Chem Soc Rev

View Article Online

REVIEW ARTICLE View Journal | View Issue
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

The recent progress on metal–organic

frameworks for phototherapy
Cite this: Chem. Soc. Rev., 2021,
50, 5086
Qiyao Zheng,a Xiangmei Liu,*b Yufeng Zheng, c
Kelvin W. K. Yeung,d
Zhenduo Cui,a Yanqin Liang, a Zhaoyang Li, a
Shengli Zhu, a
Xianbao Wang b and Shuilin Wu *a

Some infectious or malignant diseases such as cancers are seriously threatening the health of human
beings all over the world. The commonly used antibiotic therapy cannot effectively treat these diseases
within a short time, and also bring about adverse effects such as drug resistance and immune system
damage during long-term systemic treatment. Phototherapy is an emerging antibiotic-free strategy to
treat these diseases. Upon light irradiation, phototherapeutic agents can generate cytotoxic reactive
oxygen species (ROS) or induce a temperature increase, which leads to the death of targeted cells. These
two kinds of killing strategies are referred to as photodynamic therapy (PDT) and photothermal therapy
(PTT), respectively. So far, many photo-responsive agents have been developed. Among them, the metal–
organic framework (MOF) is becoming one of the most promising photo-responsive materials because its
structure and chemical compositions can be easily modulated to achieve specific functions. MOFs can
have intrinsic photodynamic or photothermal ability under the rational design of MOF construction, or serve
as the carrier of therapeutic agents, owing to its tunable porosity. MOFs also provide feasibility for various
Received 24th January 2021 combined therapies and targeting methods, which improves the efficiency of phototherapy. In this review, we
DOI: 10.1039/d1cs00056j firstly investigated the principles of phototherapy, and comprehensively summarized recent advances of MOF
in PDT, PTT and synergistic therapy, from construction to modification. We expect that our demonstration
rsc.li/chem-soc-rev will shed light on the future development of this field, and bring it one step closer to clinical trials.

1. Introduction therapeutic agents under appropriate light irradiation. There are

two distinct killing strategies: reactive oxygen species (ROS)
The history of light treatment dates back to 3000 years ago, generation and temperature increase, and the former route is
when the ancient civilizations utilized light to treat skin cancer, named photodynamic therapy (PDT) while the latter is called
rickets and vitiligo.1 In 1903, Niels Finsen was awarded the photothermal therapy (PTT). Compared to traditional treatment
Nobel Prize for using red light and ultraviolet (UV) light to treat strategies such as chemotherapy and radiotherapy, phototherapy
smallpox pustules and cutaneous tuberculosis, which was is less invasive, highly selective and causes minimum damage to
known as the beginning of ‘‘phototherapy’’.1,2 In 1975, Thomas normal tissues.4,5 In recent decades, more and more researchers
Dougherty firstly demonstrated complete tumor eradication of have been trying to apply phototherapy in the treatment of
mice, which aroused extensive investigation in this field.1,3 cancer and nonmalignant diseases, and have achieved good
Nowadays, phototherapy refers to cell-killing treatment by certain therapeutic efficacy.4,6 Therefore, phototherapy is showing great
promise as an emerging type of therapeutic method.
In the case of PDT, a photosensitizer (PS) is irradiated by a
a
School of Materials Science & Engineering, The Key Laboratory of Advanced certain wavelength of light, leading to the generation of highly
Ceramics and Machining Technology by the Ministry of Education of China,
toxic ROS, including the superoxide anion radical ( O2),
Tianjin University, Tianjin 300072, China. E-mail: shuilinwu@tju.edu.cn
b
Ministry-of-Education Key Laboratory for the Green Preparation and Application
hydroxyl radical ( OH), hydrogen peroxide (H2O2) and singlet
of Functional Materials, Hubei Key Laboratory of Polymer Materials, School of oxygen (1O2). These ROS, especially 1O2 and  OH, are cytotoxic
Materials Science & Engineering, Hubei University, Wuhan 430062, China. oxidizing agents that can diffuse through the membrane and then
E-mail: liuxiangmei1978@163.com directly react with many biomolecules in cells such as proteins
c
State Key Laboratory for Turbulence and Complex System and Department of
and DNA, thus inducing cell death and tissue lesions.4,7 The
Materials Science and Engineering, College of Engineering, Peking University,
Beijing 100871, China
superoxide anion radical is more like an intermediate for the
d
Department of Orthopaedics & Traumatology, Li Ka Shing Faculty of Medicine, generation of H2O2 and  OH via Type I process in PDT, or by
The University of Hong Kong, Pokfulam, Hong Kong, China reacting with metal ions (e.g. Fe2+), namely the Fenton reaction.7

5086 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

However, as the lifetime of ROS is short, the direct influence of inactivation and protein denaturation.29 Hence, the therapeutic
ROS has a limited radius of action. For example, the lifetime of temperature should be appropriately designed to achieve desirable
1
O2 is around 48 ns and its diffusion in cells is approximately efficacy. PTAs have several distinct types: carbon material,30–32 noble
20 nm.8 Therefore, targeted cells should be in proximity, to where metal material,33–35 organic dyes,36,37 conjugated polymers,38,39
ROS is generated. Besides direct cytotoxic effects, PDT also causes manganese dioxide,40 magnetic nanoparticles,20 etc. Similar to
vascular injury and robust immune response, which depends on PSs, limited penetration depth and lack of selectivity are the
the interval between drug administration and light irradiation.6,9
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

main drawbacks of PTAs. To utilize sensitization, prevent thermo-

A longer interval provides the PS with sufficient time to diffuse resistance and reduce damage to healthy tissues, PTT requires
through tissues, while a shorter interval renders the accumulation a more elaborate design and a new therapeutic platform for
of PS in vessels, thus inducing vascular stasis and thrombus combined therapy.
formation.1 Porphyrin and its derivatives are the most commonly Metal–organic frameworks (MOFs) are emerging materials
used PS in clinical settings, owing to their low dark toxicity and that are widely utilized in gas storage, gas separation, catalysis
high ROS quantum yield.10 Hematoporphyrin derivative (HPD) is and medical treatment.41 The main components of the MOF
the first approved PS for clinical usage by the US Food and Drug are metal nodes (or metal clusters) and organic linkers (mostly
Administration (FDA), followed by Foscans, Levulans, and carboxyl- or nitrogen-containing agents), which are joined by
Radachlorins, etc. because of their excellent photochemical coordination bonds, forming one-, two- or three-dimensional
properties.2,5 Besides that, organic dyes,11 quantum dots,12 black networks.41–43 Numerous combinations of metals and linkers
phosphorus,13 red phosphorus,14 copper sulfide,15 zinc oxide,16 give rise to diverse MOFs with different structures and properties.
graphic carbon nitride (g-C3N4)17 can also act as PSs. To date, PSs Moreover, MOF is highly porous, with pore size ranging from the
have been developed from bare HPD and its derivatives to those micro- to the meso-scale, providing space for guest molecules.44
carrier-loaded PSs with long-wavelength absorption and selec- The well-defined structure of the MOF has a clear relationship
tive accumulation. However, these properties still need further with their properties, which offers guidance for future modifica-
improvement.10 Other drawbacks of the existing PSs are tions.41 In terms of biomedical applications, MOFs have shown
hydrophobicity-induced aggregation, limited diffusion of ROS, great promise in chemotherapy, phototherapy, diagnosis and
oxygen dependence and undesirable penetration depth, etc. imaging. Due to the tunable pore size, MOFs have been reported
Different from PDT, the mechanism of PTT is like causing a as efficient drug carriers. The weak coordination bonds in MOFs
localized ‘‘fever’’ in the therapeutic sites. During this course, render the stimuli-responsive release of drugs.45 MOFs have a
the temperature increase is triggered by light irradiation on large number of reactive sites,46 which are suitable for further
photothermal agents (PTAs) through the nonradiative relaxation modifications such as targeting molecule attachment, thus greatly
of excited electrons. The efficacy of PTA was evaluated by improving the selectivity of therapeutic nanoparticles (NPs).
photothermal conversion efficiency, which refers to the ratio Under rational design, some MOFs are nontoxic and biodegrad-
between the absorption (sabs) and extinction (sext) of light, able and can be eliminated from the body with low side effects.
according to the following equation:18 The study of MOFs in phototherapy started in the last decade
(2010–2020) and showed a rapidly growing tendency in recent
m = sabs/sext (1) years (Fig. 1). The major functions of MOFs in phototherapy are
schematically summarized in Fig. 2. MOFs can directly act as PSs
Based on the final temperature, PTT can be categorized into
diathermia (o41 1C), hyperthermia (41–46 1C) and thermal
ablation (446 1C).19–23 Diathermia is a relatively mild treatment
that can promote tumor oxygenation by increasing blood flow,
thereby sensitizing cells to radiotherapy and chemotherapy.22
Diathermia is also applied in physiotherapy for rheumatic treat-
ment and muscle relaxation.19,20 In hyperthermia, heat stress
causes protein denaturation and aggregation, cell membrane
loosening and DNA cross-linking, which strongly affect cellular
functions and finally lead to cell inactivation.20,21,24 However, in
this temperature range, the expression of heat shock protein is
higher, which is a molecular chaperone that repairs thermal
damage to cells, leading to the formation of thermo-
resistance.25,26 Moreover, hyperthermia is also able to induce
the sensitization of cells towards heat,22 antibiotics,27 and other
therapies.23 Thermal ablation can directly cause coagulative necrosis
of cells in a few minutes, which is irreversible damage.19 Despite the
high efficiency of thermal ablation, it can also affect healthy cells, Fig. 1 The number of publications in the last decade that focused on the
induce inflammation and cancer metastasis.28 When the tempera- topic of phototherapy by using a metal–organic framework based on a
ture is above 60 1C, rapid necrosis of cells will occur due to enzyme Web of Science search conducted on March 31st 2020.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5087
 View Article Online

Chem Soc Rev Review Article

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 2 Summary of the major functions of MOFs in phototherapy.

or PTAs by using photo-responsive building units,47–49 thus con- to the modified MOFs, and then categorize the typical methods
structing intrinsic photodynamic or photothermal MOFs. They for therapy optimization. We mainly focus on how MOFs
can also realize photo-responsive ability by loading photothera- function and make a difference in each therapeutic system.
peutic agents,50,51 or forming core–shell structures and other Finally, we address the existing drawbacks and challenges in
composites.52,53 The structural design and modification of MOFs this rapidly growing research field.
can tune the light absorption54,55 and electron transition path-
way56 for improving the efficiency of ROS generation or tempera-
ture increase, and also help therapeutic particles to better adapt to 2. Principles of phototherapy
the physiological environment through hypoxia alleviation57 or
targeting molecule attachment.58 MOFs are also excellent plat- Phototherapy is initiated by light irradiation on therapeutic
forms for combined therapy such as chemotherapy,59–61 starva- agents. During this process, some incident photons impinge on
tion therapy62,63 and gas therapy,64,65 which greatly improve the chromophores and then undergo scattering, transmission, or
therapeutic efficacy as compared to phototherapy alone. Due to absorption.3 As only absorbed photons can take effect in
the stimuli-responsive degradation of MOF, the loaded agents can phototherapy, we use the Jablonski diagram to describe this
be released and function in certain environments, preventing the process, as shown in Fig. 3a.19,67–70 Absorption occurs when an
loss during transportation.51,66 electron in the ground state (S0) interacts with a photon with
In this review, we firstly illustrate the principles of photo- energy equal to the difference between two electronic states.
therapy, then comprehensively summarize the recent advances The energy of the photon is then transferred to the electron,
of MOFs in phototherapy (including PDT, PTT and PDT–PTT bringing the electron from the ground state to the short-lived
synergistic therapy) from the intrinsic photo-responsive MOFs (B106 s) higher-energy singlet excited state (S1), and then the

5088 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 3 Photophysics and photochemistry processes of phototherapy: (a) Jablonski diagram; (b) Type I and Type II mechanism of PDT; schematic
illustration of (c) the LSPR effect, (d) relaxation of electron–hole pairs, and (e) rare earth ion doped material in PTT.

electron relaxes to the lowest vibrational level of the excited diseased tissues.73 Hence, guidance by fluorescence imaging can
state through internal conversion.5,67 Internal conversion is a provide important information about the disease and treatment.
nonradiative process that occurs between energy states of the It is also helpful in understanding the mechanism and adjusting
same spin multiplicity, which is a rapid process.71 Finally, the the parameters of therapy.73
excited electron goes back to the ground state by three pathways: Different from fluorescence, phosphorescence emission
fluorescence, phosphorescence, and vibrational relaxation.2,67 occurs by electron transitions between energy states of different
Fluorescence refers to a radiative electron transition from the spin multiplicity. The electron firstly enters the long-lived
lowest vibrational excited state to the ground state, which has (B102 s) triplet state (T1) through a process called intersystem
the same spin multiplicity, inducing the release of a photon.3,71 crossing, leading to the change in the electron spin orienta-
In medical fields, fluorescence is mostly applied to imaging tion.3,5 Intersystem crossing is a forbidden transition caused by
guidance including diagnostics, drug delivery, monitoring, and the interaction between the orbital angular momentum of the
surgery imaging.9,72 For example, fluorescence imaging can be electron and the magnetic dipole related to the electron
used for defining the margin between a tumor and healthy spin.3,74 Afterwards, the electron usually transfers energy by
tissues, or serve as the marker of the body’s response to phosphorescence emission and then relaxes to the ground state.
treatment.9,73 After drug administration, fluorescence can monitor However, as the lifetime of T1 is long, many kinds of quenchers
the location of drugs and determine the degree of drug uptake by can react with the T1 electron before phosphorescence emission.73

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5089
 View Article Online

Chem Soc Rev Review Article

In high partial oxygen solvents, the electron in T1 can transfer surface plasmon polariton.18,68 The surface plasmon then
energy by reacting with a nearby substrate or triplet state mole- decays by light emission or nonradiative transition.82 In the
cules such as molecular oxygen (3O2), thus producing ROS.1,3 latter case, energy is dissipated in the form of heat. This
To be specific, ROS are generated through two distinct phenomenon is called localized plasmon surface resonance
routes, namely, the Type I and Type II mechanism (Fig. 3b). (LSPR), which is shown in Fig. 3c. As LSPR is related to the
In the Type I mechanism, PS in the triplet state (referred to as energy redistribution in the conduction band, it is also referred
3
to as ‘‘intraband transition’’.68 The primary control of LSPR is
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

PS*) transfers a proton or an electron to the nearby substrate

(e.g., cell membrane, protein and lipid, etc.) or other PS mole- free electron density, which determines whether a material can
cules, which leads to the formation of a radical anion or radical have the LSPR effect in the optical region.83 The LSPR fre-
cation (eqn (2) and (3)).1,10,75,76 quency is slightly influenced by particle size and shape. By
changing the particle morphology or introducing coupling
3
PS* + 3PS* -  PS+ +  PS (2)
materials, the LSPR peak can be tuned from the visible to the
3
PS* + Substrate -  Substrate+ +  PS (3) near-infrared (NIR) range.68,80 The most popular LSPR materials
are noble metals, such as gold and silver, because of their
These radicals can further react with O2, thus producing  O2
3
suitable response of light at optical frequencies. However, LSPR
(eqn (4)). is not confined to metals. Semiconductors, metal oxides and

PS + 3O2 - PS +  O2 (4) quantum dots with high carrier density can also have a photo-
thermal effect through LSPR.83 It is worth noting that enhancing
Although  O2 is not the main reactive cytotoxic agent and the electron density through modifications can change the
cannot cause much oxidative damage, it is an important inter- major electron transition from bandgap to LSPR, thus inducing
mediate in biological systems.  O2 can produce H2O2 through the photothermal effect.54,84
dismutation in the presence of superoxide dismutase or by one- Another mechanism is based on the bandgap transition of
electron reduction (eqn (5)). electrons, which mainly appears in low electron density semi-

O2 + 2H+ + e - H2O2 (5) conductors. Due to the interaction of molecular orbitals, semi-
conductors have two broad energy bands, namely the
Finally, H2O2 can generate the highly cytotoxic hydroxyl radical conduction band and valence band. The former is filled with
( OH) through one-electron reduction (eqn (6)).6,7 electrons while the latter has no electrons. There are no energy
H2O2 + e -  OH + OH (6) levels between these two bands. The energy gap between the
lowest energy of the conduction band and the highest energy of
In the Type II mechanism, the energy of the triplet state PS is the valence band is called the bandgap. If the semiconductor is
directly transferred to 3O2 due to the same electron spin multi- irradiated by photons with energy higher than or equal to the
plicity, which brings the 3O2 from the ground state to the excited bandgap, the electron will be excited to a higher energy level in
singlet state, thus producing 1O2.77,78 The energy gap between the conduction band, leaving a hole in the valence band. The
3
O2 and 1O2 is 0.98 eV, which is the energy threshold for 3PS* in electrons and holes are referred to as ‘‘hot carriers’’, due to
PDT.3 The two types of mechanism can work simultaneously, their higher temperature than the lattice. When the energy of
and the ratio depends on the oxygen concentration, substrate the photon is higher than the bandgap, the electron and hole
and PS type. Generally, most PSs are based on the Type II will be ‘‘above-bandgap’’, as shown in Fig. 3d. Then, in process 1,
mechanism, which results in the dependence of PDT on oxygen. the electron and hole will relax to the band edges through
The reaction product 1O2 is the main cytotoxic component in vibrational relaxation, which causes the thermalization process,
PDT due to its high electrophilicity. The Type I mechanism is thus converting energy into heat.69,85 Hence, narrowing the
more effective in oxygen-deficient environments.6,76,78,79 bandgap can improve the photothermal effect through a longer
Vibrational relaxation is the main mechanism of light-to- relaxation pathway. Afterwards, the electron–hole pair will
heat conversion. During this process, electrons in the excited recombine near the band edge or at deep-level defect (DLD) sites
state relax to the ground state in a nonradiative way, causing (process 2).70 During the recombination process, the excessive
the collision between the chromophore and the surrounding energy is transferred to phonon generation for the equilibrium
environment. Hence, part of the energy is released as heat.3,80 of hot carriers and lattice, resulting in heat generation due to the
Generally, most PTAs such as carbon-based PTAs, organic dyes crystal lattice vibration. DLD sites serve as the centers of non-
and photothermal polymers generate heat via this mechanism, radiative recombination. More DLDs lead to higher photon
under light irradiation at the appropriate wavelength.19,36,81 generation, thus increasing the photothermal effect.70
However, as for metals, semiconductors, and rare earth materials, The last type of photothermal mechanism is based on the
the light-to-heat conversion mechanism is more complex. ladder-like energy level of rare earth ions (Fig. 3e). Similarly, the
For high carrier density materials such as metals, the move- nonradiative relaxation of excited electrons to the ground state
ment of free conduction electrons can be guided by incident can generate heat. In rare earth doped nanocrystals, when the
light, causing the polarization of electrons. The oppositely rare earth ion content is increased, the distance to the same
accumulated charges, in turn, form a depolarization field, ions is reduced, forming electron pathways between the same
which leads to the collective oscillation of electrons, namely, ions, which is called cross-relaxation (CR).86 CR results in the

5090 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

excitation and de-excitation of two neighboring ions, and then (1000–1400 nm).19 The longer wavelength of light provides deep-
the two excited electrons relax to the ground state to produce tissue treatment because of minimal tissue absorption and
heat, thus improving the photothermal effect of PTAs.56 The scattering.90 The absorption of PSs or PTAs is a critical factor
energy of excited electrons can also migrate between the same when choosing the light source. When the irradiation wavelength
ions. Once the energy is transferred to a nonradiative center, is the same as the wavelength of the absorption peak, the photon
the energy will be released in the form of heat.19 This process is absorption is enhanced. It is worth mentioning that for coordina-
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

called energy migration. tion structures, besides the interband transition mentioned
The light’s wavelength is a critical factor that is related to the above, their absorption is also influenced by various charge
depth of light penetration, photon adsorption and the subsequent transfer processes such as metal-to-ligand charge transfer,
electron behaviors. The frequently used wavelength of light in ligand-to-metal charge transfer, etc.91 The absorptions of typical
phototherapy is red light (620–750 nm), NIR-I (750–1000 nm), and PSs, PTAs and some intrinsic photodynamic or photothermal
NIR-II (1000–1350 nm). For PSs, the wavelength is usually in the MOFs are listed in Table 1, and are summarized according to
range of 650–800 nm, which is called the ‘‘tissue transparent the published literature. The absorption of a material is not
window’’. Tissue penetration depth in this wavelength region is constant and can be tuned by different sizes and shapes or by
3–10 mm.2 Shorter wavelengths (o600 nm) have limited thera- modifications. The disease location and tissue properties should
peutic depth and will cause skin photosensitivity, while longer also be considered when selecting the appropriate wavelength of
wavelengths (4800 nm) cannot provide enough energy for 3O2 light.6
excitation.7 Since longer wavelengths of light have deeper pene- Whether an excited electron goes through fluorescence
tration, the ideal wavelength of PDT is located in the deep-red emission, nonradiative relaxation or enters into the triplet state
region.6 To further improve the tissue penetration of PDT, the PS depends on many factors, which requires deep insight into the
can be modified with upconversion nanoparticles that can absorb mechanism of photo-induced electron transitions. The ratio
NIR light and emit ultraviolet (UV)-vis light for PS activation.87–89 between radiative and nonradiative processes depends on the
In this way, NIR can be used as the light source. For PTAs, the distance and relative orientation of chromophores.3 When the
desired wavelength is longer, which is located in the first bio- fluorescence quantum yield is low, the ratio of the other two
logical window (700–980 nm) and the second biological window processes is high, which means higher efficacy of phototherapy.

Table 1 Typical PSs and PTAs responding to light of specific wavelength

PSs or PTAs Absorption under specific wavelengths of light Ref.

Porphyrin Strong absorption band at 400 nm (Soret band) 5
A set of absorption bands at 600–800 nm (Q band)
Chlorin 630–650 nm (Q band)
Bacteriochlorin 700–800 nm
Methylene blue 550–700 nm 10
Rose bengal 480–550 nm
IR780 Absorption peak at 780 nm 92
IR825 Absorption peak at 825 nm 93
Indocyanine green (ICG) Absorption peak at 777 nm 94
1,3,6,8-Tetrakis(p-benzoic acid)pyrene (H4TBAPy) Absorption peaks at B420 nm and B330 nm 95
Tetrakis(4-carboxyphenyl)-porphyrin (TCPP) Soret band at 419 nm 96
Q bands at 513, 548, 589, 645 nm
5,15-Di(p-benzoato)porphyrin (H2DBP) Soret band at 402 nm 55
Q bands at 505, 540, 566, 619 nm
5,15-Di(p-methylbenzoato)chlorin (H2DBC) Soret band at 408 nm 47
Q bands at 504, 534, 591, 643 nm
5,10,15,20-Tetra(p-benzoato) (H4TBC) Soret band at 420 nm 97
Q bands at 518, 546, 600, 652 nm
Au nanorod LSPR peak at 520 nm 98
Longitudinal peak in the NIR region
Pb nanocube Surface plasmon resonance absorption at 220–700 nm 99
Prussian blue 500–900 nm with an absorption peak at 712 nm 100
Graphene Strong absorption in the NIR region 31
Graphene oxide Absorption peak at 227 nm 101
Polypyrrole 700–1200 nm 102
Polydopamine Absorption peak centered at 500–800 nm 53
PCN-224 (composed of TCPP ligand and Zr node) Q bands at 515, 550, 591, 646 nm 96
DBP-UiO MOF (composed of H2DBP ligand and Hf node) Q bands at 510, 544, 579, 634 nm 55
DBC-UiO MOF (composed of H2DBC ligand and Hf node) Q bands at 508, 545, 592, 646 nm 47
TBC–Hf MOF (composed of H4TBC ligand and Hf node) Q bands at 520, 548, 600, 653 nm 97
Zr-Ferrocene MOF (composed of Fc(COOH)2 ligand and Zr node) 350–1350 nm 103
5,10,15,20-Tetrakis(4-pyridyl)-21H,23H-porphine (TPyP) Soret band at 418 nm 104
Pd–TPyP MOF (composed of TPyP ligand and Pd node) Soret band at 440 nm 104
Absorption band at 800–1000 nm

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5091
 View Article Online

Chem Soc Rev Review Article

Intersystem crossing can be promoted by combining PSs with and aggregation of PS in common solvents also add difficulty to
heavy atoms due to enhanced intramolecular spin–orbit coupling the MOF synthesis and porosity control.49,109 To date, many
or intermolecular collision, namely the heavy atom effect.3 Another porphyrin-based MOFs have been synthesized and applied in
requirement of PDT is that the energy of the triplet state should be PDT as PSs, which are given in Table 2.
higher than 0.98 eV, which corresponds to the energy between 1O2 The first report of porphyrin-based MOFs for PDT was made
and 3O2.3 On the other hand, to improve the photothermal effect, by Lin’s group.55 They synthesized DBP-UiO constructed by
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

intersystem crossing and fluorescence should be reduced. Chela- 5,15-di(p-benzoato)porphyrin (H2DBP) ligands and Hf nodes,
tion with metals such as Mn and Cu can quench the fluorescence with a H2DBP loading of 77 wt%. DBP-UiO has nanoplate
and ROS generation of PSs,105,106 thus promoting electrons to go morphology, which was B100 nm in diameter and B10 nm
through nonradiative decay pathways.80 Moreover, aggregates of in thickness. The high-Z ion, Hf4+, promoted the intersystem
PSs formed by p-stacking can decrease the triplet state lifetime and crossing of the PDT system. Hence, the absorption peaks of
ROS quantum yield of PSs, which improves the ratio of nonradia- DBP-UiO all red-shifted and the 1O2 generation of DBP-UiO was
tive relaxation.5,80 As PDT is dependent on oxygen, reducing the at least 2 times higher as compared to H2DBP. In vivo tests
oxygen content by clamping also hinders the PDT effect of PSs.1 manifested the excellent antitumor effect of DBP-UiO, i.e., the
tumor decreased in size by 50 times, or was completely eradicated
after PDT treatment. However, the lowest-energy Q band absorp-
3. Photodynamic therapy tion of Hf-DBP was close to the edge of the ‘‘tissue transparent
window’’, and its extinction coefficient was 2200 M1 cm1, which
3.1 Intrinsic photodynamic MOF
means the light source had limited penetration in the tissue and
3.1.1 Porphyrin-based MOFs. Using porphyrin and its deri- the light absorption of DBP-UiO was not enough as well. There-
vatives as MOF ligands is a straightforward way to fabricate intrinsic fore, Lin’s group then partially reduced H2DBP to generate H2DBC
photodynamic MOFs. The frequently-used porphyrin ligands are (H2DBC refers to 5,15-di(p-methylbenzoato)chlorin), which was
presented in Fig. 4, which are mostly carboxyl-containing derivatives used to fabricate DBC-UiO MOF (Fig. 5a).47 DBC-UiO also had
of porphyrin, chlorin and bacteriochlorin, etc.47,55,62,97,107,108 The nanoplate morphology, which was 100–200 nm in diameter and
number of p electrons in porphyrin, chlorin and bacteriochlorin is 3.3–7.5 nm in thickness (Fig. 5b). The lowest-energy Q band of
22, 20 and 18, respectively.5 The absorption of porphyrins red-shifts DBC-UiO red-shifted to 646 nm (Fig. 5c), and the extinction
with the decrease in p electrons.10 Hence, the photo-response of coefficient reached 24 600 M1 cm1, which was 11 times higher
MOFs can be tuned by using different porphyrin linkers. Compared than DBP-UiO. Moreover, DBC-UiO exhibited 3-fold higher PDT
to traditional porphyrin PSs, porphyrin-based MOFs are more efficacy and even more potent tumor eradication than DBP-UiO in
efficient and potent. Firstly, porphyrins are regularly arranged in the CT26 and HT28 model. Afterwards, Lin’s group fabricated
the MOF structure with high loading content, which prevents PSs TBC–Hf MOF (TBC refers to 5,10,15,20-tetra(p-benzoato)), and
aggregation. The channels and pores in MOFs render facile diffu- loaded IDO inhibitor (INCB24360) in it for PDT and immune-
sion of 3O2 and 1O2. Using metal nodes such as Hf can improve combined therapy.97 The lowest-energy Q band of TBC–Hf was
intersystem crossing through the heavy atom effect, thus increasing 653 nm, which was 7 nm red-shifted as compared to the DBC-UiO.
1
O2 yields. However, despite these advantages, the morphology of The extinction coefficient also increased to 38 500 M1 cm1.
metal–organic nanocomposites is hard to control. The insolubility Their results showed that TBC–Hf had the highest ROS generation
among its analogues.
After these reports from Lin’s group, a growing number of
reports about MOF-based PSs came out. Among them, the
porous coordination network (PCN) is an important branch,
which is composed of Zr clusters (mostly octahedral Zr6 clusters)
and the tetrakis(4-carboxyphenyl)-porphyrin (TCPP) ligand. Due
to its excellent ROS generation ability, biodegradability and
stability in aqueous solution, etc.,115 this kind of MOF is becom-
ing the most popular photodynamic MOF as indicated in
Table 2. Park et al. successfully fabricated PCN-224 of various
sizes from 30 to 190 nm by adding different concentrations of
benzoic acid (Fig. 5d).58 When the particle size of PCN-224 was
90 nm, the cellular uptake in HeLa cells was the highest (Fig. 5e).
Under the irradiation of 420 nm and 630 nm lasers, PCN-224 of
90 nm in size also had the highest PDT efficacy of around 80% in
cells due to the better contact of intracellular O2 with PSs
(Fig. 5f). Then, the Zr6 cluster was further functionalized with
folic acid by coordination between the carboxyl of the folic acid
Fig. 4 Typical porphyrins and their derivatives, which have been used as and Zr6 clusters for active targeting (Fig. 5g). In folate receptor-
MOF ligands. abundant cells such as ovarian tumors, the attachment of folic

