Chem Soc Rev

View Article Online

TUTORIAL REVIEW View Journal | View Issue
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

Understanding and controlling the nucleation

and growth of metal–organic frameworks
Cite this: Chem. Soc. Rev., 2023,
52, 6918
Brooke P. Carpenter, A. Rain Talosig, Ben Rose, Giuseppe Di Palma and
Joseph P. Patterson *

Metal–organic frameworks offer a diverse landscape of building blocks to design high performance
materials for implications in almost every major industry. With this diversity stems complex crystallization
mechanisms with various pathways and intermediates. Crystallization studies have been key to the
advancement of countless biological and synthetic systems, with MOFs being no exception. This
review provides an overview of the current theories and fundamental chemistry used to decipher MOF
crystallization. We then discuss how intrinsic and extrinsic synthetic parameters can be used as tools to
Received 26th April 2023 modulate the crystallization pathway to produce MOF crystals with finely tuned physical and chemical
DOI: 10.1039/d3cs00312d properties. Experimental and computational methods are provided to guide the probing of MOF crystal
formation on the molecular and bulk scale. Lastly, we summarize the recent major advances in the field
rsc.li/chem-soc-rev and our outlook on the exciting future of MOF crystallization.

Key learning points

1. Current theories and fundaments for deciphering MOF crystallization.
2. Intrinsic synthetic parameters for modulating MOF crystallization pathways.
3. Extrinsic synthetic parameters for modulating MOF crystallization pathways.
4. Experimental and computation methods for monitoring MOF crystal formation.
5. New advances and outlooks for MOF crystallization.

1. Introduction crystals with varying chemical functionality and porosity, allow-

ing small molecules and biomolecules to be selectively
Metal–organic frameworks (MOFs) have become one of the adsorbed or encapsulated.6–8 Because of this synthetic diversity
largest and fastest growing research fields since the material and ease of tunability, MOFs are now being studied for applica-
was first introduced in the late 1990’s.1 Composed of metal tions in almost every major industry, including catalysis,9 gas
nodes connected with organic ligands, MOFs have attracted storage,10 sensors,11 and drug delivery.12,13
popularity due to their incredibly high surface area, with some While high throughput screening methods can be credited
obtaining surface areas greater than the area of a football field for accelerating the synthesis of new MOFs (490 000 recorded
per gram of material (B7800 m2 g1).2 These high surface areas MOFs),14 currently only a limited number of MOFs are actually
make MOFs perfect for storage of large amounts of gaseous in industrial usage due to limited knowledge in how synthetic
guest species such as hydrogen3 or carbon dioxide.4 MOFs also scaling and how high temperature and pressure processing
enjoy greater synthetic flexibility when compared to other techniques affect MOF formation and final structure.15 The
porous materials like zeolites, which are currently used in zeolite community—which has exploded in catalysis applications
many applications but have limited catalytic usage due to a in the past two decades—has demonstrated that mechanistic
narrow range of synthetic precursors.5 In contrast, MOF coor- studies on how a material forms enables its advancement, even
dination networks can be synthesized from many different with the limited catalytic applications this material has when
transition metals and linkers and can be easily tuned to obtain compared to MOFs.16–18 Additionally, the commercialization of
most pharmaceutical drugs can be largely credited to mechanistic
studies of drug formation as the studies were used to tailor drugs
Department of Chemistry, University of California, Irvine, Irvine, CA 92697-2025, with specific surface chemistries and properties.19,20 Most studies
USA. E-mail: [email protected] on the nucleation and growth of MOFs have been from the

6918 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

perspective of crystal growth, focusing on the rate of crystal than a molecular chemistry approach, focusing on bonding.21
formation, precursor phases, and crystal morphologies, rather Extensive mechanistic studies have been performed only on a
limited number of MOF systems, and they have found MOFs
formations to often be complex, not fitting conventional
crystallization models.22,23 Large-scale studies need to be per-
Brooke Carpenter obtained her PhD
formed to generalize these findings, which is challenging due
in chemistry from the University of
to multiple intermediate phases occurring for each MOF syn-
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

California, Irvine (UCI) in 2023

thetic condition. These phases need to be identified and
while under the advisement of Prof.
characterized to better predict dominant MOF pathways. Addi-
Joseph Patterson. Her research
tionally, how the reaction solution influences the formation
interests at UCI pertained to
and dynamics of the various intermediate phases has yet to be
studying the nucleation and growth
fully realized. Thus, a combination of ex situ and in situ studies
mechanisms of protein@MOFs with
are essential to gain a holistic understanding of the reaction
a particular focus on the protein
dynamics and mechanisms of MOF synthesis on both the
secondary structure. Prior to UCI,
molecular and bulk scale. Such mechanistic studies will enable
she received her undergraduate
tighter control of MOF synthetic conditions and provide trans-
degree in Polymer Science from the
latable findings for other materials. This review encompasses
Brooke P. Carpenter University of Southern Mississippi in
how we currently understand, control, and monitor MOF
2018 while under the advisement of
crystallization mechanisms, specifically focusing on zeolitic
Prof. Sarah Morgan. Her research interests at USM pertained to
imidazolate frameworks (ZIFs), a class of MOFs which are
synthesizing polymers via aqueous RAFT polymerization for
structurally similar to zeolites and are widely studied.
determination of structure/binding interactions with food-based proteins.

Rain Talosig is a PhD candidate in Ben Rose is a 4th year under-

the chemistry department at the Uni- graduate researcher double major-
versity of California, Irvine (UCI) ing in chemistry and molecular
and is working under the advise- biology at the University of
ment of Dr Joseph Patterson. Her California, Irvine. He studies in
research focuses on understanding the Patterson Group, where he
the polymorph control of MOF has been working since 2021,
systems with a focus on probing investigating the effects of
the initial stages of nucleation and protein surface chemistry and
growth. Rain received her BS in metal-linker ratio on enzyme@
chemistry from Ithaca College in MOF crystal structure, morpho-
2019. While there, she worked logy, and activity.
A. Rain Talosig under the advisement of Dr Anna Ben Rose
Larsen on synthesis and charac-
terization of rhodium pincer ligands.

Giuseppe Di Palma is a postdoc Joe Patterson is an Assistant Prof-

in the Patterson Lab at the essor in the Chemistry Department
University of California, Irvine. at the University of California,
His interests are in metal- Irvine (UCI). He obtained his
organic frameworks, particularly undergraduate and PhD at the
oriented to cyclodextrin-based University of York and Warwick
MOFs. He obtained his PhD in respectively in the UK. He did
Chemical Engineering in 2020 at postdoctoral research at both the
the University of Birmingham, University of California, San Diego
UK, working on designing new and the Eindhoven University of
biosensors. Technology. His main research
interest is the development of
Giuseppe Di Palma Joseph P. Patterson electron microscopy methods to
provide a better understanding of
molecular self-assembly.

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6919
 View Article Online

Chem Soc Rev Tutorial Review

Additionally, we will provide our outlook on the future direc- and precursors remain in solution or bulk phase. If the inter-
tions of the field. facial energy is reduced, the nucleation barrier becomes man-
ageable, and crystal growth can proceed. Crystal growth occurs
on the downhill slope of the energy barrier and involves the
2. Understanding MOF nucleation and addition of monomers to the surface of the growing crystal
growth lattice. The key principle for crystal growth is that monomers
will tend to add to different crystal faces at different rates,
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

2.1. Nucleation and growth theories and models which controls the morphology of the crystal as it grows. This
Crystallization studies have proven to be an important area of growth often occurs through Ostwald Ripening, which refers to
research in synthetic materials such as MOFs and across many a phenomenon in which smaller particles of MOF precursors
natural systems occurring in biology and geology. Through dissolve and add onto the surface of the growing crystalline
these studies, scientists have extracted key thermodynamic phase.28
and kinetic factors that unlock how atoms, ions, and molecules 2.1.2. Nonclassical nucleation models. While CNT is essen-
can pack to obtain final crystals with targeted properties. Three tial to understanding how MOFs form, it has only been able to
broad approaches are used, either alone or in combination, to describe a few MOF systems.29–31 Most MOF crystallizations are
model MOF nucleation and growth: classical nucleation theory, categorized as non-classical, as they include intermediates and
nonclassical nucleation models, and the SBU model. irregular final crystal structures and morphologies that cannot
2.1.1. Classical nucleation theory. Since the early 20th be explained with CNT. Additionally, MOF crystals often form
century, classical nucleation theory (CNT) has served as a metastable phases, which are a local energy minima in the MOF
theoretical framework for understanding the rates and formation reaction (Fig. 1).22,32 These metastable phases have
mechanisms during crystallization.24–26 The theory breaks broad definitions as they can be characterized as molecular
crystallization down into two main categories: nucleation and clusters,33 crystalline nanoparticles,31 liquids,22 kinetic poly-
crystal growth. Nucleation describes the assembly of monomers morphs (which later evolve to a more thermodynamically
(atoms, ions, or molecules) into the smallest thermodynami- stable, often nonporous polymorph),34 and amorphous
cally stable structure which possesses a crystalline lattice, species.23 The behavior of this metastable phase has significant
known as a crystal nucleus. The key principle for nucleation consequences of the final MOF’s characteristics, as it aids in
is that there is a significant energy barrier to the formation of establishing a degree of local supersaturation for the critical
stable nuclei. This barrier is determined by two competing nucleus to form and start the nucleation process. Metastable
factors: surface or interfacial energy and bulk energy.27 The phases have been observed to form when precursors concen-
interfacial energy is always a positive term as it is the origin of trate and phase separate from the bulk solution to form solute-
the nucleation barrier due to the fact that it requires energy to rich and solute-poor areas. The solute-rich area then condenses
make an interface. The bulk energy is always a negative term as into species such as amorphous particles and/or dense-liquid
it denotes a release in energy or stabilization of the nuclei. The phases,23 which serve as the metastable phase (Fig. 1). Once the
idea is to use enthalpy and entropic strategies such as manip- metastable phase is formed, an energy barrier must be over-
ulating bond strengths and degrees of freedom of MOF pre- come for the metastable phase to transition into a critical
cursors to minimize the interfacial energy term. If the nucleus for the final crystalline phase. While this transition is
interfacial energy is too high, a nucleus cannot be formed challenging to probe, the critical nucleus can form through the

Fig. 1 Free energy diagram comparing monomer addition (classical nucleation theory) and nucleation through metastable phase (nonclassical
nucleation theory) mechanisms.

6920 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

aggregation of the metastable phase, which is based on attrac- and mixed geometry of SBUs found in solution during
tive forces of phases at close distances. If the metastable phase nucleation.50
is of amorphous character, this transition is often called the SBUs can arrange into a large library of geometries which
amorphous-to-crystalline (ATC) pathway, which is a common have direct implications on the structural properties of a
mechanism observed in zeolites,35 biomimetic polymers,36 and crystal.41 The stability, structure, and rigidity of SBUs can be
magnetite.37 Additionally, this transition into a critical nucleus tailored using factors further explored in section III, offering a
can occur when disordered metal–ligand–metal bonds rear- promising area of MOF research. As the smallest possible
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

range into an ordered, crystalline structure. In this case, assembly of MOF precursors, SBUs dictate the fundamental
nucleation occurs within the rearrangement of metal–ligand– structure of the MOFs they make up. Thus, through the design
metal bonds. Once a critical nucleus has been reached, MOFs of SBUs with appropriate geometries and sizes, final MOF
have been observed to grow through particle attachment,23 properties such as network topology, stability and porosity
where amorphous particles attach to the growing crystal in no can be controlled. This rational design of MOFs is a departure
particular orientated preference. Additionally, they can grow from previous high-throughput brute force methods which
through oriented attachment (OA),31 where crystalline precur- have been responsible for many MOF discoveries.53 As an
sors attach to the critical nucleus through alignment of their example, Park et al. used the principles of SBU design to
lattices to the growing crystal.38 produce MOFs with a new topology, ssp, by manipulating the
2.1.3. The secondary building unit (SBU) model. The structure of the organic linker to include a delocalized electron
nucleation and growth of MOFs can also be described by the hole, showing that rational design of MOFs for specific uses is
secondary building unit (SBU) approach.39–41 SBUs are the possible.54 This concept of rational MOF design has already
molecular complexes from which the extended frameworks of transformed the understanding of surface effects on MOF
MOFs are derived, and as such, they are the smallest possible nucleation and growth, where the SBU approach has been used
assemblies of MOF precursors, consisting of single or multiple in BSA@ZIF-8 studies to model the interactions between MOF
metal nodes coordinated to organic linkers.42 In these coordi- precursors and biomolecules in order to inform encapsulation.55
nation complexes, the metal node acts as the vertex and the MOF growth from SBUs can be described by two mechan-
organic linkers act as the edges that join the vertices together, isms analogous to polymer growth: chain or step growth.31
locking the metal node into a fixed geometry which lends itself In chain growth mechanisms, single SBUs add like monomers
to the rigid framework. SBUs can be classified by points of to the growing ends and facets of a crystal. In step growth
extension, defined as the number of possible metal connec- mechanisms, SBU dimers, longer oligomers, and/or smaller
tions between the organic linkers.43 SBUs are also sometimes crystals add to each other through oriented attachment.24 The
referred to as prenucleation clusters (PNCs) – which can refer latter mechanism has been probed in the ZIF-67 system
to any species present in MOF synthesis reactions before through molecular dynamics simulation, optical spectroscopy,
nucleation – but here, PNCs which contain only metal nodes and ESI-MS, giving insight into the molecular interactions that
and organic linkers are referred to as SBUs. take place during SBU addition.50 This study also found that
In the SBU model of MOF nucleation and growth, SBUs form complexes which can initiate nucleation are different from
via coordination chemistry between nodes and linkers and those which can initiate growth, both in how they form and in
aggregate via coordination linkages into a critical nucleus. This structure.50 However,this is not always the case. In the UiO-66
critical nucleus then grows into the final MOF crystal through system, SBUs form a critical nucleus upon heating, and add to
the incorporation of more SBUs. This model is supported by a the growing nucleus, ordering themselves into the final MOF
study of the MIL-n MOF family by Fe K-edge extended X-ray with time.48 Understanding how SBUs form, aggregate and
absorption fine structure (EXAFS), which revealed that SBUs are incorporate into MOFs is crucial to describing the growth
present during both the nucleation and growth of MOFs, and, especially nucleation of these materials, and we expect
retaining their connectivity and structure throughout.44 The SBU formation to continue to play a key role in discovery of new
formation of the SBUs themselves has been studied in systems MOFs and control of established ones.
including Mg-carboxylate MOFs,45 ZIF-8,46 MIL-101,47 UiO-66,48
In-MIL-68,49 ZIF-67,50 MOF-2/MOF-5,51 and SIFSIX-3-Zn52 using 2.2. Kinetic models of MOFs
a variety of techniques from electrospray ionization-mass spec- While the previously mentioned models describe the mecha-
trometry (ESI-MS) and UV-visible spectroscopy to pair distribu- nics of MOFs nucleation and crystal growth, it is also necessary
tion function analysis and single molecule electron microscopy. to investigate their crystallization kinetics. X-ray diffraction techni-
These studies have shown that SBUs can assemble through a ques (XRD) (described in Section 4) are frequently used to exploit
variety of mechanisms, such as ligand exchange and elimina- MOF crystallization kinetics. Data from this technique are coupled
tion at a single metal node,50 chain growth from a single metal to fit two quantitative kinetic models: Avrami and Gualtieri.
node and linker,49 or organic linker coordination to inorganic Originally, Avrami was a model used to describe crystallization by
metal clusters.48 These studies have also shown that not all solid-state transformation, and it later evolved to describe sol-
MOFs follow the SBU model; specifically, the critical nucleus of vothermal crystallization. The Avrami equation is expressed as:
ZIFs like ZIF-67 is not likely formed from SBU aggregation due
to the tetrahedral geometry of the metal node in the final MOF a = 1  exp[(kt)n] (1)

