works (MOFs)-derived porous carbon materials. RSC Adv.
2014;7(12):3382-3388. 2017;7(3):1266-1270.
77. Cui P, Ma Y-G, Li H-H, et al. Multipoint interactions enhanced 95. Banerjee R, Phan A, Wang B, et al. High-throughput synthe-
CO2 uptake: a zeolite-like zinc–tetrazole framework with 24- sis of zeolitic imidazolate frameworks and application to CO2
nuclear zinc cages. J Am Chem Soc. 2012;134(46):18892-18895. capture. Science. 2008;319(5865):939-943.
78. Bahamon D, Anlu W, Builes S, Khaleel M, Vega LF. Effect of 96. Nguyen NTT, Furukawa H, Gándara F, Nguyen HT, Cordova
amine functionalization of MOF adsorbents for enhanced CO2 KE, Yaghi OM. Selective capture of carbon dioxide under
capture and separation: a molecular simulation study. Front humid conditions by hydrophobic chabazite-type zeolitic imi-
Chem. 2021;8:574622. dazolate frameworks. Angew Chem Int Ed. 2014;53(40):10645-
79. Siegelman RL, Milner PJ, Forse AC, et al. Water enables 10648.
efficient CO2 capture from natural gas flue emissions in an 97. Chen YF, Babarao R, Sandler SI, Jiang JW. Metal−organic
oxidation-resistant diamine-appended metal–organic frame- framework MIL-101 for adsorption and effect of terminal water
work. J Am Chem Soc. 2019;141(33):13171-13186. molecules: from quantum mechanics to molecular simulation.
80. Lin J-B, Nguyen TTT, Vaidhyanathan R, et al. A scalable metal– Langmuir. 2010;26(11):8743-8750.
organic framework as a durable physisorbent for carbon dioxide 98. Yazaydın AÖ, Benin AI, Faheem SA, et al. Enhanced CO2
capture. Science. 2021;374(6574):1464-1469. adsorption in metal–organic frameworks via occupation of
81. Sagastuy-Breña M, Mileo PGM, Sánchez-González E, et al. open-metal sites by coordinated water molecules. Chem Mater.
Humidity-induced CO2 capture enhancement in Mg-CUK-1. 2009;21(8):1425-1430.
Dalton Trans. 2018;47(44):15827-15834. 99. Kim EJ, Siegelman RL, Jiang HZH, et al. Cooperative car-
82. Sánchez-Serratos M, Bayliss PA, Peralta RA, González-Zamora bon capture and steam regeneration with tetraamine-appended
E, Lima E, Ibarra IA. CO2 capture in the presence of water metal–organic frameworks. Science. 2020;369(6502):392-396.
vapour in MIL-53(Al). New J Chem. 2016;40(1):68-72. 100. McDonald TM, Mason JA, Kong X, et al. Cooperative inser-
83. Peralta RA, Alcántar-Vázquez B, Sánchez-Serratos M, tion of CO2 in diamine-appended metal–organic frameworks.
González-Zamora E, Ibarra IA. Carbon dioxide capture in Nature. 2015;519(7543):303-308.
the presence of water vapour in InOF-1. Inorg Chem Front. 101. Lin Y, Kong C, Chen L. Amine-functionalized metal–organic
2015;2(10):898-903. frameworks: structure, synthesis and applications. RSC Adv.
84. Choe JH, Park JR, Chae YS, et al. Shaping and silane coating of 2016;6(39):32598-32614.
a diamine-grafted metal–organic framework for improved CO2 102. Sharifzadeh Z, Morsali A. Amine-functionalized metal–organic
capture. Commun Mater. 2021;2:3. frameworks: from synthetic design to scrutiny in application.
85. Oschatz M, Antonietti M. A search for selectivity to enable CO2 Coord Chem Rev. 2022;459:214445.
capture with porous adsorbents. Energy Environ Sci. 2018;11:57- 103. Lin Y, Kong C, Chen L. Direct synthesis of amine-
70. functionalized MIL-101(Cr) nanoparticles and application
86. Ding M, Cai X, Jiang H-L. Improving MOF stability: approaches for CO2 capture. RSC Adv. 2012;2(16):6417-6419.
and applications. Chem Sci. 2019;10(44):10209-10230. 104. Mahdipoor HR, Halladj R, Ganji Babakhani E, Amjad-Iranagh
87. Xie L-H, Xu M-M, Liu X-M, Zhao M-J, Li J-R. Hydropho- S, Sadeghzadeh Ahari J. Synthesis, characterization, and CO2
bic metal–organic frameworks: assessment, construction, and adsorption properties of metal organic framework Fe-BDC. RSC
diverse applications. Adv Sci. 2020;7(4):1901758. Adv. 2021;11(9):5192-5203.