5092 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

Table 2 Summary of recent porphyrin-based MOFs in PDT. Particle size was measured by SEM (scanning electron microscopy), TEM (transmission
electron microscopy) or DLS (dynamic light scattering)

Metal Irradiation
nodes Organic ligands Materials for decoration Particle size wavelength Additional functions Ref.
Zr4+ TCPP Folic acid 30–190 nm 420 nm and Targeting 58
630 nm
Zr4+ TCPP NaFY4:Yb/Er, thiol-PEG 52.1  9.8 nm for PCN-224 980 nm Upconversion 96
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Zr4+ TCPP Glucose oxidase, catalase, 227.5 nm 660 nm Starvation therapy, targeting 62
cancer cell membrane
Zr4+ TCPP PDA, Pt NPs, folic acid 250 nm 660 nm O2 supply, targeting 110
Zr4+ TCPP Tirapazamine, cancer cell 154.0 nm 660 nm Bioreductive therapy, targeting, 111
membrane immune escape
Zr4+ TCPP Pt NPs, HOOC–PEG– 90 nm 638 nm O2 supply 48
COOH
Zr4+ TCPP L-Arg, cancer cell 105 nm for PCN-224 660 nm Gas therapy 112
membrane
4+
Zr TCPP Ag ions, HA 85 nm for PCN-224 Visible light Bacteria killing 113
Zr4+ TCPP Apatinib, MnO2, cancer 140.5 nm 660 nm Angiogenesis inhibition, GSH 60
cell membrane depletion, immune escape,
targeting
Zr4+ TCPP CeO2 190 nm 638 nm Biofilm inhibition 114
Zr4+ TCPP DOX, HA B250  20 nm 640 nm Targeting, enzyme responsive 115
drug release
Zr4+ TCPP — 4.0 nm 650 nm Renal clearance 116
Zr4+ TCPP PCL 100 nm 510 nm Bacteria killing 117
Zr4+ TCPP DOX, galactose, PEG 121.2  3.7 nm 660 nm Targeting, drug loading 118
Zr4+ TCPP a-Cyano-4- 152.4 nm 660 nm Hypoxia alleviation, targeting, 119
hydroxycinnamate, HA lactate-fueled respiration
Zr4+ TCPP DOX, ZnO, AS1411 161.7 nm 650 nm Drug loading, targeting 120
Zr4+ TCPP MnFe2O4, PEG B110 nm without PEG coating 660 nm O2 supply, GSH depletion, MRI 57
Zr4+ TCPP PAH, MnO2, DOX 72 nm for PCN-222 655 nm Drug loading, MRI, GSH depletion 121
Zr4+ TCPP Pt NPs, Au NPs, folic acid 147.5 nm 671 nm Starvation therapy, targeting 63
Zr4+ TCPP Alkaloid piperlongumine, 200 nm for PCN-222 660 nm Thioredoxin reductase inhibition, 61
cancer cell membrane targeting
Zr4+ TCPP MnO2 nanosheet, cell 105 nm without cell 409 nm O2 supply, MRI, targeting 122
membrane membrane coating
4+
Zr TCPP Acriflavine, cytosine– 105.4 nm 670 nm Immunotherapy 123
phosphate–guanine, HA
Zr4+ PtTCPP Cancer cell membrane 150.5 nm 532 nm Homologous targeting, O2 sensing 124
Zr4+ MnTCPP — 300  130 nm 660 nm MRI, O2 supply 125
(length  width)
Zr4+ FeTCPP NaYb0.92F4:Er0.08@NaYF4, B126.4 nm 980 nm Upconversion, O2 supply, starva- 89
Au NPs tion therapy
Hf4+ H2DBP — 76.3 nm 640 nm — 55
Hf4+ H2DBC — 128.5 nm 640 nm — 47
Hf4+ H4TBC IDO inhibitor 83.2 nm for Hf–TBC 650 nm Immunotherapy 97
Hf4+ H2DBBC — 220 nm in size, 4.6 nm in 750 nm Photoacoustic imaging 108
thickness
Zr4+ TBP PEG B50 nm 660 nm Immunotherapy 107
Zr4+ 5,10,15,20-Tetra(p- — 117.9  1.4 nm 740 nm — 126
benzoato)bacteriochlorin
Hf4+ TCPP PEG B130 nm 661 nm Radiotherapy 127
Cu2+ ZnTCPP — 105 nm 600 nm H2S response 106
Al3+ CuTCPP — — 650 nm GSH depletion 128
Hf4+ TCPP Tirapazamine, DOPA– 163  5 nm 635 nm Drug loading, computed topo- 129
PIMA–mPEG graphy imaging
Gd3+ TCPP — 240.7 nm 660 nm MRI 130
Fe3+ TCPP Bovine serum albumin, 122 nm 660 nm O2 supply, MRI 131
sulfadiazines, MnO2
Mn3+ TCPP — 170 nm in length, 50 nm in 660 nm MRI, GSH depletion, optical 105
width and 100 nm in thickness imaging
4+
Hf TCPP Human serum albumin 105 nm Visible light Biofilm eradication 132
coated MnO2
Cu2+ ZnTCPP — — 635 nm H2S response 133
Zn2+ TCPP DOX, PEG 2 nm in thickness 660 nm Drug loading 134
Gd3+ ZnTCPP ZnO, MnO, MgO, Fe2O3, — 660 nm — 135
CuO
Fe3+ TCPP Dihydroartemisinin, 382  23 nm in length and 182 655 nm Drug loading, MRI, oncosis 136
CaCO3  37 nm in width therapy
2+
Mn TCPP — 150 nm 650 nm O2 supply 137
Sm3+ TCPP Pt NPs, B100 nm in diameter and less 660 nm O2 supply, mitochondrion 138
triphenylphosphine than 10 nm in thickness targeting

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5093
 View Article Online

Chem Soc Rev Review Article

Table 2 (continued)

Metal Irradiation
nodes Organic ligands Materials for decoration Particle size wavelength Additional functions Ref.
6+
W TBP CpG 114.0  6.7 nm 650 nm Immunotherapy 139
Fe3+ TCPP Fe2O3, red blood cell 110–140 nm, depending on the 660 nm Targeting, Fenton reaction 140
membrane, AS1411 reaction time
Fe3+ TCPP Bovine serum albumin, 160 nm 660 nm PTT, MRI 141
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

sulfonamide
Zr4+ TCPP Gold nanorod, 53.8  2.1 nm in length and 808 nm PTT, PTT, drug loading, fluorescence 142
camptothecin 25.2  1.7 nm in width 660 nm PDT imaging
Zr4+ TCPP CuS NPs, folic acid, ICG, 112 nm for PCN-224 650 nm PTT, fluorescence imaging, pho- 143
PAH tothermal imaging
Cu2+ TCPP — 330 nm in size and 808 nm PTT, PTT, MRI, infrared thermal 144
5.1  0.3 nm in thickness 660 nm PDT imaging
4+
Zr FeTCPP siRNA 210 nm in length 635 nm Low-temperature PTT, photo- 145
thermal
imaging, photoacoustic imaging,
computed tomography imaging
Zr4+ CuTCPP — 100 nm 660 nm PTT 146

acid can increase the delivery efficiency of PCN-224 due to the mPPt from immune clearance due to the antigenic profile on the
higher cellular uptake (Fig. 5h). After folic acid modification, the membrane. In an O2-sufficient environment, the triplet phosphore-
PDT efficacy reached as high as 90% (Fig. 5i). Since then, PCN- scence of mPPt was quenched because the triplet PSs mainly
224 has been widely applied as a carrier to combine PDT and reacted with 3O2 to produce 1O2. In an O2-deficient environment,
other functions, which will be further discussed in Section 3.3. the PDT effect was limited, giving rise to a higher phosphorescence
Although TCPP is extensively applied as a MOF ligand, por- emission (Fig. 6b). The authors manifested that mPPT in a 1% O2
phyrin derivatives such as H2DBP, H2DBC and H4TBC reported by atmosphere had a 6.22-fold higher phosphorescence intensity than
Lin’s group also showed high ROS generation and therapeutic in 100% O2, and can give a quick response to O2 fluctuation (Fig. 6c
efficacy.47,55,97 Expanding the category of porphyrin-based ligands and d). This property can be used to detect O2 in vivo and realize
can be a promising direction for developing MOF-based PDT. Zeng phosphorescence imaging guidance. Another research group
et al. coordinated the tetrabenzoporphyrin (TBP) ligand with Zr6 fabricated PCN-222 with Mn-porphyrin as ligands.125 Mn chela-
clusters, which was named TBP-nMOF (Fig. 5j).107 Since benzopor- tion rendered PCN-222 with magnetic resonance imaging (MRI)
phyrin had a larger p conjugation than porphyrin, the TBP-nMOF ability due to the high spin quantum number and long electronic
exhibited a 50 nm-red-shifted lowest-energy Q band and a 100 nm- relaxation time. The longitudinal relaxivity was B35.3 mM1 s1
red-shifted lowest-energy emission as compared to other porphyrin (1.0 T). Moreover, the Mn-porphyrin can also decompose the
MOFs (Fig. 5k). The coordination with Zr6 clusters also shortened surrounding H2O2, thus providing O2 for 1O2 generation.
the fluorescence lifetime of TBP from 3.16 ns to 2.12 ns, due to The most-reported Hf- and Zr-based porphyrin-MOF is
enhanced intersystem crossing (Fig. 5l). Therefore, the 1O2 yield of stable, and can be directly applied as a PS. Other metal nodes
TBP-nMOF was about 8 times higher (Fig. 5m). Zhang and such as Cu, Mn and Fe can quench the ROS generation of PS
coworkers fabricated DBBC-UiO, which was composed of Hf4+ ligands, resulting in much less ROS generation than free PS
nodes and 5,15-di(p-benzoato)bacteriochlorin (H2DBBC) ligands.108 ligands. These metals can respond to different stimuli in the
The DBBC ligand can generate 1O2 in an O2-sufficient environment microenvironment, which means that the MOF structure will
and also produce  O2 in hypoxia through the Type I mechanism. decompose after the reaction, releasing active metal ions, PS
In the presence of superoxide dismutase,  O2 was partially con- ligands and drugs. The photodynamic activity of ligands is then
verted into the highly cytotoxic  OH through the H2O2 dispropor- restored. Hence, this kind of MOF can be used for stimulus-
tionation reaction. DBBC-UiO showed high cancer cell inhibition in triggered PDT. Besides, Mn105 and Gd130 nodes can also endow
both normoxia and hypoxia environments, proving that this MOF MOF with imaging ability. Ma et al. reported Cu-MOF with
was O2-independent, which is desirable in tumor suppression. ZnTCPP as ligands.128 Cu2+ can respond to H2S, which is of
The tetrapyrrole structure in porphyrin has a strong tendency to high content in human colon adenocarcinoma cells, and Cu2+
coordinate metals at the central site, forming metalloporphyrin, can also completely quench the fluorescence of TCPP ligands.
which is denoted as MTCPP (M is the central-coordinated metal) in When Cu2+ encountered H2S, the Cu nodes separated from the
this review.147 Metalloporphyrin ligands endow the MOF with MOF structure, restoring the photosensitivity of TCPP. Wan
more metal sites for synergistic functions. For example, Li et al. and coworkers fabricated a Mn(III)–TCPP MOF (Fig. 6e).105 The
reported a MOF that can monitor O2 fluctuation, which is fluorescence and ROS generation of TCPP were temporarily
composed of PtTCPP and Zr6 clusters, and coated with a cancer blocked by Mn nodes. Glutathione (GSH) is an antioxidative
cell membrane (referred to as mPPt) (Fig. 6a).124 The cancer cell ROS scavenger in tumor cells, which can weaken the potency of
membrane gave the mPPt immune escape ability, which protected PDT. Mn(III)–TCPP can react with intracellular GSH, thus

5094 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 5 (a) Schematic description of 1O2 generation by DBC-UiO under LED light irradiation. (b) TEM image of DBC-UiO showing nanoplate morphology.
(c) UV-vis absorption spectra of H2DBC, DBC-UiO, H2DBP and DBP-UiO in DMF or 0.67 mM PBS. Reprinted with permission from ref. 47. Copyright 2015
American Chemical Society. (d) 3D nanoporous structure of PCN-224 composed of a Zr6 cluster and H2TCPP ligand. (e) Size-dependent cellular uptake
of PCN-224 after 24 h of incubation with HeLa cells. (f) PDT efficacy of different sized PCN-224 and free TCPP molecules. (g) Schematic illustration of FA
functionalized PCN-224 and the proposed internalization. (h) ICP analysis of the cellular uptake of unfunctionalized and FA functionalized 90 nm-PCN-
224 at various concentrations in HeLa and A549 cells. Incubation time = 24 h. (i) In vitro PDT efficacy of FA functionalized (FA equivalent of 0, 1/8, 1/4, 1/2)
90 nm-PCN-224 at various concentrations in HeLa cells. Irradiation at 420 nm for 30 min for PDT. Reprinted with permission from ref. 58. Copyright
2016 American Chemical Society. (j) Schematic representation of TBP-MOF, in which the 10-connected Zr6 cluster and TBP ligand are simplified as a blue
node and red plane, respectively. (k) UV-vis spectra of TBP and TBP-nMOF. Inset shows expanded views of the Q-band regions. (l) Time-resolved
fluorescence decay traces of TBP and TBP-nMOF. (m) 1O2 generation by PpIX, TCPP, PCN-224, TBP, and TBP-nMOF with a 660 nm LED irradiation
(30 mW cm2). Reprinted with permission from ref. 107. Copyright 2018 American Chemical Society.

releasing Mn(II), free TCPP and glutathione disulfide (GSSG) by Fe–TCPP was further covered with a CaCO3 mineralized layer
GSH oxidation. Hence, the fluorescence and ROS generation of that could prevent drug leakage and TCPP toxicity during
TCPP was restored, and the PDT efficacy was protected as well transportation. The ROS generation of TCPP was temporarily
due to GSH depletion. Mn nodes can also be used for MRI. Wan quenched by Fe nodes. In an acidic environment such as the
et al. reported an Fe–TCPP MOF that was loaded with the tumor area, CaCO3 decomposes and releases Ca2+. In a high
anticancer drug dihydroartemisinin (DHA).136 In their work, GSH environment, the exposed Fe3+ on Fe–TCPP is reduced to

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5095
 View Article Online

Chem Soc Rev Review Article

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 6 (a) Preparation processes and the proposed mechanism of mPPt for cancer-targeted and phosphorescence image-guided PDT. (b) 1O2 generation
of mPPt at various O2 pressures. (c) Phosphorescence spectra of mPPt (5 mg mL1) under various O2 levels. (d) The O2 concentration-related
phosphorescence changes in mPPt (5 mg mL1). Inserted: The macroscopic images of mPPt (50 mg mL1) under the irradiation of hand-held UV lamp.
Reprinted with permission from ref. 124. Copyright 2018 Elsevier. (e) Schematic illustration of an endocytosis Mn(III)-sealed MOF nanosystem for MRI- and
optical imaging-guided PDT by controlled ROS generation and GSH depletion after being unlocked by overexpressed GSH in tumor cells. Reprinted with
permission from ref. 105. Copyright 2019 American Chemical Society.

Fe2+, releasing free TCPP. Fe2+ could activate DHA, which then coupled plasma-mass spectrometer (ICP-MS) tests, around
generates free radicals for cancer cell killing. Moreover, as DHA 12.5% of BDC had been replaced, and the final content of
could affect the Ca2+ pump ATPase, the Ca2+ from the CaCO3 I2-BDP was 31.4 wt%. After the ligand exchange, UiO-PDT
coating could enter cancer cells, leading to the increase of showed an absorption peak at 524 nm, which was slightly
intracellular Ca2+ concentration, which could induce oncosis- shifted as compared to I2-BDP (528 nm). UiO-PDT presented
like cell death. Free TCPP also restored its ROS generation. This moderate but lower 1O2 generation than that of I2-BDP because
strategy combines pH- and GSH-controlled drug activation, of the heterogeneous structure of the former. In another study,
PDT activation and oncosis-like therapy at the same time. Zhao and coworkers used TCPP to substitute for the ligand of
3.1.2 Other photodynamic building blocks. In terms of NU-1000, which was referred to as NT.150 NU-1000 was
photodynamic metal nodes, Cai and coworkers fabricated the composed of Zr clusters and 1,3,6,8-tetrakis(p-benzoic acid)pyrene
CuTz-1 MOF composed of Cu(I) nodes and 3,5-diphenyl-1,2,4- (H4TBAPy) ligand, which featured large pore sizes. About 20% of
triazole ligands, which was loaded with O2.148 F127 was coated the original ligand was substituted with TCPP. After the replace-
on CuTz-1 to increase its biocompatibility. The Cu(I) and Cu(II) ment, the maximum absorption was red-shifted to 571 nm as
mixed-valence structure of MOF could induce the intervalence compared to H4TBAPy (410 nm), due to the narrowed energy gap
charge-transfer effect and d–d transition, which resulted in its between the highest occupied molecular orbital (HOMO) and the
absorption in the visible and NIR region. Under 808 nm NIR lowest unoccupied molecular orbital (LUMO). Hence, the ROS
light irradiation, this MOF went through Type I PDT, producing generation of NT was triggered under 650 nm light irradiation in

OH and  O2 through a Fenton-like reaction. Besides, the Cu(I) their work. Park et al. in situ inserted BCDTE (1,2-bis(5-(4-
could react with GSH, thus reducing the ROS loss. The loaded O2 carbonxyphenyl)-2-methylthien-3-yl)cyclopent-1-ene) and TCPP
can also help to alleviate the hypoxia in the tumor environment. into UiO-66 by a thermodynamic-controlled synthesis process
Partially substituting the normal ligands of MOF with PS to fabricate a photochromic switch of 1O2 generation (Fig. 7a).151
agents is also a feasible way to endow MOFs with photodynamic The photoisomerization of BCDTE could switch TCPP from the
1
properties. This strategy leads to mixed-ligand MOFs, and even O2 quenching state to the activated state. Under UV light
more than one PSs can be incorporated into the non-intrinsic irradiation, BCDTE had two isomers, namely, the open form
MOF structure through the ligand-exchange process. Wang and the closed form, which had different absorptions (Fig. 7b
et al. substituted the 2-hydroxyterephthalic acid (H2BDC) ligand and c). When BCDTE was in the open form, TCPP in the triplet
of UiO-66 with I2-BDP (carboxyl-functionalized diiodo- state (T1) could react with 3O2 to produce 1O2, while the BCDTE
substituted BODIPYs) by solvent-assisted ligand exchange.149 in the closed form had another energy transfer pathway with
The final product was called UiO-PDT. According to inductively TCPP that could hinder the reaction between TCPP and 3O2, thus

5096 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

through two strategies: loading during MOF synthesis and post-

synthetic loading.109 The advantages of PS encapsulation
include the following: (i) preventing PS aggregation; (ii) blocking
the contact of PSs and O2, thus the ROS generation of PSs can be
inhibited before reaching the tumor area; (iii) preventing pre-
leakage of PSs during transportation; (iv) improving the tumor
specificity of PSs, etc.50,157,168 In 2015, Zhang et al. encapsulated
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

tetrakis(1-methylpyridinium-4-yl)porphyrin (TMPyP) in a Zn-

based MOF as a PS with a loading efficiency of 32.8%.152 The
final product was named PS@MOF. With methylene blue as a
standard agent, the 1O2 yield of PS@MOF was 0.61  0.05, which
was higher than that of TMPyP (0.10  0.02). The unshifted
fluorescence emission spectra and blue-shifted phosphore-
scence emission spectra of PS@MOF indicated that compared
to free TMPyP, PS@MOF showed the same energy gap between
the higher-energy singlet excited state (S1) and the ground state
(S0), and widened the energy gap between the triplet state (T1)
and the S0 energy level. Hence, the energy gap between S1 and T1
was narrowed, which resulted in enhanced intersystem crossing.
The phosphorescence quantum yield of encapsulated TMPyP
was B34%, which means that most of the triplet state PS was
used for reacting with 3O2, leading to more 1O2 generation.
Fig. 7 (a) Defective structure of UiO-66 with inserted TCPP and BCDTE, Furthermore, compared to free TMPyP, the PS@MOF composite
and a proposed binding scheme of TCPP and BCDTE to a Zr6 cluster. showed much less cytotoxicity.
(b) UV/vis absorption spectra of BCDTE (30 mM) and TCPP (2 mM). Inset: More and more PSs have been reported to be encapsulated
Photographs of solutions of BCDTE open (left) and closed (right) isomers. in MOFs for PDT, such as chlorine e6 (Ce6), zinc phthalocyanines
(c) Chemical structures and photoisomerization of BCDTE. (d) Proposed
(ZnPc), methylene blue and rose bengal, as shown in Table 3.
scheme of 1O2 control via competitive energy-transfer pathways upon
photoisomerization. Reprinted with permission from ref. 151. Copyright Taking Ce6 as an example, which is an efficient PS but suffers
2016 Wiley. from aggregation-caused quenching,50 Wang and coworkers
fabricated MOF-199 (composed of Cu nodes and 1,3,5-
benzenetricarboxylic acid (H3BTC) ligands) and loaded Ce6 in
quenching 1O2 generation (Fig. 7d). More importantly, the ratio the MOF with a loading efficiency of 49 wt%.168 The Ce6 was in
of BCDTE to TCPP can be tuned. When the ratio of BCDTE to an inert state in MOF-199. After the internalization of cells,
TCPP was the highest, the quenching ability was the best as well. Cu(II) nodes of MOF-199 reacted with GSH, leading to the
This strategy was proved to be effective in the cell model, which collapse of the MOF structure. Hence, the loaded Ce6 was
could serve as a protective system in PS delivery. released and contacted the intracellular O2 to generate ROS.
This GSH-controlled PS release strategy can minimize ROS loss
3.2 Modifications via photodynamic agents and scavenge intracellular GSH at the same time. Xie et al.
Fabricating composites using PSs and non-intrinsic photo- fabricated a O2–Cu/ZIF-8@Ce6/ZIF-8@F127 hybrid MOF
dynamic MOF is another way to make MOF-based PDT agents. (Fig. 9a),171 in which the Cu2+-doped ZIF-8 was encapsulated
The reported structures include PSs encapsulation, surface by Ce6/ZIF-8@F127. The former was used as an O2 carrier and
attachment and core–shell structure. In this strategy, PSs are GSH scavenger, while the latter was employed for PDT. As a
not limited to porphyrin and its derivatives. The choices of typical pH-responsive MOF, ZIF-8 will collapse in the mild acid
MOF are also more varied, such as ZIF-8, UiO-66 and MIL-101, environment of the tumor area, releasing O2, Ce6 and Cu2+.
etc. (ZIF, UiO, MIL refer to the zeolitic imidazolate framework, Due to a sufficient O2 supply, the PDT efficacy of Ce6 was
Universitetet i Oslo, Materials Institute Lavoisier, respectively), enhanced. Moreover, Cu2+ could go through a Fenton-like
which are frequently used in biomedical applications. Recent reaction, thus reducing GSH content and the product Cu1+
advances are summarized in Table 3, and the chemical structures can further generate cytotoxic  OH for better PDT efficacy.
of typical organic linkers and encapsulated PSs are summarized in Wang et al. encapsulated Ce6-functionalized DNAzyme into
Fig. 8.148,151–153,155,157,158,160,161,164,179 ZIF-8 to realize imaging-guided gene-photodynamic synergistic
3.2.1 PS encapsulation. PS encapsulation is the most therapy (Fig. 9b).51 ZIF-8 can improve the cellular uptake of
widely used strategy to cause non-intrinsic MOF photodynamic DNA through enhanced permeability and the retention effect.
activity. MOF here serves as the carrier of PSs, creating a ‘‘ship In the tumor area, ZIF-8 would degrade and release Zn2+, which
in a bottle’’ structure. The loaded PSs should have certain served as the DNAzyme cofactor to trigger the RNA-hydrolytic
functional groups, such as –COOH and –SO3H, or have the ability of DNAzyme. The mRNA, human early growth response-1,
opposite charge to the MOF.66,188 The loading can be achieved acted as the substrate for gene therapy. After the treatment of

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5097
 View Article Online

Chem Soc Rev Review Article

Table 3 Summary of MOFs modified by PSs in PDT. Particle size was measured by SEM, TEM or DLS