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6921
 View Article Online

Chem Soc Rev Tutorial Review

where a is the extent of crystallization, k is the crystal growth The model breaks the mechanisms down into 3 main steps:
rate constant, t is the reaction time, and n is the Avrami initiation of SBUs, chain and step growth, and termination.
exponent related to the dimensionality and the crystal growth. By breaking the mechanisms into 3 different steps, the activity
A significant limitation of this model is that it does not make a energy barrier for initiation can be solved along with the rate
distinction between the nucleation and the growth step. The constants for both the initiation and growth reactions.31 A key
Avrami model assumes that crystallization occurs by adding benefit of the microkinetic model is that it can further relate
individual molecules at a constant rate.56 In contrast to the the nucleation rate to reaction conditions such as temperature
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

Avrami model, Gualtieri separates the rates of nucleation and and concentration of precursors using the following equation:
crystal growth, as it was originally developed for zeolite synthe-  
sis analysis.57,58 The Gualtieri model can expressed by the _ TÞ ¼ k0;n xAn ð1  xÞBn exp DGn
Nðx; (4)
RT
equation:
:
1  
 In eqn (4), N represents the rate of nucleation in units of
a¼    1  exp  kg t nG (2) per second while x is the reaction extent. T represents the
t  aN
1 þ exp  temperature in units of Kelvin, and k0,n is the pre-exponential
bN
factor of the nucleation rate constant kn in units of per second.
The Gualtieri expression describes the extent of crystallization An and Bn are rate orders with An being with respect to the
(a) as a function of time (t), growth rate constant (kg), and reaction extent and Bn with respect to the limiting reactant. DGn
crystal growth dimensionality (n). For example, if a crystal with (J mol1) represents the activation energy required for nuclea-
tetrahedral geometry forms, the growth dimensionality (n) tion, and R (J mol1 K1) is the gas constant.
would equal 3 as this geometry is 3-dimensional. The fitting For example, the authors who created this equation demon-
parameters aN, which is the reciprocal of the nucleation rate strated that by increasing the temperature of a system, the ratio
(kN), and bN, which is variance of the probability of nucleation of the chain to initiation rate can be used to predict the
distribution, can then be used to calculate the nucleation and induction period and concentration of SBUs that can be used
crystallization probability.59 A bN value below 15 indicates in the growth step. Additionally, the equation can be used to
heterogeneous nucleation; a value of 20 indicates homogenous determine how increasing or decreasing the concentration of
nucleations; and a value greater than 20 indicates autocatalytic MOF precursors affects the initiation rate and concentration of
nucleations.21 The nucleation probability is determined using SBUs that can be used in the growth step. The microkinetic
the fitting parameters aN and bN using the following equation: model can further extract information regarding rate of MOF
  growth and the reaction order with the following equation:
ðt  aÞ2
PN ¼ exp  (3)  
2b2 DGg
Gðx; TÞ ¼ k0;n xAg ð1  xÞBg exp  (5)
RT
While the majority of the experimental data used to fit into
these models are obtained from XRD techniques, a study by In eqn (5), G represents the rate of growth in units of meters
Filez et demonstrated that the Gualtieri model could be fit to per second. k0,g is the pre-exponential factor of the growth rate
curves in an in situ VIS spectra. Through the model, the constant kg in units of per second. Ag and Bg are rate orders with
nucleation distributions could be predicted. Additionally by Ag being with respect to the reaction extent and Bgwith respect
fitting the model to curves from different synthetic ratios of the to the limiting reactant. DGg (J mol1) represents the activation
ligand:metal, trends relating to how change in ligand:metal barrier required for growth.
affects nucleation and growth rates could be determined.50 By extracting the growth rates from this equation, crystal
In summary, while the Avrami and Gualtieri models have been size and dispersity can be predicted. Furthermore, the micro-
extensively applied to MOF crystallization, Avrami is often kinetic model can be used to describe mechanisms leading to
inadequate in characterizing MOF kinetics due to it being too crystal termination. For example, the authors who created this
general. Gualtieri, on the other hand, is more specific and can equation demonstrated that the termination step was caused by
make distinctions between the nucleation and crystal growth decreasing the reactivity of the MOF surface. While the micro-
rates.21 However, information relating to the nucleation energy kinetic model definitely provides significant mechanistic
barrier and to how synthetic parameters such as concentration insights, Flory’s approximation is used to predict the bond
and temperature affect the barriers cannot be obtained from strength for the growth step of the model, which does not take
either model. into account size dependence of the clusters or aggregates as
Within the last few years, the microkinetic model has been the size of aggregates can influence overall binding strengths.
introduced, which takes into account the SBU approach rather Additionally, as MOFs can develop crystal defects, future work
than the traditional classical and nonclassical nucleation and is needed to develop a kinetic model that can integrate defect
growth theories. While the Avrami and Gualtieri models focus mechanisms.
on nucleation and crystallization rate and probability, and still
provide useful information for MOF nucleation and growth, 2.3. Polymorphism in MOFs
the microkinetic model extrapolates fundamental mechanisms MOFs can exhibit polymorphism, where MOF crystals can
occurring at the different steps in a crystal’s formation. consist of identical chemical composition but differ in their

6922 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

network topology.60 As a result, multiple crystalline phases 2.4. Summary

of MOFs can be produced with different symmetries and The field of nucleation and growth is well established, and
densities. By tuning the activation energy of the MOF through while significant progress has been made in understanding the
changing synthesis conditions, polymorphs of varying nucleation and growth of MOFs, much work still needs to be
stabilities can be formed; thus polymorphs can be found at done. While all MOFs are assembled via coordination chem-
different energy minimas of a free energy landscape istry, the mechanism by which this occurs can be different for
diagram.61 Since polymorphs can display different crystal
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

each MOF and can change with different synthesis conditions.

structures while retaining identical molecular composition, Furthermore, the nucleation and growth of some MOFs can be
the structure–function relationship based on structural para- described using classical nucleation theory, while others
meters can be investigated.62,63 Polymorph transition can be require different theoretical frameworks like SBU addition
induced post-synthesis through physical steps like additional and/or OA, or a combination of several. In a similar fashion,
heating,64 pressure, or template mediation.65 Typically, the mathematical models used to properly describe both MOF
post-synthetic phase transitions result in the formation of nucleation and growth are not universal; different MOF systems
more dense, thermodynamic crystalline phases through can be described by different equations. Understanding the
reconstruction or displacement of internal structures.62 For different mechanisms of different MOF systems is essential in
example, washing of ZIF-8 MOFs synthesized in water with the development of a unifying theory describing the nucleation
alcohol solvents can initiate phase change from ZIF-C and and growth of these exciting materials. Specifically, more work
ZIF-L to ZIF-8 (sod).66,67 Additionally, intrinsic and extrinsic is needed to understand the molecular chemistry for all MOF
synthetic conditions can control the competition among poly- classes, and more work should be done to expand nucleation
morph phases during the MOF formation.68 For example, the and growth for different classes of MOFs. In the case of ZIFs,
use of template vapors in chemical vapor deposition enables while much is known about the nucleation and growth of these
control over the formation of a porous or nonporous ZIF composites, many unanswered questions still exist. The for-
polymorph.65 mation mechanisms of many ZIF polymorphs—such as the
While polymorphism is often not screened in MOF studies, thermodynamically formed dia and carbonate-incorporating
mechanistic studies of zeolites, which have B250 different ZIF-C—are not understood. We predict that further examina-
polymorphs,69 have aided in polymorph control of MOFs, tion of the nucleation and growth of ZIFs specifically, perhaps
as the metal–ligand coordination in ZIFs are similar to the guided by previous work studying zeolites, will help in the
building units of inorganic zeolites. Whereas zeolites are understanding of similar the nucleation and growth of other
composed of SiO4 and AlO4 units, ZIFs have tetrahedral metal MOF systems like metal azolate frameworks (MAFs).
centers, commonly divalent metals such as Zn2+ or Co2+, that
are bridged with imidazolate ligands. Consequently, ZIFs share
many of the same network topologies and rich polymorphisms 3. Controlling MOF nucleation and
as zeolites.70 In all formation reactions of zeolite polymorphs, growth
primary amorphous phases form, followed by secondary amor-
phous phases. Recent computational studies have determined MOF nucleation and growth reactions can be controlled
the rate in which this secondary amorphous phase grows into a through manipulating phase changes during synthesis. Phase
crystalline phase determines the final polymorph selection.71 change of a system occurs when the total free energy of the
Polymorphs with distinct physical properties can be produced proceeding phase is less than the total free energies of the prior
by altering the packing density of the building units as well as phase. Intrinsic and extrinsic parameters can dictate whether a
the orientation of the imidazolate units. For example, ZIF-8 phase transformation is energetically favorable through
sodalite (sod), which is the most widely studied kinetic poly- increasing or decreasing the free energy barrier to nucleate a
morph of ZIF-8, is the least densely packed, resulting in large particular phase. Intrinsic factors refer to the internal proper-
pores (B11.4 Å) that are ideal for encapsulating guest ties within a MOF system, such as the chemical composition
molecules.72 In contrast, ZIF-8 diamondoid (dia), which is the and concentration of the precursors, along with reaction tem-
most thermodynamically stable polymorph, is the most densely perature and medium. Extrinsic factors include the presence of
packed and not ideal for encapsulating guest molecules. Poly- foreign surfaces, small molecule modulators, and confinement,
morphs of ZIFs can be revealed through in situ monitoring which can directly affect the surface energy barrier to nucleation.
techniques such as in situ XRD, which allowed for the observa- Extrinsic factors frequently control the rates and mechanisms of
tion of a new metastable ZIF-8 polymorph kat, generated MOF by favoring a type of nucleation called heterogeneous
through mechanochemical milling.34 Additionally, electron nucleation, which occurs outside of the bulk solution such as
microscopy studies have shown amorphous particles emerge on a foreign surface. Heterogeneous nucleation enables nuclea-
for the formation of ZIF-8 (sod) and dia polymorphs.23 While tion to occur at lower supersaturation conditions, resulting in
these microscopy studies have also shown that the amorphous increased rates of nucleation compared to nucleation occurring
particles undergo a redissolution–crystallization mechanism in the bulk solution (homogeneous nucleation).73 Much has
for sod, a separate mechanism is suggested for dia formation been written about the effects of different synthetic conditions
that requires further exploration. on final MOF size and quality. In this section, we use these final

