Layered metal-organic frameworks and metal-organic nanosheets as functional materials

Layered metal-organic frameworks and metal-organic nanosheets have been regarded as an important series of low-dimensional materials. Their ﬂexible designability and intrinsic porosity allow layered metal-organic frameworks and metal-organic nanosheets to be differentiated from inorganic counter-parts. The present review article is devoted to collecting pieces of research aiming at exploiting layered metal-organic frameworks and metal-organic nanosheets as innovative nanomaterials. Spearheaded


Introduction
Two-dimensional (2D) materials have collected substantial recent attention of researchers, the scientific field of which has been spearheaded by inorganic nanosheets, such as graphene and other monoelemental 2D materials (Xenes), metal oxide nanosheets, transition metal dichalcogenides (TMDCs), and MXenes. This series of inorganic nanomaterials shows distinctive physical properties based on their 2D structures, resulting in significant anticipation of future breakthrough applications in various fields. The bright futurity for inorganic nanosheets has prompted scientists to shed spotlight also on molecule-based nanosheets that are constructed from organic monomers [1][2][3]. Metal complexes afford an important series of molecular superstructures created from organic ligand molecules and metal sources, and some of them indeed possess 2D nanosheet motifs. Such layered metal-organic frameworks (MOFs) and metal-organic nanosheets (MONs), possess two major virtues over inorganic nanosheets. One is their flexible designability achieved by customizing and selecting organic ligand molecules and metal components, respectively. The other virtue lies in the porous structures derived from organic ligand molecules with a certain degree of size. The two features allow layered MOFs and MONs to enjoy a wide variety of chemical and physical properties, as well as to be distinguished from inorganic nanosheets. In fact, a steep increase in the number of publications has been found for layered MOFs and MONs in recent years ( Fig. 1). Their next challenge for lies in the demonstration of functionalities and potential applications.
Preparation methods for layered MOFs and MONs have been assembled recently in our previous review articles [1,3]. Thus, this aspect is not treated as an individual section in the present one, but just indicated in several samples of layered MOFs and MONs materials discussed herein. Instead, the present review article focuses on this hot topic, layered MOFs and MONs as potential nanomaterials. It covers physical properties such as mechanical strengths, electrical and proton conductivities, exotic electrical properties, magnetism, redox, luminescence. This review article also includes various applications, such as electrocatalysis, separation and filtration, energy devices, and sensors. Theoretical predictions for unveiled properties are also summarized. The outlook for layered MOFs and MONs is given in the last section.

Mechanical properties
Many applications require materials that are morphologically designed to exhibit mechanical properties suitable for specific applications. Therefore, research to control the mechanical properties of new materials such as layered MOFs and MONs is also of critical importance. Zamora and coworkers reported a simple yet powerful top-down method to produce large layered MOF flakes with an area of several hundred square microns. This method also allowed them to produce layers from a few to 50 layers with controllable thicknesses [4]. These large MOF layers of [Cu(l-pym 2 S 2 ) (l-Cl)] n ÁnMeOH showed red emission when deposited on SiO 2 ( Fig. 2a,b). Importantly, the large lateral dimensions allowed the layers to be fabricated in a freestanding state, allowing the authors to evaluate their mechanical properties. The observed Young modulus of 5 GPa was in good agreement with the values calculated using density functional theory (3.4-4.1 GPa; Fig. 2c). The authors also prepared layered MOF films of 4 nm thickness based on copper (I) iodothioacetamide by a one-pot method using two commercially available inexpensive reagents [5]. The Young modulus of these films was also measured in a freestanding configuration and was 11 GPa.