Metal Organic Photodynamic Irradiation Additional

MOF nodes ligands agents PS loading Materials for decoration Particle size wavelength functionality Ref.
2+
— Zn H3BTC TMPyP 32.8% GPTS, Cy3-labelled caspase-3 — 660 nm Targeting, 152
peptide, H2N–PEG–FA fluorescence
imaging
ZIF-8 Zn2+ 2-mIm g-C3N4 — DOX 60 nm Visible Drug loading, 153
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

light fluorescence
imaging
ZIF-8 Zn2+ 2-mIm AlPcS4 10.6% Catalase, cancer cell 110 nm 660 nm O2 supply, 154
membrane targeting
MIL- Fe3+ H2BDC Ce6–peptide 32.3 mg g1 camptothecin, COOH–PEG–FA 95 nm 660 nm Fluorescence 155
101(Fe) imaging
ZIF-8 Zn2+ 2-mIm g-C3N4 — NaGdF4:Yb, Tm@NaGdF4, PEI, B150 nm 980 nm Upconversion 156
carbon dots
ZIF-8 Zn2+ 2-mIm Methylene blue 1.97% NaYF4:60%,2%Er, catalase B450 nm for 980 nm Upconversion, 157
ZIF-8 O2 supply
4+
UiO-66 Zr H2BDC ZnPc — Erlotinib 100 nm for 670 nm Targeting, drug 158
UiO-66 loading
MIL-88 Fe3+ H2BDC Methylene blue 4.3 wt% DOX 152 nm 635 nm Drug loading 159
— Hf4+ BATA Ce6 80% (w/w) DOX, PEG B90 nm 660 nm Drug loading, 160
CT imaging
ZIF-90 Zn2+ IcaH 2I-BodipyPhNO2 25.4  1.0 wt% — o80 nm 540 nm — 161
ZIF-8 Zn2+ 2-mIm ZnPc–COOH 0.296 g g1 CTAB 83.5 nm 670 nm Drug loading 162
ZIF-8 Zn2+ 2-mIm ZnPc 5.9 mg mg1 — 255 nm (in 650 nm — 163
PBS for 72 h)
MIL- Fe3+ H3BTC Ce6, TPEDC, 49, 42 and F127 340 nm White light O2 supply 50
100 TPETCF 58 wt% (average)
ZIF-8 Zn2+ 2-mIm Rose bengal 35.3 mg mg1 SiO2 93.8  532 nm — 164
17.3 nm
ZIF-90 Zn2+ IcaH Rose bengal 5.6% SiO2, NaYF4:Yb/ 140 nm 808 nm Drug loading, 165
Er@NaYbF4:Nd@NaGdF4,DOX, upconversion,
PEGFA MRI, targeting
UiO-66 Zr4+ H2BDC Photochlor 0.38% AQ4N, PEG, p- 95 nm 671 nm Drug loading, 166
azidomethylbenzoic acid fluorescence
imaging
ZIF-8 Zn2+ 2-mIm Ce6 modified 10 wt% — 167 nm 660 nm Gene therapy 51
DNAzyme
2+
ZIF-8 Zn 2-mIm Ce6 28.3 wt% Bovine serum albumin–MnO2 100 nm for 650 nm O2 supply, MRI 167
NPs Ce6@ZIF-8
MOF- Cu2+ H3BTC TPAAQ, Ce6 58 wt% and F-127 150 nm White light GSH depletion 168
199 49 wt%
2+
MOF- Cu H3BTC PEG conjugated — Pt(IV) NPs 160 nm White light GSH depletion, 169
199 TBD drug loading,
O2 supply
ZIF-8 Zn2+ 2-mIm Ce6 67.4% PVP, Au NPs 106  7.3 nm 660 nm O2 supply 170
ZIF-8 Zn2+, 2-mIm Ce6 3.34 wt% F127 95 nm 650 nm Drug loading, 171
Cu2+ oxygen deliv-
ery, GSH
depletion
ZIF-8 Zn2+ 2-mIm Ce6 — Si–Gd NPs, DOX, HOOC– 70 nm 630 nm Drug loading, 172
PDMAEMA–SH, PEG–FA fluorescence
imaging, MRI
MIL- Fe3+ H2BDC Methylene blue 54.5% DHA, PLA, PEG 120 nm 650 nm MRI, drug 173
101 loading, O2
supply
ZIF-8 Zn2+ 2-mIm Methylene blue 5% Catalase, PDA NPs — 660 nm PTT, O2 supply 174
ZIF-8 Zn2+ 2-mIm TAPP — Catalase, GOx, NaYF4:Yb,Tm — 980 nm O2 supply, 175
starvation
therapy
— Fe3+ Tannic Self-assembled 13.89% for Ce6 Catalase — — O2 supply 176
acid Ce6 and
rapamycin core
ZIF-8 Zn2+ 2-mIm Ammonium From 1.5  0.2 wt% Polyacrylic acid, AgNPs, 150 nm 650 nm Bacteria 177
methylbenzene to 99.5  6.0 wt% vancomycin/NH2–PEG killing, biofilm
blue inhibition
ZIF-8 Zn2+ 2-mIm Ce6 76.80% HA 150 nm 660 nm — 178
Bio- Zn2+ H2BPDC [Ru(bpy)3]2+, — — 490 and Two-photon 179
MOF-1 [Ru(phen)3]2+, 800 nm activation PDT
[Ru(phen)2hipp]2+
UiO-67 Zn2+ H2BPDC [Ru(bpy)3]2+ 13.85 wt% — B92 nm White light Two-photon 180
activation PDT

5098 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

Table 3 (continued)

Metal Organic Photodynamic Irradiation Additional

MOF nodes ligands agents PS loading Materials for decoration Particle size wavelength functionality Ref.
2+ 18
ZIF-8 Zn 2-mIm Au25(Capt) — Fe3O4 100 nm 808 nm PTT 181
PB Fe3+, –CN AlPc — PDA, bovine serum albumin B108 nm 660 nm PTT, MRI, 182
Fe2+ photoacoustic
imaging,
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

fluorescence
imaging
UiO-66 Zr4+ H2BDC Ce6 5.95% Pt, Au shell 60 nm 808 nm PTT 183
without Au PTT,
shell 670 nm
PDT
UiO-66 Zr4+ H2BDC ICG — Red blood cell membrane 64.7 nm 808 nm PTT, immune 184
escape
MIL-53 Fe3+ H2BDC Cypate — PEG, transferrin 250 nm 785 nm PTT, targeting 185
MIL- Fe3+ H2BDC Black phosphorus 15.8 wt% HOOC–PEG–folic acid 140 nm 808 nm PTT, targeting, 186
101 PTT, O2 supply
660 nm
PDT
UiO-68 Zr4+ TPDC Protoporphyrin 29.1 wt% — B120 nm 635 nm PTT 187
IX (diameter),
B20 nm
(thickness)

DNAzyme@ZIF-8, the substrate RNA showed down-regulated imaging guidance and ROS generation. Under light irradiation,
expression in MCF-7 cells (Fig. 9c). Ce6 provided the MOF the early apoptotic ratio increased to 44.9%, which was much
higher than individual PDT (33.6%) and gene therapy (19.85%).
3.2.2 Surface attachment. Besides encapsulation, PSs can
also be attached to the exterior of the MOF. There are two types
of surface attachment: covalent and coordinative modification,
both of which are post-synthetic modifications.189–191 In the
former strategy, the MOF should be modified with certain
functional groups (e.g. –NH2 and –COOH), which can bind
the targeted agents on the surface of MOF.190,192 The attach-
ment of PSs is irreversible, which can prevent the pre-leakage of
agents.109,193 The attached functional agents should maintain
their properties while on the MOF, or be released in certain
environments due to the cleavage of the linkage with MOF, and
then restore their function.109 The latter strategy refers to
binding targeted agents to the unsaturated metal sites or
linkers of MOF, such as metal nodes that coordinate with
solvent molecules and linkers containing functional groups
that do not participate in the MOF formation.189,190 This
strategy is more straightforward than the former and usually
does not affect the topology of the MOF.189 Although these two
strategies have been widely applied in MOF modification, the
reported surface-attached PSs on the MOF are few. Here are
some examples.
Liu and coworkers firstly modified MIL-101(Fe) MOF with –NH2
(referred to as MOF-NH2). The amino groups were used as an
anchor to attach the Ce6-labelled cathepsin B (CaB) substrate
peptide.155 The fluorescence of Ce6 was inhibited because the
electron of the excited Ce6 was transferred to MOF-NH2. CaB is a
lysosomal cysteine endopeptidase that exhibits higher expression
in cancer cells. In this therapeutic system, CaB acted as an
intracellular target to trigger PDT. When in contact with CaB,
the substrate peptide was cut off, then the photodynamic activity
Fig. 8 Chemical structures of (a) MOF ligands and (b) PSs that can be used of Ce6 was restored. Moreover, MIL-101(Fe) was loaded with the
for MOF modification. anticancer drug camptothecin. This drug-release combined PDT

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5099
 View Article Online

Chem Soc Rev Review Article

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 9 (a) Schematic illustration of the fabrication process of the tumor microenvironment-responsive O2–Cu/ZIF-8@Ce6/ZIF-8@F127 nanoplatform
for enhanced PDT and chemotherapy through GSH depletion and O2 replenishment. Reprinted with permission from ref. 171. Copyright 2019 American
Chemical Society. (b) The pH-triggered release of DNAzyme and the corresponding Zn2+-ion cofactors for efficient gene silencing therapy, and an all-in-
one DNAzyme@ZIF-8 nanosystem for fluorescence imaging-guided combined gene/photodynamic therapy. Reprinted with permission from ref. 51.
Copyright 2019 Wiley.

can reduce drug resistance and PDT’s dependence on O2. Nian generation. Therefore, the ROS generation of the final product
et al. attached the carboxyl substituted ZnPc and anticancer drug was much higher than that of free g-C3N4. It was noted that this
erlotinib on N3-UiO-66-NH2 by covalent modification.158 ZnPc dual-model PDT system was protected by the ZIF-8 shell. As a
was connected to –N3 via amidation and erlotinib was connected result, the composite showed little change in particle size, even
to –NH2 via click chemistry. The final product, E-UiO-66-Pc, had when immersed in solution for 5 days. Moreover, the outside shell
higher ROS generation than ZnPc alone. This strategy provides of ZIF-8 also provided space to store O2 and H2O as the raw
accessibility for functional group-guided PS and drug surface material for ROS production.
attachment.
3.2.3 Core–shell structure. The core–shell structure refers to
growing the MOF shell on the PS. The core–shell structure can 3.3 Optimizing the efficacy of photodynamic therapy
combine the properties of the core and shell material while We have discussed how to synthesize a MOF-based PS based on
keeping the stability and activity of the inner PS core by the MOF construction and PS agent selection. To successfully apply
MOF shell protection.142 For example, Chen and coworkers these PSs in clinical trials with potent therapeutic efficacy, the
synthesized a ZIF-8 shell on the g-C3N4 core, and loaded therapeutic system is not limited to ROS generation. On the
doxorubicin (DOX) in ZIF-8.153 Under visible light irradiation, other hand, is hard to cure advanced cancer with PDT alone
the g-C3N4@ZIF-8 nanocomposite showed similar 1O2 genera- without recurrence.194 Hence, more elaborate designs of MOF-
tion to that of g-C3N4, indicating that the MOF shell did not based PSs and more complex combined systems have been
influence the properties of the g-C3N4 shell, but facilitated the developed by researchers, which will be discussed in this
diffusion of 1O2 through the porous structure of the MOF. section.
Moreover, the MOF shell was loaded with DOX. The drug 3.3.1 Utilizing the properties of the therapeutic environment.
release was responsive to an acid environment, which added PDT is usually applied to the tumor or bacterial infected area.
more potency to the therapy. However, g-C3N4 was excited Hypoxia is one typical feature of the tumor microenvironment and
under visible light, which has limited penetration depth. To biofilms,110,195 however, excessive acidity, highly expressed GSH
solve this problem, Yang et al. decorated g-C3N4 with upconver- levels and the presence of adenosine-50 -triphosphate (ATP) also
sion nanoparticles and carbon dots (CDs), which were protected exist.196 It is known that PDT is strongly dependent on oxygen
by the ZIF-8 shell.156 Since the absorption of g-C3N4 in the visible concentration.2,67 In an oxygen insufficient area, PDT will have
light region was stronger than that in NIR light region, when much less efficacy. GSH is an important antioxidant in tumor
irradiated by a 980 nm laser, upconversion nanoparticles cells, which can protect cells from the attack of free radicals
converted NIR light to UV-vis light, thus triggering g-C3N4 to including ROS. Thus, the PDT efficacy will also be affected.197,198
produce ROS. Simultaneously, CDs converted the UV-vis light One straightforward way to solve these problems is by increasing
generated from upconversion nanoparticles into blue visible the amount of PS, however, a PS overdose could arouse serious
light, which once again triggered g-C3N4 for a second ROS side effects.128 As such, researchers have made efforts to design

5100 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

specific structures of MOFs or make targeted modifications to MnFe2O4 was not consumed, which leads to its continuous
achieve self-oxygen generation and GSH depletion. catalytic ability. In hypoxia, the 1O2 generation of MnFe2O4@MOF
As mentioned in Sections 3.1 and 3.2, Mn(III)105 and was improved due to the O2 supply. In the presence of GSH,
Cu(II)128,169 can reduce GSH to oxidized GSH (GSSG); Fe(II),57 after 3 min of irradiation, only 17.6% of 1O2 generated by
Mn(II),137 Fe(III)50,173 and Cu(I)148,199 can react with both intra- MnFe2O4@MOF was depleted by GSH, while the 1O2 depletion
and extra-cellular H2O2 through the Fenton or Fenton-like percentage for bare Zr–TCPP was 57.4%. The MnFe2O4 core can
action to produce O2. Taking Fe3+ and Fe2+ as examples, the
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

also be used for MRI. Recently, Wang and coworkers annealed

related Fenton reaction (eqn (7)) and Fenton-like reaction Mn3[Co(CN)6]2 MOF to fabricate the mesoporous Mn1.8Co1.2O4
(eqn (8)) processes are presented below:200 nanoenzyme, which was loaded with Ce6 as PS.202 The nano-
enzyme was coated with polydopamine (PDA) and polyethylene
Fe2+ + H2O2 - Fe3+ +  OH + OH (7)
glycol (PEG) for better biocompatibility. The final product was
Fe3+ + H2O2 - Fe2+ +  OH2 + H+ (8) named MCOPP–Ce6. The Mn atoms in the nanoenzyme served as
active sites that can be used to coordinate with Ce6 molecules and
The MOF mentioned above is H2O2 or GSH-responsive, decompose H2O2 by the Fenton reaction without self-consump-
which means that the material will decompose after reaching tion. Because of the highly porous structure of the MOF-derived
the therapeutic area. On the other hand, modifications based nanoenzyme, the H2O2 decomposition rate of the composite was
on these elements and other catalytical agents have also been 0.0438 min1, which was much higher as compared to MnO2
reported, such as MnO2 NPs and coating,121,122,132,201 Pt NPs,48,63 (0.0233 min1). The porosity of the nanoenzyme also assisted the
catalase,154,157,175 etc., which have similar functions. For instance, O2 diffusion to contact Ce6. Hence, MCOPP–Ce6 showed around
Liu and coworkers. Synthesized a MnO2 shell on the Zr–TCPP core 95.2% tumor cell killing in an O2 insufficient environment.
by mixing the MOF solution with KMnO4 solution under vigorous Besides the above-mentioned strategies, Meng et al. found
stirring.201 The MnO2 coating quenched the ROS generation and that the disulfide-containing ligand of MOF can also deplete
the fluorescence of the inner core due to the fluorescence GSH through the disulfide-thiol exchange reaction.203 In their
resonance energy transfer. When in contact with GSH, the work, the MOF (referred to as Ce6@RMOF) was composed of a
MnO2 shell was reduced to Mn2+, resulting in GSH depletion disulfide-bearing imidazole ligand and Zn node. Ce6 was
and GSSG generation. Therefore, the properties of Zr–TCPP were loaded in the MOF with a loading efficiency of 14.9  2.7%,
recovered. As the consumption of MnO2 shell was fast, it was which acted as the PS. Compared to the disulfide-free MOF,
unable to offer consistent catalyzing ability. In this regard, Yin Ce6@RMOF showed obvious GSH depleting ability. The ligand
et al. fabricated a core–shell structure composed of a MnFe2O4 was also responsive to low pH and 1O2 due to ligand ionization.
core and Zr–TCPP shell, which was termed as MnFe2O4@MOF Therefore, Ce6 can be released in the presence of GSH, tumor
(Fig. 10a).57 MnFe2O4 features both catalase-like and GSH acidity or under light irradiation. More importantly, the
peroxidase-like activities. Both Fe2+ and Mn2+ firstly went through authors mentioned that glutathione peroxidase 4 can repair
the Fenton reaction to generate Fe3+ and Mn4+, which were further lipid peroxidation, which requires the participation of
reduced by GSH and H2O2 to produce GSSG and O2, creating a GSH. The GSH depletion (both by reacting with ROS and
reaction cycle. Therefore, the total content of Fe3+ and Mn2+ in ligand) can inhibit this process, thus leading to cell death.

Fig. 10 (a) Schematic illustration of MnFe2O4@MOF for persistently providing O2 and consuming GSH to efficiently enhance PDT. Reprinted with
permission from ref. 57. Copyright 2019 Wiley. (b) Schematic illustration of the rational design of MOF QDs and their usage as renal-clearable nanoagents
for enhanced PDT of cancer. (c) Size distribution and TEM image of MOF QDs. (d) Light-induced ROS generation from PCN-224 NMOFs, PEG-NMOFs,
PCN-224 QDs and MOF QDs. (e) Pharmacokinetics of MOF QDs from tumor-bearing mice after intravenous injection. Reprinted with permission from
ref. 116. Copyright 2019 American Chemical Society.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5101
 View Article Online

Chem Soc Rev Review Article

This phenomenon is called ferroptosis, which is a newly-found

contributor to PDT efficacy. Moreover, ferroptosis can be tuned
by ligand content. Hence, their work inspired future researchers
to consider the coexisting mechanism of PDT and find out more
about the relation between PDT and the immune system.
3.3.2 Changing particle size. Particle size is another impor-
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

tant factor that deals with ROS generation, cellular uptake,

clearance from the body, and biofilm penetration. In general,
due to the abnormal tumor vascularization, therapeutic NPs with
a size of 5–200 nm can preferentially enter tumor tissue and be
retained in it because of poor lymphatic clearance.204–206 This is
called enhanced permeability and retention effect, which endows
nano-sized therapeutic agents with a much higher drug delivery
efficiency by passive targeting.207 When the particle size is smaller
than 5 nm, the enhanced permeability and retention effect
become negligible. This kind of smaller-sized particles can be
Fig. 11 Chemical structures of typical drugs and other modifying agents
cleared by the kidneys, which means they have fewer side effects, for MOF modification.
but their targeting ability is also poorer.206,208 On the other hand,
lager-sized NPs (4300 nm) will be recognized as ‘‘foreign sub-
stances’’ by the reticuloendothelial system organs such as the liver ideal platform for combined therapy through modification
and spleen, and are retained in the body for a long time, thus methods such as encapsulation, surface attachment, and core–
increasing the risk of side effects.116,205 Hence, exploring the shell structure, etc. Hence, in this part, we mainly focus on the
influence of particle size is necessary to simultaneously achieve therapeutic systems composed of MOF-based PDT and other
higher therapeutic efficacy and fewer potential side effects. therapies. The chemical structures of typical drugs and other
Park et al. modulated the size of PCN-224 by analyzing the modifying agents are presented in Fig. 11.61,111,112,115,142,266
equilibrium of MOF formation.58 Using benzoic acid as a Chemotherapy functions by using anticancer drugs to stop
competing agent that can form coordination bonds with the or slow the growth of tumor cells.209 In MOF-based PDT
Zr6 cluster, the size of PCN-224 increased with higher concen- therapeutic systems, drugs are usually loaded in photodynamic
trations of benzoic acid. They also suggested that other factors MOF. The loaded drugs are not limited to traditional anticancer
included in the formation equation of PCN-224 such as TCPP drugs such as DOX and camptothecin (Fig. 11). Li et al. loaded
concentration can also influence the diameter of the product. hypoxia-activated tirapazamine (TPZ) in PCN-224, and coated
Their research shed light on controlling the particle size of PCN-224 with the cancer cell membrane.111 Unlike the design
MOF, which is beneficial for investigating the influence of of hypoxia alleviation mentioned above, this strategy further
particle size on PDT efficacy and cellular response. More decreased the intracellular O2 content by PDT under irradia-
recently, Wang and coworkers fabricated PCN-224 nanodots tion. This resulted in extremely low-oxygen environment-
by sonication (PCN-224 QDs), which were further PEGylated activated TPZ, which released transient oxidizing radicals to
and referred to as MOF QDs (Fig. 10b).116 The sonication kill cancer cells. Therefore, TPZ@PCN@Mem showed better
process could attack the defect sites of PCN-224, thus resulting cancer cell inhibition in hypoxia, and the cancer cell viability
in smaller fragments with new defect sites, which were further decreased with increasing the content of TPZ. Min and coworkers
attacked by ultrasound. The final MOF QDs had an average incorporated apatinib in PCN-224, and coated it with MnO2 and
hydrodynamic diameter of 4.5 nm (Fig. 10c), which was smaller cancer cell membrane.60 Angiogenesis in the tumor area will lead
than the renal filtration threshold. Moreover, the PCN-224 QDs to tumor regrowth and metastasis. Apatinib is a vascular
presented 2-fold ROS yield as compared to the nanosized PCN- endothelial growth factor inhibitor that can effectively inhibit
224 (Fig. 10d), due to the better utilization of 1O2 from the angiogenesis. Apatinib was released in the presence of GSH due
interior PCN-224 nanodots. Animal tests manifested that the to the decomposition of MnO2. Their results showed that
MOF QDs could be excreted from the mouse bodies mostly apatinib had little influence at the early-stage of treatment but
through renal clearance within 1 week, with a blood circulation several days later, the groups that were not treated with apatinib
half-life of 2.66  0.19 h and high tumor accumulation resumed the tumor growth, indicating that apatinib-assisted
(Fig. 10e). This strategy offered insight into the fabrication of PDT had long-term tumor inhibition. Cheng et al. reported a
ultrasmall MOF with high therapeutic efficacy and biosafety. cancer cell membrane-coated PCN-222 as the carrier of alkaloid
3.3.3 Combined therapy. Since it is hard to cure advanced piperlongumine (PL) (Fig. 12a).61 The thioredoxin/thioredoxin
cancers without recurrence with PDT alone, PDT can be combined reductase (Trx/TrxR) system is another way that cancer cells
with various therapies such as chemotherapy, gas therapy, starva- develop to confront oxidative stress, which maintains the cellular
tion therapy and immunotherapy. Due to the synergistic effect redox homeostasis but also leads to the ROS resistance of cancer
between PDT and these therapies, the combined therapy usually cells. The dithiol group in Trx can react with ROS through the
shows a ‘‘1 + 1 4 2’’ effect. In this regard, MOFs can act as an thiol-disulfide exchange reaction, and the reduced Trx can take

5102 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

electrons from NADPH catalyzed by TrxR.210 In this regard, PL produced by glucose decomposition could further react with
can inhibit the activity of TrxR, thus breaking the redox homeostasis catalase for O2 production. Therefore, this strategy presented
in tumor cells. Hence, the ROS level in cells was promoted by enhanced PDT efficacy and tumor proliferation inhibition.
1.6 times after the treatment with PCN-PL@CM, which proved Another nano-
that the effect of ROS was guaranteed, and tumor cells also reactor was made by Liu et al.63 They made a sandwich structure
became more sensitive to ROS. where the Pt NPs was embedded between the outer PCN shell and
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Starvation therapy refers to cutting off the nutrition supply of the inner PCN core. The outer PCN was further incorporated with
cells by glucose decomposition,25,211 which is usually achieved Au NPs (Fig. 12b). Afterwards, folic acid was incorporated for
by glucose oxidase (GOx) or enzyme-like agents according to the active targeting. The Pt NPs and Au NPs showed catalase-like and
following equation:25,62 glucose oxidase-like activity, respectively. The tumor inhibition of
the final product was 90.88%, much higher than that of folic acid
Glucose + O2 + H2O - Gluconic acid + H2O2 (9)
attached PCN-224 (41.93%), due to the combined photodynamic
Similar to PDT, starvation therapy also requires an O2 supply. and starvation therapy.
Therefore, Li et al. fabricated a cascade bioreactor mCGP by Gas therapy is an emerging therapy that uses gaseous trans-
encapsulating glucose oxidase (GOx) and catalase in PCN-224.62 mitters to kill cancer cells with negligible negative effect.212 Gas
PCN-224 was further camouflaged with the cancer cell membrane. molecules are cytotoxic under appropriate concentrations with-
Endogenous H2O2 was consumed by catalase to generate O2, out harming normal cells.213 Taking NO as an example, NO not
which was important for ROS production and glucose consump- only participates in many physiological processes, but also
tion. Simultaneously, glucose was decomposed by GOx, breaking shows a dose-dependent anti-tumor effect. When the concen-
the glucose metabolism balance of tumor cells. The H2O2 tration of NO is higher than 1 mM, it can directly kill cancer cells

Fig. 12 (a) Schematic illustration of interfering with redox homeostasis in cancer cells for improved PDT. Reprinted with permission from ref. 61.
Copyright 2019 Elsevier. (b) Schematic illustration of the catalytic cascades-enhanced synergistic cancer therapy driven by dual-inorganic nanozymes-
engineered PCNs. Reprinted with permission from ref. 63. Copyright 2019 American Chemical Society. (c) Schematic illustration of L-Arg@PCN@Mem
preparation and lethal mechanism of gas therapy and sensitized PDT against tumors. Reprinted with permission from ref. 112. Copyright 2018 Elsevier.
(d) W-TBP/CpG promoted antigen presentation via immunogenic PDT and CpG delivery synergizes with checkpoint blockade immunotherapy to afford
systemic antitumor immunity. Concentration-dependent IFN-a (e) and IL-6 (b) levels by ELISA, n = 3. Reprinted with permission from ref. 139. Copyright
2020 Wiley.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5103
 View Article Online