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6923
 View Article Online

Chem Soc Rev Tutorial Review

properties to extrapolate information about MOF nucleation and Different linkers and metals will display different nucleation
growth mechanisms. Some properties mentioned in this review, and growth kinetics, the understanding of which varies
which can be tied to nucleation and growth, include metal– between MOF systems.
ligand bond strength (which relates to both nucleation prob- Metal–ligand bond strength cannot be discussed in a
ability and to growth),74 crystal size (which can relate to the vacuum; the reaction solution, and specifically the solvent, is
relative nucleation and crystal growth rates),75 crystalline defects critical for dictating the metal and ligand coordination
(which indicate errors in SBU formation and/or arrested growth), environments.86 Furthermore, MOF’s can only self-assemble
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

and crystal polymorph (which can help determine whether a into crystalline units if the metal–ligand bond formation is
reaction is thermodynamically or kinetically controlled).23 Both more energetically favorable than the solvent–ligand exchange
intrinsic and extrinsic factors will be discussed in more detail in rate. Solvent can play various roles in the self-assembly process.
this section. It can govern the rate and pathway of crystal formation, be
incorporated into the MOF by coordinating with the metal ions,
3.1. Chemical composition and medium or both. In the case of water-stable ZIFs, the DG of hydrolysis
Both MOF composition and solvent play significant roles in must be a positive value, meaning that metal–ligand binding is
both the nucleation and growth of MOFs. The strength of the more energetically favorable than the metal or ligand binding
metal–ligand bond, which is essential to the assembly of stable with water.82 By modulating the hydrophobicity of the ligand,
SBUs, can be modulated based on electron sharing between the the rate of ligand/solvent binding can further be tuned, as
ligand and metal. When choosing a ligand, factors to consider hydrophobic ligands have a lower rate of binding with water
are the electron-withdrawing and electron-donating effects. For molecules.
example, ZIF-8 has a stronger Zn–N coordination than ZIF-108, The solubility of the ligand in the reaction medium plays an
as the ligand in ZIF-108, 2-nitroimidazole, has a greater important role in the crystallization rate as a more soluble
electron-withdrawing effect on the imidazole ring, and thus ligand facilitates increased nucleation and growth rate. For
greater Zn–N bond length than the ligand in ZIF-8, 2-methyl- example, a study demonstrated this concept by using a ligand
imidazole.76 The different strength of the Zn–N bond in differ- more soluble in DMF than in water and showed how the MOF
ent Zn-based imidazolate MOFs provides different applications crystallized much faster in DMF than water due to the ligand’s
for the materials; while stronger Zn–N bonds are desired for high solubility in DMF.87 The solubility of the ligand also can
storage applications, weaker bonds can facilitate reversible phase impact the network dimension of a MOF as it influences the
transformation, which broadens the material’s applicability.77 binding affinity of the solvent molecules to the metal centers.
Indeed, ZIF-108 undergoes metal substitution much easier than For example, a study showed that through increasing ligand
ZIF-8 does, the latter of which requires a base to prevent phase solubility, the solvent–metal binding consequently decreased,
transformation during this process.76,78 This suggests that the which resulted in an increase of the crystal dimensionality
rate of growth should be higher for ZIF-8, as hydrolysis-resistant (i.e. 1D, 2D, 3D).88
metal–ligand bonds lead to higher crystal growth rates. When The rate and extent of deprotonation of the ligand further
choosing a metal, the Irving–Williams series can be used to affects the self-assembly and final MOF structure and size.
determine the strength of divalent metals complexes in A study by Kim et al. observed larger particles formed in more
water.79,80 This series is often explained based on the ionic polar solvents as more polar solvents have higher proton
radius of a divalent metal and the crystal field stabilization acceptance power.89 As a result, the interfacial energy of the
energy of a complex. For example, Cu2+ is expected to form a system decreases as the concentration of available ligands for
more stable metal complex than Zn2+ based on the smaller metal coordination increases. With a higher concentration of
ionic radius of Cu2+. Of course, to measure MOF nucleation and available ligands, larger MOFs with high crystallinity can form.
growth, a given metal and linker must be able to form a MOF. This is not only seen in systems with changing solvent polarity;
While this may seem obvious, many metal–ligand combina- in MOFs, lower metal–ligand binding (growth) rate equals
tions will not produce MOFs, meaning selection of a linker for a larger particle size (Fig. 2).
corresponding metal node must be done with care. When Functionalizing organic ligands allows for electronic and
designing the metal–ligand combination for a MOF, Pearsons’s steric effects during self-assembly and in the final MOF product
hard-soft acids and bases (HSAB) theory must be utilized, as the to be finely tuned. For example, by combining two complemen-
binding strength and stability of the metal and ligand bond tary imidazole ligands, a study showed the organization and
drives the self-assembly process and dictates the final perfor- conformation of ligands in the initial building units could be
mance of a MOF.81 Hard metals such as Zr4+ form strong tuned based on ligand size, as this influences ligand–ligand
metal–ligand complexes with strong bases such as carboxylate interactions and, consequently, sterics.90 Ligands can further
ligands, and soft metals such as Zn2+ form strong metal–ligand be functionalized in a way that does not alter MOF self-
complexes with weak bases such as azolate ligands.82 Border- assembly, yet plays a greater role in the final properties of a
line metals such as Cu2+ can interact with both strong and weak MOF.91 As shown in early ZIF studies, increasing the length of
bases, examples being benzene-1,3,5-tricarboxylate, a strong the ligand directly increases the pore size without affecting
base, and 2,2 0 -dipyridyl, a weak base, in the case of the nucleation and growth.5 Furthermore, a study done by Kim
HKUST-1 and [Cu(4,4 0 -DP)0.5Cl]n frameworks respectively.83–85 et al. suggested that ligand rigidity directly affects final MOF

6924 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

after which ZIF-8 (dia) crystals formed. This further supports

the idea that the crystallization rate decreases at lower ligand to
metal ratios, and that ZIF-8 (dia) is the thermodynamic product
of the ZIF-8 reaction.

3.2. Temperature
Temperature is another key intrinsic factor for controlling MOF
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

nucleation and growth. The temperature of the MOF synthesis

reaction dictates the ligand solubility and conformation as well
as its extent of coordination to the metal node. Systems
synthesized under high temperatures increase the chemical
potential, Dm, within a system and have greater solubility of
reactants and rates of metal–ligand bond formation as given by:
Fig. 2 Particle size dependence on solvent polarity and reaction
temperature.
Dm = kBTs (6)

s = ln(AP/Ksp) (7)
stability and selectivity. In this study, researchers examined six
ligands with varying rigidity, and found that the MOF synthe- where kB is the Boltzmann constant, T is temperature, s is
sized with the most rigid ligand exhibited the best stability in supersaturation, Ksp is the MOF reaction’s equilibrium con-
water.92 stant, and AP is the activity product of the reactant.27 It is
In addition to solvent polarity, the purity and ionic strength important to note that there are other ways of defining both
of a solvent affects the arrangement and size of the initial Dm and s, with some definitions of s being given as
assembled clusters. For example, a pure aqueous solvent has approximations.27,96,97 Additionally, the temperature can dic-
been shown to facilitate fast nucleation and produce larger tate the probability of overcoming an energy barrier and
clusters compared to an aqueous solution with ions.33 Addi- stability of the final crystal. By increasing temperatures, the
tionally, free ions, such as Na+ and F, in solution can dictate thermal energy within the system increases, resulting in a
the extent of defects in a crystal structure.93 This occurs as a greater driving force for crystallization, specifically of the
result of the ions interacting with the MOF building units thermodynamic product, to occur. On the other hand, at lower
during growth, preventing ligands from rotating and conse- temperatures, the thermal energy within a system is low,
quently forming defects. Defect formation is most prominent at allowing for intermediates and kinetic products to be formed.
lower ion concentrations. As ion concentration increases, the A study compared ZIF-8 synthesized in methanol at 130 1C and
probability of crystalline unit formation is decreased, due to an 25 1C, and found the higher synthesis temperature facilitated a
increased stability between the metal and counterion. Ions also greater percent yield of MOFs, which indicates a greater rate of
can stabilize complexes during growth, raising the energy ligand/metal coordination.95 Target synthesis of MOF crystals
barrier to add more monomers or SBUs, which leads to defects with desired size, porosity, and morphology can be accom-
and trapping of kinetic products. Ions can also add themselves plished in part by obtaining a holistic understanding of how
into the crystal lattice of a MOF, as is the case with ZIF-C— temperature influences MOF growth. The number of nucleating
where carbonate ions incorporate themselves into the ZIF-8 species and nucleation rate is greater at higher temperatures
lattice during formation—changing the nucleation and growth facilitating the formation of more nuclei and, consequently,
mechanism in a way that is still not understood. crystals with smaller diameters. This theory is supported by a
The concentration of precursors and ratio of ligand to metal study that found that increasing the synthesis temperature
is further key in MOF self-assembly mechanisms. Excess ligand from 15 1C to 60 1C caused the average diameter of ZIF-8
drives nucleation in some ZIF systems through the deprotona- crystals to decrease from 78 nm to 26 nm.98 In addition to
tion of building units, reducing the rate of hydrolysis. Increas- controlling the nucleation and growth and crystal size, reaction
ing the extent of excess ligands speeds up the reaction rate, temperatures can dictate the dimensionality of the final MOF
producing smaller kinetic products with low dispersity. Ogata crystal by controlling the rate of metal binding to ligand and
et al. found that at high ligand-to-metal ratios, ZIF-8 is formed solvent molecules. By increasing the temperature, crystals with
by adding amorphous nanoparticles. These precursors then greater dimensionality can be obtained as coordination to
follow a dissolution–recrystallization mechanism to form ZIF-8 solvent molecules decrease, often resulting in the coordination
(sod) crystals. Without excess ligands, protonated ligands number of the metal center to increase.99
remain in solution, causing the crystallization rate to decrease
and the formation of thermodynamic phases such as dia.23,94,95 3.3. Surfaces
In the same study by Ogata et al., it was found that ZIF-8 Surface substrates have shown great potential for nanoscale
nanoparticles initially formed at low ligand-to-metal ratios, but control of crystal formation,100 aiding in the discovery of new
that the nanoparticles appeared to be relatively stable, as no types of materials for CO2 separation,101 water splitting,102 and
particle aggregation could be observed for the first 24 hours, field-effect transistors.103 Surfaces control the self-assembly

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6925
 View Article Online

Chem Soc Rev Tutorial Review

pathway of crystal formation by modifying the interfacial free

energy barrier and offering a site for crystals to nucleate
(Fig. 4).104 The effects of surfaces themselves on crystal growth
are usually not as pronounced, although confinement by sur-
faces can noticeably affect growth of MOFs; this is discussed in
a separate section. The extent of which a surface can modify
crystal nucleation is largely dependent on two factors: (1) atomic
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

structure of the surface and (2) the strength of the bonds

formed by the MOF precursors and the surface. By matching
the atomic structure of a surface to a desired lattice place of a
crystal, surfaces can be used to control the orientation of the
crystal lattice by minimizing the lattice strain in a desired
orientation. Planar substrates include flat, smooth surfaces
such as alumina105 and silica106,107 and provide a uniform Fig. 3 Comparison of free energy barrier in homogeneous (without sur-
surface for MOFs to grow. These substrates often include self- face) and heterogeneous crystallization (with surface).123
assembled monolayers (SAMs),108,109 that can be functionalized
with different terminating groups to dictate the orientation of
With this insight, molecular modifications of proteins have
the MOF crystal. In a study by Shekhah et al., the authors
shown to be promising methods for tailoring the surface
describe studying the formation of metal–organic frameworks
chemistry to promote nucleation and growth of a particular
(MOFs) in a sequential fashion on a functionalized SAMs.110
MOF phase and morphology.122 Furthermore, proteins are
The surface of the SAMs was functionalized with mercaptohex-
believed to add to the growing crystal through particle addition.
adecanoic acid (MHDA) as the carboxylic acid groups matched
Through addition, the proteins create defects in the crystal as
with the (100) plane in the crystal lattice. As a result, the
they are too complex to match the lattice planes of the
orientation and growth of the crystal could be controlled,
MOF.23,119 By controlling the size and concentration of protein
resulting in homogeneous crystalline films. Furthermore SAMs
particles, such defects can be tailored.
can be functionalized with multiple different terminating
Along with biomolecules, polymers can also serve as tem-
groups which promotes nucleation of different lattice orienta-
plates to regulate the crystalline phase, shape, size, and
tions. Such control of the nucleation enables the engineering of
porosity of MOFs.124 Ionic polymers offer many coordination
novel MOF structures with finely tuned lattice patterns.111
sites for MOF nucleation, while instead, non-ionic polymers
Another aspect to consider when choosing a surface is the
may self-assemble to influence the final MOF structure and
chemical composition of the surface interface. The interfacial
porosity through electrostatic interactions. For example,
chemistry can be used to modulate the binding strength of
through a coordination–modulation mechanism, a non-ionic
the MOF precursors to the surface, and consequently, the
block copolymer was able to trap a kinetic crystalline phase
probability of nucleation. By the MOF precursors forming
through stabilization of prenucleation aggregates.125 In sum-
stronger bonds with the surface rather than with the solvent
mary, each type of surface has its own benefits for tailoring the
molecules, the interfacial energy barrier required for nucleation
nucleation and growth of MOF formation. For example, if a
decreases.27 The available surface area for strong-bonds to form
planar substrate is being used, a fabrication method may be
can also impact the probability of nucleation. Non-planar
developed that takes advantage of the flat and smooth surface
substrates, which are surfaces that are more complex and have
of the substrate to help control the growth and formation of
three-dimensional structures, provide a larger surface area
the MOFs in a precise and uniform manner. On the other
for the MOFs to grow on, leading to faster formation and higher
hand, if a non-planar substrate is being used, a fabrication
yields.112–114 Examples of non-planar substrates include
method may be developed that can adequately grow the MOFs
foams,115 fibers,116 nanowires,117 proteins,7,114,118,119 and por-
on the more complex surface of the substrate, potentially
ous metals.120,121 A study by Ogata et al. used bovine serum
leading to faster formation, higher yields, and trapping of
albumin as a non-planar substrate to lower the interfacial
kinetic products.
energy barrier and tune the crystal morphology and struc-
ture.23 In the study, cryo-TEM observed that a protein was
capable of concentrating MOF precursors, forming amorphous 3.4. Small molecules
protein/Zn/HmIm particles and, subsequently, ZIF-8(sod) crys- The idea of using small organic molecules as structure-
tals through a dissolution recrystallization mechanism. Other- directing agents (SDA) to design porous frameworks has been
wise, in the absence of protein, ZIF-8(dia), the nonporous and well-established in the zeolite community.126,127 Additionally,
thermodynamically stable ZIF-8 polymorph, formed (Fig. 3). SDAs are widely credited for the structural diversity of zeolites
The ability for the protein to drive crystal formation of ZIF-8 as they can embed into zeolite structures and influence the
(sod)was strictly dependent on its negatively charged surface pores based on the chemical structure and size of the
as other studies have shown that biomolecules with high additive.35 Small molecules have been similarly used as modu-
isoelectric points (47) cannot promote crystal formation.114 lators during the self-assembly of metal–organic frameworks