Electrical conductivity
A series of electronically conductive layered MOFs and MONs shares features in their chemical structures, p-conjugated planar ligands and square planar coordination spheres [ML 4 ] (M = transition metal; L = S, NH, OH) [6][7][8][9][10][11][12][13][14][15][16][17][18]. This series of features allows electrons to delocalize over the 2D nanosheet framework, facilitating their transport. Bis(dithiolato)nickel 2D coordination polymer (NiDT) prepared from benzenehexathiol (BHT) and nickel acetate is the first conductive layered MOFs and MONs of this series reported by Nishihara and coworkers (Fig. 3a) [6][7][8]. NiDT comprises bis(dithiolato)Ni units fused by phenylene bridges, thereby boasting developed 2D p-conjugation. NiDT exhibited semiconducting behavior in the temperature-conductivity characteristic, where the conductivity decreases (resistivity increases) upon lowering the temperature (Fig. 3b). The conductivity of as-prepared NiDT at room temperature was 2.8 S cm À1 , and reached 160 S cm À1 when oxidized chemically. Thus, NiDT demonstrated a conductivity control by chemical redox. Multilayered NiDT is predicted to be a metal [7], suggesting that eliminating the inter-domain resistance by increasing the size of single domain crystals is important to understand the intrinsic electronic nature of NiDT. Chen, Xu, and Zhu prepared CuBHT [9,10], a NiDT analogue but had a different composition ([Cu 3 (C 6 S 6 )] n instead of [M 1.5 (C 6 S 6 )] n ; with additional metal ions in the interstitial coordination sites) (Fig. 3b,c). CuBHT exhibited metallic behavior, where the resistivity decreased upon refrigeration, dropping finally to zero resistance at 250 mK ( Fig. 3e). This report corresponded to the first layered MOFs featuring superconducting transition. Note that Zhao and Liu predicted theoretically that monolayer CuBHT should possess a critical temperature (T C ) of 4.43 K, while T C of bulk CuBHT should be 1.58 K [11]. Pedersen and Clérac reported on another type of conductive layered MOFs, CrCl 2 (pyz) 2 , the formula of which was CrCl 2 (pyrazine) 2 ( Fig. 4a-c) [19]. A simple reaction between CrCl 2 and pyrazine at 200°C produced microcrystalline CrCl 2 (pyz) 2 with a 2D structure network closely approaching the symmetry of a square lattice. Electrical conductivity measurements demonstrated that CrCl 2 (pyz) 2 reached a conductivity of 0.0032 S cm À1 at room temperature (Fig. 4d). The projected density of states (pDOS) indicated the frontier orbitals of the valence and conduction bands were dominated by C and N 2p states (Fig. 4e). Moreover, CrCl 2 (pyz) 2 underwent a ferrimagnetic order below 55 K. The conductivity and magnetism should be derived from the interaction between the redox active pyrazine ligand and the Cr center with reduction ability: for example, a series of analysis indicated the valence of the Cr center was closer to +3 than original +2.
After the first report by Nishihara and coworkers, a number of conductive BHT-based layered MOFs and MONs were reported by employing various metal ions such as cobalt [13], silver [20], gold [20], and platinum [21]. Nishihara and coworkers developed recently a series of mixed-metal layered MOFs (Ni x Cu 1-x /BHT) by blending two different metal ions, nickel and copper [22]. Surprisingly, Ni x Cu 1-x /BHT grown at a liquid/liquid interface showed improved crystallinity compared to pure NiBHT and CuBHT, revealing the preferential formation of the layered MOF with a metal composition of Ni/Cu = 1/2. Through a multifaceted structural analysis, the authors proposed a plausible model for the preferred structure of the mixed-metal layered MOF, in which the inplane vacancies of the NiDT network are filled with Cu 6 BHT units (Fig. 5a). As a result, a significant enlargement of the domain size was achieved, together with an increase in electrical conductivity ( Fig. 5b-e).

Proton conductivity
Proton conducting materials have been explored intensively with accelerating demands, such as solid electrolytes for fuel cells. To surpass conventional but still prevailing proton conductors such as Nafion Ò , various materials have been investigated, including a layered MOF presented by Kitagawa and coworkers (Fig. 6) [23]. They prepared a 350 nm-thick CuTCCP sheet (H 2 TCCP = 5,10,15,2 0-tetrakis(4-carboxyphenyl)porphyrin), which exhibited a high proton conductivity of 3.9 Â 10 -3 S cm À1 , superior to other hydrated 3D-MOFs. The prominent performance of the layered MOF may be attributed to dangling carboxyl groups and water molecules adsorbed to Cu open sites distributed on the surface of the layered MOF film, which served as a proton pathway in addition to those existing in the bulk moiety.