Chem Soc Rev Review Article

by mitochondria and DNA nitrosation.212 Moreover, NO molecules For checkpoint blockade immunotherapy, the commonly used
can react with  O2 to produce cytotoxic peroxynitrite (ONOO), checkpoints are T-lymphocyte-associated protein 4 (CTLA4), pro-
which has efficient oxidizing and nitrating ability.214 Generally, gas grammed cell death 1 (PD-1), programmed cell death 1 ligand
therapy requires the loading of gas donors or the attachment of (PD-L1), and indoleamine 2,3-dioxygenase (IDO), etc.215,221 Lu
gas molecules, and the release of gas is stimuli-triggered. Light et al. loaded IDO inhibitor in a TBC–Hf MOF with a content of
irradiation is a suitable stimulus due to minimal side effects, 4.7 wt%.97 IDO is highly expressed in the tumor microenviron-
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

and it also provides the possibility for combining gas therapy ment, which can inhibit the clonal expansion of T cells, thus
and PDT to enhance the therapeutic potency. As an ideal gas promoting the immune escape of tumors. The IDO inhibitor in
carrier, MOF is also appropriate for loading the gas donor and MOF can effectively block the activity of IDO, and the combination
storing the generated gas. Wan and coworkers encapsulated of the TBC–Hf and the IDO inhibitor under light irradiation
biocompatible L-arginine (L-Arg) in PCN-224 as the NO donor showed the best tumor suppression. Notably, when TBC–Hf and
for combined gas therapy (Fig. 12c).112 Upon irradiation, the the IDO inhibitor were applied without light irradiation, the
donor L-Arg can react with H2O2 and ROS to generate NO with a tumor growth was slightly inhibited, indicating that the PDT-
long half-life and wide diffusion range. More importantly, in a induced tumor cell death and the following release of tumor-
hypoxic environment, due to the free diffusion and penetration associated antigens are necessary for immunotherapy. Moreover,
of NO, NO can sensitize tumor cells to ROS. Although NO in this combined therapy, the immune cells can migrate to other
generation consumed part of the ROS, their results showed tumor areas, causing a strong abscopal effect. Hence, the tumor at
that the combination of the released NO and PDT achieved the untreated site can also be attacked by immune cells.
much better therapeutic efficacy than PDT alone, almost com- As for improving the efficiency of antigen presentation, Ni
pletely eliminating the tumor. Guan et al. firstly synthesized the et al. fabricated a cationic W-TBP MOF (TBP refers to 5,10,15,20-
UiO-66-OH(Hf) MOF composed of Hf nodes and H2BDC tetra(p-benzoato)porphyrin), which can adsorb anionic cytosine-
ligand.64 Afterwards, 2I-BODIPY, which acted as a PS, was phosphate-guanine (CpG) (Fig. 12d).139 CpG is a toll-like receptor
attached to UiO-66-OH(Hf) by etherification, and the gas donor, agonist. When internalized by dendritic cells, CpG can bind to
MnCO, was further coordinated to Hf cluster nodes. In the toll-like receptor-9, thus promoting the maturation of dendritic
presence of oxidizing agents such as H2O2 and 1O2, the Mn(I) in cells. The dendritic cells then release cytokine (e.g., immunosti-
MnCO can be oxidized to Mn(II), thus releasing CO molecules. mulatory cytokines type I Interferon (IFN-a) and interleukin-6
Hence, light irradiation not only triggered the PDT effect (IL-6)) as the marker of maturation. Hence, the activated dendritic
but also CO release through 1O2 generation. CO can hinder cells will present more tumor-associated antigens to T cells, which
the ATP generation in tumor cells by activating the oxidative then prime and traffic to activate other T cells. In their work, W-TBP
phosphorylation process, thus preventing tumor cell prolifera- absorbed CpG with an efficiency of 89.9%, which is attributed to
tion and tumor metastasis. The selectivity of the MOF and light- the high positive zeta-potential of W-TBP (37.2  0.6 mV). The MOF
controlled release of CO also prevented the side effects of CO. carrier endowed CpG with higher internalization by dendritic cells,
Besides directly releasing cytotoxic agents to kill cancer cells as evidenced by the highest dendritic maturation in Fig. 12e (IFN-a)
as mentioned above, the response of the immune system is and Fig. 12f (tested by IL-6). The combination of PDT and CpG-
worth considering. In this regard, immunotherapy was put induced immunotherapy showed 96.6% of tumor regression, much
forward, which harnesses immune cells in the tumor microen- higher than free CpG and PDT alone. Afterwards, the authors
vironment or host lymphoid tissues to target and eradicate demonstrated that this immunotherapy can be combined with
tumor cells, thus preventing tumor metastasis and facilitating checkpoint blockade immunotherapy by injecting the mice with
systemic immune surveillance.215,216 Generally, during PDT, the the a-PD-L1 antibody. This combination caused a strong and
death of tumor cells will result in the release of tumor-assisted consistent abscopal effect, which can take effect on distant tumors.
antigens, which are then presented to T cells by antigen- 3.3.4 Enhancing the penetration depth. Limited penetra-
presenting cells (especially dendritic cells).139,216,217 However, tion depth is one of the biggest shortcomings of phototherapy.
the tumor can release immunosuppressive signals through the Especially for PDT, the frequently used wavelength of light was
expression of proteins or interfering with the receptors on T 650–800 nm, which can only penetrate the tissue for 3–10 mm.2
cells, thus leading to T cell apoptosis and preventing the Hence, traditional PDT can only treat lesions that are superficial or
response of the immune system.107,216,218 On the other hand, within 1 cm of depth.6,222 Two factors cause this problem. One is
tumor-assisted antigens are deficient in the tumor microenviron- that the absorption peak of most PSs is located in the visible range,
ment, and the antigen-presenting efficiency of dendritic cells is which means that the longer-wavelength but deeper-penetrating
also poor, which all hinder the immune response.123,139,219 light is not suitable for the excitation of these PSs.222 Another
Hence, several strategies have been put forward, such as blocking concern is that when the wavelength of light is longer than
the inhibitory signals of T cells or improving the efficiency of 800 nm, the light cannot provide enough energy for the excitation
antigen presentation.216 The former is referred to as checkpoint of 3O2.7 In this regard, the use of two-photon activated PSs and
blockade immunotherapy. These strategies usually utilize immuno- upconversion nanoparticles was put forward.
stimulatory agents, such as antibodies and small molecules, which Most PSs are single-photon excitation PS, which means the PS
can be loaded in the MOF, thus preventing their degradation and can only absorb one photon under irradiation. By comparison,
improving their targeting ability.215,216,220 simultaneously absorbing two photons, namely two-photon

5104 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

activation, features long-wavelength two-photon absorption, lipoteichoic acid, which is easy for PSs to cross. Whereas, besides
which renders deep-tissue treatment and prevents the energy the cytoplasmic membrane, Gram-negative bacteria have
loss of light.223 One typical example of two-photon-excited PS is another outer membrane that acts as a barrier to prevent the
the polypyridyl ruthenium complex (RCs). Zhang and coworkers penetration of PSs.229 Therefore, traditional PDT has limited
encapsulated cationic RCs ([Ru(bpy)3]2+, [Ru(phen)3]2+ and efficacy against Gram-negative bacteria. In view of this, PS
[Ru(phen)2hipp]2+) in anionic bio-MOF-1 by means of ion requires modifications that increase the outer membrane per-
exchange.180 The MOF carrier hindered the intramolecular rota-
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

meability. It is known that bacteria usually bear a negative

tion and p–p interactions of RCs, which reduced their aggrega- charge, hence, cationic PSs have greater affinity for bacteria.
tion. Moreover, bio-MOF-1 also provided an extended Another concern is the formation of biofilms, which is a
p-conjugation system, which improved the light harvesting ability consortium of bacteria growing compactly on living tissues or
of RCs. The single-photon activation of bio-MOF-1&RCs was mea- implant materials.230,231 The extracellular polymeric sub-
sured under 490 nm light irradiation. The singlet oxygen yields of stances form a dense and protective shelter that hinders the
bio-MOF-1&[Ru(bpy)3]2+ under single-photon activation was 0.86, contact of PSs and the inner bacteria.232 The biofilm area also
which was higher than that of [Ru(bpy)3]2+ (0.81) due to the reduced features acidity (pH B 5.5) and hypoxia.132 Therefore, many
aggregation of RCs. In terms of two-photon activation, the strategies have been put forward to accommodate MOF-based
maximum two-photon activation wavelength was in the range PSs for antibacterial applications.
of 800–820 nm. Under 800 nm light irradiation, owing to the Ag ions are an important biocide that can react with bio-
stronger electron delocalization conjugate system of bio-MOF- molecules such as DNA and peptides due to their affinity for thiols,
1&[Ru(phen)3]2+ and bio-MOF-1&[Ru(phen)2hipp]2+, they had amines and phosphates, thus causing the irreversible aggregation
higher 1O2 generation than bio-MOF-1&[Ru(bpy)3]2+. Therefore, of these biomolecules and finally leading to the inactivation of
the therapeutic wavelength shifted to the longer region, which bacteria.233 Zhang et al. loaded Ag ions into PCN-224 and sealed it
rendered deeper light penetration. with surface-adaptive HA.113 The HA shell can be decomposed by
Another way to improve the penetration depth of PDT is by hyaluronidase secreted by Gram-positive bacteria. PCN-224-Ag+ was
using upconversion nanoparticles (UCNPs), which can absorb two able to attack bacteria due to charge interaction, leading to
or more low-energy photons and generate one high-energy bacterial death. Their results showed that the antibacterial abilities
photon.224,225 In this way, NIR light can be converted into visible of PCN-224 and PCN-224–HA against methicillin-resistant Staphy-
light, thus activating PSs. The frequently used UCNPs are lococcus aureus (MRSA) were similar under light irradiation, sug-
lanthanide-doped UCNPs224,226 and quantum dots.227,228 In MOF- gesting that the HA coating was degraded by MRSA. After the
based therapeutic systems, UCNPs can be attached to the surface encapsulation of Ag+, PCN-224–Ag–HA exhibited obvious bacterial
of MOF,96 or covered by MOF as the core material of the core–shell inhibition, which was more potent than PCN-224–HA and AgNO3.
structure.89,165 For example, He et al. firstly fabricated a core– The results confirmed that loading Ag+ in PCN-224 was effective for
shell-shell UCNP (NaYb0.92F4:Er0.08@NaYF4) with a diameter of bacteria-killing by the synergistic action of both Ag+ and PDT. It
29.8  2.2 nm.89 Under 980 nm laser irradiation, this UCNP can also provided insight for loading other kinds of antibacterial agents
convert the light source into 524, 542 and 660 nm light, with a in MOF-based PSs, such as Cu2+ and Zn2+.234–237
quantum yield of 6.55  0.34%, which is much higher than that of Bacterial adhesion is the initial stage of biofilm formation,
traditional UCNPs. The UCNP was then coated with the PCN-222 which is promoted by ATP stimulated cell lysis and extracellular
MOF shell (composed of Zr nodes and FeTCPP ligands) and further DNA release. At this time point, the bacteria are easily attacked
decorated with Au NPs. The absorption of PCN-222 was located at by PS because of the absence of biofilm protection. In view of
554 and 646 nm, which matches well with the emission of UCNPs. this, Qiu et al. decorated PCN-224 with CeO2 NPs of 4 nm in size
Au NPs serve as glucose oxidases to trigger glucose starvation to form a shell structure on PCN-224 with L-arg as the capping
therapy and generate a large amount of H2O2. The chelated Fe agents to perform ATP deprivation and PDT (Fig. 13).114 The
atom in the FeTCPP further reacted with H2O2 for O2 generation. Ce3+ and Ce2+ in CeO2 can react with ATP through the nitrogen
Hence, the Au NPs and PCN-222 together caused a cascade and oxygen on adenine and triphosphate of ATP. Therefore, the
reaction, and the UCNPs also deepened the therapeutic depth of biofilm formation was inhibited because of ATP deprivation,
PDT. After 8 days of 980 nm light treatment, the tumor-bearing and the planktonic bacteria were subsequently killed by PDT.
mice showed complete tumor eradication without recurrence. Results showed that after incubation with different contents
3.3.5 Typical modifications towards antibacterial applications. of MOF@CeO2, the ATP content decreased with the higher
Bacteria-killing is another important application of PDT, and the concentration of MOF@CeO2, while the ATP content showed
mechanism is similar to the anticancer mechanism. Under light no obvious variation with different concentrations of PCN-224.
irradiation, bacteria are killed by ROS due to the damage of the Therefore, the decoration of CeO2 could effectively eliminate
bacterial cell membrane and the DNA inside. However, the strate- ATP, which led to 40% of biofilm inhibition for MOF@CeO2
gies in PDT are not completely suitable for antibacterial applica- (50 mg mL1). When light irradiation (638 nm) was introduced,
tions due to the following reasons. Firstly, bacteria can be divided the biofilm inhibition was 90% for MOF@CeO2 at the concen-
into Gram-positive (e.g., Staphylococcus aureus) and Gram-negative tration of 200 mg mL1.
(e.g., Escherichia coli) bacteria. The cytoplasmic membrane of Different from the initial stage, once the biofilm forms, the
Gram-positive bacteria is covered by porous peptidoglycan and inner bacteria are protected by the biofilm.238 In this case, the

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5105
 View Article Online

Chem Soc Rev Review Article

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 14 Basic cubic network structure of PB. Reprinted with permission

from ref. 239. Copyright 2015 Royal Society of Chemistry.

4.1.1 Prussian blue-based MOF. PB is one of the oldest

synthetic MOFs that has been intensively studied.239 The chemical
structure of PB is presented in Fig. 14. The carbon atom and
nitrogen atom of –CRN– in PB is coordinated with Fe(II) and
Fe(III), respectively, creating a face-centered cubic structure.239,240
The PB has two categories: soluble PB (KFeIII[FeII(CN)6]) and
Fig. 13 Schematic illustration of the preparation process of CeO2- insoluble PB (FeIII II
4 [Fe (CN)6]3nH2O).
240
As for insoluble PB,
decorated PCN-224 and the inhibition of biofilm formation through the
depending on the coordination site of water molecules, it can
synergy of adhesion-related molecule deprivation and cytotoxic ROS
generation. Reprinted with permission from ref. 114. Copyright 2019 Wiley. be divided into coordinative type (water molecules coordinate to
iron ions) and zeolitic type (water molecules occupy cavities).240
In biomedical applications, PB has been approved by FDA to
diffusion of PS and ROS should be taken into consideration. treat radioactive exposure as clinical medicine with good bio-
Reducing the particle size of therapeutic agents is a feasible way to compatibility and biosafety.100,241 Owing to the charge transfer
penetrate the biofilm. Deng and coworkers fabricated a MOF dot between Fe(II) and Fe(III), PB can generate heat under NIR
composed of Hf nodes and TCPP ligands, which was further irradiation for tumor ablation and bacteria disinfection.100
covered with human serum albumin and MnO2.132 The final The Fe(III) nodes in PB can also react with H2O2 and generate
product was named MMNPs. The diameter of MMNPs was about 
OH through the Fenton reaction.242 Moreover, the coordina-
105 nm. The outer MnO2 would degrade in the mildly acidic tion between Fe(III) and water molecules in insoluble PB renders
environment or in the presence of H2O2, releasing porphyrin- an inner-sphere longitudinal relaxation time with a longitudinal
based MOF and O2 for hypoxia alleviation. The ultrasmall MOF relaxivity of 0.14 mM1, which can be used for MRI-guided
dot was about 5 nm in size and positively charged, which had therapy.243
excellent penetration ability through the biofilm and a high PB can directly serve as PTA in PTT. Yue et al. fabricated PB
affinity for the negatively charged bacteria. Therefore, the MMNPs by mixing a solution of FeCl3 and K4[Fe(CN)6] with citric acid as
showed 99% and 90% of the antibacterial ratio against E. coli and the surface capping agent.100 In their work, the size of PB could
S. aureus, respectively. Compared to the bulk MOF, this MOF dot be controlled by different concentrations of citric acid, ranging
showed 12 times higher accumulation in the biofilm. Animal from 10 to 50 nm. In addition, their PB had a broad absorption
testing manifested that the abscesses of S. aureus-infected mice band at 500–900 nm, and the absorption peak was located at
healed with no obvious inflammation in 5 days after MMNPs 712 nm. Under 808 nm irradiation, the molar extinction
treatment and visible light irradiation. coefficient was 1.09  109 M1 cm1, which was slightly lower
than that of Au nanorods. The temperature of PB rose to 43 1C
under less than 3 min of irradiation. After the PTT treatment by
PB, the viability of HeLa cells was lower than 10%. However, the
4. Photothermal therapy absorption peak of PB was near the edge of the near-infrared
4.1 Intrinsic photothermal MOF region (700–900 nm), and the photothermal conversion efficiency
Intrinsic photothermal MOF refers to MOF that can be directly was merely 20% under 808 nm irradiation.54,56 The subsequent
used as PTAs without the need for extra PTA decoration. One of research mainly focused on improving the therapeutic efficacy by
the most important branches of intrinsic photothermal MOF is doping, etching, drug loading and auxiliary methods such as
Prussian blue (PB) and its analogues. Other intrinsic photo- imaging guidance.
thermal MOF is based on (i) photothermal building units and Doping is a common modifying method for PB. Up to now,
(ii) the ligand-to-metal charge transfer mechanism to have the doped metal ions include Mn2+,244 Zn2+,54,245 Cu2+ (ref. 246)
photothermal ability. Hence, in this section, we will introduce and Gd2+,247–249 etc., which are usually located at the interstitial
the composition as well as typical modification methods for site or lattice site (Fig. 15a).249 The dopant size, distribution
intrinsic photothermal MOF. and concentration will affect the chemical properties of PB,

5106 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

rendering PB with tunable photothermal behavior.54 Besides, the authors added Mn2+ and [Fe(CN)6]2 as sources to fabricate
imaging guidance can also be introduced through doping. Zhu a Mn-containing PB analogue shell on both the exterior and
et al. doped Mn2+ in PB, and coated PB with poly(allylamine interior surfaces of HMPB (Fig. 15b). The final product was
hydrochloride) (PAH), polyacrylic acid (PAA) and PEG through termed as HMPB–Mn. DOX was loaded in HMPB–Mn with an
the layer-by-layer method.244 The dopant changed the electron efficiency of 97.5%, which was attributed to its coordination
density and orbital energy of PB. Thus, the absorption peak of with Fe2+ and Mn2+ nodes. In an acidic environment, the
Mn–CRN–Fe structure will decompose, thus releasing Mn2+
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

PB red-shifted to 718, 730, and 768 nm, corresponding to the

Mn2+ doping ratio of 5%, 15% and 25%, respectively. On the and DOX. The release efficiency of Mn2+ and DOX at pH 5.0 was
other hand, Mn2+ also increased the longitudinal relaxivity of 95% and 34.7%, respectively. The pH-induced Mn2+ release
PB, which showed much-enhanced contrast in MRI. Shou et al. facilitated the MRI ability of HMPB–Mn, which showed the
doped 10% Zn2+ in PB (SPBZn(10%)), which resulted in ultra- brightest MRI images at pH 5.0 with a molar relaxivity of
small PB particles (3.8  0.90 nm).245 The SPBZn(10%) showed 7.43 mM1 s1. Therefore, this pH-triggered smart therapeutic
a 1.3 times higher absorption peak than the non-doped PB. agent can be used for MRI-monitored drug release, which was
When the doping concentration changed, the photothermal synergized with PTT. In another work, Cai et al. loaded DOX
conversion efficiency varied from 37.73% to 47.33%. Besides, and perfluoropentane (PFP) in HMPB.252 HMPB had a higher
the doped Zn2+ could replace Fe2+, giving rise to higher Fe3+ intrinsic molar extinction coefficient at 808 nm, which could be
concentration, and the magnetic saturation value was also used for photoacoustic (PA) imaging. When the temperature
improved, which was beneficial for the MRI guidance. increased, the bubbles generated by PFP could act as the
The pore size is another important factor that affects the contrast agent of ultrasound (US) imaging. The combination
photothermal performance of PB MOF and the drug-loading of PA and US imaging was beneficial for the early diagnosis of
efficacy. Generally, the pore size of PB is smaller than 1 nm,250 cancer. Therefore, this nanoplatform combined PTT, chemo-
and the loading capacity is also limited, which is not suitable therapy and diverse imaging on the basis of HMPB, showing
for drug loading.251 Therefore, many researchers have reported almost complete eradiation of tumor.
the hollow mesoporous PB (HMPB) structure by chemical The Fe2+, Fe3+ and –CN in PB can be active sites for coordina-
etching or the hydrothermal method. Under light irradiation, tion with therapeutic agents. One of the major applications of
the elevated local temperature can accelerate the drug diffu- coordination modification is gas combined therapy. Gas therapy
sion. For example, Zhou and coworkers used hydrochloric acid utilizes gas molecules (e.g., CO,253 NO212,254 and H2104) as cytotoxic
to etch PB crystals.25 Nitrogen absorption isotherms mani- agents, which can be attached to a carrier such as MOFs and then
fested that the Brunauer–Emmett–Teller (BET) surface area of released under certain environment conditions. Besides ROS (as
PB increased from 302.56 m2 g1 to 922.14 m2 g1 with a pore mentioned in Section 3.3.3), light-induced temperature increase is
diameter of 10.12 nm. Cai and coworkers synthesized HMPB by also an ideal trigger for gas release. For example, Li et al.
the hydrothermal method and chemical etching.241 Afterwards, coordinated Fe(CO)5 on mesoporous PB by replacing –CN with

Fig. 15 Different modification methods of PB. (a) Doping: schematic of Gd3+ simultaneously optimizing the properties of PB nanocrystals. Reprinted
with permission from ref. 249. Copyright 2016 American Chemical Society. (b) Hollow mesoporous PB: schematic illustration of the synthetic procedure
of HMPB–Mn. Reprinted with permission from ref. 241. Copyright 2015 Wiley. (c) Coordination modification: schematic illustration of PB–CO–TPZ NPs
with enhanced bioreductive chemotherapy and CO-mediated pro-apoptotic gas therapy. Reprinted with permission from ref. 255. Copyright 2019
Elsevier. (d) PB-based composites: schematic illustration of HSP70 promoter-based PB theranostic platform for gene therapy/PTT. Reprinted with
permission from ref. 257. Copyright 2018 Wiley.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5107
 View Article Online

Chem Soc Rev Review Article

one of the CO moieties of Fe(CO)5.46 CO can lead to cancer cell

apoptosis by inducing mitochondria disorder. However, CO can
also bind with hemoglobin and oxygen, showing acute toxicity to
normal tissues. In order to avoid CO poisoning, the CO release
amount was controlled by irradiation time and intensity. The CO Fig. 16 Typical photothermal ligands for intrinsic photothermal MOF.
release was suppressed in blood circulation. Only under irradia-
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

tion can the photothermal effect of PB cleave the Fe–CO bond for
CO release. Since CO treatment can accelerate mitochondrial computed topography (CT), etc.) of PB. The core material also
respiration in the second stage, which facilitates O2 consumption, provides synergistic functions. Peng et al. coated PB on MnO2
Li et al. introduced hypoxia-activated TPZ in Fe(CO)5 and PB and encapsulated the NP with red cell membrane (RCM).272
therapeutic system, and coated it with PAH, PAA and NH2- DOX was loaded within RCM by coextruding. The MnO2 core
modified PEG (Fig. 15c).255 Another coordination method was decomposed H2O2 at the tumor site with O2 generation, which
reported by Feng et al. During the fabrication process of PB, the resulted in RCM disruption and DOX release. Moreover, the
authors doped PB with sodium nitroprusside (SNP), which is a NO RCM prolonged the circulation time of NPs and increased the
donor.256 As SNP was embedded in the crystal structure of PB, aggregation of NPs at the tumor site, which resulted in a higher
some of the Fe3+ nodes in PB were coordinated with NO. After- PTT temperature (59.6 1C) as compared to the non-coated samples
wards, the PB was loaded with the anticancer drug docetaxel (49.3 1C). This group also demonstrated that the PB/MnO2
(DTX), generating DTX@m-PB–NO. Upon irradiation, the heat led nanocomposite has a high diamagnetic transverse relaxation time
to the cleavage of Fe–NO bonds, while no NO release was detected (T2) signal intensity, which was promising in MRI and PAI.271
at 37 1C, indicating its safety in blood circulation. The highest NO 4.1.2 Utilizing photothermal agents as building units.
release amount was 10.43 mM. When the NO content reached the MOFs with intrinsic photothermal effect can be synthesized
micromolar level, it could inhibit tumor growth and metastasis by by employing various kinds of PTAs as ligands, as schematically
nitrosation. shown in Fig. 16.49,93,103
Besides the above PB-based PTAs, some other PB-based For instance, Yang et al. synthesized a MOF using Mn as
composites can be derived by coating specific materials on nodes and IR825 as ligands, and they covered this MOF with
the surface of PB or employing PB as the coating. The former PDA and PEG for better biocompatibility.93 On the one hand, as
can be categorized into polymer coating (PEI,257 PEG,258,259 a common NIR dye with NIR light absorption peak at 825 nm,
chitosan,260 gelatin,261 and PDA,262 etc.), inorganic coating IR825 endowed the composite with photothermal performance.
(SiO2,251,263,264 etc.), and MOF coating (ZIF-8,265 MIL-100(Fe),169,266 Under 808 nm light irradiation, the mass extinction coefficient
UiO-66,267 etc.), which is mainly used for stability enhancement, of MOF at 808 nm was 78.2 L g1 cm1, and the temperature
drug loading, and other combined therapy. Wang et al. fabricated quickly rose to B52 1C within 5 min. In addition, their results
the PB@MIL-100(Fe) dual-MOF structure.266 The outer MOF shell showed that this material possessed photothermal cycling
was used as a carrier of anticancer drug artemisinin (ART) with a ability. On the other hand, the Mn nodes also offered the
loading efficiency of 848.4 mg g1. ART was released under an MOF with MRI guidance. More recently, Lü et al. reported a
acidic environment because of MIL-100 shell decomposition. More- Zr-PDI (PDI refers to perylenediimide) MOF with high photo-
over, as the MIL-100 shell increased the dielectric constant of the thermal conversion efficiency.49 PDI can be reduced into PDI ,
composite NPs, the absorption peak of PB red-shifted 35 nm. which is a delocalized radical anion with red-shifted absorption
Hence, the temperature of PB@MIL-100(Fe) increased 30 1C at a compared to the PDI molecule. However, PDI  is unstable in
concentration of 0.2 mg mL1 with excellent photostability. Liu ambient conditions. To solve this problem, the authors coordi-
et al. coated PB with PEI, which was loaded with pDNA by nated PDI with Zr6 clusters, forming Zr-PDI MOF (Fig. 17a–c).
electrostatic interactions (Fig. 15d).257 The pDNA was HSP70– The Zr6 cluster was coordinated with 12 carboxyl groups, and
p53–GFP, wherein HSP70 has higher expression in the moderate the PDIs showed axial chirality (Fig. 17b). Zr-PDI can obtain
temperature region (39–43 1C) and acted as a promoter to activate electrons from electron donors (e.g., amine vapors) through
gene therapy, while p53 and GFP was the targeted tumor- photo-induced electron transfer (PET), generating radical anion
suppressive gene and reporter, respectively. After the endocytosis Zr-PDI . In this work, the PET process was performed by
of PB@PEI/pDNA NPs, the low heat (r43 1C) can facilitate the irradiating Zr-PDI with blue light (455 nm) in amine vapor
endosomal escape of material by destroying the endosomal (Fig. 17d). The absorption of Zr-PDI  was obviously red-shifted
membrane, thus releasing the pDNA, which was then expressed (Fig. 17e). Due to the shielding effect of the MOF structure, the
to the nucleus. When the temperature was about 41 1C, HSP70 quenching of radical anions was efficiently prevented. Zr-PDI 
initiated the expression of p53, which resulted in tumor apoptosis. is the first reported stable isolated radical anion, which
When the temperature exceeded 50 1C, the hyperthermia led to cell retained the high light-to-thermal conversion of PDI  with a
necrosis. Therefore, the gene therapy and PTT showed a synergistic photothermal conversion efficiency of 52.3% (Fig. 17f), showing
effect, which was controlled by PB-induced heat generation. great promise in PTT. More recently, Deng et al. used 1,1 0 -
In the latter case, the most reported strategies were to coat ferrocenedicarboxylic acid (Fc(COOH)2) as a ligand that was
PB on contrast agents such as Au,268,269 Fe3O4,270 and MnO2,271 coordinated with Zr nodes.103 The final product was termed as
etc., to enhance the imaging properties (MRI, PA imaging, and Zr-Fc MOF, which is a 16.4 nm thick MOF nanosheet. The PTT

5108 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 17 (a) Structures of the Zr-cluster and P-2COOH. (b) Connection mode of the Zr-cluster (the Zr-cluster is fully coordinated by 12 carboxylate units)
and the molecular arrangement of chiral P-2COOH. (c) c-Axis crystal structure of Zr-PDI. (d) UV-vis-NIR absorption of Zr-PDI, Zr-PDI  and Zr-PDI 
after a month. The insets show the photographs of color changes. (e) Illustration of the colored PDI  formation and photothermal conversion of
Zr-PDI . (f) Photothermal conversion curves of Zr-PDI  films on quartz glass under laser irradiation (808 nm, 0.7 W cm2). Reprinted with permission
from ref. 49, Copyright 2019 Nature.