6926 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

(MOFs). These molecules, which can be organic or inorganic, the rate of monomers diffusing and depositing onto the facet
can be added during the synthesis of the MOFs to stabilize decreases. Such mechanisms can facilitate the formation of
kinetic products and control the resulting crystals’ size, struc- higher energy, kinetic products. Examples of modulators used
ture, and shape. to tune the crystal size and morphology of ZIF-8 include organic
Acids and bases are common small molecule modulators for amines134 and surfactants.135,136 A study used by Pan et al. used
polymorph selection and crystal size, as they can either increase the surfactant cetyltrimethylammonium bromide (CTAB) to
or decrease the rate of crystallization. For example, MAF-7, trap kinetic morphologies of ZIF-8.137 In the study, CTAB
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

which uses triazolate linkers, only forms rhombic dodecahe- preferentially adsorbed onto the {100} face causing a reduction
dron morphologies upon addition of ammonia. Otherwise, the of crystal growth in that direction. By increasing the concen-
interfacial energy required for crystallization is too high, and tration of CTAB, the capping effect was amplified, causing
only amorphous products are formed.128 The addition of a smaller particles with truncated cube morphologies to form
strong base greatly increases the supersaturation of the depro- rather than the typical rhombic dodecahedron morphology
tonated ligand (which is required for MOF formation) and (Fig. 4).
decreases the interfacial energy required for nucleation. While small molecules can serve as capping agents, they can
In contrast, Brønsted acids such as HCl or monocarboxylic also serve as competing agents which promote the formation of
acids can be used to increase the interfacial energy of a system larger crystals and the most thermodynamically stable mor-
as it increases the concentration of protonated ligands, result- phology (i.e. the Wulff polygon). A study by Cravillon et demon-
ing in a decrease in the rate of crystallization.129 Thus, fewer strated that small molecules, specifically monodentate ligands,
nuclei are formed promoting the growth of large, highly crystal- can serve as competing agents that compete with the bidentate
line MOFs. For example, a study used benzoic acid to modulate ligands to bind to the metal ions. In the study, target diameters
the rate of crystallization of UiO-68-NH2 and obtained crystals of ZIF-8 nanoparticles could be achieved with addition of
as large as 100 mm.130 monodentate ligands (1-methylimidazole, sodium formate, or
Capping agents are types of small molecules that can n-butylamine).134 Through addition of 1-methylimidazole or
selectively bind to a facet of a crystal. This type of modulator sodium formate, larger crystals (1 mm) with narrow size dis-
is able to affect both thermodynamic and kinetic factors of tribution were formed, as the competitive binding decreases
MOF growth. From a thermodynamics perspective, the struc- the rate of nucleation and growth. Furthermore, the particle
tural morphology of a system can be estimated by the Wulff size can be dictated by tuning the similarity of the ligand and
construction principle, where the crystal shape with the lowest modulator pKa as it is key to dictating the rate of binding to the
surface energy will be formed.131,132 By adding onto a preferred metal node. Additionally, the presence of the monodentate
facet of a crystal, capping agents can minimize the surface ligands resulted in preference of the thermodynamic morphol-
energy of that facet and alter the most stable structural topol- ogy (rhombic dodecahedra) over the kinetic morphology
ogy. From a kinetics perspective, the morphology of a system (cubes). In summary, by understanding how modulators can
is dependent on the rate of diffusion of monomers that are be used to control the activation energy barriers for nucleation
deposited onto a growing crystal.133 Capping agents can act as a and the growth rates, we can use them as tools to design MOFs
physical barrier for a particular facet, resulting in a decrease in with targeted shapes and sizes.

Fig. 4 Schematic diagram of various kinetic and thermodynamic products associated with small molecules either serving as capping or competing
agents during the crystal formation. With a small molecule acting as a capping agent, it is adsorbed onto a facet of a crystal resulting in square-shaped
crystals. Depending on the binding strength of the capping agent, crystals of varying sizes can be produced. Without a small molecule, hexagon-shaped
crystals form. With a competing agent, MOF precursors are competing with a small molecule to bind to the growing crystal resulting in larger hexagon-
shaped crystals.

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6927
 View Article Online

Chem Soc Rev Tutorial Review

3.5. Confinement and synthesis volume research is still needed to use confinement as a method to
Nanoscopic confinement has been widely used to design high- discover and trap metastable MOF products. Additionally, on
performance materials with hierarchical structure in the bio- the other side of the spectrum, nucleation and growth mechan-
mineralization community. Nanoscopic confinement influ- isms that occur during macroscopic scaling of a MOF synthe-
ences occur in restricted volumes such as membranes and sis are currently unexplored. Such mechanistic studies are
pores which can limit crystal growth in the x, y, or z dimen- crucial for designing large-scale MOF synthesis using green
approaches, as many current synthetic approaches in industry
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

sions. The confinement’s character can significantly influence

the nucleation rate and stability of intermediate phases. use toxic solvents and MOF precursors along with pressure-
By increasing the degree of confinement on a system---lowering sealed reaction vessels, all of which pose high financial
the volume where synthesis and crystallization can take costs.144,145 Recently, advances towards this challenge include
place—studies have observed a decrease in the nucleation rate a study by Deacon et al. which investigated kilogram-scale
of crystallization. This decrease is due to a decrease in any production of ZIF-8 while using water as the solvent. The
impurities that could act as nucleators and volume is propor- authors note that the crystals observed in the scale-up synthesis
tional to nucleation probability.138 have distinctly different sizes than those reported in literature
Nanoscopic confinement has been shown to stabilize meta- from small-scale synthesis.66 Such findings emphasize the
stable products and phases for non-MOF systems. For example, impact that scaling has on final MOF crystals and the need
a study found that by increasing the degree of confinement for mechanistic studies.
during the crystallization of amorphous calcium carbonate, the
amorphous phase was stabilized.139 This is due to the energy 3.6. Summary
barrier for the thermodynamic product increasing. In addition Intrinsic and extrinsic factors are commonly used to control
to the degree of confinement affecting growth, the surface area MOF formation by manipulating the MOF nucleation energy
of the synthesis vessel can have a further effect. For example, barrier and growth rate. Intrinsic factors include chemical
when a system is assembling in a porous confinement, a greater composition, medium and reaction temperature, and extrinsic
fraction of the nucleus will be in contact with the substrate and factors include surfaces, modulators and confinement. For
thus have a greater nucleation rate than in a system that occurs informed decisions to be made when designing the experi-
within flat confinement.138 Confinement methods using var- mental setup for MOFs, a comprehensive understanding of the
ious interfaces, such as colloforms140 and droplets,141 have fundamental chemistries driving each intrinsic factor, such as
synthesized MOF nanoparticles with high uniformity and low HSAB theory, solubility rules, and electronic effects is critical.
size distribution. The key to obtaining such products lies in Currently, intrinsic factors are better understood than extrinsic
having fast nucleation but slow growth. Synthesizing a MOF in factors. Heterogeneous nucleation is a powerful way to control
a confined environment allows this to occur as it decreases the nucleation and growth, especially for applications in uses of
probability for nuclei to aggregate. MOFs like drug delivery or enzyme immobilization. Extrinsic
Confinement interfaces have shown the potential to serve as factors are also extremely useful for finely controlling things
templates for controlling the growth direction of crystals by like crystal size distribution, which is vital for applications like
restricting growth in designated orientations. Many studies catalysis. We predict that the future studies of extrinsic factors
have succeeded with this method by using emulsion interfaces on MOF nucleation and growth will become more important
to confine MOF growth within a specific geometric mold. By as the aforementioned applications of MOFs become more
varying the length of the surfactant in a reverse micelle system, widespread.
ZIF-8 crystals with targeted sizes could be achieved.105 One
specific study used an aqueous–organic liquid interface to
mold HKUST-1, ZIF-8, and LnBT crystals into a wide range of 4. Monitoring MOF nucleation and
crystal geometries such as hearts, circles, and rectangles.142 growth
The ability to obtain each crystal geometry was based on
localizing the nucleation and growth within molds of specific The complexity of MOF nucleation and growth requires strate-
geometries. This mold was stationed at an opening in the gic experimental and instrumental methods to obtain a thor-
interface surface where its shape could be altered by varying ough understanding of the mechanisms and kinetics driving
buoyancy and capillarity forces. In addition to emulsion inter- each intermediate and final phase. While no single instrumen-
faces being used to control crystal shapes, hard confinement tation method can capture the broad spectrum of the crystal-
surfaces can template MOFs with hierarchical structures and lization process from initial building units to final bulk crystal,
porosity. Synthesizing such hierarchical materials can be each instrument has its own strength as some instrumental
advantageous for catalytic applications as demonstrated with techniques enable characterization of a specific, localized
PCN-224 which was synthesized within the confinement of region and provide qualitative insight into individual crystal
macroporous melamine foam to create geometry that promoted size and morphology, while other techniques may provide
greater catalytic activity.143 ensemble information and quantitative insight into the crystal-
While studies have used nanoscopic confinement methods lization kinetics. Thus, multiple complementary instrumenta-
to synthesize MOFs with desired physical properties, future tion methods are required to obtain a holistic overview of the

6928 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

Fig. 5 Summary of experiential and computational techniques used to study nucleation and growth of MOFs and of the phase regime that they are
useful for probing.

crystallization by providing descriptions of each phase and mechanisms and dynamics in which metastable phases such
phase transition (Fig. 5). Each instrument will be discussed as particles form. Electrospray ionization (ESI) is a soft ioniza-
in more detail in this section, and examples of how each tion technique that is frequently paired with MS (i.e. ESI-MS)
technique can be used to monitor MOF nucleation and growth to minimize any damage, such as fragmentation, to the
will be provided. prenucleation species. In a study by Filez et al., ESI-MS was
used to determine the composition of the prenucleation
4.1. Mass spectrometry and spectroscopy techniques clusters in ZIF-67, which allowed for a greater understanding
Mass spectrometry and spectroscopy techniques are the most of the quantity and composition of the initial SBUs.50 By taking
widely used techniques among organic and organometallic measurements of sample aliquots, the mechanism in which
chemists to provide characterization on chemical structure these building units evolve to form a stable nuclei was deter-
and interactions of small molecules. Spectroscopy operates mined to be through monomer-by-monomer addition. In a
through radiating a sample with photons and measuring the different study by Salionov et al., high resolution ESI-MS was
absorption of the radiation, which can then be used to deter- used to monitor the growth of MIL-53(Al), which revealed how
mine properties of nuclei, bonds, or atomic orbitals, depending solvent byproducts play a role in the nuclei formation.146 The
on the energy of the photon. Mass spectrometry measures the study further uncovered the crystallization mechanism to occur
interactions between the sample and radiation, allowing the through nuclei aggregation.
mass of a sample to be determined through its mass-to-charge 4.1.2. Nuclear magnetic resonance (NMR) spectroscopy.
ratio. Such measurements enable the study of the initial Nuclear magnetic resonance (NMR) is typically used to investi-
coordination among prenucleation phases as well as the local gate prenucleation and metastable intermediates, but can also
bonding environments in metastable and stable phases. While be used to study final MOF crystals and interactions with guest
both mass spectrometry and spectroscopy are unable to differ- species.147 As crystallization processes are complex, producing
entiate between localized regions of differing chemical interac- both liquid and solid phases, different NMR methods
tions, they provide an averaged overview of an entire system, (i.e. solution state NMR or solid state NMR) must be used for
which is powerful for monitoring nucleation and growth. each phase. Solution state NMR is advantageous for studying
4.1.1. Mass spectrometry (MS). Mass spectrometry (MS) the solvent-assisted ligand exchange that occurs within the
provides compositional and structural information of the liquid phase during MOF formation. A study by Yuan et al.
initial building units in solutions that occur prenucleation. used in situ solution-state NMR to compare the rate of linker-
Such studies enable a molecular understanding into the solvent exchange during a MOF formation in various solvents