Exotic electrical properties
Memristors emerged around 2010 as a fourth circuit component that can store information more conveniently than conventional electronics technology. Thus far, memristors were demonstrated in metal-insulator-metal structures, mainly using metal oxides and chalcogenides in sandwich or stacked structures. In contrast, Zamora and Gómez-Navarro, in a new effort, proposed a distinctive approach that incorporated memristive layered MOF [Cu 2 I 2 (TAA)] n in planar device architectures (Fig. 7a,b) [5]. Most memristive devices to date possessed stacked configurations, which was a major obstacle to understanding the intrinsic mechanisms that produced memristive behavior and hindered controllability and reproducibility. In contrast, the Zamora and Gómez-Navarro devices had a planar configuration, which made it possible to elucidate the mechanisms of electrical conduction and resistive switching (i.e., memristive properties) beyond grain boundaries. Furthermore, devices with stacked MOF films exhibited inherent memristic behavior and high 2D electrical conductivity (up to 50 S cm À1 ) (Fig. 7c-e).

Magnetism
Layered MOFs and MONs may exhibit magnetism when there is appropriate interaction among spins derived from metal centers and/or ligand molecules. Of note, p-conjugated planar ligand molecules may enhance such magnetic interaction, which was verified by Harris and coworkers in a layered MOF system, [Fe 2 (Cha) 3 ] n ( Fig. 8a; H 2 Cha = chloranilic acid) [24,25]. Therein the ligand moiety bore an anion radical state (Cha 3-• ), thereby interacting magnetically with the Fe 3+ center, resulting in ferrimagnetism below 105 K (Fig. 8b,c). Its analogous layered MOF created by Long and coworkers with 4-fold symmetry also exhibited ferromagnetic behavior under 8 K [26]. Feng and coworkers reported a ferromagnetic planar layered MOF comprising a coronene derivative (Fig. 8d) [27]. The coronene bridge contributed to a p electron system delocalized over the 2D sheet motif, mediating ferromagnetic interaction among the Fe 3+ ions. This led to spontaneous magnetization below 20 K (Fig. 8e), and at the same time, a high electrical conductivity of 10 S cm À1 was realized. Such magnetic and electrical feature is ideal for a kind of spintronics application.
A distinctive approach, observations for the magnetic order of the atomically thin layers was performed by Espallargas and Coronado [28]. They fabricated MON MUV-1-Cl with a formula of [Fe(bimCl) 2 ] (HbimCl = 5-chlorobenzimidazole) (Fig. 9a,b). Bulk crystalline MUV-1-Cl underwent antiferromagnetic coupling between the high-spin Fe(II) center, with Curie-Weiss temperature, Néel temperature, and exchange parameter of, respectively, h = À80.6 K, T N = 20 K, and J = À 22.9 cm À1 (Fig. 9c,d). A 5.7-nm   thick MUV-1-Cl was then subjected to low-temperature magnetic force microscopy (LT-MFM; Fig. 9e). The MUV-1-Cl MON enjoyed attractive tip-sample interaction below T N (red region), while it was cancelled above T N . The series of LT-MFM results was associated with the antiferromagnetism in the bulk form.