effect was attributed to the ferrocene ligand. Zr-Fc MOF showed In terms of photothermal nodes, Zhou et al. reported a
broad absorption in the region of 350–1350 nm. Under 808 nm porphyrin-palladium MOF (Pd-MOF) composed of Pd nodes
light irradiation, the temperature of Zr-Fc reached 92 1C in and 5,10,15,20-tetrakis(4-pyridyl)-21H,23H-porphine (TPyP) ligands
3 min, while for Fc(COOH)2, the highest temperature was merely for hydrogen-thermal therapy.104 The photothermal effect was
46.8 1C. This difference was derived from the greater stability of attributed to a single-atom Pd unit, and the photothermal
Zr-Fc MOF. Besides, the radiative decay of Zr-Fc MOF was efficacy was up to 44.2%. More importantly, Pd nodes have
inhibited as manifested by the fluorescence emission, which specific coordination ability with hydrogen. After hydrogenation,
indicates that more excited photo-electrons decayed through the obtained PdH-MOF had high hydrogen loading. The continuous
nonradiative pathways, thus more heat was generated for PTT. hydrogen release could effectively scavenge  OH and ONOO,
Moreover, Zr-Fc can also act as a Fenton catalyst, which showed causing the disorder of tumor energy metabolism. Hydrogen
3.3 times higher  OH production at 45 1C as compared to that at therapy also lowered the required intensity of light in PTT.274
25 1C. Hence, when incorporated in the MOF structure, the 4.1.3 Photothermal effect by ligand-to-metal charge transfer.
ferrocene ligand could provide chemo- and photothermal- For coordination structures such as MOF, the electrons go
synergistic therapy, which was simple and straightforward. through not only interband transition, but also various charge
As for mixed-ligand MOFs, Zheng et al. incorporated tetra- transfer processes such as metal-to-ligand charge transfer and
topic porphyrin (TCPC) into Hf-UiO-66 with a content of 8 wt% ligand-to-metal charge transfer, etc.91 Ligand-to-metal charge
(referred to as TCPC-UiO).273 As a porphyrin derivative, TCPC transfer (LMCT) occurs when the metal nodes are in a relatively
had photodynamic activity. However, the ROS yield of TCPC- low valence state. When the material is excited, the electron will
UiO was much less than the free TCPC molecule, which was move from the donor orbital of the ligand to the accepting
probably ascribed to the weakened confinement of TCPC and orbital of the metal nodes.91 LMCT endows materials with strong
the resulting aggregation. Instead, the photothermal ability of absorption in the low-energy regions, such as the visible light
TCPC played a dominant role in this therapeutic system. Under and NIR light region, which facilitates photothermal ability.91,275
630 nm light irradiation, the temperature increased to 42 1C The reported combinations of metal nodes and linkers used in
within 5 min at the lowest NPs concentration. The tumor PTT are given in Fig. 18.275–277
inhibition also reached up to 90% after treatment. Therefore, Liu et al. fabricated Fe-CPND composed of Fe3+ nodes and
the photothermal ability of porphyrin-based MOF is also worth gallic acid ligand, which was protected by PVP.275 The Fe-CPND
exploring. was 5.3 nm in size, which can be cleared by the kidneys, thus

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5109
 View Article Online

Chem Soc Rev Review Article

encapsulation in MOF can improve the photothermal ability of

ICG, which was ascribed to the better photostability of ICG
under the protection of the MOF. Moreover, Fe nodes and HA
also rendered the integration of fluorescence imaging, PA
imaging and MRI guidance. Zhu et al. encapsulated Pd
Fig. 18 Typical combinations of intrinsic photothermal MOFs based on
ligand-to-metal charge transfer. nanosheets and DOX in ZIF-8 through a one-pot strategy, and
coated ZIF-8 with PDA.279 The Pd nanosheets acted as a PTA,
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

which was 17 nm in size. After the encapsulation of Pd

lowering the toxicity of the material. In an acidic environment, nanosheets, the particle size of ZIF-8 increased from B150 to
the gallic acid ligand gradually disassociated from Fe-CPND B300 nm. After the coating of PDA, the photothermal conver-
and the Fe3+–gallic acid complex changed from tris- to bis- sion efficiency of the final product was 45%. Besides the
coordination, which was attributed to the change in the coor- photothermal effect, this material also has pH-controlled drug
dination number of gallic acid. This pH-induced structural release ability. Zhang and coworkers used IR820 as the bond
change endowed Fe-CPND with higher longitudinal relaxivity between the anticancer drug cytarabine (Ara) and ZIF-8 to
at pH = 5.0. Hence, the MRI ability of Fe-CPND can be triggered improve the interaction between the model drug and MOF.66
by tumor acidity. The LMCT effect was attributed to the ZIF-8 was further covered with HA to realize active targeting
phenolic oxygen on gallic acid. Under 808 nm light irradiation, through enhanced permeability and the retention effect. Due to
the temperature of Fe-CPND quickly rose to 50 1C, which the small molecular size and lack of strong bonding functional
significantly inhibited tumor growth. Moreover, the absorption groups, Ara was hard to load in the MOF without leaking. In
of Fe-CPND did not change after being irradiated for 60 min, their work, the amino bond of Ara was firstly bonded with the
which indicates the good photostability of Fe-CPND. More carboxyl bond of IR820, forming a prodrug. Then, the prodrug
recently, it has been reported that MOFs composed of Fe3+ was loaded in ZIF-8 by coordinating with the sulfonic group of
metal nodes and ellagic acid277 or hydrocaffeic acid ligands276 IR820 with a loading content of 39.8%. In the acidic tumor
can be used as PTAs due to the LMCT effect. Similar to the area, ZIF-8 would decompose, after which the amide linkage
gallic acid ligand mentioned above, both ellagic acid and was hydrolyzed by amidase, releasing Ara for chemotherapy.
hydrocaffeic acid can form complexes with Fe3+. The photo- ZIF-8 also improved the photostability of IR820, which resulted in
thermal ability of this kind of MOF originated from the higher photothermal conversion efficacy and high-temperature
Fe-phenol structure, which resulted in the LMCT effect. Hence, increase (26.7 1C).
these findings could inspire the design of MOF compositions in
the future exploration of intrinsic photothermal MOF.
4.3 MOF-derived carbon materials
4.2 Modifications via photothermal agents MOF-derived carbon materials are fabricated by MOF pyrolysis,
Photothermal MOFs can be fabricated by the encapsulation of which possess high porosity and controllable structures.280–282
PTAs, such as indocyanine green (ICG) and its derivatives. The During pyrolysis, the metal cations in the MOF are reduced to
chemical structures are presented in Fig. 19.66,185,278 isolated metal sites, and the organic linkers are changed into
Cai et al. loaded indocyanine green (ICG) in HA-coated carbon supports due to carbonization.283 The heteroatoms on
MIL-100(Fe) with a loading content of 40%.278 In vitro tests organic linkers (e.g., N atoms on the 2-mIm linkers of ZIF-8)
manifested that under 808 nm light irradiation for 3 min, the remained on the carbon framework.280,284 To add more metal
temperature could reach up to 70 1C, and the temperature sites to this carbon material, metal precursors can be pre-
could remain for 10 min, while free ICG solutions reached encapsulated in the MOF before pyrolysis.285 Moreover, the
65.47 1C after 3 min of irradiation but soon decreased due to carbon defects, metal sites and the graphitic contents can be
the poor photostability of ICG. This result indicates that the tuned by different pyrolysis conditions.284 Like the traditional
carbon materials, MOF-derived carbon materials also have
light-to-thermal conversion ability, which can be used for
PTT. More importantly, the doped metal atoms and heteroatoms
can be used for single-atom catalysts.283,285,286 To be specific,
if the organic linker contains nitrogen, a M–N–C structure
(M refers to Co, Fe, Mn, Zn, etc.) can be generated after pyrolysis
under certain environments, wherein the metal centers are
coordinated with several N atoms, forming the M–Nx structure,
and each metal site is isolated.287 Compared to the bulk or
nanocatalysts, single-atom catalysts are on the atom scale, which
means they have much higher efficiency than traditional
catalysts.283 Therefore, single-atom catalysts have been widely
applied in the O2 reduction reaction288 and CO2 reduction
Fig. 19 Chemical structures of PTAs that can be encapsulated in MOFs. reaction, etc.289 In biomedical applications, it can also generate

5110 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev


OH through the Fenton reaction, which is comparable to structure, and then cleaved by the homolytic path, releasing a
natural enzymes.287,290 
OH and leaving a hydroxyl group on the Fe atom. In an acidic
Huo et al. fabricated a single-atom catalyst based on an environment, the residual hydroxyl group can be desorbed from
Fe-decorated ZIF-8 precursor that could combine PTT with  OH the Fe atom in the form of a H2O molecule by reacting with a
generation chemotherapy (Fig. 20a).290 The authors firstly protonated hydrogen atom. Hence, this Fe–N4 structure can be
encapsulated FeIII acetylacetone in ZIF-8 by a precursor isola- used for decomposing another H2O2 molecule. However, in a
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

tion strategy, and the composite was then subjected to pyrolysis neutral environment the desorption of OH* is much harder
at 800 1C. After the pyrolysis, Fe atoms from FeIII acetylacetone (Fig. 20d) due to a lack of protonated hydrogen atoms. Although
were coordinated with neighboring N atoms, forming the Fe1– the residual hydroxyl group can be attached to another H2O2
Nx structure, which was used for single-atom Fe nanocatalysts, group through hydrogen bonding, the  OOH formation and
termed as SAF NCs (Fig. 20b). The loading of the Fe single-atom desorption is hard to occur due to a high energy barrier. Hence,
was 1.54 wt%. The  OH production rate of this material is much in a neutral environment, the catalytic activity of the single-atom
higher than that of Fe3O4, and the  OH production is accelerated catalyst is inhibited. To improve its biocompatibility, SAF NCs
in an acid environment. The  OH generation mechanism is was further PEGylated, referred to as PSAF NCs, which could
shown in Fig. 20c and d. When an H2O2 molecule approached induce the ferroptosis of tumor cells. The resulting lipid peroxi-
SAF NCs, it was firstly absorbed on the Fe atom of the Fe–N4 dation also contributed to tumor cell death. The tumor

Fig. 20 (a) Schematic illustration of SAF NCs for PTT and the Fenton reaction in the tumor microenvironment. (b) Schematic diagram of the isolation-
pyrolysis approach to synthesizing SAF NCs. Proposed reaction mechanism schematics for SAF NCs in the heterogeneous Fenton reaction toward
generating  OH under (c) acidic (protonated) and (d) neutral catalytic milieu. Reprinted with permission from ref. 290. Copyright 2019 American Chemical
Society.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5111
 View Article Online

Chem Soc Rev Review Article

inhibition rate of PSAF NCs without light irradiation was up to maximum temperature under irradiation was 57.2 1C, and its
63.5%. As for PTT, under 808 nm light irradiation, the tempera- photothermal conversion efficiency was 21.6%.
ture of PSAF NCs increased by 45 1C. With the assistance of PTT, The small size of the polymer monomer makes it easy to be
PSAF NCs could achieve complete eradication of tumor. introduced and subsequently polymerized in the nanopores of
For ion release-related applications, Yang et al. annealed MOF; this method is called in situ polymerization. By this
ZIF-8 under an Ar atmosphere, 800 1C and O2 atmosphere, method, it is easy to control the polymer size and generate a
uniform hybrid of MOF and polymers.293 Huang et al. fabricated
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

200 1C, respectively, after which the ZIF-8 was turned into ZnO-
doped carbon NPs with a diameter of 50 nm.291 Afterwards, MIL-53, which has mixed-valent Fe nodes.297 The unsaturated
poly(N-isopropylacrylamide) (PNIPAM) was coated, which coordination of the Fe3+ node can be used to oxidize the
served as a thermo-responsive gel layer. The final product was polymerization of Py. After the in situ polymerization of PPy,
called ZnO-CNP-TRGL. Under 5 min of 808 nm light irradiation, MIL-53 retained its structure and drug loading capacity. The
the temperature of ZnO-CNP-TRGL increased to 55 1C, confirming DOX loading content in MIL-53 was up to 90%, and the release
the photothermal ability of this ZIF-8-derived carbon material. of DOX was triggered in an acidic environment. Under 808 nm
Moreover, when immersed in phosphate buffer solution (PBS), light irradiation, the temperature increased by 22.1 1C at
about 0.63 wt% of Zn2+ from ZnO-CNP-TRGL was released, which 0.50 mg mL1 with good photostability. The PPy@MIL-53 com-
can destroy the membranes of the bacteria and cause the posite can also act as a contrast agent for MRI. Wang et al. firstly
denaturation of proteins and enzymes. Moreover, the thermo- blended dopamine monomer with manganese acetate, which
responsive gel layer changed from highly hydrophilic to hydro- formed coordination bonds through phenolic hydroxyl groups.53
phobic when the temperature increased from room temperature Afterwards, the ligand precursor, K3[Co(CN)6], was added to the
to higher than the lower critical solution temperature of PNIPAM. mixture. Hence, the dopamine monomer was loaded in the
Hence, at a higher temperature, ZnO-CNP-TRGL can form bacterial pores of MnCo MOF, which is a PB analogue. After 24 h of
aggregations due to its hydrophobicity and the hydrophilic stirring, dopamine was in situ polymerized, forming the com-
adhesin proteins on bacteria. The bacteria trapping process posites of PDA and MnCo, referred to as MCP, which was further
occurred at 40 1C, and the material showed complete bacterial attached with PEG and tumor-targeting peptide cRGD-SH (cyclic
eradication of S. aureus when the temperature was higher than arginine–glycine–aspartic acid). Due to the p–p stacking
45 1C, due to the synergistic effect of PTT, Zn2+ release and between –CRN and dopamine, the absorption of MCP showed
bacteria trapping. Fan and coworkers in situ fabricated ZIF-8 on an extra absorption at 550–700 nm, which indicates that the
graphene oxide nanosheets, which was then carbonized under absorption of MCP was enhanced. The photothermal conver-
an Ar and O2 atmosphere, successively.292 The as-obtained ZnO- sion efficiency of MCP was 41.3%, which was higher than that
doped graphene was grafted with transformable thermal-responsive of PDA NPs (36.9%). Hence, in situ polymerization could
brushes, giving rise to TRB-ZnO@G. When the temperature rose enhance the therapeutic efficacy of photothermal polymers,
to 56 1C by the PTT effect, the polymer brushes changed from and the MnCo structure also provided feasibility for MRI and
hydrophilic to hydrophobic, which can be used for bacteria targeting molecule attachment.
trapping. The 2D morphology of graphene nanosheets can also 4.4.2 Core–shell structure. Traditional photothermal NPs
kill bacteria via physical cutting. such as Au-based particles (Au nanorods,298 Au nanostars,52,299
and Au NPs300–302), Pd nanocubes,99 magnetic carbon,303 PB,
4.4 MOF-based composites and photothermal polymers, etc., usually lack homogeneous
4.4.1 MOF-photothermal polymer composites. Polymers targeting and drug loading capacity. Moreover, the particles
have been widely applied in MOF-based therapeutic systems also suffer from aggregation. Coating the MOF shell on the
for better stability and biocompatibility.293 Besides this property, photothermal core not only makes up for these shortcomings
polypyrrole (PPy),39,294 PDA295 and polyaniline (PAN),296 etc. have but also improves the crystallinity of the composite.304 As PB-
additional photothermal ability. In the hybridization of photo- and polymer-based core–shell structures have been discussed
thermal polymers and MOF, polymers are usually in the form of above, we firstly take Au-based particles as an example. The
surface coating, core structure, decoration NPs or in situ poly- photothermal ability of Au particles mainly stems from the
merized in the pores of MOF. Zhu et al. synthesized PPy NPs, and LSPR effect under light irradiation.82 Li et al. fabricated a single
introduced PVP on PPy NPs to facilitate the nucleation of gold nanorod (AuNR) of 47 nm in diameter and 12 nm in
MIL-100.294 Afterwards, DOX was loaded in MIL-100. The release length.298 Afterwards, ZIF-8 was synthesized on PVP-stabilized
of DOX was influenced by both low-pH-induced MOF decomposi- gold nanorods and loaded with DOX. The LSPR absorption
tion and NIR irradiation-induced temperature increase. In their peak of AuNR@ZIF-8 shifted to B810 nm as compared to the
work, PPy served as the PTA, and its combination with MIL-100 sole AuNR (B790 nm), which was attributed to the influence of
provided feasibility for drug loading-assisted PTT. Wang et al. the ZIF-8 shell. Owing to the AuNR core, the nanocomposite
utilized UiO-66 to absorb aniline monomer through electrostatic showed potent photothermal efficacy under 808 nm NIR light
interaction. Afterwards, the oxidizing agent, ammonium persul- irradiation. The release of DOX was triggered by the acidic
fate, was added to trigger the polymerization. PAN was polymerized environment and NIR light irradiation because of ZIF-8
on the surface of UiO-66, giving a smooth surface of the particle. decomposition, thus enhancing the potency of therapy. Deng
Due to the strong absorption of PAN in the NIR region, the and coworkers reported a ZIF-8-covered Au nanostar, which was

5112 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

further etched by tannic acid.52 The cavity between ZIF-8 and Additionally, as a typical contrast agent, Gd3+ improved the MRI
the Au nanostar was utilized to store DOX. The Au nanostar has efficacy of PB. Moreover, the Gd-doped PB can also reduce
strong absorption in the second NIR light (NIR-II) region (1000– oxidative stress by scavenging ROS.
1350 nm). Owing to the deeper penetration and larger max- Li et al. doped PB with Zn with increasing doping levels,
imum permissible exposure, the NIR-II-based PTT has better which were called ZnPB-1, -2, and -3.54 Fe2+ in PB was replaced
therapeutic efficacy. The composite had higher photothermal by Zn2+, forming the Fe–CRN–Zn structure. As shown in
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

conversion efficiency under 1064 nm light (48.5%) as compared Fig. 21a, the structure of PB was simplified into two kinds of
to 808 nm light irradiation (30.2%). The generated heat also octahedrons with Fe located at the center (Fe(III)–N and Fe(III)–C).
dissociated the Zn–O coordination bond, thus accelerating the The Zn doping gave rise to another octahedron that was centered
DOX release. Moreover, the strong absorption also provided PA with the Zn atom (Zn(II)–C). Along with increasing the Zn dopant
imaging and infrared photothermal imaging. concentration, the bandgap of ZnPB decreased from 1.72 to 1.65 eV
Besides Au-based NPs, Deng et al. coated MIL-100 on magnetic according to density functional theory, which was also proved by
carbon NPs.303 The MOF shell was PEGylated and subsequently experimental calculations (Fig. 21b). On the other hand, the
modified with Mn carbonyl. Finally, DOX was loaded in MIL-100. electronic density also increased with more Zn doping, which gave
The magnetic carbon core served as a PTA for PTT, and it also rise to the major electron transition changing from bandgap to
provided PA imaging and MRI response for imaging guidance. LSPR. Due to the hybridization of Zn dopant and crystal structure
The [Mn(CO)5]+ moiety was used as the CO source for gas therapy, interaction, the NIR absorption peak of PB red-shifted to the lower
and the CO capacity was 1.16% (w/w). Under 808 nm light irradia- energy region (Fig. 21c). Owing to the narrowed bandgap and red-
tion, the temperature elevation not only accelerated the DOX shifted LSPR, the photothermal conversion efficiency of Zn-doped
release, but also triggered CO therapy. The combined CO therapy PB reached up to 39.79% (ZnPB-3). Under 808 nm light irradiation,
facilitated the escape of the composite from lysosomes, guaran- the temperature of ZnPB-3 could increase to above 50 1C, which is
teeing the efficacy of the drug, and also made up for the deficiency the required temperature for bacteria-killing. Moreover, the heat
of the relatively low photothermal temperature. generated from the enhanced photothermal effect also facilitated
the diffusion of interstitial Zn2+, which improved the collagen
4.5 Enhancing the efficacy of photothermal therapy deposition for wound healing. To sum up, ZnPB-3 had excellent
4.5.1 Improving photothermal conversion efficiency. Generally, short-term and long-term antibacterial ability against E. coli,
for PTAs, high photothermal conversion efficiency can decrease S. aureus, and MRSA biofilm.
the irradiation time, laser power density and PTA dose, thus Yu et al. coated PB on NaNdF4 particles, forming a core–shell
minimizing tissue damage.305 Therefore, many methods have structure (NdNP@PB).56 During fabrication, citric acid was applied
been devoted to increasing the photothermal conversion efficiency. as a surfactant to provide growth sites for PB and improve the
As mentioned in Section 2, two processes are related to the stability of NP (Fig. 21d). The ladder-like energy levels of Nd3+ have
photothermal effect: light absorption and nonradiative relaxation cross-relaxation (CR) pathways between Nd3+ ions. Since the CR
of electrons. To improve the photothermal conversion efficiency, process and the subsequent photon relaxation to the ground state
researchers can optimize the absorption of the material or create can contribute to heat generation (Fig. 21e), NaNdF4 can be used
more nonradiative pathways. To date, several strategies have been as PTA, but its photothermal conversion efficiency was merely
put forward as follows: (i) increasing the density of DLDs;70 8.7%. One way to improve the photothermal conversion efficiency
(ii) introducing more electron circuit loops through the hetero- was by increasing the Nd3+ concentration to generate more CR
structure;306 (iii) narrowing the bandgap;307 (iv) introducing more pathways between the same lanthanide ions (Fig. 21f). However,
CR pathways;56 (v) optimizing LSPR absorption;249 (vi) transferring increasing the doping content of Nd3+ to even 100% cannot
electron transition from the bandgap to the LSPR,54 etc. Here, we provide enough photothermal effect. Therefore, new CR pathways
listed some examples of enhancing the photothermal conversion needed to be introduced. PB has a continuous energy band. When
efficiency of MOF-based materials. the energy band of PB was in close contact with that of NaNdF4,
Cai et al. doped Gd3+ in PB for the tunable LSPR and new CR pathways with shorter distances were generated (Fig. 21g).
enhanced MRI ability.249 It is known that LSPR is based on Therefore, the nonradiative relaxation distance was much longer,
the collective oscillations of free charge carriers. In PB, the free leading to better PTT effects. Hence, NdNP@PB showed a much
charge carrier mainly refers to [Fe(CN)6]. The position of the enhanced photothermal conversion efficiency of up to 60.8%, and
Gd3+ dopant was firstly at the interstitial site, which had no the tumor growth inhibition of NdNP@PB was 76.7%.
impact on the [Fe(CN)6] vacancy. With the increase of Gd3+ 4.5.2 Low-temperature photothermal therapy. Another
concentration, the dopant position changed to the lattice site, important factor of PTT is therapeutic temperature. As high-
forming the Fe–CRN–Gd structure. The number of [Fe(CN)6] temperature PTT will result in serious damage to ambient healthy
vacancies was therefore decreased. Simultaneously, the electron tissue, researchers put forward the low-temperature PTT
density and orbital energies of –CRN– were also affected by (e.g., 43 1C) strategy. The main concern of low-temperature
lattice Gd3+. As a result, the LSPR absorption peak red-shifted PTT is the cell’s resistance to hyperthermia, which is related to
from 710 to 910 nm. The authors adjusted the absorption peak a higher expression of heat shock proteins (HSPs) in the low-
to around 808 nm, which was beneficial for applying 808 nm temperature region. Zhang and coworkers encapsulated siRNA,
irradiation, giving rise to a much enhanced photothermal efficacy. which is an HSP 70 inhibitor, in the Zr-ferriporphyrin MOF

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5113
 View Article Online

Chem Soc Rev Review Article

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 21 (a) Simplified geometrical structure of PB and ZnPB with various doping levels (color coding: Fe(III) – red, Fe(II) – yellow, Zn – blue, N – grayish
and C – dark gray). (b) Scheme of bandgap-narrowing effect with the increase of Zn-doped density by theory and experiment. (c) UV-vis-NIR spectra of
PB, ZnPB-1, ZnPB-2 and ZnPB-3. Reprinted with permission from ref. 54. Copyright 2019 Nature. (d) Synthesis of NdNP@PB. Simplified diagrams of the
(e) radiative and nonradiative processes in a single Nd3+ ion, (f) CR between Nd3+ ions, (g) generation of new CR pathways between Nd3+ ions and PB.
Reprinted with permission from ref. 56. Copyright 2019 Wiley.