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6929
 View Article Online

Chem Soc Rev Tutorial Review

which could then be linked to the extent of defect formation in gas adsorption in ZIF-8 due to larger windows between the
the final MOF crystals.148 methyl-imidazole rings in the MOF building blocks.156 Using a
Solid-state NMR (SSNMR) can be used to study both the combination of both Raman and IR is most impactful, as each
short-range and medium-range order. Short-range order is is sensitive to different vibrations, providing greater insight
when atoms are ordered only over a short distance, such as into the chemical reactions occurring in solution, and this is
during nucleation, and medium-range order is when atoms are applicable to all MOFs. Embrechts et al. used in situ Raman and
ordered over a medium distance, such as in metastable amor- FTIR to study the nucleation and growth of MIL-53 (Al) in DMF
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

phous phases.149 Additionally, information on local defects and through evolution in the vibrational bands.157 Through this
disorders on the bulk crystal can be obtained.150 SSNMR is study, the transformation of the atomic ordering in the meta-
also particularly useful for providing information pertaining to stable phase to the stable crystalline nucleus could be probed.
the amount and position of guest molecules—such as CO2
gas—within a framework, in addition to the chemical bonds 4.2. Scattering techniques
between the MOF and guest molecule, as discussed in a review Scattering techniques can be performed on MOFs to obtain
by Witherspoon et al.147 Advanced NMR techniques such as averaged ensemble data pertaining to the nucleation and
high-field-SSNMR allow both liquid and solid phases to be growth kinetics of a crystallization as well as the evolution of
monitored simultaneously on the molecular scale, providing particle size, morphology, and distribution. Data of a sample
insight into the nucleation and growth processes. Such is obtained through collecting scattered radiation off a sample
methods are currently challenging for ZIFs, with SSNMR used as a function of angle and/or time. Based on the radiation
only for characterization of the structural ordering in final ZIF wavelength, different length scales can be obtained.
structures. This is due to multiple reasons, with the three 4.2.1. Light scattering. Dynamic light scattering (DLS) and
biggest being the low gamma and therefore low sensitivity, static light scattering (SLS) use light or photon radiation
high quadrupole moment–which can lead to peak broadening sources, allowing for nanometer resolution, of initial MOF
and improper assignment–and low natural abundance of 67Zn, species.158 DLS measures the scattering of light at a single
the isotope of zinc needed for NMR spectroscopy.151,152 angle to determine the hydrodynamic radius (Rh), which relates
Applications of SSNMR to non-ZIF MOFs abound however; the diffusion of a particle in solution to the diameter of a
an in situ high-field SSNMR study by Jones and coworkers sphere. SLS measures radiation at several angles to obtain the
was successfully performed on a nickel-phosphonate MOF, radius of gyration (Rg), or the radius based on a particle’s center
as multiple NMR spectras could be monitored.153 Through of mass. Rg allows the molecular weight of particles to be
the study, NMR was used to monitor time dependent changes determined. Using a mathematical approximation, Saha and
that occurred during MOF formation and collected kinetic coworkers fit time-resolved SLS data from the growth of ZIF-71
information for reactions at various temperatures. nanoparticles to a Nucleation and Growth (NG) model.159,160
4.1.3. Infrared and Raman spectroscopy. Infrared (IR) and By determining the mass values of the intermediates, the NG
Raman spectroscopy both provide information on the mole- model then provided information on how the mass of the
cular bindings occurring in solution and in various phases in nanoparticles evolve over time. Additionally, by using the ratio
MOF formation through measurement, by means of either of Rg/Rh, particle morphology and shape evolution of MOF
scattering or absorption of the vibrational modes of the particles can be studied.160
chemical bonds. Thus, influences by intrinsic or extrinsic Light scattering techniques can further be used to measure
factors can be monitored through changes in the vibration solution turbidity by measuring the scattering intensity,
mode. IR spectroscopy relies on atoms being free to vibrate in enabling the rate of particle formation to be determined.23
order to detect different functional groups with known absorp- Zeta potential, which measures the surface charge of particles,
tions. Atoms in MOFs are locked in a crystal lattice, having measurements can be performed with light scattering techni-
different IR absorption than the free atoms. Time-resolved ques. Surface charge measurements are particularly important
Fourier transform infrared spectroscopy (FTIR) has been used to understanding the electrostatics driving heterogeneous
to measure the folding of enzymes within ZIF-8, specifically the nucleations on surfaces such as biomolecules114 and
Amide I, II, and III peaks.154 FTIR measurements can be used to zeolites.161 It is important to note that light scattering techni-
determine the presence of enzymes in a p-MOF, as well as ques are limited in that they cannot obtain accurate particle or
presence of CO2, in the case of ZIF-C.67,154 intensity measurements in very heterogeneous and/or turbid
Whereas IR is based on absorption of IR light, Raman MOF solutions. Thus, controls are needed for all lights scatter-
spectrometry measures vibrational modes by means of photon ing methods to determine the best working concentrations of
scattering. Raman is particularly useful for studying MOF– MOF samples.
guest interactions in aqueous solutions, as Raman is not 4.2.2. X-ray scattering. X-ray scattering techniques use
sensitive to water vibration which could overlap the MOF–guest X-rays as radiation sources, enabling higher resolution than
vibrations.155 In situ Raman has been used to study how tem- previously mentioned light scattering techniques due to the
perature plays a role in molecular packing and, consequently, short wavelengths of X-rays compared to visible light. These
ability to integrate guest molecules. Furthermore, a study by methods can provide insight into the nucleation and growth
Kumari et al. found that lower temperatures promoted greater mechanisms of MOFs through morphological and structural

6930 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

evolution studies of the amorphous and crystalline phases. sources can and have mitigated these issues, obtaining experi-
X-ray diffraction is a subclass of X-ray scattering where the ment time at these sources is difficult, and consequently, few
scattered radiation is elastic (i.e. the same energy as the studies of this type have been conducted. WAXS has been used
incident X-rays) and is limited to measuring crystalline struc- mainly to monitor the growth of characteristic crystalline peaks
tures. While the arrangement of amorphous species cannot be and measure the kinetic growth. For example, it has been used
determined with XRD, the presence of the species can still be to understand the formation mechanisms of a multicomponent
determined through broad peaks occurring in the spectra MOFs, revealing phase transitions throughout the synthesis.166
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

background. The rate of crystallinity can then be monitored The kinetic information gained from these experiments are
through changes in peak intensities over time, as it gives crucial in the development of scaled up reactions and optimi-
insight into reaction pathways and observes intermediate zation of the synthesis.
crystalline and amorphous phases. A study by Katsensis and 4.2.3. Neutron scattering. Small angle neutron scattering
coworkers used in situ XRD to monitor the structural evolution (SANS) has the highest resolution of the scattering techniques,
of ZIF-8 while applying a mechanical stress (i.e. mechanochem- allowing nanometer resolution. As it is sensitive in measuring
ical synthesis).162 By monitoring the XRD peaks, they were able lighter elements such as carbon and oxygen, SANS is useful in
to discover new metastable phases and a deeper understanding monitoring the encapsulation and structure of biomolecules in
of how structural density plays a role in polymorph transforma- MOFs. Using SANS, the spatial arrangement of a protein was
tion. Additionally, pair distribution function analysis (PDF) of able to be determined within the pores of MOF-919, which is
in situ XRD measurements is a powerful tool for measuring the not possible with other scattering or microscopy techniques.167
interatomic distances of species present during MOF prenu- In the study, the protein was deuterated to enhance the con-
cleation, nucleation, and growth, and has been used to study trast between the protein and MOF. Additionally, SANS has
the UiO-66 system.48 been used to determine the arrangement of ions in the pores of
Small angle X-ray scattering (SAXS) and wide angle X-ray a conductive MOF, providing insight into how to enhance the
scattering (WAXS) are both types of X-ray scattering techniques performance of the system as a capacitor.168 While SANS is a
which can identify crystallographic information as well as relatively new technique to studying MOFs, we foresee it having
information pertaining to particle shape and size. SAXS is a huge impact in deciphering how guest species influence
extremely versatile and can monitor the size and shape of nucleation and growth mechanisms as well as further insight
initial MOF phases during nucleation and growth of MOF in into how crystal structure influences the performance of guest
the 1–100 nm range. Samples can be amorphous, semi crystal- species.
line, or crystalline, allowing for information on the initial
stages of nucleation and growth that are not typically observed 4.3. Microscopy techniques
in XRD to be collected. In situ SAXS is also used to track the Microscopy methods provide images of localized regions of a
increase in crystalline peak area over time after nucleation, sample enabling phases that occur simultaneously to be dis-
offering insight into the growth rate of MOF systems and tinguished. Additionally, information on the dynamics and
polymorphs.159,163 SAXS applications are not limited to identifi- character of individual phases and crystals can be obtained.
cation of crystalline size and shape; this technique can also be The resolution or viewing range of a sample is dependent on
used to understand the shape and function of enzymes when the wavelength of the beam source and the voltage at which the
they are encapsulated in MOFs.164 In a recent study, SAXS was beam is emitted. For example, microscopes that use photons
used to track the microstructural changes of hollowed MOFs to (a.k.a. visible light) rather than electrons for a beam source
hold different enzymes and showed that the microporous have lower resolution as photons have a longer de Broglie
region would shrink, allowing for the macropores to expand wavelength than electrons.
to hold the enzymes. 4.3.1. Optical microscopy. Optical microscopes use visible
WAXS measurements are taken with the detector in a closer light enabling samples to be viewed within the microns to
position than with SAXS, so larger diffraction angles can be millimeters range. This viewing range limits usage of this
monitored, which is used to look at crystalline phases.163,165 technique to characterization of final crystal size measure-
WAXS has been utilized to studying the change and formation ments and morphologies as the prenucleation and amorphous
of transitory MOF structures and monitor the kinetics of MOF phase regimes occur within the nanometer range. Despite
crystallization and growth. MOF synthesis reactions often these limitations, optical microscopes are more accessible to
absorb significant amounts of X-rays, limiting the amount researchers as their cost and maintenance is more affordable
and accuracy of data that can be collected via X-ray scattering than other microscopy techniques and are more user-friendly,
techniques. Additionally, many laboratory X-ray sources lack making them a powerful technique to study final MOF crystals.
the brilliance, beam coherence, and tunable wavelength ideal Using optical microscopy techniques, the orientation and
for studying material formation kinetics in situ. SAXS and shape of final ZIF-8, MIL-88A, and MIL-96-1 crystals could be
WAXS also suffer from poor temporal resolution, as laboratory described, providing insight into how the crystals pack onto
X-ray source measurements—especially ones with low signal-to- certain substrates.169 In the case when large (B100 mm) MOF
noise—often take minutes to hours, during which impor- crystals are produced, it is even possible to monitor the growth
tant kinetic information can become lost. While synchrotron kinetics, such as in a study with MOF-5.170 Additionally,

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6931
 View Article Online

Chem Soc Rev Tutorial Review

confocal scanning laser microscopy can be used to determine MOFs, cryogenic TEM (cryo-TEM) has been used to improve
surface functionalization171 and localization of guest species beam stability of the specimen. In this technique, aliquots of
such as drugs172 or biomolecules.7 samples are brought to cryogenic temperatures, providing a
4.3.2. Atomic force microscopy. Atomic force microscopy snapshot of the reaction at certain time points.175 To provide
(AFM) uses a laser probe that allows resolution 1000 times further information into the kinetics and dynamics of a crystal-
greater than that of an optical microscope. The technique lization in real-time, liquid phase TEM (LPTEM) can be used to
produces images through direct interactions or scannings of visualize the growth of MOFs while in their native solution.178
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

the probe tip with the MOF surface, allowing for topological Using LPTEM, Liu et al. captured three key nucleation steps
and morphological growth studies of the metastable phases during the formation of MOF nanocrystals.22 The LPTEM
and final crystals. Advancements in in situ AFM have enabled videos display a homogeneous solution of MOF precursors that
AFM measurements to be performed in the native, solution phase separate to form a dense liquid phase. The dense liquid
environment of MOF crystallization.173 Such studies have dis- phase then condenses into an amorphous cluster that then
covered, through measuring the change in surface heights, that undergoes crystallization. The direct observations in this study
nucleation of ZIF-8 occurs on the surface followed by spreading provides insight into how to better control crystallization and
of a metastable phase to form a stable, crystalline phase.86 should be generalizable to many other MOF systems.
Additionally the study provided insight into how non- Scanning electron microscopy (SEM) uses lower energy
framework species, such as solvent molecules, can play a role electrons compared to TEM, allowing external features such
in the stabilization and ordering of monomeric species during as the surface morphology of the MOF crystals to be imaged,
crystal growth. A follow up study revealed that ZIF-67, which with resolution as low as 1 nm.179 SEM has proved to be
has the same crystal structure as ZIF-8 but has a cobalt metal particularly useful for studying how the size, shape, and
center rather than zinc, follows the same growth mechanism as morphology of metastable and stable phases evolve over time.
ZIF-8.174 A study by Jian et al. demonstrated that by modulating the rate
4.3.3. Electron microscopy. Electron microscopes utilize of nucleation through precursor concentration, the shape and
beams of electrons to image samples, allowing atomic scale size of ZIF-8 crystals could be controlled.180 At a set concen-
resolution. The two main categories of electron microscopes tration, the authors monitored the morphological evolution
are transmission electron microscopes (TEM) and scanning over the course of 24 hours and observed crystal growth
electron microscopes (SEM). These microscopes are largely through Ostwald ripening as small ZIF-8 crystals disappeared
differentiated based on the energy of electrons emitted. Trans- overtime while larger crystals appeared. Because MOFs are not
mission electron microscopy uses high energy electrons to pass conductive materials, electrons often build up on the surface
through the sample and provide internal information such of the crystals, called ‘‘charging’’, producing blurry images.181
as pores and defects in the crystal structure. TEM is unique To reduce charging, thin coatings of conductive elements such
among all techniques because the lattice structure of individual as iridium or gold are coated onto the sample using a plasma
phases and products can be imaged using high resolution TEM sputter coater.182
(HRTEM).175 This allows amorphous phases to be distin-
guished from crystalline phases and provide insight into the 4.4. Computational techniques
evolution of the system.23 Video imaging using HRTEM has Computational techniques have the ability to complement each
also been used to elucidate molecular level information about of the previously mentioned experimental techniques and can
prenucleation species in MOF formation in the case of MOF be used to predict the mechanistic pathways of each phase and
5.51 While HRTEM provides crystal structure in real-space phase transition. The most common computational techniques
imaging, micro electron diffraction (microED) images a sample used to study MOF formation are molecular dynamic (MD)
in reciprocal space and obtains diffraction patterns, similar to simulations. MD simulations can simulate individual atoms
that in X-ray diffraction, on selected phases of interest.176 of the MOF precursors along with the atoms in the solution,
While previously mentioned TEM methods provide 2D which can provide dynamic and kinetic information on initial
images of a sample, electron tomography (ET) allows the metal–ligand coordinations and how phase separation from the
sample to be tilted at various angles within the TEM to provide bulk solution occurs.22 Such information can provide valuable
a three dimensional view of the crystal structure. This method insight into polymorph selection. MD simulations have been
is advantageous for studying how guest molecules such as commonly paired with density functional theory (DFT) calcula-
drugs and biomolecules interact with the MOF structure. For tions to extract mechanistic information on nucleation. A study
example, ET was used to understand the distribution of doxor- by Filez and coworkers determined through pairing of MD and
ubicin, a cancer treatment drug, within ZIF-8.6 Additionally, ET DFT, the initial metal–ligand complexes that occurred during
has been used to determine the mechanisms which small ZIF-67 formation as well as the coordination environment of
molecules cap onto growing crystal facets to dictate growth amorphous metastable species.50 Balestra and coworkers also
rate and final crystal morphology.135 used DFT calculations in combination with MD and a calcu-
MOFs can be a challenge to image with TEM, especially lated force field to simulate the coordination bonds occurring
at high resolution, as the high electron energy can damage during nucleation and crystal decomposition of ZIF-8.183 This
the crystal structure.175,177 To combat the beam sensitivity of simulation found that ZIF-8 first forms small clusters, then