Redox
Redox active layered MOFs and MONs may be prepared with the aid of metal centers with such ability. Nishihara and coworkers reported the redox activity of MONs 1-Fe and 1-Co featuring the bis(terpyridine)metal complex motif (Fig. 10a) [29]. The MONs were synthesized from three-way terpyridine ligand 1 and Fe(II) or Co(II) metal salt using a liquid/liquid interfacial reaction. Typical thickness of MONs 1-Fe and 1-Co were, respectively, 200 and 120 nm. The color of as-prepared MON 1-Fe was purple (Fig. 10c), which was derived from a metal-to-ligand charge transfer (MLCT) transition. The MON physisorbed on an electrode underwent a reversible redox reaction in an organic electrolyte solution (Fig. 10b), ascribable to the Fe 3+ /Fe 2+ redox couple. The redox reaction was fast with a faradaic current decay of 0.35 s, which was associated with a color change to yellow (Fig. 10c). The electrochromic behavior was persistent in a solidified device configuration (Fig. 10d,e). Finally, the authors fabricated a solidified dual electrochromic device using both MONs 1-Fe and 1-Co as active materials (Fig. 10f). The cobalt center of MON 1-Co experienced a reversible Co 2+ /Co + redox reaction with a color change between orange (Co 2+ ) to blue (Co + ). As a result, the dual electrochromic system featured operation shown in Fig. 10g. Some metal complexes undergo ligand-centered redox reactions. As mentioned in section 2.2, NiDT featured a conductivity change, which was in fact induced by the ligand-centered redox reaction localized on the dithiolato ligand part.

Luminescence
Luminescent films are of interest to researchers because of their potential use in displays, electronic and optical devices, lightemitting diodes, solar cells, photodetectors, flat panel displays, chemical and biological sensors, and many other applications. Among light-emitting materials, soluble metal complexes based on closed-shell d 10 metals have shown attractive processability and photophysical properties. Sakamoto and Nishihara replaced the metal centers of MONs 1-Fe and 1-Co mentioned in Section 2.6 with Zn 2+ to produce MON 1-Zn (Fig. 11a) [30]. This modification resulted in the loss of the redox ability, but at the same time gained luminescence property. The 65-nm-thick MON 1-Zn was colorless and transparent, but upon UV irradiation it showed blue emission centered at 480 nm (Fig. 11b,c). Thus, it was demonstrated that functional MONs can be diversified by the design and selection of ligands and metal centers. The luminescence was attributed to  Fig. 11d,f,g).When G1@1-Zn was excited with UV light, the red emission from the guest dye was found to be dominant over the native blue emission from the 1-Zn backbone ( Fig. 11e,h). Comparing the luminescence quantum yields of the host 1-Zn and the guest G1 dye, the authors concluded that a quasi-quantitative energy transfer from the former to the latter took place. Crystalline, freestanding and luminescent silver thiocarboxylate layered MOF thin films were reported by Zamora et al [31]. In this study, they demonstrated a simple and green procedure for the direct synthesis of crystalline two-dimensional coordination polymers with a formula of [Ag(TB)] n at room temperature ( Fig. 12a; TB = thiobenzoate). The layered MOF material exhibited electrical semiconductivity and intense luminescence property. Note that silver coordination polymers are usually non-luminescent. Furthermore, upon cooling from 300 K to 110 K, the material showed a green to yellow thermochromic behavior (Fig. 12b). The crystalline layered MOF thin film was prepared by a simple interface chemistry technique. The generated thin films could be transferred and supported on various surfaces and even isolated as freestanding films. The thermochromic emission observed in the bulk was retained in the thin film form (Fig. 12c).
Amo-Ochoa and coworkers synthesized a MON of [Cu 2 I 2 (2aminopyrazine)] n with an average thickness of 51 ± 21 nm (Fig. 13a) using ultrasonic liquid-phase exfoliation, and the MON showed an emission response to temperature (Fig. 13b) [32]. The thermochromism was investigated by confocal microscopy and AFM of nanosheets deposited on SiO 2 (Fig. 13c,d). At 300 K, emission from the MON was observed at around 550 nm upon excitation with 351 and 364 nm with a laser power of 9.36 lW ( Fig. 13e). At the same temperature, emission was no longer observed when the laser power was decreased to 6.34 lW ( Fig. 13f), but it was revived at the same laser power when the temperature dropped to 80 K (Fig. 13g). The authors suggested that this series of thermochromic behaviors should lead to nanosensors that are low cost and easy to fabricate.
Furthermore, Zamora and coworkers reported that a solvothermal reaction of Cu(BF 4 ) 2 ÁxH 2 O and 4-mercaptobenzoic acid, a commercially available inexpensive building block, produced two types of Cu(I)-thiophenolate-based coordination polymers (MOF1 and MOF2; Fig. 14a) [33]. The microcrystals of MOF1 and MOF2 featured lamellar structures (Fig. 14b), which were readily dispersible and integrated with polyvinylidene fluoride (PVDF) to form homogeneous, reversible thermochromic thin films. The thermo-stimuliresponsive thin film of MOF1 was flexible, freestanding, free of macroscopic defects, and resistant to mechanical bending stress (Fig. 14c,d). This set of attractive properties suggested the potential of layered MOFs as 2D imaging sensors.