(siRNA/Zr-FeP) with a loading capacity of 76.13%.145 Under maximum temperature of the composite rose to 59.2 1C, which
635 nm light irradiation, the temperature increased to 45.4 1C, was above the melting point of lauric acid (44.7 1C), thus
and the photothermal conversion efficiency was 33.7%. The triggering the release of DOX and losartan. After 21 min of
release of siRNA was triggered by tumor acidity and heat. It irradiation, the release rate of DOX was up to 21.9%, while
was observed that the expression of HSP 70 was inhibited after almost no release was observed under 3 min of irradiation.
treatment with siRNA/Zr-FeP. Animal testing manifested that Moreover, the released losartan could enhance the penetration
siRNA/Zr-FeP had the best therapeutic efficacy due to the inhibited of DOX by degrading extracellular matrices. Therefore, the
thermal resistance. Another strategy is inhibiting the generation of combination of controlled-drug release and PTT gave rise to a
ATP, which is related to the expression of HSP. This strategy can be tumor inhibition rate of up to 81.3%. Wu and coworkers
used for combining PTT with starvation therapy. Zhou et al. loaded covered ZIF-8 with PCM and PDA after loading DOX in the
GOx in HMPB and coated them with HA. Upon irradiation for MOF.295 The melting point of the PCM, tetradecanol, was
5 min, the temperature reached 45 1C by HMPB.25 After endo- 38–40 1C. The photothermal ability was attributed to PDA
cytosis, the intracellular GSH can cleave the disulfide bond coating. The DOX release was controlled by both NIR irradia-
between HMPB and HA, triggering the release of GOx. HMPB tion and pH value, and NIR light-triggered PCM degradation
then decomposed H2O2 to increase the O2 content for the was more critical. Within 5 min of 808 nm light irradiation, the
oxidization of glucose. Results showed that after the treatment, temperature increase in aqueous solution was 34.6 1C (500 mg mL1),
the ATP level of HepG2 and HL-7702 cells decreased and the which was above the melting point of the PCM shell. Additionally,
expressions of both HSP70 and HSP90 were dramatically inhibited. the tumor acidity-induced ZIF-8 decomposition also accelerated
Therefore, the efficacy of low-temperature PTT was improved, and the DOX release. Hence, the drug release amount under both NIR
the combined starvation therapy also contributed to tumor growth irradiation and pH 5.0 was up to 78%.
suppression.
4.5.3 Drug release controlled by phase-change material.
Drug loading is a ubiquitous combined therapy for MOF- 5. Photodynamic and photothermal
based PTT, as MOF is an ideal drug carrier and photothermal synergistic therapy
heat can accelerate the diffusion of the drug. PTT also enhances
the sensitivity of chemotherapy.252 Based on these properties, Although both PDT and PTT have been reported to be efficient
precise control over drug release was achieved by introducing strategies in clinical applications, there are still some drawbacks
phase-change materials (PCM) such as 1-pentadecanol,308,309 in each therapy. First of all, due to the hypoxia in the therapeutic
tetradecanol295 and lauric acid.310 Only when the temperature area, it is hard to satisfy the O2 demand of PDT, which hinders
rises above the melting point of PCM will the loaded drug be the ROS generation.67 As for PTT, improving photothermal
released. Therefore, in blood circulation, this strategy can conversion efficiency requires complex modification.141 Another
realize ‘‘zero drug leakage’’. Zhang et al. incorporated DOX concern is that to achieve sufficient ROS yield or temperature
and losartan with lauric acid, and encapsulated them in increase, the laser power density is usually high, which might
HMPB.310 After 10 min of 808 nm NIR light irradiation, the cause lesions in healthy tissues.187 To solve these problems,

5114 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

combined therapies have been put forward, such as chemo- the 808 nm laser. The singlet oxygen quantum yield of PMCS was
therapy,263,311,312 gas therapy,46,112,255,256 etc., which adds to the comparable to that of ICG, and the photothermal conversion
complexity of the design. As PDT and PTT can be simultaneously efficiency was 33.0%. Due to the strong absorption, PMCS can
triggered by light irradiation, combining PDT and PTT in one serve as the contrast agent for PA imaging. The combination of
therapeutic system is a promising way to overcome the dis- imaging guidance, PTT, and PDT showed complete eradication
advantages of PDT and PTT. Moreover, the heat generated by against tumors.
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

PTT can improve the blood flow, and consequently increase the Some dyes have been reported to have both ROS generation
ambient O2 content to alleviate hypoxia.145 On the other hand, and light-to-heat conversion ability. Gao et al. applied UiO-66 as
targeted cells also become more sensitive to heat due to the the O2 carrier and coordinated ICG on the surface of UiO-66.184
interference of PDT.145 Therefore, PDT and PTT can work Afterwards, the nanoparticles were coated with red blood cell
synergistically. In this section, we will discuss how to construct membrane. ICG served as both PS and PTA. Under 808 nm light
PDT and PTT synergic therapeutic systems based on MOFs. irradiation, the temperature increased to 43.5 1C. The generated
The ‘‘all-in-one’’ strategy means incorporating both photo- heat accelerated the diffusion of O2 from the inner UiO-66 core,
thermal and photodynamic ability in one MOF without further which in turn facilitated the PDT efficacy of ICG. Cypate is
modification. Han et al. introduced Cu2+ into the porphyrin another organic dye that features photodynamic, photothermal,
ligands of PCN-224 with various doping levels (5%, 10%, 15% and versatile imaging abilities. Yang and coworkers introduced
and 25%).146 The Cu2+ dopant could trap electrons, thus cypate during the fabrication process of MIL-53, denoting as
hindering the recombination of electron–hole pairs. After the CMNP.185 CMNP was then coated with PEG and transferrin
electron trapping, Cu2+ was reduced to Cu, which could react (referred to as CMNP-Tf). During fabrication, a cypate-Fe3+
with holes and decrease the photocatalytic ability. Therefore, precursor was firstly formed owing to the coordination bond
the dopant concentration should be moderate. Their results between the carboxyl of cypate and Fe3+. Then after the addition
showed that PCN-224 with 10% Cu2+ dopant (denoted as of ligand solution, MIL-53, with defects, was created. The defects
Cu10MOF) had the highest 1O2 generation and relatively higher can be used to control the pore size of MIL-53, which increased
absorption at 660 nm. Moreover, the d–d transition of the with increasing the concentration of cypate. This method
chelated Cu2+ provided the photothermal effect. Under increased the druggability and bioavailability of cypate, and also
660 nm light irradiation, the temperature of Cu10MOF was avoided its photobleaching. Due to the successful loading and
the highest (48.4 1C). Therefore, the synergistic therapy of versatility of cypate, CMNP-Tf showed complete tumor ablation
Cu10MOF exhibited 99.71% and 97.14% of bacteria-killing with good biocompatibility.
against S. aureus and E. coli, respectively. Li et al. reported a Lastly, to synthesize MOF-based composites with PDT and PTT
Cu–TCPP MOF nanosheet.144 The thickness of the nanosheet dual functions, researchers usually combine the above-mentioned
was 5.1  0.3 nm, which rendered a quicker response to light as PSs (dye, porphyrin-based MOF, etc.) and PTAs (Au-based PTA and
compared to the bulk material. The PDT effect of the TCPP photothermal polymer, etc.), which all achieved superior thera-
ligand was triggered by 660 nm laser. Owing to the d–d peutic efficacy.142,182,186,267 Here, we mainly illustrate examples
transition of Cu2+ nodes, Cu–TCPP also efficiently transforms of multi-MOF core–shell structure. Liu et al. fabricated a dual
light into heat. Under 5 min of 808 nm light irradiation, the MIL-101 core–shell structure by the in situ growth method, and
temperature quickly increased 34.5 1C. Moreover, due to the decorated the outer shell with PEG-FA and cyanine 3-labelled
unpaired 3d electrons of Cu2+ nodes, the nanosheets could be peptide (referred to as BQ-MIL@cat-fMIL) (Fig. 22a).186 The inner
used for MRI and infrared thermal imaging. MIL-101 was loaded with black phosphorus dots (BQ), while the
As mentioned in Section 4.3, the high-temperature pyrolysis outer MIL-101 was loaded with catalase. Because of quantum
of MOFs can generate MOF-derived carbon materials, which confinement and edge effects, BQ is promising for combining
features photothermal conversion ability. If the ligand is PDT and PTT. The photosensitivity of MIL-101 resulted in a
N-containing, the MOF-derived carbon material can show a widened gap between the triplet state and the ground state of
porphyrin-like M–N–C structure (M denotes metal center) the composite, indicating that the excited BQ-MIL@cat-fMIL
under appropriate pyrolysis conditions, wherein the metal mainly reacted with 3O2 rather than went through phosphorescence
cations are dispersed and each of them is coordinated with 4 emission. Therefore, the 1O2 generation of BQ-MIL@cat-fMIL
pyridinic nitrogen, which is similar to the coordination was increased to 88.3%. Moreover, the catalase in the outer shell
environment in porphyrin.313–315 Therefore, the porphyrin- of MIL-101 efficiently provided O2 for the inner BQ by decom-
like structure further endows this material with PDT ability. posing H2O2. Under 660 nm light irradiation, the apoptotic
Wang and coworkers coated ZIF-8 with mesoporous silica, percentage was 52.1% by PDT, which was 8.7 times higher than
followed by high-temperature pyrolysis.313 Afterwards, mSiO2 that without catalase. On the other hand, the photothermal
was removed by NaOH etching. This surface protection prevented effect was triggered by 808 nm light, and the photothermal
the aggregation of the M–N–C structure during pyrolysis. The conversion efficiency was 23.3%, which provided 28.7% of cell
diameter of the as-obtained material (referred to as PMCS) was apoptosis. After applying dual-light irradiation, the combination
about 140 nm, and it retained the porosity of the ZIF-8 precursor. of PDT and PTT led to 75.6% of cell apoptosis.
Similar to other carbon materials, PMCS had strong absorption in Luo et al. coated PB with the UiO-66 MOF shell (Fig. 22b),
the NIR region. PDT and PTT were simultaneously triggered by which was used for bacteria-infected wound healing under

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5115
 View Article Online

Chem Soc Rev Review Article

Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Fig. 22 (a) The stepwise assembly of BQ and catalase in MOF layers and its application as a tandem catalyst for enhanced therapy against hypoxic tumor
cells. Reprinted with permission from ref. 186. Copyright 2019 Wiley. (b) Schematic illustration of the core–shell structure of PB@MOF. (c) Schematic
illustration of the bacteria killing processes with the PB@MOF under dual light irradiation. (d) Schematic illustration of the rational photocatalytic
mechanism for PB@MOF heterojunction photocatalysts. Reprinted with permission from ref. 267. Copyright 2019 American Chemical Society.

dual-light irradiation (660 nm light for PDT and 808 nm light MOF in phototherapy. However, to push this field forward, we
for PTT (Fig. 22c)).267 Taking advantage of the defects in the have listed some issues that need to be addressed.
MOF structure, the outer UiO-66 was doped with TCPP, which In spite of the advantages mentioned in this review, photo-
served as PS. The core–shell structure not only combined two therapy still has many limitations such as O2-dependence,
therapies, but also enhanced the photocatalytic ability by inhomogeneous distribution of heat, and limited tissue pene-
forming a heterojunction structure. Both PB and TCPP-doped tration, etc. Future research on phototherapy may be combined
UiO-66 (referred to as MOF) were n-type semiconductors. Their with various therapies, including but not limited to chemotherapy,
results showed that PB had a lower CB than the MOF (Fig. 22d). radiotherapy, gas therapy, starvation therapy and immunotherapy.
Hence, the photo-induced electrons of PB could move to MOF, Another concern is that the limited penetration is a big problem
accelerating the photo-electron transfer and inhibiting photo- for phototherapy. Although the use of upconversion nanoparticles
electron–hole recombination. Though the photothermal effect and two-photon activated PSs can improve the therapeutic depth,
of PB was partially hindered by the UiO-66 shell, the tempera- it is still difficult for phototherapy to cure deep-sited tumors or
ture still exceeded 50 1C within 5 min of irradiation. Moreover, infections. Imaging guidance is another important auxiliary
the composite released trace amounts of Fe and Zr element method in phototherapy, which can determine the location and
during degradation, which facilitated the wound healing. morphology of lesions and monitor the distribution of therapeutic
Therefore, this core–shell dual-MOF composite is superior for agents in order to provide the appropriate irradiation. More
bacterial infection treatment. Under dual-light irradiation, the importantly, with the guidance of fluorescence, it is desirable to
PB@MOF composite showed more than 99% of antibacterial implant a light source in solid organs, which can be used for
efficiency against S. aureus and E. coli. treating deep lesions.6 In view of these concerns, the MOF is an
ideal carrier for these functions. To date, the combined therapy
and imaging-assisted therapy based on MOFs have achieved good
results. However, improvements are still needed for precise
6. Conclusions and perspectives control, potent efficacy, resistance prevention and lowering
tissue lesions before being applied in clinical settings.
In the last decade, the applications of MOFs in phototherapy Many researchers have mainly focused on improving the
have presented a booming trend, as illustrated in this review. ROS yield and photothermal conversion efficiency of PSs and
Due to the tunability of the MOF structure, MOFs could directly PTAs. However, other physiochemical properties of materials
serve as PSs or PTAs by applying photo-responsive building need to be taken into consideration as well, such as crystallinity,
units, or act as the carrier of phototherapeutic agents. The periodic aqueous stability and degradability, etc. The specific mechanisms
array of MOFs prevent the aggregation and self-quenching of of photodynamic and photothermal effects should be further
PSs and PTAs, which greatly enhances their efficacy. Moreover, clarified, such as the electron transition pathway, and the
the active sites and cavities of MOFs make numerous modifica- correlation between radiative and nonradiative decays, which
tions feasible, such as combined therapy, active targeting and will shed light on the future design of MOFs with desirable ROS
imaging guidance, etc. To date, many complex and elaborate yields, photothermal conversion efficiency and other photo-
designs have been put forward, indicating the great potential of related functions. On the other hand, the requirement of

5116 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

combined therapy gives rise to the integration of various reported phototherapy in cancer treatment, researchers can
therapeutic agents, and each of them has separate functions. consider using these materials in other medical applications.
More studies towards the synergistic effect of these agents are In conclusion, the applications of MOFs in phototherapy
needed, from synthesis to treatment mechanisms. For example, have been put forward and rapidly developed in the last decade.
the influence on MOF morphology and crystallinity, the interface Under rational design, MOFs can improve the efficacy of
interactions of core–shell structure, the synergistic effects of the traditional phototherapy with multiple combined functions,
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

drug and ROS, etc. After the in-depth understanding of the which is now attracting more and more attention. However,
underlying mechanisms, researchers can build up a comprehen- although we have seen the elaborate design of MOFs and good
sive correlation between composition, structure, properties, and therapeutic efficacy in recent publications, the research on
efficacy, which is beneficial for fabricating highly effective rather MOFs in phototherapy is in its infancy. Numerous efforts are
than purely structurally sophisticated medical materials. still needed before it reaches clinical trials. We believe that the
In terms of clinical applications, the most important concern applications of MOFs in phototherapy will continue to expand,
is the biosafety of materials. Unfortunately, up to now, none of and finally become an important part of medical treatment.
these therapeutic agents have been approved by the FDA,
indicating that more attempts towards improving biosafety are
needed. Although there have been reports about different surface Abbreviations
coatings to improve the stability and biocompatibility of materials,
researchers still need to consider every process of materials in the ROS Reactive oxygen species
body, such as drug administration, circulation in the blood flow, PDT Photodynamic therapy
degradation, and elimination from the body. The related proper- PTT Photothermal therapy
ties including biodegradability, particle size and the release of PS Photosensitizer
various agents during degradation should be carefully designed. 
O 2 Superoxide anion radical

Many researchers have reported the features of the therapeutic OH Hydroxyl radical
area, such as hypoxia and the generation of acidity, GSH, ATP and H2O2 Hydrogen peroxide
1
H2S, etc., and they have put forward various responsive methods. O2 Singlet oxygen
However, the understanding of the therapeutic area is not enough, HPD Hematoporphyrin derivative
as the real microenvironment is more complex. The immune FDA Food and Drug Administration
response and physiological properties of different tumors, bacteria, g-C3N4 Graphic carbon nitride
biofilms and other diseases still need further investigation. PTA Photothermal agent
Although most reported MOF-based PSs or PTAs are aiming MOF Metal–organic framework
at cancer treatment, ROS generation and temperature increase NP Nanoparticle
3
induced by light irradiation are promising in many other fields O2 Molecular oxygen
as well. For example, many researchers have applied PDT and LSPR Localized plasmon surface resonance
PTT in antibacterial applications, which can be divided into NIR Near infrared
water disinfection,316,317 medical device sterilization,318 wound CB Conduction band
healing,319,320 and implant modification,321–323 with light source VB Valence band
ranging from the visible to NIR range. Moreover, the combi- DLD Deep-level defect
nation of PTAs and PCMs has been applied to fabricate the CR Cross relaxation
anticorrosion coating of magnesium alloy implants, which has UV Ultraviolet
light-induced self-healing ability.324 Therefore, the applications SEM Scanning electron microscope
of PDT and PTT can be extended in various fields, which have TEM Transmission electron microscope
different requirements for light penetration, ROS yield, and DLS Dynamic light scattering
temperature increase. Besides the traditional material fabrica- H2DBP 5,15-Di(p-benzoato)porphyrin
tion methods such as encapsulation and surface coating, some H2DBC 5,15-Di(p-methylbenzoato)chlorin
novel techniques such as biomineralization,325–327 which com- H4TBC 5,10,15,20-Tetra(p-benzoato)
bines materials with bacteria and viruses, can be used in PCN Porous coordination network
phototherapy as well. As the penetration depth of light is limited, TPDC Terphenyl-4,400 -dicarboxylic acid
researchers can focus on superficial diseases such as dental TCPP Tetrakis(4-carboxyphenyl)-porphyrin
antibacterial applications, which can better utilize the light TBP Tetrabenzoporphyrin or
source. MOFs have been successfully fabricated in the form of 5,10,15,20-tetra(p-benzoato)porphyrin
nanoparticles, membranes,328 mixed-matrix membranes117 and H2DBBC 5,15-Di(p-benzoato)bacteriochlorin
coatings,329,330 which make it possible for MOF-based PSs and MRI Magnetic resonance imaging
PTAs to adapt to different therapeutic conditions. Due to the GSH Glutathione
unique structure of the MOF, it can be used as a precursor or H4TBAPy 1,3,6,8-Tetrakis(p-benzoic acid)pyrene
template for fabricating new porous materials such as MOF- DHA Dihydroartemisinin
based carbon single-atom catalysts. Hence, on the basis of the H2BDC 2-Hydroxyterephthalic acid

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5117
 View Article Online

Chem Soc Rev Review Article

BODIPY Boron-dipyrromethene PNIPAM Poly(N-isopropylacrylamide)

ICP-MS Inductively coupled plasma-mass spectrometer PBS Phosphate buffer solution
H4TBAPy 1,3,6,8-Tetrakis(p-benzoic acid)pyrene PPy Polypyrrole
HOMO Highest occupied molecular orbital PAN Polyaniline
LUMO Lowest unoccupied molecular orbital cRGD-SH Cyclic arginine–glycine–aspartic acid
BCDTE 1,2-Bis(5-(4-carbonxyphenyl)-2-methylthien- HSP Heat shock protein
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

3-yl)cyclopent-1-ene PCM Phase-change material

ZIF Zeolitic imidazolate framework HA Hyaluronic acid
UiO Universitetet i Oslo GPTS (3-Glycidyloxypropyl)trimethoxysilane
MIL Materials Institute Lavoisier FA Folic acid
TMPyP Tetrakis(1-methylpyridinium-4-yl)-porphyrin AlPcS4 Al(III) phthalocyanine chloride tetrasulfonic acid
ZnPc Zinc phthalocyanines 2-mIm 2-Methylimidazole
Ce6 Chlorine e6 PEI Poly(ethylenimine)
H3BTC 1,3,5-Benzenetricarboxylic acid BATA Bis-(alkylthio) alkene
CaB Cathepsin IcaH Imidazole-2-carboxaldehyde
DOX Doxorubicin TPEDC 2-((4 0 -(2,2-Bis(4-methoxyphenyl)-1-phenylvinyl)-
CD Carbon dot [1,1 0 -biphenyl]-4-yl)(phenyl)methylene)
ATP Adenosine triphosphate malononitrile
GSSG Oxidized glutathione TPETCF (E)-2-(4-(4-(2,2-Bis(4-methoxyphenyl)-1-
PDA Polydopamine phenylvinyl)styryl)-3-cyano-5,5-dimethylfuran-
PEG Polyethylene glycol 2(5H)-ylidene)malononitrile
TPZ Tirapazamine AQ4N Banoxantrone
PL Piperlongumine TPAAQ 2-(4-(Diphenylamino)phenyl)anthracene-9,
Tex Thioredoxin 10-dione
TexR Thioredoxin reductase PDMAEMA Poly(2-(diethylamino)ethylmethacrylate)
GOx Glucose oxidase TPyP 5,10,15,20-Tetrakis(4-pyridyl)-21H,23H-porphine
ONOO Peroxynitrite PLA Polylactic acid
L-Arg L-Arginine PVP Polyvinyl pyrrolidone
CpG Cytosine-phosphate-guanine TAPP 5,10,15,20-Tetrakis(4-aminophenyl)porphyrin
IFN-a Immunostimulatory cytokines type I Interferon H2BPDC 2,2 0 -Bipyridine-5.5 0 -dicarboxylic acid
IL-6 Interleukin-6 AlPc Aluminum phthalocyanine
CTLA4 T-lymphocyte-associated protein 4 PCL Polycaprolactone
PD-1 Programmed cell death 1 PEGFA NH2-poly(ethylene glycol) modified folic acid
PD-L1 Programmed cell death 1 ligand CTAB Cetyltrimethylammonium bromide
IDO Indoleamine 2,3-dioxygenase E. coli Escherichia coli
RC Polypyridyl ruthenium complex S. aureus Staphylococcus aureus
MRSA Methicillin-resistant Staphylococcus aureus
PB Prussian blue
PAH Poly(allylamine hydrochloride) Conflicts of interest
PAA Polyacrylic acid There are no conflicts to declare.
HMPB Hollow mesoporous Prussian blue
BET Brunauer–Emmett–Teller
PFP Perfluoropentane Acknowledgements
PA Photoacoustic
US Ultrasound This work is jointly supported by the National Science Fund for
SNP Sodium nitroprusside Distinguished Young Scholars (no. 51925104), National Natural
DTX Docetaxel Science Foundation of China (no. 51871162, 51671081), Natural
ART Artemisinin Science Fund of Hubei Province, (no. 2018CFA064), RGC/NSFC
CT Computed tomography (N_HKU725-1616), Hong Kong ITC (ITS/287/17, GHX/002/14SZ),
RCM Red cell membrane as well as Health and Medical Research Fund (no. 03142446).
PDI Perylenediimide
Fc(COOH)2 1,1 0 -Ferrocenedicarboxylic acid References
TCPC Tetratopic porphyrin
LMCT Ligand-to-metal charge transfer 1 D. E. J. G. J. Dolmans, D. Fukumura and R. K. Jain, Nat.
ICG Indocyanine green Rev. Cancer, 2003, 3, 380–387.
Ara Cytarabine 2 G. Lan, K. Ni and W. Lin, Coord. Chem. Rev., 2019, 379, 65–81.

5118 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

3 K. K. Ng and G. Zheng, Chem. Rev., 2015, 115, 11012–11042. 26 Y. Yang, W. Zhu, Z. Dong, Y. Chao, L. Xu, M. Chen and
4 L. Cheng, C. Wang, L. Feng, K. Yang and Z. Liu, Chem. Rev., Z. Liu, Adv. Mater., 2017, 29, 1703588.
2014, 114, 10869–10939. 27 L. Tan, Z. Zhou, X. Liu, J. Li, Y. Zheng, Z. Cui, X. Yang,
5 M. Ethirajan, Y. Chen, P. Joshi and R. K. Pandey, Chem. Y. Liang, Z. Li, X. Feng, S. Zhu, K. W. K. Yeung, C. Yang,
Soc. Rev., 2011, 40, 340–362. X. Wang and S. Wu, Adv. Sci., 2020, 7, 1902070.
6 P. Agostinis, K. Berg, K. A. Cengel, T. H. Foster, 28 D. Yoo, H. Jeong, S.-H. Noh, J.-H. Lee and J. Cheon, Angew.
Chem., Int. Ed., 2013, 52, 13047–13051.
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

A. W. Girotti, S. O. Gollnick, S. M. Hahn, M. R. Hamblin,

A. Juzeniene, D. Kessel, M. Korbelik, J. Moan, P. Mroz, 29 K. F. Chu and D. E. Dupuy, Nat. Rev. Cancer, 2014, 14, 199–208.
D. Nowis, J. Piette, B. C. Wilson and J. Golab, Ca-Cancer 30 H. Y. Mao, S. Laurent, W. Chen, O. Akhavan, M. Imani,
J. Clin., 2011, 61, 250–281. A. A. Ashkarran and M. Mahmoudi, Chem. Rev., 2013, 113,
7 A. P. Castano, T. N. Demidova and M. R. Hamblin, Photo- 3407–3424.
diagn. Photodyn. Ther., 2004, 1, 279–293. 31 L. Feng, L. Wu and X. Qu, Adv. Mater., 2013, 25, 168–186.
8 J. S. Dysart and M. S. Patterson, Phys. Med. Biol., 2005, 50, 32 Y. Zou, Y. Zhang, Q. Yu and H. Chen, Biomater. Sci., 2021, 9,
2597–2616. 10–22.
9 J. P. Celli, B. Q. Spring, I. Rizvi, C. L. Evans, K. S. Samkoe, 33 B. Li, L. Tan, X. Liu, Z. Li, Z. Cui, Y. Liang, S. Zhu, X. Yang,
S. Verma, B. W. Pogue and T. Hasan, Chem. Rev., 2010, 110, K. W. Kwok Yeung and S. Wu, J. Hazard. Mater., 2019,
2795–2838. 380, 120818.
10 M. C. DeRosa and R. J. Crutchley, Coord. Chem. Rev., 2002, 34 Y. Wang, T. Wei, Y. Qu, Y. Zhou, Y. Zheng, C. Huang,
233–234, 351–371. Y. Zhang, Q. Yu and H. Chen, ACS Appl. Mater. Interfaces,
11 X. Li, S. Kolemen, J. Yoon and E. U. Akkaya, Adv. Funct. 2020, 12, 21283–21291.
Mater., 2017, 27, 1604053. 35 Y. Qu, T. Wei, J. Zhao, S. Jiang, P. Yang, Q. Yu and H. Chen,
12 J. Ge, M. Lan, B. Zhou, W. Liu, L. Guo, H. Wang, Q. Jia, J. Mater. Chem. B, 2018, 6, 3946–3955.
G. Niu, X. Huang, H. Zhou, X. Meng, P. Wang, C.-S. Lee, 36 H. S. Jung, P. Verwilst, A. Sharma, J. Shin, J. L. Sessler and
W. Zhang and X. Han, Nat. Commun., 2014, 5, 4596. J. S. Kim, Chem. Soc. Rev., 2018, 47, 2280–2297.
13 L. Tan, J. Li, X. Liu, Z. Cui, X. Yang, K. W. K. Yeung, H. Pan, 37 T. Wei, Q. Yu and H. Chen, Adv. Healthcare Mater., 2019,
Y. Zheng, X. Wang and S. Wu, Small, 2018, 14, 1703197. 8, 1801381.
14 Q. Zhang, X. Liu, L. Tan, Z. Cui, Z. Li, Y. Liang, S. Zhu, 38 C. Mao, Y. Xiang, X. Liu, Y. Zheng, K. W. K. Yeung, Z. Cui,
K. W. K. Yeung, Y. Zheng and S. Wu, Chem. Eng. J., 2020, X. Yang, Z. Li, Y. Liang, S. Zhu and S. Wu, ACS Appl. Mater.
383, 123088. Interfaces, 2019, 11, 17902–17914.
15 P. Yu, Y. Han, D. Han, X. Liu, Y. Liang, Z. Li, S. Zhu and 39 X. Chen, M. Zhang, S. Li, L. Li, L. Zhang, T. Wang,
S. Wu, J. Hazard. Mater., 2020, 390, 122126. M. Yu, Z. Mou and C. Wang, J. Mater. Chem. B, 2017, 5,
16 Y. Xiang, C. Mao, X. Liu, Z. Cui, D. Jing, X. Yang, Y. Liang, 1772–1778.
Z. Li, S. Zhu, Y. Zheng, K. W. K. Yeung, D. Zheng, X. Wang 40 X. Wang, K. Su, L. Tan, X. Liu, Z. Cui, D. Jing, X. Yang,
Y. Liang, Z. Li, S. Zhu, K. W. K. Yeung, D. Zheng and S. Wu,
and S. Wu, Small, 2019, 15, 1900322.
ACS Appl. Mater. Interfaces, 2019, 11, 15014–15027.
17 Y. Li, X. Liu, L. Tan, Z. Cui, D. Jing, X. Yang, Y. Liang, Z. Li,
41 L. Jiao, J. Y. R. Seow, W. S. Skinner, Z. U. Wang and
S. Zhu, Y. Zheng, K. W. K. Yeung, D. Zheng, X. Wang and
H.-L. Jiang, Mater. Today, 2019, 27, 43–68.
S. Wu, Adv. Funct. Mater., 2019, 29, 1900946.
42 S. Kitagawa, R. Kitaura and S.-i. Noro, Angew. Chem., Int.
18 R. Long, Y. Li, L. Song and Y. Xiong, Small, 2015, 11, 3873–3889.
Ed., 2004, 43, 2334–2375.
19 D. Jaque, L. Martı́nez Maestro, B. del Rosal, P. Haro-
43 A. Kirchon, L. Feng, H. F. Drake, E. A. Joseph and
Gonzalez, A. Benayas, J. L. Plaza, E. Martı́n Rodrı́guez
H.-C. Zhou, Chem. Soc. Rev., 2018, 47, 8611–8638.
and J. Garcı́a Solé, Nanoscale, 2014, 6, 9494–9530.
44 T. Simon-Yarza, A. Mielcarek, P. Couvreur and C. Serre,
20 C. S. S. R. Kumar and F. Mohammad, Adv. Drug Delivery
Adv. Mater., 2018, 30, 1707365.
Rev., 2011, 63, 789–808.
45 M.-X. Wu and Y.-W. Yang, Adv. Mater., 2017, 29, 1606134.
21 J. Beik, Z. Abed, F. S. Ghoreishi, S. Hosseini-Nami, S. Mehrzadi,
46 W.-P. Li, C.-H. Su, L.-C. Tsao, C.-T. Chang, Y.-P. Hsu and
A. Shakeri-Zadeh and S. K. Kamrava, J. Controlled Release, 2016,
C.-S. Yeh, ACS Nano, 2016, 10, 11027–11036.
235, 205–221.
47 K. Lu, C. He and W. Lin, J. Am. Chem. Soc., 2015, 137,
22 A. Hervault and N. T. K. Thanh, Nanoscale, 2014, 6, 11553–11573.
7600–7603.
23 N. R. Datta, S. G. Ordóñez, U. S. Gaipl, M. M. Paulides,
48 Y. Zhang, F. Wang, C. Liu, Z. Wang, L. Kang, Y. Huang,
H. Crezee, J. Gellermann, D. Marder, E. Puric and S. Bodis,
K. Dong, J. Ren and X. Qu, ACS Nano, 2018, 12, 651–661.
Cancer Treat. Rev., 2015, 41, 742–753.
49 B. Lü, Y. Chen, P. Li, B. Wang, K. Müllen and M. Yin, Nat.
24 X. Huang, P. K. Jain, I. H. El-Sayed and M. A. El-Sayed,
Commun., 2019, 10, 767.
Lasers Med. Sci., 2007, 23, 217.
50 F. Hu, D. Mao, Kenry, Y. Wang, W. Wu, D. Zhao, D. Kong
25 J. Zhou, M. Li, Y. Hou, Z. Luo, Q. Chen, H. Cao, R. Huo,
and B. Liu, Adv. Funct. Mater., 2018, 28, 1707519.
C. Xue, L. Sutrisno, L. Hao, Y. Cao, H. Ran, L. Lu, K. Li and
51 H. Wang, Y. Chen, H. Wang, X. Liu, X. Zhou and F. Wang,
K. Cai, ACS Nano, 2018, 12, 2858–2872. Angew. Chem., Int. Ed., 2019, 58, 7380–7384.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5119
 View Article Online