6932 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

dissolves into an amorphous precursor from which larger and bulk scale. Large experimental and instrumental efforts
clusters form, which has been verified with TEM studies.23 have been made to determine how reaction conditions such as
Monte Carlo simulations can also be used to investigate solvent and temperature tailor the mechanistic pathways that
individual atoms of the MOF precursors and solution, similar govern the size, dimensionality, and structure of crystals. We
to MD, but varies in its simulation approach. A study by Wells anticipate innovation in expanding the current pressure and
and coworkers developed an algorithm using Monte Carlo temperature limitations in instrumentation—specifically TEM
simulations and discovered different phases with varying and X-ray scattering—to enable systems that form under
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

topologies could be formed as a result of manipulating the broader ranges of synthetic conditions to be better understood.
metal–ligand ratio.184 To understand the mechanisms of sec- Such studies would provide essential, transferable studies to
ondary building units, large-scale MD simulations have been industry which often use high pressure and temperature pro-
performed using graph theory models to understand the for- cessing conditions. As we continue to develop high-throughput
mation mechanisms in various reaction conditions. Kollias characterization methods, such as with the recent development
and coworkers have simulated the initial nucleation stages of of high-throughput TEM,186 the size in which data sets can be
MIL-101(Cr) and have demonstrated how solvents and ions generated will advance, allowing translatable findings to a
affect the early cluster formation, which dictates the configuration greater range of systems.
and complexity of the final MOF lattice.33 By modeling the Mechanistic studies using small molecules and surfaces
interactions of the building units with the free metal ion and have revealed new mechanistic pathways in which the activa-
organic linker, we can better understand MOF formation.185 tion energy to nucleation can be adjusted to trap metastable
and kinetic products. Additionally, countless structure–func-
4.5. Summary tion properties have been discovered pertaining to MOF–guest
A wide range of tools have been deployed to study MOF species interactions, with broad implications in applications
nucleation and growth at all stages, from the molecular chem- ranging from catalysis to gas storage. While MOF crystallization
istry of the precursors to the final bulk crystals. Depending pathways can be exceedingly complex, especially with the
on the type of information required and which stage of the introduction of a guest molecule, we anticipate large efforts
nucleation and growth process is of interest, different tools in methodology and instrumentation to differentiate the
should be selected. Spectroscopy, mass spectrometry, and compu- multiple phases that frequently form simultaneously during
tational methods are commonly deployed to understand the crystallization. This is particularly important for enzyme incor-
molecular chemistry of MOF nucleation and growth while micro- poration as it is essential to distinguish which intermediate
scopy and scattering methods are more commonly used to study and final phase(s) contain the enzyme to understand the
the crystal growth mechanisms and kinetics. However, it is influence of the enzyme in the nucleation of each phase, to
important to note that all methods can be used to study a wide understand how it drives the growth of each phase, and to
range of aspects in nucleation and growth. We predict that understand how (location, orientation etc.) the enzyme is
some of these instruments, specifically electron microscopy, will incorporated into each phase. With each new discovery pertain-
become more important to the study of MOF nucleation and ing to how intrinsic and extrinsic factors tailor crystal for-
growth from a molecular perspective as the resolution and mation, we encourage researchers to further explore how the
sensitivities increase. We also expect that mass spectrometry will synthetic volume, both in confined and large bulk volumes,
play a bigger role in understanding molecular chemistry in the alter the crystallization process. Such studies will further accel-
early stages of MOF nucleation. Importantly, a combination of erate how we understand synthetic scaling of MOF crystals,
instrumentation techniques should be used to elucidate high which is key to mass production. As we are excited for the
quality information about MOF nucleation and growth, and we future of MOF crystallization, we anticipate relying heavily on
do not predict this to change as the field advances. nucleation and growth fundamentals to tap into the unexplored
design space and engineer MOFs with enhanced performances.

5. Outlook and future directions

Conflicts of interest
Compelling progress in MOF crystallization has been made in
the last several years, unraveling new mechanistic insights. There are no conflicts to declare.
By combining fundamental chemistry theories with experi-
mental and computational experiments, progress has been
made in determining the structural composition of prenuclea- References
tion species and the mechanisms in which these species evolve
to become a bulk crystal. Progress can be largely credited to 1 H. Li, M. Eddaoudi, M. O’Keeffe and O. M. Yaghi, Nature,
advancement of characterization methodology and instrumen- 1999, 402, 276–279.
tation. As temporal and spatial resolution continue to improve 2 I. M. Hönicke, I. Senkovska, V. Bon, I. A. Baburin,
among instrumentation, we look forward to greater insight into N. Bönisch, S. Raschke, J. D. Evans and S. Kaskel, Angew.
the nucleation and growth of crystals on both the molecular Chem., Int. Ed., 2018, 57, 13780–13783.

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6933
 View Article Online

Chem Soc Rev Tutorial Review

3 M. P. Suh, H. J. Park, T. K. Prasad and D.-W. Lim, Chem. 29 F. Millange, M. I. Medina, N. Guillou, G. Férey, K. M.
Rev., 2012, 112, 782–835. Golden and R. I. Walton, Angew. Chem., Int. Ed., 2010, 49,
4 R. Aniruddha, I. Sreedhar and B. M. Reddy, J. CO2 Util., 763–766.
2020, 42, 101297. 30 X.-G. Wang, Q. Cheng, Y. Yu and X.-Z. Zhang, Angew.
5 K. S. Park, Z. Ni, A. P. Côté, J. Y. Choi, R. Huang, F. J. Uribe- Chem., Int. Ed., 2018, 57, 7836–7840.
Romo, H. K. Chae, M. O’Keeffe and O. M. Yaghi, Proc. Natl. 31 A. V. Dighe, L. Huelsenbeck, R. R. Bhawnani, P. Verma,
Acad. Sci. U. S. A., 2006, 103, 10186–10191. K. H. Stone, M. R. Singh and G. Giri, JACS Au, 2022, 2,
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

6 H. Zheng, Y. Zhang, L. Liu, W. Wan, P. Guo, A. M. Nyström 453–462.

and X. Zou, J. Am. Chem. Soc., 2016, 138, 962–968. 32 S. L. Anderson, A. Gładysiak, P. G. Boyd, C. P. Ireland,
7 K. Liang, R. Ricco, C. M. Doherty, M. J. Styles, S. Bell, P. Miéville, D. Tiana, B. Vlaisavljevich, P. Schouwink,
N. Kirby, S. Mudie, D. Haylock, A. J. Hill, C. J. Doonan and W. van Beek, K. J. Gagnon, B. Smit and K. C. Stylianou,
P. Falcaro, Nat. Commun., 2015, 6, 7240. CrystEngComm, 2017, 19, 3407–3413.
8 F. Lyu, Y. Zhang, R. N. Zare, J. Ge and Z. Liu, Nano Lett., 33 L. Kollias, R. Rousseau, V.-A. Glezakou and M. Salvalaglio,
2014, 14, 5761–5765. J. Am. Chem. Soc., 2022, 144, 11099–11109.
9 J. Lee, O. K. Farha, J. Roberts, K. A. Scheidt, S. T. Nguyen 34 A. D. Katsenis, A. Puškarić, V. Štrukil, C. Mottillo, P. A.
and J. T. Hupp, Chem. Soc. Rev., 2009, 38, 1450–1459. Julien, K. Užarević, M.-H. Pham, T.-O. Do, S. A. J. Kimber,
10 R. B. Getman, Y.-S. Bae, C. E. Wilmer and R. Q. Snurr, P. Lazić, O. Magdysyuk, R. E. Dinnebier, I. Halasz and
Chem. Rev., 2012, 112, 703–723. T. Friščić, Nat. Commun., 2015, 6, 6662.
11 G. Lu and J. T. Hupp, J. Am. Chem. Soc., 2010, 132, 35 S. L. Burkett and M. E. Davis, J. Phys. Chem., 1994, 98,
7832–7833. 4647–4653.
12 Q. Wang, Y. Sun, S. Li, P. Zhang and Q. Yao, RSC Adv., 36 L. B. Gower and D. J. Odom, J. Cryst. Grow., 2000, 210,
2020, 10, 37600–37620. 719–734.
13 S. Peng, B. Bie, Y. Sun, M. Liu, H. Cong, W. Zhou, Y. Xia, 37 L. Kuhrts, S. Prévost, D. M. Chevrier, P. Pekker, O. Spaeker,
H. Tang, H. Deng and X. Zhou, Nat. Commun., 2018, 9, 1293. M. Egglseder, J. Baumgartner, M. Pósfai and D. Faivre,
14 S. M. Moosavi, A. Nandy, K. M. Jablonka, D. Ongari, J. Am. Chem. Soc., 2021, 143, 10963–10969.
J. P. Janet, P. G. Boyd, Y. Lee, B. Smit and H. J. Kulik, 38 R. L. Penn and J. F. Banfield, Science, 1998, 281(5379),
Nat. Commun., 2020, 11, 4068. 969–971, DOI: 10.1126/science.281.5379.969.
15 A. Zuliani, N. Khiar and C. Carrillo-Carrión, Anal. Bioanal. 39 A. Aerts, C. E. A. Kirschhock and J. A. Martens, Chem. Soc.
Chem., 2023, 1–19. Rev., 2010, 39, 4626–4642.
16 Y. Li and J. Yu, Nat. Rev. Mater., 2021, 6, 1156–1174. 40 M. Eddaoudi, D. B. Moler, H. Li, B. Chen, T. M. Reineke,
17 Y. Li, H. Cao and J. Yu, ACS Nano, 2018, 12, 4096–4104. M. O’Keeffe and O. M. Yaghi, Acc. Chem. Res., 2001, 34,
18 J. Yu and R. Xu, Acc. Chem. Res., 2010, 43, 1195–1204. 319–330.
19 A. Budiman, Z. G. Citraloka, M. Muchtaridi, S. Sriwidodo, 41 M. J. Kalmutzki, N. Hanikel and O. M. Yaghi, Sci. Adv.,
D. L. Aulifa and A. Rusdin, Pharmaceutics, 2022, 14, 2386. 2018, 4, eaat9180.
20 F. Artusio, F. Fumagalli, A. Valsesia, G. Ceccone and 42 G. Férey, J. Solid State Chem., 2000, 152, 37–48.
R. Pisano, ACS Appl. Mater. Interfaces, 2021, 13, 15847–15856. 43 A. Schoedel, Metal–Organic Frameworks for Biomedical
21 M. J. Van Vleet, T. Weng, X. Li and J. R. Schmidt, Chem. Applications, ed. M. Mozafari, Woodhead Publishing,
Rev., 2018, 118, 3681–3721. 2020, pp. 11–44.
22 X. Liu, S. W. Chee, S. Raj, M. Sawczyk, P. Král and 44 S. Surblé, F. Millange, C. Serre, G. Férey and R. I. Walton,
U. Mirsaidov, Proc. Natl. Acad. Sci. U. S. A., 2021, 118, 10, Chem. Commun., 2006, 1518–1520.
DOI: 10.1073/pnas.2008880118. 45 J. A. Rood, W. C. Boggess, B. C. Noll and K. W. Henderson,
23 A. F. Ogata, A. M. Rakowski, B. P. Carpenter, D. A. J. Am. Chem. Soc., 2007, 129, 13675–13682.
Fishman, J. G. Merham, P. J. Hurst and J. P. Patterson, 46 M. W. Terban, D. Banerjee, S. Ghose, B. Medasani,
J. Am. Chem. Soc., 2020, 142, 1433–1442. A. Shukla, B. A. Legg, Y. Zhou, Z. Zhu, M. L. Sushko,
24 J. J. De Yoreo, P. U. P. A. Gilbert, N. A. J. M. Sommerdijk, J. J. D. Yoreo, J. Liu, P. K. Thallapally and S. J. L. Billinge,
R. L. Penn, S. Whitelam, D. Joester, H. Zhang, J. D. Rimer, Nanoscale, 2018, 10, 4291–4300.
A. Navrotsky, J. F. Banfield, A. F. Wallace, F. M. Michel, 47 D. C. Cantu, B. P. McGrail and V.-A. Glezakou, Chem.
F. C. Meldrum, H. Cölfen and P. M. Dove, Science, 2015, Mater., 2014, 26, 6401–6409.
349, aaa6760. 48 H. Xu, S. Sommer, N. L. N. Broge, J. Gao and B. B. Iversen,
25 D. Kashchiev, J. Chem. Phys., 2003, 118, 1837–1851. Chem. – Eur. J., 2019, 25, 2051–2058.
26 D. G. Vlachos and K. F. Jensen, Surf. Sci., 1992, 262, 49 T. Lee, Y. H. Chang and H. L. Lee, CrystEngComm, 2017, 19,
359–370. 426–441.
27 J. J. De Yoreo and P. G. Vekilov, Rev. Mineral. Geochem., 50 M. Filez, C. Caratelli, M. Rivera-Torrente, F. Muniz-
2003, 54, 57–93. Miranda, M. Hoek, M. Altelaar, A. J. R. Heck, V. Van
28 Y. Liu, Y. Yang, Y. Sun, J. Song, N. G. Rudawski, X. Chen Speybroeck and B. M. Weckhuysen, Cell Rep. Phys. Sci.,
and W. Tan, J. Am. Chem. Soc., 2019, 141, 7407–7413. 2021, 2, 100680.