Hydrogen evolution reaction (HER)
HER comprises one of the most fundamental but important reaction steps in electrochemical water splitting. From the viewpoint of application, HER is expected to play a vital role in the hydrogen economy, converting electricity into hydrogen as chemical energy and fuel. Feng and coworkers demonstrated [34] the HER catalytic activity of a single-layer MON based on the bis(dithiolene)metal complex motif known to show HER activity [35] (Fig. 15). The authors employed Ni and triphenylenehexathiol (THT) as the metal center and ligand (THTNi; Fig. 15a), and gas/liquid interfacial synthesis [36] was adopted to fabricate the single-layer MON. A series of comprehensive analyses was provided: For example, AFM quantified the thickness of the singlelayer THTNi as 0.7 nm (Fig. 15b), while a freestanding nanosheet domain with a lateral size of $100 lm was disclosed in TEM ( Fig. 15c). The HER activity of THTNi was investigated by linear sweep voltammetry (LSV) using a glassy carbon (GC) rotating disk electrode modified with the MON (Fig. 15d,e). In 0.5 M H 2 SO 4 aq, hydrogen started to evolve at an overpotential of ca. 110 mV, above which the current density associated with H 2 production increased steeply, and the operating potential at 10 mA cm À2 was attained at 333 mV. The Tafel slope was 80.5 mV decade -1 , indicative of the Volmer type reaction where proton adsorption onto the modified electrode surface is the rate-determining step. The exchange current density (i 0 ) was calculated to be ca. 6 Â 10 -4 mA cm À2 . In sharp contrast, the bare GC electrode featured a negligible current density in the potential sweep range. THTNi was also catalytically active in other electrolyte solutions, such as 0.025 M H 2 SO 4 aq (with 413 mV overpotential to achieve 10 mA cm À2 ) and 0.05 M KOH aq (with 574 mV overpotential). The series of HER performance of THTNi went one better than molecule-based catalysts supported on carbon nanotubes and N-, P-, or S-doped graphene. Prior to this report, Marinescu clarified the HER activity of Co analogues with benzene or triphenylene linkers [37]. The materials were synthesized by liquid/liquid interfacial synthesis [36], with a typical thickness of 360 ± 40 nm. Their HER activity was better in acidic media, with overpotentials of 340 and 530 mV to reach current densities of 10 m A cm À2 at pH 1.3. Nishihara and coworkers demonstrated a 10 mA cm À2 overpotential of 370 mV at pH 1.3 with a [Ni(NH 2 ) 2 S 2 ] type of analogous MON [17].