88. Zhu N-X, Wei Z-W, Chen C-X, et al. Self-generation of sur- 105. Flaig RW, Osborn Popp TM, Fracaroli AM, et al. The chem-
face roughness by low-surface-energy alkyl chains for highly istry of CO2 capture in an amine-functionalized metal–organic
stable superhydrophobic/superoleophilic MOFs with multiple framework under dry and humid conditions. J Am Chem Soc.
functionalities. Angew Chem Int Ed. 2019;58(47):17033-17040. 2017;139(35):12125-12128.
89. Schaate A, Dühnen S, Platz G, Lilienthal S, Schneider AM, 106. Lee WR, Hwang SY, Ryu DW, et al. Diamine-functionalized
Behrens P. A novel Zr-based porous coordination polymer con- metal–organic framework: exceptionally high CO2 capacities
taining azobenzenedicarboxylate as a linker. Eur J Inorg Chem. from ambient air and flue gas, ultrafast CO2 uptake rate, and
2012;2012(5):790-796. adsorption mechanism. Energy Environ Sci. 2014;7:744-751.
90. Qian X, Sun F, Sun J, et al. Imparting surface hydrophobicity 107. Fracaroli AM, Furukawa H, Suzuki M, et al. Metal–organic
to metal–organic frameworks using a facile solution-immersion frameworks with precisely designed interior for carbon diox-
process to enhance water stability for CO2 capture. Nanoscale. ide capture in the presence of water. J Am Chem Soc.
2017;9(5):2003-2008. 2014;136(25):8863-8866.
JIANG et al. 29

108. Li L, Jung HS, Lee JW, Kang YT. Review on applications 127. Liu X, Wang X, Kapteijn F. Water and metal–organic
of metal–organic frameworks for CO2 capture and the per- frameworks: from interaction toward utilization. Chem
formance enhancement mechanisms. Renewable Sustainable Rev. 2020;120(16):8303-8377.
Energy Rev. 2022;162:112441. 128. Birsa Čelič T, Mazaj M, Guillou N, et al. Study of hydrothermal
109. Derouane EG. Shape selectivity in catalysis by zeolites: the nest stability and water sorption characteristics of 3-dimensional
effect. J Catal. 1986;100(2):541-544. Zn-based trimesate. J Phys Chem C. 2013;117(28):14608-14617.
110. Liu R-S, Xu S, Hao G-P, Lu A-H. Recent advances of porous 129. Gul-E-Noor F, Jee B, Pöppl A, Hartmann M, Himsl D, Bertmer
solids for ultradilute CO2 capture. Chem Res Chin Univ. M. Effects of varying water adsorption on a Cu3 (BTC)2 metal–
2022;38(1):18-30. organic framework (MOF) as studied by 1 H and 13 C solid-state
111. Banerjee A, Nandi S, Nasa P, Vaidhyanathan R. Enhancing NMR spectroscopy. Phys Chem Chem Phys. 2011;13(17):7783-
the carbon capture capacities of a rigid ultra-microporous 7788.
MOF through gate-opening at low CO2 pressures assisted by 130. Kang IJ, Khan NA, Haque E, Jhung SH. Chemical and thermal
swiveling oxalate pillars. Chem Commun. 2016;52(9):1851-1854. stability of isotypic metal–organic frameworks: effect of metal
112. Vaidhyanathan R, Iremonger SS, Shimizu GK, Boyd PG, Alavi ions. Chem Eur J. 2011;17(23):6437-6442.
S, Woo TK. Direct observation and quantification of CO2 bind- 131. Colombo V, Galli S, Choi HJ, et al. High thermal and chemical
ing within an amine-functionalized nanoporous solid. Science. stability in pyrazolate-bridged metal–organic frameworks with
2010;330(6004):650-653. exposed metal sites. Chem Sci. 2011;2(7):1311-1319.