6934 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

51 J. Xing, L. Schweighauser, S. Okada, K. Harano and 74 S. Yuan, L. Feng, K. Wang, J. Pang, M. Bosch, C. Lollar,
E. Nakamura, Nat. Commun., 2019, 10, 3608. Y. Sun, J. Qin, X. Yang, P. Zhang, Q. Wang, L. Zou,
52 B. B. Skjelstad, Y. Hijikata and S. Maeda, Inorg. Chem., Y. Zhang, L. Zhang, Y. Fang, J. Li and H.-C. Zhou, Adv.
2023, 62, 1210–1217. Mater., 2018, 30, 1870277.
53 N. Stock and S. Biswas, Chem. Rev., 2012, 112, 933–969. 75 M. M. Mitrović, A. A. Žekić and M. M. Napijalo, J. Cryst.
54 S. S. Park, C. H. Hendon, A. J. Fielding, A. Walsh, Grow., 2000, 216, 437–442.
M. O’Keeffe and M. Dincă, J. Am. Chem. Soc., 2017, 139, 76 Y. Ban, Y. Li, Y. Peng, H. Jin, W. Jiao, X. Liu and W. Yang,
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

3619–3622. Chem. – Eur. J., 2014, 20, 11402–11409.

55 A. Mittal, S. Gandhi and I. Roy, Sci. Rep., 2022, 12, 10331. 77 Y. Zhou, Y. Ban and W. Yang, Inorg. Chem., 2022, 61,
56 M. Ohring, in Engineering Materials Science, ed. M. Ohring, 17342–17352.
Academic Press, San Diego, 1995, pp. 431–500. 78 J. Sun, L. Semenchenko, W. T. Lim, M. F. Ballesteros Rivas,
57 G. Zahn, P. Zerner, J. Lippke, F. L. Kempf, S. Lilienthal, V. Varela-Guerrero and H.-K. Jeong, Microporous Meso-
C. Schroeder, A. M. Schneider and P. Behrens, CrystEngComm, porous Mater., 2018, 264, 35–42.
2014, 16, 9198–9207. 79 H. Irving and R. J. P. Williams, J. Chem. Soc., 1953, 3192–3210.
58 S. J. I. Shearan, J. Jacobsen, F. Costantino, R. D’Amato, 80 H. Irving and R. J. P. Williams, Nature, 1948, 162, 746–747.
D. Novikov, N. Stock, E. Andreoli and M. Taddei, Chem. – 81 R. G. Pearson, J. Am. Chem. Soc., 1963, 85, 3533–3539.
Eur. J., 2021, 27, 6579–6592. 82 X. Zhang, B. Wang, A. Alsalme, S. Xiang, Z. Zhang and
59 S. D. Bagi, A. S. Myerson and Y. Román-Leshkov, Cryst. B. Chen, Coord. Chem. Rev., 2020, 423, 213507.
Growth Des., 2021, 21, 6529–6536. 83 S. Ahrland, J. Chatt and N. R. Davies, Q. Rev., Chem. Soc.,
60 A. J. Cruz-Cabeza and J. Bernstein, Chem. Rev., 2014, 114, 1958, 12, 265–276.
2170–2191. 84 D. Sharma, S. Rasaily, S. Pradhan, K. Baruah, S. Tamang
61 E. Balog, G. Varga, Á. Kukovecz, Á. Tóth, D. Horváth, and A. Pariyar, Inorg. Chem., 2021, 60, 7794–7802.
I. Lagzi and G. Schuszter, Cryst. Growth Des., 2022, 22, 85 M. N. Ahamad, M. Shahid, M. Ahmad and F. Sama, ACS
4268–4276. Omega, 2019, 4, 7738–7749.
62 D. Aulakh, J. R. Varghese and M. Wriedt, Inorg. Chem., 86 P. Y. Moh, P. Cubillas, M. W. Anderson and M. P. Attfield,
2015, 54, 8679–8684. J. Am. Chem. Soc., 2011, 133, 13304–13307.
63 Q. Ke, Y. Duan, Y. Ji, D. Zhao, H. Zhang, C. Duan, L. Li and 87 F. Ragon, H. Chevreau, T. Devic, C. Serre and P. Horcajada,
Y. Wei, J. Chem. Eng. Data, 2021, 66, 3483–3492. Chem. – Eur. J., 2015, 21, 7135–7143.
64 R. N. Widmer, G. I. Lampronti, S. Chibani, C. W. Wilson, 88 D. Banerjee, J. Finkelstein, A. Smirnov, P. M. Forster,
S. Anzellini, S. Farsang, A. K. Kleppe, N. P. M. Casati, L. A. Borkowski, S. J. Teat and J. B. Parise, Cryst. Growth
S. G. MacLeod, S. A. T. Redfern, F.-X. Coudert and Des., 2011, 11, 2572–2579.
T. D. Bennett, J. Am. Chem. Soc., 2019, 141, 9330–9337. 89 Y. J. Kim, M.-Z. Kim, S. F. Alam, A. ur Rehman,
65 M. Tu, D. E. Kravchenko, B. Xia, V. Rubio-Giménez, A. Devipriyanka, P. Sharma, H. R. Lee and C.-H. Cho,
N. Wauteraerts, R. Verbeke, I. F. J. Vankelecom, Mater. Chem. Phys., 2021, 259, 124021.
T. Stassin, W. Egger, M. Dickmann, H. Amenitsch and 90 T. Wu, X. Bu, J. Zhang and P. Feng, Chem. Mater., 2008, 20,
R. Ameloot, Angew. Chem., Int. Ed., 2021, 60, 7553–7558. 7377–7382.
66 A. Deacon, L. Briquet, M. Malankowska, F. Massingberd- 91 T. R. Cook, Y.-R. Zheng and P. J. Stang, Chem. Rev., 2013,
Mundy, S. Rudić, T. L. Hyde, H. Cavaye, J. Coronas, 113, 734–777.
S. Poulston and T. Johnson, Commun. Chem., 2022, 5, 1–10. 92 M. Kim, J. F. Cahill, H. Fei, K. A. Prather and S. M. Cohen,
67 F. Carraro, M. de, J. Velásquez-Hernández, E. Astria, J. Am. Chem. Soc., 2012, 134, 18082–18088.
W. Liang, L. Twight, C. Parise, M. Ge, Z. Huang, 93 L. Kollias, D. C. Cantu, V.-A. Glezakou, R. Rousseau and
R. Ricco, X. Zou, L. Villanova, C. O. Kappe, C. Doonan M. Salvalaglio, Adv. Theory Simul., 2020, 3, 2000092.
and P. Falcaro, Chem. Sci., 2020, 11, 3397–3404. 94 W. Liang, R. Ricco, N. K. Maddigan, R. P. Dickinson, H. Xu,
68 J. Lyu, X. Gong, S.-J. Lee, K. Gnanasekaran, X. Zhang, Q. Li, C. J. Sumby, S. G. Bell, P. Falcaro and C. J. Doonan,
M. C. Wasson, X. Wang, P. Bai, X. Guo, N. C. Gianneschi Chem. Mater., 2018, 30, 1069–1077.
and O. K. Farha, J. Am. Chem. Soc., 2020, 142, 4609–4615. 95 M. Malekmohammadi, S. Fatemi, M. Razavian and
69 Database of Zeolite Structures, https://www.iza-structure. A. Nouralishahi, Solid State Sci., 2019, 91, 108–112.
org/databases/, accessed January 27, 2023. 96 A. Hussein, in Essentials of Flow Assurance Solids in Oil and
70 M. Eddaoudi, D. F. Sava, J. F. Eubank, K. Adil and Gas Operations, ed. A. Hussein, Gulf Professional Publish-
V. Guillerm, Chem. Soc. Rev., 2014, 44, 228–249. ing, 2023, pp. 143–197.
71 A. A. Bertolazzo, D. Dhabal and V. Molinero, J. Phys. Chem. 97 J. W. Mullin, in Crystallization, ed. J. W. Mullin, Butterworth-
Lett., 2022, 13, 977–981. Heinemann, Oxford, 4th edon, 2001, pp. 86–134.
72 Y. Lo, C. H. Lam, C.-W. Chang, A.-C. Yang and D.-Y. Kang, 98 C.-W. Tsai and E. H. G. Langner, Microporous Mesoporous
RSC Adv., 2016, 6, 89148–89156. Mater., 2016, 221, 8–13.
73 S. Karthika, T. K. Radhakrishnan and P. Kalaichelvi, Cryst. 99 Y.-X. Sun and W.-Y. Sun, Chin. Chem. Lett., 2014, 25,
Growth Des., 2016, 16, 6663–6681. 823–828.

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6935
 View Article Online

Chem Soc Rev Tutorial Review

100 D. J. Lewis, L. Z. Zornberg, D. J. D. Carter and R. J. Nanocomposite Systems, ed. S. Thomas, R. Shanks and
Macfarlane, Nat. Mater., 2020, 19, 719–724. S. Chandrasekharakurup, William Andrew Publishing,
101 S. R. Venna and M. A. Carreon, Chem. Eng. Sci., 2015, 124, Boston, 2016, pp. 75–99.
3–19. 124 M. Kalaj, K. C. Bentz, S. Ayala, J. M. Palomba, K. S. Barcus,
102 S. Liu, W. Xiao, C. Jin, S. Xia, W. Wang, X. Jiang, L. Li, Y. Katayama and S. M. Cohen, Chem. Rev., 2020, 120,
S. Wang and C. Chen, Appl. Surf. Sci., 2022, 605, 154697. 8267–8302.
103 S. Kumar, Y. Pramudya, K. Müller, A. Chandresh, S. Dehm, 125 J. Hwang, T. Heil, M. Antonietti and B. V. K. J. Schmidt,
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