Chem Soc Rev Review Article

52 X. Deng, S. Liang, X. Cai, S. Huang, Z. Cheng, Y. Shi, M. Pang, 77 S. Kwiatkowski, B. Knap, D. Przystupski, J. Saczko,
P. A. Ma and J. Lin, Nano Lett., 2019, 19, 6772–6780. E. Ke ˛dzierska, K. Knap-Czop, J. Kotlińska, O. Michel,
53 D. Wang, H. Wu, J. Zhou, P. Xu, C. Wang, R. Shi, H. Wang, K. Kotowski and J. Kulbacka, Biomed. Pharmacother.,
H. Wang, Z. Guo and Q. Chen, Adv. Sci., 2018, 5, 1800287. 2018, 106, 1098–1107.
54 J. Li, X. Liu, L. Tan, Z. Cui, X. Yang, Y. Liang, Z. Li, S. Zhu, 78 D. Van Straten, V. Mashayekhi, H. S. De Bruijn, S. Oliveira
Y. Zheng, K. W. K. Yeung, X. Wang and S. Wu, Nat. and D. J. Robinson, Cancers, 2017, 9, 19.
Commun., 2019, 10, 4490. 79 C. A. Robertson, D. H. Evans and H. Abrahamse, J. Photochem.
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

55 K. Lu, C. He and W. Lin, J. Am. Chem. Soc., 2014, 136, Photobiol., B, 2009, 96, 1–8.
16712–16715. 80 Y. Liu, P. Bhattarai, Z. Dai and X. Chen, Chem. Soc. Rev.,
56 Z. Yu, W. Hu, H. Zhao, X. Miao, Y. Guan, W. Cai, Z. Zeng, 2019, 48, 2053–2108.
Q. Fan and T. T. Y. Tan, Angew. Chem., Int. Ed., 2019, 58, 81 G. Hong, S. Diao, A. L. Antaris and H. Dai, Chem. Rev.,
8536–8540. 2015, 115, 10816–10906.
57 S.-Y. Yin, G. Song, Y. Yang, Y. Zhao, P. Wang, L.-M. Zhu, 82 P. K. Jain, X. Huang, I. H. El-Sayed and M. A. El-Sayed, Acc.
X. Yin and X.-B. Zhang, Adv. Funct. Mater., 2019, Chem. Res., 2008, 41, 1578–1586.
29, 1901417. 83 J. M. Luther, P. K. Jain, T. Ewers and A. P. Alivisatos, Nat.
58 J. Park, Q. Jiang, D. Feng, L. Mao and H.-C. Zhou, J. Am. Mater., 2011, 10, 361–366.
Chem. Soc., 2016, 138, 3518–3525. 84 D. Choi, B. Yoon, D.-K. Kim, H. Baik, J.-H. Choi and
59 H. Guo, S. Yi, K. Feng, Y. Xia, X. Qu, F. Wan, L. Chen and K. S. Jeong, Chem. Mater., 2017, 29, 8548–8554.
C. Zhang, Chem. Eng. J., 2021, 403, 126432. 85 J. Li, X. Liu, L. Tan, Y. Liang, Z. Cui, X. Yang, S. Zhu, Z. Li,
60 H. Min, J. Wang, Y. Qi, Y. Zhang, X. Han, Y. Xu, J. Xu, Y. Li, Y. Zheng, K. W. K. Yeung, X. Wang and S. Wu, Small
L. Chen, K. Cheng, G. Liu, N. Yang, Y. Li and G. Nie, Adv. Methods, 2019, 3, 1900048.
Mater., 2019, 31, 1808200. 86 X. Wang, H. Li, F. Li, X. Han and G. Chen, Nanoscale, 2019,
61 Q. Cheng, W. Yu, J. Ye, M. Liu, W. Liu, C. Zhang, C. Zhang, 11, 22079–22088.
J. Feng and X.-Z. Zhang, Biomaterials, 2019, 224, 119500. 87 C. Liu, B. Liu, J. Zhao, Z. Di, D. Chen, Z. Gu, L. Li and
62 S.-Y. Li, H. Cheng, B.-R. Xie, W.-X. Qiu, J.-Y. Zeng, C.-X. Li, Y. Zhao, Angew. Chem., Int. Ed., 2020, 59, 2634–2638.
S.-S. Wan, L. Zhang, W.-L. Liu and X.-Z. Zhang, ACS Nano, 88 Y. Shao, B. Liu, Z. Di, G. Zhang, L.-D. Sun, L. Li and C.-
2017, 11, 7006–7018. H. Yan, J. Am. Chem. Soc., 2020, 142, 3939–3946.
63 C. Liu, J. Xing, O. U. Akakuru, L. Luo, S. Sun, R. Zou, Z. Yu, 89 L. He, Q. Ni, J. Mu, W. Fan, L. Liu, Z. Wang, L. Li, W. Tang,
Q. Fang and A. Wu, Nano Lett., 2019, 19, 5674–5682. Y. Liu, Y. Cheng, L. Tang, Z. Yang, Y. Liu, J. Zou, W. Yang,
64 Q. Guan, L.-L. Zhou, Y.-A. Li and Y.-B. Dong, Chem. O. Jacobson, F. Zhang, P. Huang and X. Chen, J. Am. Chem.
Commun., 2019, 55, 14898–14901. Soc., 2020, 142, 6822–6832.
65 J. Mu, L. He, W. Fan, W. Tang, Z. Wang, C. Jiang, D. Zhang, 90 V. J. Pansare, S. Hejazi, W. J. Faenza and R. K. Prud’homme,
Y. Liu, H. Deng, J. Zou, O. Jacobson, J. Qu, P. Huang and Chem. Mater., 2012, 24, 812–827.
X. Chen, Small, 2020, 16, 2004016. 91 C. Gao, J. Wang, H. Xu and Y. Xiong, Chem. Soc. Rev., 2017,
66 H. Zhang, Q. Li, R. Liu, X. Zhang, Z. Li and Y. Luan, Adv. 46, 2799–2823.
Funct. Mater., 2018, 28, 1802830. 92 S. Li, S. Zhou, Y. Li, X. Li, J. Zhu, L. Fan and S. Yang, ACS
67 M. Lismont, L. Dreesen and S. Wuttke, Adv. Funct. Mater., Appl. Mater. Interfaces, 2017, 9, 22332–22341.
2017, 27, 1606314. 93 Y. Yang, J. Liu, C. Liang, L. Feng, T. Fu, Z. Dong, Y. Chao, Y. Li,
68 A. Gellé, T. Jin, L. de la Garza, G. D. Price, L. V. Besteiro and G. Lu, M. Chen and Z. Liu, ACS Nano, 2016, 10, 2774–2781.
A. Moores, Chem. Rev., 2020, 120, 986–1041. 94 M. Zheng, P. Zhao, Z. Luo, P. Gong, C. Zheng, P. Zhang,
69 J. Wang, Y. Li, L. Deng, N. Wei, Y. Weng, S. Dong, D. Qi, C. Yue, D. Gao, Y. Ma and L. Cai, ACS Appl. Mater.
J. Qiu, X. Chen and T. Wu, Adv. Mater., 2017, 29, 1603730. Interfaces, 2014, 6, 6709–6716.
70 Y. Cheng, Y. Chang, Y. Feng, H. Jian, Z. Tang and 95 E. Caballero-Mancebo, B. Cohen, S. Smolders, D. E. De Vos
H. Zhang, Angew. Chem., Int. Ed., 2018, 130, 252–257. and A. Douhal, Adv. Sci., 2019, 6, 1901020.
71 C. M. Marian, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 96 L. He, M. Brasino, C. Mao, S. Cho, W. Park, A. P. Goodwin
2012, 2, 187–203. and J. N. Cha, Small, 2017, 13, 1700504.
72 J. Yao, M. Yang and Y. Duan, Chem. Rev., 2014, 114, 97 K. Lu, C. He, N. Guo, C. Chan, K. Ni, R. R. Weichselbaum
6130–6178. and W. Lin, J. Am. Chem. Soc., 2016, 138, 12502–12510.
73 J. F. Lovell, T. W. B. Liu, J. Chen and G. Zheng, Chem. Rev., 98 X. Huang, I. H. El-Sayed, W. Qian and M. A. El-Sayed, J. Am.
2010, 110, 2839–2857. Chem. Soc., 2006, 128, 2115–2120.
74 S. Mai, P. Marquetand and L. González, Int. J. Quantum 99 Q. Yang, Q. Xu, S.-H. Yu and H.-L. Jiang, Angew. Chem., Int.
Chem., 2015, 115, 1215–1231. Ed., 2016, 55, 3685–3689.
75 Á. Juarranz, P. Jaén, F. Sanz-Rodrı́guez, J. Cuevas and 100 G. Fu, W. Liu, S. Feng and X. Yue, Chem. Commun., 2012,
S. González, Clin. Transl. Oncol., 2008, 10, 148–154. 48, 11567–11569.
76 W. M. Sharman, C. M. Allen and J. E. van Lier, Drug 101 Y. Zhou, Q. Bao, L. A. L. Tang, Y. Zhong and K. P. Loh,
Discovery Today, 1999, 4, 507–517. Chem. Mater., 2009, 21, 2950–2956.

5120 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

102 K. Yang, H. Xu, L. Cheng, C. Sun, J. Wang and Z. Liu, Adv. 128 W. Zhang, J. Lu, X. Gao, P. Li, W. Zhang, Y. Ma, H. Wang
Mater., 2012, 24, 5586–5592. and B. Tang, Angew. Chem., Int. Ed., 2018, 57, 4891–4896.
103 Z. Deng, C. Fang, X. Ma, X. Li, Y.-J. Zeng and X. Peng, ACS 129 M. Liu, L. Wang, X. Zheng, S. Liu and Z. Xie, ACS Appl.
Appl. Mater. Interfaces, 2020, 12, 20321–20330. Mater. Interfaces, 2018, 10, 24638–24647.
104 G. Zhou, Y. S. Wang, Z. Jin, P. Zhao, H. Zhang, Y. Wen and 130 Y. Zhao, Y. Kuang, M. Liu, J. Wang and R. Pei, Chem.
Q. He, Nanoscale Horiz., 2019, 4, 1185–1193. Mater., 2018, 30, 7511–7520.
105 S.-S. Wan, Q. Cheng, X. Zeng and X.-Z. Zhang, ACS Nano, 131 W. Zhu, L. Zhang, Z. Yang, P. Liu, J. Wang, J. Cao, A. Shen,
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

2019, 13, 6561–6571. Z. Xu and J. Wang, Chem. Eng. J., 2019, 358, 969–979.
106 Y. Ma, X. Li, A. Li, P. Yang, C. Zhang and B. Tang, Angew. 132 Q. Deng, P. Sun, L. Zhang, Z. Liu, H. Wang, J. Ren and
Chem., Int. Ed., 2017, 56, 13752–13756. X. Qu, Adv. Funct. Mater., 2019, 29, 1903018.
107 J.-Y. Zeng, M.-Z. Zou, M. Zhang, X.-S. Wang, X. Zeng, 133 P. Ling, C. Qian, J. Yu and F. Gao, Chem. Commun., 2019,
H. Cong and X.-Z. Zhang, ACS Nano, 2018, 12, 4630–4640. 55, 6385–6388.
108 K. Zhang, Z. Yu, X. Meng, W. Zhao, Z. Shi, Z. Yang, H. Dong 134 W. Zhu, Y. Yang, Q. Jin, Y. Chao, L. Tian, J. Liu, Z. Dong
and X. Zhang, Adv. Sci., 2019, 6, 1900530. and Z. Liu, Nano Res., 2019, 12, 1307–1312.
109 C. He, D. Liu and W. Lin, Chem. Rev., 2015, 115, 11079–11108. 135 Y. Zhao, X. Cai, Y. Zhang, C. Chen, J. Wang and R. Pei,
110 X.-S. Wang, J.-Y. Zeng, M.-K. Zhang, X. Zeng and Nanoscale, 2019, 11, 12250–12258.
X.-Z. Zhang, Adv. Funct. Mater., 2018, 28, 1801783. 136 X. Wan, H. Zhong, W. Pan, Y. Li, Y. Chen, N. Li and
111 S.-Y. Li, H. Cheng, W.-X. Qiu, L. Zhang, S.-S. Wan, J.-Y. Zeng B. Tang, Angew. Chem., Int. Ed., 2019, 58, 14134–14139.
and X.-Z. Zhang, Biomaterials, 2017, 142, 149–161. 137 J. Lu, L. Yang, W. Zhang, P. Li, X. Gao, W. Zhang, H. Wang
112 S.-S. Wan, J.-Y. Zeng, H. Cheng and X.-Z. Zhang, Biomaterials, and B. Tang, Chem. Commun., 2019, 55, 10792–10795.
2018, 185, 51–62. 138 Z. Gao, Y. Li, Y. Zhang, K. Cheng, P. An, F. Chen, J. Chen,
113 Y. Zhang, P. Sun, L. Zhang, Z. Wang, F. Wang, K. Dong, Z. Liu, C. You, Q. Zhu and B. Sun, ACS Appl. Mater. Interfaces,
J. Ren and X. Qu, Adv. Funct. Mater., 2019, 29, 1808594. 2020, 12, 1963–1972.
114 H. Qiu, F. Pu, Z. Liu, Q. Deng, P. Sun, J. Ren and X. Qu, 139 K. Ni, T. Luo, G. Lan, A. Culbert, Y. Song, T. Wu, X. Jiang
Small, 2019, 15, 1902522. and W. Lin, Angew. Chem., Int. Ed., 2020, 59, 1108–1112.
115 K. Kim, S. Lee, E. Jin, L. Palanikumar, J. H. Lee, J. C. Kim, 140 Y. Zhao, J. Wang, X. Cai, P. Ding, H. Lv and R. Pei, ACS
J. S. Nam, B. Jana, T.-H. Kwon, S. K. Kwak, W. Choe and Appl. Mater. Interfaces, 2020, 12, 23697–23706.
J.-H. Ryu, ACS Appl. Mater. Interfaces, 2019, 11, 27512–27520. 141 W. Zhu, Y. Liu, Z. Yang, L. Zhang, L. Xiao, P. Liu, J. Wang,
116 H. Wang, D. Yu, J. Fang, C. Cao, Z. Liu, J. Ren and X. Qu, C. Yi, Z. Xu and J. Ren, J. Mater. Chem. B, 2018, 6, 265–276.
ACS Nano, 2019, 13, 9206–9217. 142 J.-Y. Zeng, M.-K. Zhang, M.-Y. Peng, D. Gong and
117 M. Liu, L. Wang, X. Zheng and Z. Xie, ACS Appl. Mater. X.-Z. Zhang, Adv. Funct. Mater., 2018, 28, 1705451.
Interfaces, 2017, 9, 41512–41520. 143 X. Hu, Y. Lu, L. Zhou, L. Chen, T. Yao, S. Liang, J. Han,
118 J. Hu, W. Wu, Y. Qin, C. Liu, P. Wei, J. Hu, P. H. Seeberger C. Dong and S. Shi, J. Mater. Chem. B, 2020, 8, 935–944.
and J. Yin, Adv. Funct. Mater., 2020, 30, 1910084. 144 B. Li, X. Wang, L. Chen, Y. Zhou, W. Dang, J. Chang and
119 Z.-X. Chen, M.-D. Liu, M.-K. Zhang, S.-B. Wang, L. Xu, C. Wu, Theranostics, 2018, 8, 4086.
C.-X. Li, F. Gao, B.-R. Xie, Z.-L. Zhong and X.-Z. Zhang, Adv. 145 K. Zhang, X. Meng, Y. Cao, Z. Yang, H. Dong, Y. Zhang,
Funct. Mater., 2018, 28, 1803498. H. Lu, Z. Shi and X. Zhang, Adv. Funct. Mater., 2018,
120 Q.-g. Zhao, J. Wang, Y.-p. Zhang, J. Zhang, A.-n. Tang and 28, 1804634.
D.-m. Kong, J. Mater. Chem. B, 2018, 6, 7898–7907. 146 D. Han, Y. Han, J. Li, X. Liu, K. W. K. Yeung, Y. Zheng,
121 X.-T. Tian, P.-P. Cao, H. Zhang, Y.-H. Li and X.-B. Yin, Z. Cui, X. Yang, Y. Liang, Z. Li, S. Zhu, X. Yuan, X. Feng,
Chem. Commun., 2019, 55, 6241–6244. C. Yang and S. Wu, Appl. Catal., B, 2020, 261, 118248.
122 D. Zhang, Z. Ye, L. Wei, H. Luo and L. Xiao, ACS Appl. 147 S. J. Thompson, M. R. Brennan, S. Y. Lee and G. Dong,
Mater. Interfaces, 2019, 11, 39594–39602. Chem. Soc. Rev., 2018, 47, 929–981.
123 Z. Cai, F. Xin, Z. Wei, M. Wu, X. Lin, X. Du, G. Chen, 148 X. Cai, Z. Xie, B. Ding, S. Shao, S. Liang, M. Pang and J. Lin,
D. Zhang, Z. Zhang, X. Liu and C. Yao, Adv. Healthcare Adv. Sci., 2019, 6, 1900848.
Mater., 2020, 9, 1900996. 149 W. Wang, L. Wang, Z. Li and Z. Xie, Chem. Commun., 2016,
124 S.-Y. Li, B.-R. Xie, H. Cheng, C.-X. Li, M.-K. Zhang, 52, 5402–5405.
W.-X. Qiu, W.-L. Liu, X.-S. Wang and X.-Z. Zhang, Biomaterials, 150 X. Zhao, Z. Zhang, X. Cai, B. Ding, C. Sun, G. Liu, C. Hu,
2018, 151, 1–12. S. Shao and M. Pang, ACS Appl. Mater. Interfaces, 2019, 11,
125 M. He, Y. Chen, C. Tao, Q. Tian, L. An, J. Lin, Q. Tian, 7884–7892.
H. Yang and S. Yang, ACS Appl. Mater. Interfaces, 2019, 11, 151 J. Park, Q. Jiang, D. Feng and H.-C. Zhou, Angew. Chem.,
41946–41956. Int. Ed., 2016, 55, 7188–7193.
126 T. Luo, K. Ni, A. Culbert, G. Lan, Z. Li, X. Jiang, M. Kaufmann 152 L. Zhang, J. Lei, F. Ma, P. Ling, J. Liu and H. Ju, Chem.
and W. Lin, J. Am. Chem. Soc., 2020, 142, 7334–7339. Commun., 2015, 51, 10831–10834.
127 J. Liu, Y. Yang, W. Zhu, X. Yi, Z. Dong, X. Xu, M. Chen, 153 R. Chen, J. Zhang, Y. Wang, X. Chen, J. A. Zapien and
K. Yang, G. Lu, L. Jiang and Z. Liu, Biomaterials, 2016, 97, 1–9. C.-S. Lee, Nanoscale, 2015, 7, 17299–17305.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5121
 View Article Online

Chem Soc Rev Review Article

154 H. Cheng, J.-Y. Zhu, S.-Y. Li, J.-Y. Zeng, Q. Lei, K.-W. Chen, 178 X. Fu, Z. Yang, T. Deng, J. Chen, Y. Wen, X. Fu, L. Zhou,
C. Zhang and X.-Z. Zhang, Adv. Funct. Mater., 2016, 26, Z. Zhu and C. Yu, J. Mater. Chem. B, 2020, 8, 1481–1488.
7847–7860. 179 R. Chen, J. Zhang, J. Chelora, Y. Xiong, S. V. Kershaw, K. F. Li,
155 J. Liu, L. Zhang, J. Lei, H. Shen and H. Ju, ACS Appl. Mater. P.-K. Lo, K. W. Cheah, A. L. Rogach, J. A. Zapien and C.-S. Lee,
Interfaces, 2017, 9, 2150–2158. ACS Appl. Mater. Interfaces, 2017, 9, 5699–5708.
156 D. Yang, G. Yang, S. Gai, F. He, C. Li and P. Yang, ACS Appl. 180 W. Zhang, B. Li, H. Ma, L. Zhang, Y. Guan, Y. Zhang,
Mater. Interfaces, 2017, 9, 6829–6838. X. Zhang, P. Jing and S. Yue, ACS Appl. Mater. Interfaces,
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

157 H.-J. Cai, T.-T. Shen, J. Zhang, C.-F. Shan, J.-G. Jia, X. Li, W.- 2016, 8, 21465–21471.
S. Liu and Y. Tang, J. Mater. Chem. B, 2017, 5, 2390–2394. 181 D. Yang, G. Yang, S. Gai, F. He, G. An, Y. Dai, R. Lv and
158 F. Nian, Y. Huang, M. Song, J.-J. Chen and J. Xue, J. Mater. P. Yang, Nanoscale, 2015, 7, 19568–19578.
Chem. B, 2017, 5, 6227–6232. 182 Y. Wang, X. Pang, J. Wang, Y. Cheng, Y. Song, Q. Sun,
159 S. Sharma, K. Sethi and I. Roy, New J. Chem., 2017, 41, Q. You, F. Tan, J. Li and N. Li, J. Mater. Chem. B, 2018, 6,
11860–11866. 2460–2473.
160 J. Liu, G. Yang, W. Zhu, Z. Dong, Y. Yang, Y. Chao and 183 C. Liu, L. Luo, L. Zeng, J. Xing, Y. Xia, S. Sun, L. Zhang,
Z. Liu, Biomaterials, 2017, 146, 40–48. Z. Yu, J. Yao, Z. Yu, O. U. Akakuru, M. Saeed and A. Wu,
161 Q. Guan, L.-L. Zhou, Y.-A. Li and Y.-B. Dong, Inorg. Chem., Small, 2018, 14, 1801851.
2018, 57, 10137–10145. 184 S. Gao, P. Zheng, Z. Li, X. Feng, W. Yan, S. Chen, W. Guo,
162 M.-R. Song, D.-Y. Li, F.-Y. Nian, J.-P. Xue and J.-J. Chen, D. Liu, X. Yang, S. Wang, X.-J. Liang and J. Zhang, Bioma-
J. Mater. Sci., 2018, 53, 2351–2361. terials, 2018, 178, 83–94.
163 D. Xu, Y. You, F. Zeng, Y. Wang, C. Liang, H. Feng and 185 P. Yang, Y. Men, Y. Tian, Y. Cao, L. Zhang, X. Yao and
X. Ma, ACS Appl. Mater. Interfaces, 2018, 10, 15517–15523. W. Yang, ACS Appl. Mater. Interfaces, 2019, 11, 11209–11219.
164 G. Hu, L. Yang, Y. Li and L. Wang, J. Mater. Chem. B, 2018, 186 J. Liu, T. Liu, P. Du, L. Zhang and J. Lei, Angew. Chem., Int.
6, 7936–7942. Ed., 2019, 58, 7808–7812.
165 Z. Xie, X. Cai, C. Sun, S. Liang, S. Shao, S. Huang, Z. Cheng, 187 R. Chen, W.-C. Chen, L. Yan, S. Tian, B. Liu, X. Chen, C.-
M. Pang, B. Xing, A. A. A. Kheraif and J. Lin, Chem. Mater., S. Lee and W. Zhang, J. Mater. Chem. B, 2019, 7, 4763–4770.
2019, 31, 483–490. 188 Y. Wang, J. Yan, N. Wen, H. Xiong, S. Cai, Q. He, Y. Hu,
166 Z. He, Y. Dai, X. Li, D. Guo, Y. Liu, X. Huang, J. Jiang, D. Peng, Z. Liu and Y. Liu, Biomaterials, 2020, 230, 119619.
S. Wang, G. Zhu, F. Zhang, L. Lin, J.-J. Zhu, G. Yu and 189 K. K. Tanabe and S. M. Cohen, Chem. Soc. Rev., 2011, 40,
X. Chen, Small, 2019, 15, 1804131. 498–519.
167 Q. Sun, H. Bi, Z. Wang, C. Li, C. Wang, J. Xu, D. Yang, 190 S. Wang, C. M. McGuirk, A. d’Aquino, J. A. Mason and
F. He, S. Gai and P. Yang, ACS Appl. Mater. Interfaces, 2019, C. A. Mirkin, Adv. Mater., 2018, 30, 1800202.
11, 36347–36358. 191 A. Pandey, N. Dhas, P. Deshmukh, C. Caro, P. Patil,
168 Y. Wang, W. Wu, J. Liu, P. N. Manghnani, F. Hu, D. Ma, M. Luisa Garcı́a-Martı́n, B. Padya, A. Nikam, T. Mehta
C. Teh, B. Wang and B. Liu, ACS Nano, 2019, 13, and S. Mutalik, Coord. Chem. Rev., 2020, 409, 213212.
6879–6890. 192 A. Zimpel, T. Preiß, R. Röder, H. Engelke, M. Ingrisch,
169 Y. Wang, W. Wu, D. Mao, C. Teh, B. Wang and B. Liu, Adv. M. Peller, J. O. Rädler, E. Wagner, T. Bein, U. Lächelt and
Funct. Mater., 2020, 30, 2002431. S. Wuttke, Chem. Mater., 2016, 28, 3318–3326.
170 Y.-C. Ma, Y.-H. Zhu, X.-F. Tang, L.-F. Hang, W. Jiang, M. Li, 193 K. Lu, T. Aung, N. Guo, R. Weichselbaum and W. Lin, Adv.
M. I. Khan, Y.-Z. You and Y.-C. Wang, Biomater. Sci., 2019, Mater., 2018, 30, 1707634.
7, 2740–2748. 194 D. Luo, K. A. Carter, D. Miranda and J. F. Lovell, Adv. Sci.,
171 Z. Xie, S. Liang, X. Cai, B. Ding, S. Huang, Z. Hou, P. a. Ma, 2017, 4, 1600106.
Z. Cheng and J. Lin, ACS Appl. Mater. Interfaces, 2019, 11, 195 E. B. Rankin and A. J. Giaccia, Science, 2016, 352, 175–180.
31671–31680. 196 R. Mo and Z. Gu, Mater. Today, 2016, 19, 274–283.
172 Y.-T. Qin, H. Peng, X.-W. He, W.-Y. Li and Y.-K. Zhang, ACS 197 X. Zhong, X. Wang, L. Cheng, Y. Tang, G. Zhan, F. Gong,
Appl. Mater. Interfaces, 2019, 11, 34268–34281. R. Zhang, J. Hu, Z. Liu and X. Yang, Adv. Funct. Mater.,
173 C. Wang, X. Jia, W. Zhen, M. Zhang and X. Jiang, ACS 2020, 30, 1907954.
Biomater. Sci. Eng., 2019, 5, 4435–4441. 198 Q. Chen, J. Zhou, Z. Chen, Q. Luo, J. Xu and G. Song, ACS
174 J. Feng, W. Yu, Z. Xu and F. Wang, Chem. Sci., 2020, 11, Appl. Mater. Interfaces, 2019, 11, 30551–30565.
1649–1656. 199 Z. Wang, B. Liu, Q. Sun, S. Dong, Y. Kuang, Y. Dong, F. He,
175 Y. You, D. Xu, X. Pan and X. Ma, Appl. Mater. Today, 2019, S. Gai and P. Yang, ACS Appl. Mater. Interfaces, 2020, 12,
16, 508–517. 17254–17267.
176 P. Liu, X. Xie, X. Shi, Y. Peng, J. Ding and W. Zhou, ACS 200 H. Ranji-Burachaloo, P. A. Gurr, D. E. Dunstan and
Appl. Mater. Interfaces, 2019, 11, 48261–48270. G. G. Qiao, ACS Nano, 2018, 12, 11819–11837.
177 H. Chen, J. Yang, L. Sun, H. Zhang, Y. Guo, J. Qu, W. Jiang, 201 Y. Liu, C. S. Gong, L. Lin, Z. Zhou, Y. Liu, Z. Yang, Z. Shen,
W. Chen, J. Ji, Y.-W. Yang and B. Wang, Small, 2019, G. Yu, Z. Wang, S. Wang, Y. Ma, W. Fan, L. He, G. Niu,
15, 1903880. Y. Dai and X. Chen, Theranostics, 2019, 9, 2791–2799.