113. Shekhah O, Belmabkhout Y, Chen Z, et al. Made-to-order 132. Bordwell FG. Equilibrium acidities in dimethyl sulfoxide solu-
metal–organic frameworks for trace carbon dioxide removal tion. Acc Chem Res. 1988;21(12):456-463.
and air capture. Nat Commun. 2014;5(1):4228. 133. Zhang W, Hu Y, Ge J, Jiang H-L, Yu S-H. A facile and
114. Nugent P, Belmabkhout Y, Burd SD, et al. Porous materials general coating approach to moisture/water-resistant metal–
with optimal adsorption thermodynamics and kinetics for CO2 organic frameworks with intact porosity. J Am Chem Soc.
separation. Nature. 2013;495(7439):80-84. 2014;136(49):16978-16981.
115. Usman M, Iqbal N, Noor T, et al. Advanced strategies in metal– 134. Ding M, Jiang H-L. Improving water stability of metal organic
organic frameworks for CO2 capture and separation. Chem Rec. frameworks by a general surface hydrophobic polymerization.
2022;22(7):e202100230. CCS Chem. 2021;3(8):2740-2748.
116. Li Z, Liu P, Ou C, Dong X. Porous metal–organic frameworks 135. Wu T, Shen L, Luebbers M, et al. Enhancing the stability of
for carbon dioxide adsorption and separation at low pressure. metal–organic frameworks in humid air by incorporating water
ACS Sustain Chem Eng. 2020;8:15378-15404. repellent functional groups. Chem Commun. 2010;46(33):6120-
117. Wang J, Zhang Y, Su Y, et al. Fine pore engineering in a series of 6122.
isoreticular metal–organic frameworks for efficient C2 H2 /CO2 136. Bellarosa L, Gutiérrez-Sevillano JJ, Calero S, López N. How
separation. Nat Commun. 2022;13(1):200. ligands improve the hydrothermal stability and affect the
118. Cui X, Chen K, Xing H, et al. Pore chemistry and size control adsorption in the IRMOF family. Phys Chem Chem Phys.
in hybrid porous materials for acetylene capture from ethylene. 2013;15(40):17696-17704.
Science. 2016;353(6295):141-144. 137. Cavka JH, Jakobsen S, Olsbye U, et al. A new zirconium
119. Li J-R, Yu J, Lu W, et al. Porous materials with pre-designed inorganic building brick forming metal organic frameworks
single-molecule traps for CO2 selective adsorption. Nat Com- with exceptional stability. J Am Chem Soc. 2008;130(42):13850-
mun. 2013;4(1):1538. 13851.
120. Saputro A, Agusta M, Yuliarto B, Dipojono H, Maezono 138. Dan-Hardi M, Serre C, Frot T, et al. A new photoactive crys-
R. Density functional study of adsorptions of CO2 , NO2 talline highly porous titanium(IV) dicarboxylate. J Am Chem
and SO2 molecules on Zn (0002) surfaces. J Phys Conf Ser. Soc. 2009;131(31):10857-10859.
2016;739:012080. 139. Tang Y, Kourtellaris A, Tasiopoulos AJ, et al. Selective CO2
121. Zhao J, Deng S, Zhao L, et al. Synergistic and competitive adsorption in water-stable alkaline-earth based metal–organic
effect of H2 O on CO2 adsorption capture: mechanism expla- frameworks. Inorg Chem Front. 2018;5(3):541-549.
nations based on molecular dynamic simulation. J CO2 Util. 140. Zhang X, Wang B, Alsalme A, Xiang S, Zhang Z, Chen B. Design
2021;52:101662. and applications of water-stable metal–organic frameworks:
122. Low JJ, Benin AI, Jakubczak P, Abrahamian JF, Faheem SA, status and challenges. Coord Chem Rev. 2020;423:213507.
Willis RR. Virtual high throughput screening confirmed exper- 141. He T, Kong X-J, Li J-R. Chemically stable metal–organic frame-
imentally: porous coordination polymer hydration. J Am Chem works: rational construction and application expansion. Acc
Soc. 2009;131(43):15834-15842. Chem Res. 2021;54(15):3083-3094.
123. Dinh T-V, Kim J-C. Moisture removal techniques for a con- 142. Lv X-L, Yuan S, Xie L-H, et al. Ligand rigidification for enhanc-
tinuous emission monitoring system: a review. Atmosphere. ing the stability of metal–organic frameworks. J Am Chem Soc.
2021;12(1):61. 2019;141(26):10283-10293.