Oxygen evolution reaction (OER)
OER is a counterpart of HER in water splitting. The complex four-electron process in OER makes it difficult to design catalysts suitable for OER, and resultantly OER is being bottlenecked with large overpotentials and slow reaction rates in various applications, such as hydrogen production, metal-air batteries, and artificial photosynthesis. Zhou and Zhang demonstrated that a pillaredlayer 3D MOF was exfoliated into a MON form with oxidation stimuli, and the resultant MON featured better OER activity (Fig. 16) [38]. The 3D MOF (3D-Co) had a composition of (H 3 O) 2 10 ]. On the other hand, The [Co 6 O(dhbdc) 2 ] 2+ 2D layer structure was still retained in 2D-Co, while the interlayer pillar was removed (Fig. 16b,d,f). Further transformation of 3D-Co proceeded in 0.1 M KOH aqueous solution (pH 13) at room temperature for 2 h, giving rise to an amorphous phase, 2D-Co-NS. This transformation was faster with anodic polarization for microcrystals of 3D-Co grown onto a Co film (3D-Co@Co) in 0.1 M KOH aq This state was in fact a fully exfoliated MON form (Fig. 16h-j), with a minimum thickness of $2 nm. The authors ascribed the oxidation-prompted electrochemical or chemical exfoliation to oxidation of the interlayer pillar in 3D-Co, from linear H 4 dhbdc to oxocyclopenta-3,5-diene-1,3-dicarboxylic acid with a bent bridging structure that should destabilize the pillared-layer 3D MOF motif. The OER activity of 3D-Co-or 2D-Co-NS-related materials are summarized in Fig. 16k,l. In comparison with 3D-Co@Co and 2D-Co-NS@Co, the superiority of the MON form was confirmed in terms of OER. The 2D-Co MON supported on Ni foam (2D-Co-NS@Ni) featured an OER performance better than 2D-Co-NS@Co. The best OER performance was achieved with Fe doped 2D-Co-NS on Ni foam (Fe:2D-Co-NS@Ni), which was fabricated by electrolysis in an aqueous solution of FeSO 4 . It boasted low overpotential and Tafel slope of 211 mV at 10 mA cm À2 and 46 mV decade -1 . Furthermore, Fe:2D-Co-NS@Ni was confirmed to show longterm stability (electrolysis for 96 h at 10 mA cm À2 increased the overpotential only by 10 mV), good Faraday efficiency of 99 % at 10 mA cm À2 , and TOF of 30 ± 2 s À1 with an overpotential of 300 mV and optimized catalyst loading. There are several other reports on OER performances for layered MOFs and MONs [39][40][41][42]. Among those, Lin and coworkers demonstrated that a NiFe-based MON, MIL-53(FeNi), which was in situ grown on Ni foam, exhibited high current density (50 mA cm À2 ) at an overpotential of 233 mV and shallow Tafel slope of 31.3 mV decade À1 , with excellent stability in alkaline aqueous solution (1 M KOH) [39].

Other reactions
Dincȃ and coworkers fabricated an analogue of THTNi (Fig. 15a), where the S atoms are substituted with NH groups [43]. This analogue served as an oxygen reduction reaction (ORR) electrocatalyst, which is indispensable for fuel cells. Chen and Zhao achieved overall electrochemical water splitting using a NiFe heterometal MON which contained 2,6-naphthalenedicarboxylate ligand molecules, and was formed directly on Ni foam as a scaffold from Ni(OAc) 2 , Fe(NO 3 ) 3 , and the organic ligand [44].