5122 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

202 D. Wang, H. Wu, W. Q. Lim, S. Z. F. Phua, P. Xu, Q. Chen, 229 M. R. Hamblin and T. Hasan, Photochem. Photobiol. Sci.,
Z. Guo and Y. Zhao, Adv. Mater., 2019, 31, 1901893. 2004, 3, 436–450.
203 X. Meng, J. Deng, F. Liu, T. Guo, M. Liu, P. Dai, A. Fan, 230 H. Wu, C. Moser, H.-Z. Wang, N. Høiby and Z.-J. Song, Int.
Z. Wang and Y. Zhao, Nano Lett., 2019, 19, 7866–7876. J. Oral Sci., 2015, 7, 1–7.
204 A. K. Iyer, G. Khaled, J. Fang and H. Maeda, Drug Discovery 231 L. Tan, J. Li, X. Liu, Z. Cui, X. Yang, S. Zhu, Z. Li, X. Yuan,
Today, 2006, 11, 812–818. Y. Zheng, K. W. K. Yeung, H. Pan, X. Wang and S. Wu, Adv.
205 H. Kobayashi, R. Watanabe and P. L. Choyke, Theranostics, Mater., 2018, 30, 1801808.
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

2014, 4, 81–89. 232 J. Wu, C. Zhang, S. Xu, X. Pang, G. Cai and J. Wang, Colloid
206 M. Yu and J. Zheng, ACS Nano, 2015, 9, 6655–6674. Interface Sci. Commun., 2018, 24, 98–104.
207 J. Fang, H. Nakamura and H. Maeda, Adv. Drug Delivery 233 B. Le Ouay and F. Stellacci, Nano Today, 2015, 10, 339–354.
Rev., 2011, 63, 136–151. 234 E.-L. Zhang, S. Fu, R.-X. Wang, H.-X. Li, Y. Liu, Z.-Q. Ma,
208 H. Soo Choi, W. Liu, P. Misra, E. Tanaka, J. P. Zimmer, G.-K. Liu, C.-S. Zhu, G.-W. Qin and D.-F. Chen, Rare Met.,
B. Itty Ipe, M. G. Bawendi and J. V. Frangioni, Nat. 2019, 38, 476–494.
Biotechnol., 2007, 25, 1165–1170. 235 J.-M. Zhang, Y.-H. Sun, Y. Zhao, Y.-L. Liu, X.-H. Yao,
209 J. Zhou, G. Yu and F. Huang, Chem. Soc. Rev., 2017, 46, B. Tang and R.-Q. Hang, Rare Met., 2019, 38, 552–560.
7021–7053. 236 T. Yang, D. Wang and X. Liu, J. Mater. Chem. B, 2020, 8,
210 J. Zhang, X. Li, X. Han, R. Liu and J. Fang, Trends 406–415.
Pharmacol. Sci., 2017, 38, 794–808. 237 N. Khan, D. Kumar and P. Kumar, Colloid Interface Sci.
211 B. Yang, Y. Chen and J. Shi, Adv. Mater., 2019, 31, 1901778. Commun., 2020, 35, 100242.
212 W. Fan, N. Lu, P. Huang, Y. Liu, Z. Yang, S. Wang, G. Yu, 238 Y.-S. Huang and H.-H. Huang, Rare Met., 2019, 38,
Y. Liu, J. Hu, Q. He, J. Qu, T. Wang and X. Chen, Angew. 512–519.
Chem., Int. Ed., 2017, 56, 1229–1233. 239 B. Kong, C. Selomulya, G. Zheng and D. Zhao, Chem. Soc.
213 L. Chen, S.-F. Zhou, L. Su and J. Song, ACS Nano, 2019, 13, Rev., 2015, 44, 7997–8018.
10887–10917. 240 Z. Qin, Y. Li and N. Gu, Adv. Healthcare Mater., 2018,
214 Y. Li, X. Liu, B. Li, Y. Zheng, Y. Han, D.-f. Chen, 7, 1800347.
K. W. K. Yeung, Z. Cui, Y. Liang, Z. Li, S. Zhu, X. Wang 241 X. Cai, W. Gao, M. Ma, M. Wu, L. Zhang, Y. Zheng, H. Chen
and S. Wu, ACS Nano, 2020, 14, 8157–8170. and J. Shi, Adv. Mater., 2015, 27, 6382–6389.
215 J. D. Martin, H. Cabral, T. Stylianopoulos and R. K. Jain, 242 L. Hang, H. Li, T. Zhang, D. Men, C. Zhang, P. Gao and
Nat. Rev. Clin. Oncol., 2020, 17, 251–266. Q. Zhang, ACS Appl. Mater. Interfaces, 2019, 11,
216 X. Duan, C. Chan and W. Lin, Angew. Chem., Int. Ed., 2019, 39493–39502.
58, 670–680. 243 M. Shokouhimehr, E. S. Soehnlen, A. Khitrin, S. Basu and
217 L. A. Emens, P. A. Ascierto, P. K. Darcy, S. Demaria, S. D. Huang, Inorg. Chem. Commun., 2010, 13, 58–61.
A. M. M. Eggermont, W. L. Redmond, B. Seliger and 244 W. Zhu, K. Liu, X. Sun, X. Wang, Y. Li, L. Cheng and Z. Liu,
F. M. Marincola, Eur. J. Cancer, 2017, 81, 116–129. ACS Appl. Mater. Interfaces, 2015, 7, 11575–11582.
218 X. Li, C. Shao, Y. Shi and W. Han, J. Hematol. Oncol., 2018, 11, 31. 245 P. Shou, Z. Yu, Y. Wu, Q. Feng, B. Zhou, J. Xing, C. Liu,
219 L. Guo, D. D. Yan, D. Yang, Y. Li, X. Wang, O. Zalewski, J. Tu, O. U. Akakuru, Z. Ye, X. Zhang, Z. Lu, L. Zhang and
B. Yan and W. Lu, ACS Nano, 2014, 8, 5670–5681. A. Wu, Adv. Healthcare Mater., 2020, 9, 1900948.
220 H. Zhang, J. Zhang, Q. Li, A. Song, H. Tian, J. Wang, Z. Li 246 W. Wu, L. Yu, Y. Pu, H. Yao, Y. Chen and J. Shi, Adv. Mater.,
and Y. Luan, Biomaterials, 2020, 245, 119983. 2020, 32, 2000542.
221 J. Liu, S. J. Blake, M. C. R. Yong, H. Harjunpää, S. F. Ngiow, 247 L. Fétiveau, G. Paul, A. Nicolas-Boluda, J. Volatron,
K. Takeda, A. Young, J. S. O’Donnell, S. Allen, M. J. Smyth R. George, S. Laurent, R. Muller, L. Sancey, P. Mejanelle,
and M. W. L. Teng, Cancer Discovery, 2016, 6, 1382–1399. A. Gloter, F. Gazeau and L. Catala, Chem. Commun., 2019,
222 S. S. Lucky, K. C. Soo and Y. Zhang, Chem. Rev., 2015, 115, 55, 14844–14847.
1990–2042. 248 Y. Li, C. H. Li and D. R. Talham, Nanoscale, 2015, 7,
223 J. Schmitt, V. Heitz, A. Sour, F. Bolze, H. Ftouni, 5209–5216.
J.-F. Nicoud, L. Flamigni and B. Ventura, Angew. Chem., 249 X. Cai, W. Gao, L. Zhang, M. Ma, T. Liu, W. Du, Y. Zheng,
Int. Ed., 2015, 54, 169–173. H. Chen and J. Shi, ACS Nano, 2016, 10, 11115–11126.
224 S. Han, R. Deng, X. Xie and X. Liu, Angew. Chem., Int. Ed., 250 M. Ishizaki, H. Ando, N. Yamada, K. Tsumoto, K. Ono,
2014, 53, 11702–11715. H. Sutoh, T. Nakamura, Y. Nakao and M. Kurihara,
225 Z. Gu, L. Yan, G. Tian, S. Li, Z. Chai and Y. Zhao, Adv. J. Mater. Chem. A, 2019, 7, 4777–4787.
Mater., 2013, 25, 3758–3779. 251 W. Tian, Y. Su, Y. Tian, S. Wang, X. Su, Y. Liu, Y. Zhang,
226 X. Chen, D. Peng, Q. Ju and F. Wang, Chem. Soc. Rev., 2015, Y. Tang, Q. Ni, W. Liu, M. Dang, C. Wang, J. Zhang, Z. Teng
44, 1318–1330. and G. Lu, Adv. Sci., 2017, 4, 1600356.
227 X. Jia, J. Li and E. Wang, Nanoscale, 2012, 4, 5572–5575. 252 X. Cai, X. Jia, W. Gao, K. Zhang, M. Ma, S. Wang, Y.
228 J. Shen, Y. Zhu, C. Chen, X. Yang and C. Li, Chem. Zheng, J. Shi and H. Chen, Adv. Funct. Mater., 2015, 25,
Commun., 2011, 47, 2580–2582. 2520–2529.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5123
 View Article Online

Chem Soc Rev Review Article

253 X. Yao, P. Yang, Z. Jin, Q. Jiang, R. Guo, R. Xie, Q. He and 276 J. Li, C. Zhang, S. Gong, X. Li, M. Yu, C. Qian, H. Qiao and
W. Yang, Biomaterials, 2019, 197, 268–283. M. Sun, Acta Biomater., 2019, 94, 435–446.
254 J. An, Y.-G. Hu, C. Li, X.-L. Hou, K. Cheng, B. Zhang, 277 G. Zhao, H. Wu, R. Feng, D. Wang, P. Xu, P. Jiang, K. Yang,
R.-Y. Zhang, D.-Y. Li, S.-J. Liu, B. Liu, D. Zhu and H. Wang, Z. Guo and Q. Chen, ACS Appl. Mater. Interfaces,
Y.-D. Zhao, Biomaterials, 2020, 230, 119636. 2018, 10, 3295–3304.
255 Y. Li, J. Dang, Q. Liang and L. Yin, Biomaterials, 2019, 209, 278 W. Cai, H. Gao, C. Chu, X. Wang, J. Wang, P. Zhang, G. Lin,
138–151. W. Li, G. Liu and X. Chen, ACS Appl. Mater. Interfaces, 2017,
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

256 T. Feng, J. Wan, P. Li, H. Ran, H. Chen, Z. Wang and 9, 2040–2051.

L. Zhang, Biomaterials, 2019, 214, 119213. 279 W. Zhu, M. Chen, Y. Liu, Y. Tian, Z. Song, G. Song and
257 Y. Liu, G. Shu, X. Li, H. Chen, B. Zhang, H. Pan, T. Li, X. Zhang, Nanoscale, 2019, 11, 20630–20637.
X. Gong, H. Wang, X. Wu, Y. Dou and J. Chang, Adv. Funct. 280 S. Dang, Q.-L. Zhu and Q. Xu, Nat. Rev. Mater., 2017,
Mater., 2018, 28, 1802026. 3, 17075.
258 L. Cheng, H. Gong, W. Zhu, J. Liu, X. Wang, G. Liu and 281 S.-N. Zhao, X.-Z. Song, S.-Y. Song and H.-j. Zhang, Coord.
Z. Liu, Biomaterials, 2014, 35, 9844–9852. Chem. Rev., 2017, 337, 80–96.
259 M. Gautam, B. Kumar Poudel, Z. Chi Soe, K. Poudel, 282 K. Xiong, J. Li, L. Tan, Z. Cui, Z. Li, S. Wu, Y. Liang, S. Zhu
S. Maharjan, S. Kwang Ku, C. Soon Yong, S. Woo Joo, J. Oh and X. Liu, Colloid Interface Sci. Commun., 2019,
Kim and J. Hoon Byeon, Chem. Eng. J., 2020, 383, 123177. 33, 100201.
260 X.-D. Li, X.-L. Liang, F. Ma, L.-J. Jing, L. Lin, Y.-B. Yang, 283 H. Xiang, W. Feng and Y. Chen, Adv. Mater., 2020,
S.-S. Feng, G.-L. Fu, X.-L. Yue and Z.-F. Dai, Colloids Surf., 32, 1905994.
B, 2014, 123, 629–638. 284 Y. Weng, S. Guan, L. Wang, H. Lu, X. Meng,
261 P. Xue, K. K. Y. Cheong, Y. Wu and Y. Kang, Colloids Surf., G. I. N. Waterhouse and S. Zhou, Small, 2020, 16, 1905184.
B, 2015, 125, 277–283. 285 A. Han, B. Wang, A. Kumar, Y. Qin, J. Jin, X. Wang, C. Yang,
262 C. Tong, X. Zhong, Y. Yang, X. Liu, G. Zhong, C. Xiao, B. Dong, Y. Jia, J. Liu and X. Sun, Small Methods, 2019,
B. Liu, W. Wang and X. Yang, Biomaterials, 2020, 3, 1800471.
243, 119936. 286 L. Jiao and H.-L. Jiang, Chem, 2019, 5, 786–804.
263 Y. Y. Su, Z. Teng, H. Yao, S. J. Wang, Y. Tian, Y. L. Zhang, 287 B. Xu, H. Wang, W. Wang, L. Gao, S. Li, X. Pan, H. Wang,
W. F. Liu, W. Tian, L. J. Zheng, N. Lu, Q. Q. Ni, X. D. Su, H. Yang, X. Meng, Q. Wu, L. Zheng, S. Chen, X. Shi, K. Fan,
Y. X. Tang, J. Sun, Y. Liu, J. Wu, G. F. Yang, G. M. Lu and X. Yan and H. Liu, Angew. Chem., Int. Ed., 2019, 58, 4911–4916.
L. J. Zhang, ACS Appl. Mater. Interfaces, 2016, 8, 17038–17046. 288 P. Yin, T. Yao, Y. Wu, L. Zheng, Y. Lin, W. Liu, H. Ju, J. Zhu,
264 J.-J. Hu, Y. Chen, Z.-H. Li, S.-Y. Peng, Y. Sun and X. Hong, Z. Deng, G. Zhou, S. Wei and Y. Li, Angew. Chem.,
X.-Z. Zhang, Nano Lett., 2019, 19, 5568–5576. Int. Ed., 2016, 55, 10800–10805.
265 D. Wang, J. Zhou, R. Shi, H. Wu, R. Chen, B. Duan, G. Xia, 289 Q. Yang, C.-C. Yang, C.-H. Lin and H.-L. Jiang, Angew.
P. Xu, H. Wang and S. Zhou, Theranostics, 2017, 7, 4605. Chem., Int. Ed., 2019, 58, 3511–3515.
266 D. Wang, J. Zhou, R. Chen, R. Shi, G. Zhao, G. Xia, R. Li, 290 M. Huo, L. Wang, Y. Wang, Y. Chen and J. Shi, ACS Nano,
Z. Liu, J. Tian, H. Wang, Z. Guo, H. Wang and Q. Chen, 2019, 13, 2643–2653.
Biomaterials, 2016, 100, 27–40. 291 Y. Yang, Y. Deng, J. Huang, X. Fan, C. Cheng, C. Nie, L. Ma,
267 Y. Luo, J. Li, X. Liu, L. Tan, Z. Cui, X. Feng, X. Yang, W. Zhao and C. Zhao, Adv. Funct. Mater., 2019, 29, 1900143.
Y. Liang, Z. Li, S. Zhu, Y. Zheng, K. W. K. Yeung, C. Yang, 292 X. Fan, F. Yang, J. Huang, Y. Yang, C. Nie, W. Zhao, L. Ma,
X. Wang and S. Wu, ACS Cent. Sci., 2019, 5, 1591–1601. C. Cheng, C. Zhao and R. Haag, Nano Lett., 2019, 19, 5885–5896.
268 L. Jing, X. Liang, Z. Deng, S. Feng, X. Li, M. Huang, C. Li 293 T. Kitao, Y. Zhang, S. Kitagawa, B. Wang and T. Uemura,
and Z. Dai, Biomaterials, 2014, 35, 5814–5821. Chem. Soc. Rev., 2017, 46, 3108–3133.
269 W. Zhu, M.-Y. Gao, Q. Zhu, B. Chi, L.-W. Zeng, J.-M. Hu 294 Y.-D. Zhu, S.-P. Chen, H. Zhao, Y. Yang, X.-Q. Chen, J. Sun,
and A.-G. Shen, Nanoscale, 2020, 12, 3292–3301. H.-S. Fan and X.-D. Zhang, ACS Appl. Mater. Interfaces,
270 G. Fu, W. Liu, Y. Li, Y. Jin, L. Jiang, X. Liang, S. Feng and 2016, 8, 34209–34217.
Z. Dai, Bioconjugate Chem., 2014, 25, 1655–1663. 295 Q. Wu, M. Niu, X. Chen, L. Tan, C. Fu, X. Ren, J. Ren, L. Li,
271 J. Peng, M. Dong, B. Ran, W. Li, Y. Hao, Q. Yang, L. Tan, K. Xu, H. Zhong and X. Meng, Biomaterials, 2018, 162,
K. Shi and Z. Qian, ACS Appl. Mater. Interfaces, 2017, 9, 132–143.
13875–13886. 296 W. Wang, L. Wang, Y. Li, S. Liu, Z. Xie and X. Jing, Adv.
272 J. Peng, Q. Yang, W. Li, L. Tan, Y. Xiao, L. Chen, Y. Hao and Mater., 2016, 28, 9320–9325.
Z. Qian, ACS Appl. Mater. Interfaces, 2017, 9, 44410–44422. 297 J. Huang, N. Li, C. Zhang and Z. Meng, ACS Appl. Mater.
273 X. Zheng, L. Wang, M. Liu, P. Lei, F. Liu and Z. Xie, Chem. Interfaces, 2018, 10, 38729–38738.
Mater., 2018, 30, 6867–6876. 298 Y. Li, J. Jin, D. Wang, J. Lv, K. Hou, Y. Liu, C. Chen and
274 Y. Wu, M. Yuan, J. Song, X. Chen and H. Yang, ACS Nano, Z. Tang, Nano Res., 2018, 11, 3294–3305.
2019, 13, 8505–8511. 299 L. Zhang, C. Liu, Y. Gao, Z. Li, J. Xing, W. Ren, L. Zhang,
275 F. Liu, X. He, H. Chen, J. Zhang, H. Zhang and Z. Wang, A. Li, G. Lu, A. Wu and L. Zeng, Adv. Healthcare Mater.,
Nat. Commun., 2015, 6, 8003. 2018, 7, 1801144.

5124 | Chem. Soc. Rev., 2021, 50, 5086–5125 This journal is © The Royal Society of Chemistry 2021
 View Article Online

Review Article Chem Soc Rev

300 X. Yang, L. Li, D. He, L. Hai, J. Tang, H. Li, X. He and 316 C. Liu, D. Kong, P.-C. Hsu, H. Yuan, H.-W. Lee, Y. Liu,
K. Wang, J. Mater. Chem. B, 2017, 5, 4648–4659. H. Wang, S. Wang, K. Yan, D. Lin, P. A. Maraccini,
301 L. An, M. Cao, X. Zhang, J. Lin, Q. Tian and S. Yang, ACS K. M. Parker, A. B. Boehm and Y. Cui, Nat. Nanotechnol.,
Appl. Mater. Interfaces, 2020, 12, 8050–8061. 2016, 11, 1098–1104.
302 X. Liu, Y. Pan, J. Yang, Y. Gao, T. Huang, X. Luan, Y. Wang 317 Y. Ren, Y. Han, Z. Li, X. Liu, S. Zhu, Y. Liang,
and Y. Song, Nano Res., 2020, 13, 653–660. K. W. K. Yeung and S. Wu, Bioact. Mater., 2020, 5, 201–209.
303 J. Yao, Y. Liu, J. Wang, Q. Jiang, D. She, H. Guo, N. Sun, 318 B. Wu, Y. Li, K. Su, L. Tan, X. Liu, Z. Cui, X. Yang, Y. Liang,
Published on 26 February 2021. Downloaded by Nanyang Technological University on 5/15/2023 5:39:29 AM.

Z. Pang, C. Deng, W. Yang and S. Shen, Biomaterials, 2019, Z. Li, S. Zhu, K. W. K. Yeung and S. Wu, J. Hazard. Mater.,
195, 51–62. 2019, 377, 227–236.
304 W. Shang, C. Zeng, Y. Du, H. Hui, X. Liang, C. Chi, 319 Y. Li, X. Liu, L. Tan, Z. Cui, X. Yang, Y. Zheng,
K. Wang, Z. Wang and J. Tian, Adv. Mater., 2017, K. W. K. Yeung, P. K. Chu and S. Wu, Adv. Funct. Mater.,
29, 1604381. 2018, 28, 1800299.
305 B. Li, Q. Wang, R. Zou, X. Liu, K. Xu, W. Li and J. Hu, 320 M. Chen, Z. Long, R. Dong, L. Wang, J. Zhang, S. Li,
Nanoscale, 2014, 6, 3274–3282. X. Zhao, X. Hou, H. Shao and X. Jiang, Small, 2020,
306 M. Wang, K. Deng, W. Lü, X. Deng, K. Li, Y. Shi, B. Ding, 16, 1906240.
Z. Cheng, B. Xing, G. Han, Z. Hou and J. Lin, Adv. Mater., 321 J. Li, L. Tan, X. Liu, Z. Cui, X. Yang, K. W. K. Yeung,
2018, 30, 1706747. P. K. Chu and S. Wu, ACS Nano, 2017, 11, 11250–11263.
307 X. Deng, K. Li, X. Cai, B. Liu, Y. Wei, K. Deng, Z. Xie, Z. Wu, 322 B. Huang, L. Tan, X. Liu, J. Li and S. Wu, Bioact. Mater.,
P. a. Ma, Z. Hou, Z. Cheng and J. Lin, Adv. Mater., 2017, 2019, 4, 17–21.
29, 1701266. 323 S. Tavakoli, S. Nemati, M. Kharaziha and S. Akbari-
308 H. Chen, Y. Ma, X. Wang and Z. Zha, J. Mater. Chem. B, Alavijeh, Colloid Interface Sci. Commun., 2019, 28, 20–28.
2017, 5, 7051–7058. 324 L. Huang, J. Li, W. Yuan, X. Liu, Z. Li, Y. Zheng, Y. Liang,
309 J. Li, F. Zhang, Z. Hu, W. Song, G. Li, G. Liang, J. Zhou, S. Zhu, Z. Cui, X. Yang, K. W. K. Yeung and S. Wu, Corros.
K. Li, Y. Cao, Z. Luo and K. Cai, Adv. Healthcare Mater., Sci., 2020, 163, 108257.
2017, 6, 1700005. 325 Q.-W. Chen, X.-H. Liu, J.-X. Fan, S.-Y. Peng, J.-W. Wang,
310 Y. Zhang, Y. Liu, X. Gao, X. Li, X. Niu, Z. Yuan and X.-N. Wang, C. Zhang, C.-J. Liu and X.-Z. Zhang, Adv. Funct.
W. Wang, Acta Biomater., 2019, 90, 314–323. Mater., 2020, 30, 1909806.
311 Y. Zhang, L. Wang, L. Liu, L. Lin, F. Liu, Z. Xie, H. Tian and 326 R. Riccò, W. Liang, S. Li, J. J. Gassensmith, F. Caruso,
X. Chen, ACS Appl. Mater. Interfaces, 2018, 10, 41035–41045. C. Doonan and P. Falcaro, ACS Nano, 2018, 12, 13–23.
312 Z. Luo, L. Jiang, S. Yang, Z. Li, W. M. W. Soh, L. Zheng, X. J. Loh 327 S. Yan, X. Zeng, Y. Wang and B.-F. Liu, Adv. Healthcare
and Y.-L. Wu, Adv. Healthcare Mater., 2019, 8, 1900406. Mater., 2020, 9, 2000046.
313 S. Wang, L. Shang, L. Li, Y. Yu, C. Chi, K. Wang, J. Zhang, 328 S. Qiu, M. Xue and G. Zhu, Chem. Soc. Rev., 2014, 43,
R. Shi, H. Shen, G. I. N. Waterhouse, S. Liu, J. Tian, 6116–6140.
T. Zhang and H. Liu, Adv. Mater., 2016, 28, 8379–8387. 329 J. Zhao, W. T. Nunn, P. C. Lemaire, Y. Lin, M. D. Dickey,
314 F. Zhang, Y. Liu, J. Lei, S. Wang, X. Ji, H. Liu and Q. Yang, C. J. Oldham, H. J. Walls, G. W. Peterson, M. D. Losego and
Adv. Sci., 2019, 6, 1901378. G. N. Parsons, J. Am. Chem. Soc., 2015, 137, 13756–13759.
315 P. Yang, Y. Tian, Y. Men, R. Guo, H. Peng, Q. Jiang and 330 J. Zhao, B. Kalanyan, H. F. Barton, B. A. Sperling and
W. Yang, ACS Appl. Mater. Interfaces, 2018, 10, 42039–42049. G. N. Parsons, Chem. Mater., 2017, 29, 8804–8810.

This journal is © The Royal Society of Chemistry 2021 Chem. Soc. Rev., 2021, 50, 5086–5125 | 5125