S. Heidrich, A. Fediai, D. Parmar, D. Perera, M. Rommel, J. Am. Chem. Soc., 2018, 140, 2947–2956.
L. Heinke, W. Wenzel, C. Wöll and R. Krupke, Adv. Mater., 126 R. F. Lobo, S. I. Zones and M. E. Davis, J. Inclusion Phenom.
2021, 33, 2103316. Macrocyclic Chem., 1995, 21, 47–78.
104 C. Altintas and S. Keskin, ACS Sustainable Chem. Eng., 127 R. Simancas, D. Dari, N. Velamazán, M. T. Navarro,
2019, 7, 2739–2750. A. Cantı́n, J. L. Jordá, G. Sastre, A. Corma and F. Rey,
105 X. Zhao, X. Fang, B. Wu, L. Zheng and N. Zheng, Sci. China: Science, 2010, 330, 1219–1222.
Chem., 2014, 57, 141–146. 128 M. de J. Velásquez-Hernández, E. Astria, S. Winkler, W.
106 P. Horcajada, C. Serre, D. Grosso, C. Boissière, Liang, H. Wiltsche, A. Poddar, R. Shukla, G. Prestwich,
S. Perruchas, C. Sanchez and G. Férey, Adv. Mater., 2009, J. Paderi, P. Salcedo-Abraira, H. Amenitsch, P. Horcajada,
21, 1931–1935. C. J. Doonan and P. Falcaro, Chem. Sci., 2020, 11,
107 R. Makiura, S. Motoyama, Y. Umemura, H. Yamanaka, 10835–10843.
O. Sakata and H. Kitagawa, Nat. Mater., 2010, 9, 565–571. 129 S. Diring, S. Furukawa, Y. Takashima, T. Tsuruoka and
108 E. Biemmi, C. Scherb and T. Bein, J. Am. Chem. Soc., 2007, S. Kitagawa, Chem. Mater., 2010, 22, 4531–4538.
129, 8054–8055. 130 A. Schaate, P. Roy, A. Godt, J. Lippke, F. Waltz, M. Wiebcke
109 T. P. Vello, M. Strauss, C. A. R. Costa, C. C. Corrêa and and P. Behrens, Chem. – Eur. J., 2011, 17, 6643–6651.
C. C. Bof Bufon, Phys. Chem. Chem. Phys., 2020, 22, 131 G. D. Barmparis, Z. Lodziana, N. Lopez and I. N.
5839–5846. Remediakis, Beilstein J. Nanotechnol., 2015, 6, 361–368.
110 O. Shekhah, H. Wang, S. Kowarik, F. Schreiber, M. Paulus, 132 G. Wulff, Z. Kristallogr. – Cryst. Mater., 1901, 34, 449–530.
M. Tolan, C. Sternemann, F. Evers, D. Zacher, R. A. Fischer 133 T.-H. Yang, Y. Shi, A. Janssen and Y. Xia, Angew. Chem., Int.
and C. Wöll, J. Am. Chem. Soc., 2007, 129, 15118–15119. Ed., 2020, 59, 15378–15401.
111 I. E. Khalil, J. Fonseca, M. R. Reithofer, T. Eder and 134 J. Cravillon, R. Nayuk, S. Springer, A. Feldhoff, K. Huber
J. M. Chin, Coord. Chem. Rev., 2023, 481, 215043. and M. Wiebcke, Chem. Mater., 2011, 23, 2130–2141.
112 J. Liu and C. Wöll, Chem. Soc. Rev., 2017, 46, 5730–5770. 135 G. Zheng, Z. Chen, K. Sentosun, I. Pérez-Juste, S. Bals,
113 M. Lv, M. Sun, M. Wu, F. Zhang, H. Yin, Y. Sun, R. Liu, L. M. Liz-Marzán, I. Pastoriza-Santos, J. Pérez-Juste and
Z. Fan and J. Du, Nano Lett., 2022, 22, 9621–9629. M. Hong, Nanoscale, 2017, 9, 16645–16651.
114 N. K. Maddigan, A. Tarzia, D. M. Huang, C. J. Sumby, 136 X. Fan, W. Wang, W. Li, J. Zhou, B. Wang, J. Zheng and
S. G. Bell, P. Falcaro and C. J. Doonan, Chem. Sci., 2018, 9, X. Li, ACS Appl. Mater. Interfaces, 2014, 6, 14994–14999.
4217–4223. 137 Y. Pan, D. Heryadi, F. Zhou, L. Zhao, G. Lestari, H. Su and
115 O. Shekhah, L. Fu, R. Sougrat, Y. Belmabkhout, A. J. Z. Lai, CrystEngComm, 2011, 13, 6937–6940.
Cairns, E. P. Giannelis and M. Eddaoudi, Chem. Commun., 138 F. C. Meldrum and C. O’Shaughnessy, Adv. Mater., 2020,
2012, 48, 11434–11436. 32, 2001068.
116 F. Liu and H. Xu, Talanta, 2017, 162, 261–267. 139 C. J. Stephens, S. F. Ladden, F. C. Meldrum and H. K.
117 M.-S. Yao, W.-X. Tang, G.-E. Wang, B. Nath and G. Xu, Adv. Christenson, Adv. Funct. Mater., 2010, 20, 2108–2115.
Mater., 2016, 28, 5229–5234. 140 M. Pang, A. J. Cairns, Y. Liu, Y. Belmabkhout, H. C. Zeng and
118 O. Bim-Junior, Y. Alania, F. S. Tabatabaei, R. Frem, M. Eddaoudi, J. Am. Chem. Soc., 2013, 135, 10234–10237.
A. K. Bedran-Russo and P. N. Lisboa-Filho, Langmuir, 141 J. Cui, N. Gao, X. Yin, W. Zhang, Y. Liang, L. Tian, K. Zhou,
2022, 38, 1600–1610. S. Wang and G. Li, Nanoscale, 2018, 10, 9192–9198.
119 L. Tong, S. Huang, Y. Shen, S. Liu, X. Ma, F. Zhu, G. Chen 142 J.-O. Kim, K.-I. Min, H. Noh, D.-H. Kim, S.-Y. Park and
and G. Ouyang, Nat. Commun., 2022, 13, 951. D.-P. Kim, Angew. Chem., Int. Ed., 2016, 55, 7116–7120.
120 A. Huang, H. Bux, F. Steinbach and J. Caro, Angew. Chem., 143 N. Huang, H. Drake, J. Li, J. Pang, Y. Wang, S. Yuan,
Int. Ed., 2010, 49, 4958–4961. Q. Wang, P. Cai, J. Qin and H.-C. Zhou, Angew. Chem.,
121 H. T. Kwon, H.-K. Jeong, A. S. Lee, H. S. An and J. S. Lee, Int. Ed., 2018, 57, 8916–8920.
J. Am. Chem. Soc., 2015, 137, 12304–12311. 144 P. Silva, S. M. F. Vilela, J. P. C. Tomé and F. A. A. Paz, Chem.
122 B. P. Carpenter, A. R. Talosig, J. T. Mulvey, J. Merham, Soc. Rev., 2015, 44, 6774–6803.
J. Esquivel, B. Rose, A. F. Ogata, D. A. Fishman and 145 P. A. Julien, C. Mottillo and T. Friščić, Green Chem., 2017,
J. P. Patterson, Chem. Mater., 2022, 34(18), 8336–8344. 19, 2729–2747.
123 K. Sudhakar, N. N. Reddy, T. Jayaramudu, J. Jayaramudu, 146 D. Salionov, O. O. Semivrazhskaya, N. P. M. Casati,
A. B. Reddy, B. Manjula and E. R. Sadiku, in Design and M. Ranocchiari, S. Bjelić, R. Verel, J. A. van Bokhoven
Applications of Nanostructured Polymer Blends and and V. L. Sushkevich, Nat. Commun., 2022, 13, 3762.

6936 | Chem. Soc. Rev., 2023, 52, 6918–6937 This journal is © The Royal Society of Chemistry 2023
 View Article Online

Tutorial Review Chem Soc Rev

147 V. J. Witherspoon, J. Xu and J. A. Reimer, Chem. Rev., 2018, 166 B. He, L. K. Macreadie, J. Gardiner, S. G. Telfer and
118, 10033–10048. M. R. Hill, ACS Appl. Mater. Interfaces, 2021, 13,
148 N. Yuan, T. L. Church, E. G. Brandt, N. Hedin, X. Zou and 54284–54293.
D. Bernin, Sci. Rep., 2018, 8, 17530. 167 X. Wang, L. He, J. Sumner, S. Qian, Q. Zhang, H. O’Neill,
149 J. R. Helliwell, Crystallogr. Rev., 2012, 18(3), 241–245. Y. Mao, C. Chen, A. M. Al-Enizi, A. Nafady and S. Ma, Nat.
150 R. Pallach, J. Keupp, K. Terlinden, L. Frentzel-Beyme, Commun., 2023, 14, 973.
M. Kloß, A. Machalica, J. Kotschy, S. K. Vasa, P. A. 168 L. He, L. Yang, M. Dincă, R. Zhang and J. Li, Angew. Chem.,
Published on 05 October 2023. Downloaded by Université Claude Bernard Lyon on 10/18/2024 10:10:52 AM.

Chater, C. Sternemann, M. T. Wharmby, R. Linser, Int. Ed., 2020, 59, 9773–9779.

R. Schmid and S. Henke, Nat. Commun., 2021, 12, 4097. 169 D. Lyu, W. Xu, J. E. L. Payong, T. Zhang and Y. Wang, Nat.
151 R. S. K. Madsen, A. Qiao, J. Sen, I. Hung, K. Chen, Z. Gan, Commun., 2022, 13, 3980.
S. Sen and Y. Yue, Science, 2020, 367, 1473–1476. 170 M. T. Conato and A. J. Jacobson, Microporous Mesoporous
152 F. H. Larsen, J. Skibsted, H. J. Jakobsen and N. Chr. Mater., 2013, 175, 107–115.
Nielsen, J. Am. Chem. Soc., 2000, 122, 7080–7086. 171 W. Schrimpf, J. Jiang, Z. Ji, P. Hirschle, D. C. Lamb,
153 C. L. Jones, C. E. Hughes, H. H.-M. Yeung, A. Paul, O. M. Yaghi and S. Wuttke, Nat. Commun., 2018, 9, 1647.
K. D. M. Harris and T. L. Easun, Chem. Sci., 2021, 12, 172 J. Zhuang, C.-H. Kuo, L.-Y. Chou, D.-Y. Liu, E. Weerapana
1486–1494. and C.-K. Tsung, ACS Nano, 2014, 8, 2812–2819.
154 W. Liang, H. Xu, F. Carraro, N. K. Maddigan, Q. Li, S. G. 173 L. D. B. Mandemaker, M. Filez, G. Delen, H. Tan, X. Zhang,
Bell, D. M. Huang, A. Tarzia, M. B. Solomon, H. Amenitsch, D. Lohse and B. M. Weckhuysen, J. Phys. Chem. Lett., 2018,
L. Vaccari, C. J. Sumby, P. Falcaro and C. J. Doonan, J. Am. 9, 1838–1844.
Chem. Soc., 2019, 141, 2348–2355. 174 R. Wagia, I. Strashnov, M. W. Anderson and M. P. Attfield,
155 K. I. Hadjiivanov, D. A. Panayotov, M. Y. Mihaylov, E. Z. Cryst. Growth Des., 2018, 18, 695–700.
Ivanova, K. K. Chakarova, S. M. Andonova and 175 L. Liu, D. Zhang, Y. Zhu and Y. Han, Commun. Chem., 2020,
N. L. Drenchev, Chem. Rev., 2021, 121, 1286–1424. 3, 1–14.
156 G. Kumari, K. Jayaramulu, T. K. Maji and C. Narayana, 176 D. Shi, B. L. Nannenga, M. G. Iadanza and T. Gonen, eLife,
J. Phys. Chem. A, 2013, 117, 11006–11012. 2013, 2, e01345.
157 H. Embrechts, M. Kriesten, K. Hoffmann, W. Peukert, 177 C. Wiktor, M. Meledina, S. Turner, O. I. Lebedev and
M. Hartmann and M. Distaso, J. Phys. Chem. C, 2018, R. A. Fischer, J. Mater. Chem. A, 2017, 5, 14969–14989.
122, 12267–12278. 178 J. P. Patterson, P. Abellan, M. S. Denny, C. Park, N. D.
158 J. Cravillon, S. Münzer, S.-J. Lohmeier, A. Feldhoff, K. Huber Browning, S. M. Cohen, J. E. Evans and N. C. Gianneschi,
and M. Wiebcke, Chem. Mater., 2009, 21, 1410–1412. J. Am. Chem. Soc., 2015, 137, 7322–7328.
159 S. Saha, M. Wiebcke and K. Huber, Cryst. Growth Des., 179 A. E. Vladár, M. T. Postek and B. Ming, Microsc. Today,
2018, 18, 4653–4661. 2009, 17, 6–13.
160 S. Saha, S. Springer, M. E. Schweinefuß, D. Pontoni, 180 M. Jian, B. Liu, R. Liu, J. Qu, H. Wang and X. Zhang, RSC
M. Wiebcke and K. Huber, Cryst. Growth Des., 2016, 16, Adv., 2015, 5, 48433–48441.
2002–2010. 181 A. Rizvi, J. T. Mulvey, B. P. Carpenter, R. Talosig and J. P.
161 P. Netzsch, R. Ettlinger and R. E. Morris, APL Mater., 2023, Patterson, Chem. Rev., 2021, 121, 14232–14280.
11, 031115. 182 J. I. Goldstein, D. E. Newbury, P. Echlin, D. C. Joy,
162 A. D. Katsenis, A. Puškarić, V. Štrukil, C. Mottillo, P. A. A. D. Romig, C. E. Lyman, C. Fiori and E. Lifshin, Scanning
Julien, K. Užarević, M.-H. Pham, T.-O. Do, S. A. J. Kimber, Electron Microscopy and X-Ray Microanalysis, Springer, US,
P. Lazić, O. Magdysyuk, R. E. Dinnebier, I. Halasz and Boston, MA, 1992, pp. 671–740.
T. Friščić, Nat. Commun., 2015, 6, 6662. 183 S. R. G. Balestra and R. Semino, J. Chem. Phys., 2022,
163 J. Cravillon, C. A. Schröder, R. Nayuk, J. Gummel, K. Huber 157, 184502.
and M. Wiebcke, Angew. Chem., Int. Ed., 2011, 50, 8067–8071. 184 S. A. Wells, N. F. Cessford, N. A. Seaton and T. Düren, RSC
164 S.-Y. Chen, W.-S. Lo, Y.-D. Huang, X. Si, F.-S. Liao, S.-W. Adv., 2019, 9, 14382–14390.
Lin, B. P. Williams, T.-Q. Sun, H.-W. Lin, Y. An, T. Sun, 185 L. Kollias, D. C. Cantu, M. A. Tubbs, R. Rousseau,
Y. Ma, H.-C. Yang, L.-Y. Chou, F.-K. Shieh and C.-K. Tsung, V.-A. Glezakou and M. Salvalaglio, J. Am. Chem. Soc.,
Nano Lett., 2020, 20, 6630–6635. 2019, 141, 6073–6081.
165 E. Stavitski, M. Goesten, J. Juan-Alcañiz, A. Martinez- 186 X. Gong, K. Gnanasekaran, K. Ma, C. J. Forman, X. Wang,
Joaristi, P. Serra-Crespo, A. V. Petukhov, J. Gascon and S. Su, O. K. Farha and N. C. Gianneschi, J. Am. Chem. Soc.,
F. Kapteijn, Angew. Chem., Int. Ed., 2011, 50, 9624–9628. 2022, 144, 6674–6680.

This journal is © The Royal Society of Chemistry 2023 Chem. Soc. Rev., 2023, 52, 6918–6937 | 6937