124. Zuluaga S, Tan K, Nijem N, et al. Water interactions in metal 143. Wang K, Lv X-L, Feng D, et al. Pyrazolate-based porphyrinic
organic frameworks. CrystEngComm. 2015;17:247-260. metal–organic framework with extraordinary base-resistance.
125. Burtch NC, Jasuja H, Walton KS. Water stability and adsorption J Am Chem Soc. 2016;138(3):914-919.
in metal–organic frameworks. Chem Rev. 2014;114(20):10575- 144. Veneman R, Kamphuis H, Brilman DWF. Post-combustion CO2
10612. capture using supported amine sorbents: a process integration
126. Wang C, Liu X, Keser Demir N, Chen JP, Li K. Applica- study. Energy Proc. 2013;37:2100-2108.
tions of water stable metal–organic frameworks. Chem Soc Rev. 145. Bahamon D, Díaz-Márquez A, Gamallo P, Vega LF. Energetic
2016;45(18):5107-5134. evaluation of swing adsorption processes for CO2 capture in
30 JIANG et al.

selected MOFs and zeolites: effect of impurities. Chem Eng J.

2018;342:458-473. ciate, assistant professor, and lecturer positions at
146. Hossain M, Cunningham J, Becker T, et al. Impact of MOF Shanghai Jiao Tong University, Newcastle University,
defects on the binary adsorption of CO2 and water in UiO-66. Durham University, and the University of Aberdeen,
Chem Eng Sci. 2019;203:346-357. and mainly engaged in basic scientific and applied tech-
147. Young J, McIlwaine F, Smit B, Garcia S, van der Spek M. nology research in areas such as efficient heat and cold
Process-informed adsorbent design guidelines for direct air storage technology, carbon capture and resource uti-
capture. Chem Eng J. 2023;456:141035.
lization technology, efficient food drying technology,
148. Hu X, Brandani S, Benin AI, Willis RR. Testing the stability of
and waste heat power generation technology.
novel adsorbents for carbon capture applications using the zero
length column technique. Chem Eng Res Des. 2018;131:406-413.
Jinyuan Yong graduated from
149. Kizzie AC, Wong-Foy AG, Matzger AJ. Effect of humidity
on the performance of microporous coordination polymers as Chongqing University in 2021 wiht
adsorbents for CO2 capture. Langmuir. 2011;27(10):6368-6373. a a bachelor’s degree. Currently,
150. Mangano E, Kahr J, Wright PA, Brandani S. Accelerated degra- she is pursuing a master’s degree
dation of MOFs under flue gas conditions. Faraday Discuss. at Zhejiang University under the
2016;192:181-195. supervision of Dr. Long Jiang. Her
151. Li G, Xiao P, Webley P, Zhang J, Singh R, Marshall M. Cap- research focuses on carbon cap-
ture of CO2 from high humidity flue gas by vacuum swing
ture from wet flue gas using a
adsorption with zeolite 13X. Adsorption. 2008;14(2-3):415-422.
diamine-appended metal-organic framework.
152. Lesch DA. Carbon Dioxide Removal from Flue Gas Using
Microporous Metal Organic Frameworks. United States. 2010.
Zongbi Bao is a professor at Zhe-
Accessed May 5, 2023. https://www.osti.gov/servlets/purl/
1003992 jiang University. He received his
Bachelor’s and PhD degrees in
Chemical Engineering from Zhe-
jiang University in 2003 and 2008,
How to cite this article: Jiang L, Yong J, Xie R, respectively. He then worked as a
et al. Screening, preparation, and prototyping of postdoc at Zhejiang University and
metal–organic frameworks for adsorptive carbon New Mexico State University from
capture under humid conditions. SusMat. 2009 to 2010 and joined the College of Chemical and
2023;1-30. https://doi.org/10.1002/sus2.154 Biological Engineering at Zhejiang University in 2011.
His research interests include (1) synthesis and appli-
cations of MOFs, carbon-based materials, and chiral
stationary phase for adsorption and chromatography;
AU T H O R B I O G R A P H I E S
(2) adsorption processes for gas separation and purifi-
cation; (3) green separation processes in ionic liquids or
Dr. Long Jiang is currently a
supercritical fluids; and (4) extraction and purification
researcher at Zhejiang University’s
of bioactive compounds from biomass.
“Hundred Talents Program.” He
was graduated from Shanghai Jiao
Tong University in 2015 with a PhD
in Refrigeration and Cryogenic
Engineering. From 2015 to 2020, he
held postdoctoral, research asso-