Gas separation
Membrane-based gas separation offers an environmental friendly and energy-efficient way to purify gaseous materials. However, this method often faces an intrinsic problem, a trade-off limitation between gas permeability and selectivity. Recently, layered MOFs have been reported in which high gas permeance and selectivity were compatible with each other [45][46][47][48]. Yang and coworkers synthesized layered MOF crystals, [Zn 2 (benzimidazole) 3 (OH)(H 2 O)] n (abbreviated as Zn 2 (Bim) 3 , Fig. 17a-c), which was turned into a nanosheet form using a soft-physical process [46]: Zn 2 (Bim) 3 was first ball-milled in a mixture of methanol and n-propanol. This process afforded exfoliated bilayer MON Zn 2 (Bim) 3 with a thickness of 1.6 nm (Fig. 17d). The nanosheet was then deposited onto an a-Al 2 O 3 disk as a solid support, which was annealed at 200°C (Fig. 17e), and was subjected to gas permeation tests for H 2 (kinetic diameter: 0.289 nm), CO 2 (0.33 nm), N 2 (0.36 nm), and CH 4 (0.38 nm). Therein, only H 2 exhibited a high permeance: The H 2 /CO 2 ideal selectivity reached 135, which was comparable to the binary gas separation factor, and exceeded the Knudsen selectivity, with H 2 and CO 2 permeance of 1.1 Â 10 -6 mol m À2 s À1 Pa À1 and 8.1 Â 10 -9 mol m À2 s À1 Pa À1 , respectively (Fig. 17f). This selectivity was associated with the aperture size ($0.29 nm) of the honeycomb structure of Zn 2 (Bim) 3 . Zhao and coworkers reported a layered MOF, Ni 8 (5-bbdc) 6 (l-OH) 4 (bbdc: 5-tert-butyl-1,3-benzenedicarboxylate), which exhibited a thermally switchable gas permeance performance (Fig. 18a) [48]. The MON membrane of Ni 8 (5-bbdc) 6 (l-OH) 4 exhibited good permeance for He and H 2 , while that of CO 2 , O 2 , N 2 and CH 4 was poor. The gas permselectivities were quantified to be 268 for H 2 /CO 2 , 96 for H 2 /O 2 , 123 for H 2 /N 2 and 164 for H 2 /CH 4 , respectively. The authors ascribed the selectivity to the hydrophilic inner wall of the gas pathway (PW2, Fig. 18b), which afforded different affinity to each gas molecules (Fig. 18c). Of note, the MON membrane exhibited thermo-switchable H 2 permeation behavior caused by the flexibility of the framework (Fig. 18d). At around room temperature, tert-butyl groups around the pathway (PW1, Fig. 18b) could rotate freely, which allowed H 2 to permeate the MON membrane. In contrast, at higher temperatures, the interlayer distance was reduced, which hampered the free rotation of the tert-butyl group due to the steric hindrance.

Nanofiltration
Layered MOFs and MONs are also applicable to separation of organic molecules and nanoparticles. Hong reported the nanofiltration performance of 2D MÀTCP(Fe) (M = Zn, Co and Fe) that comprised Fe(III) meso-tetra(4-carboxyphenyl)porphyrin chloride (TCP(Fe)) and metal ions (Fig. 19a), and was used for removing organic dye molecules from water [49]. Four organic dyes with different molecular sizes and charges were prepared, methyl red (MR, 0.8 Â 1.1 nm, neutral), methylene blue (MnB, 1.4 Â 1.9 nm, positively charged), methyl orange (MO, 1.0 Â 1.2 nm, negatively charged), and brilliant blue G (BB, 1.6 Â 1.9 nm, negatively charged) as probe solutes in water. Pristine MÀTCP(Fe) MONs with thicknesses of ca. 65-70 nm could achieve high rejection rates of the solute greater than 90%, for MR, MO, and BB (Fig. 19b-d). In contrast, none of them could remove MnB efficiently (<50 %) because the negatively charged surface of 2D MÀTCP(Fe) confirmed by a zeta-potential measurement induced an attractive force between cationic MnB, which let the latter to slip through the nanochannel of the membrane. To solve the drawback, polycation-regulated 2D Zn-TCP(Fe) was synthesized by mixing a suspension of 2D Zn-TCP(Fe) and polycationic polymer solutions   (PEI: polyethylenimine and PDDA: poly(diallyldimethylammonium chloride 0.1 wt%; Fig. 19a). The free carboxylic groups at the peripheral edge of the 2D Zn-TCP(Fe) MON interact ionically with polycation coils to form polycation-regulated assemblies. The separation performance of the polycation-regulated 2D Zn-TCP(Fe) was in fact improved (Fig. 19e): the 48-nm-thickness membrane achieved the rejection rate of 90% for all organic dyes, with water permeance up to $2 times greater than pristine Zn-TCP(Fe).
Ruoff and coworkers synthesized a honeycomb layered MOF composed of tris(b-diketone) ligand molecules and Cu ions (Fig. 20a) [50]. The MON layer was stacked with each other in a staggered fashion, and a membrane prepared by depositing the layered MOF powder on a supporting Anodisc filter performed size-selective filtration of gold nanoparticles with a cut-off of ca. 2.4 nm (Fig. 20b).

Lithium ion batteries (LIBs)
Despite the great success of LIBs, there are still many pieces of vigorous research on them, seeking for further capacity enlargement and superior rate capability during the charge/discharge  process to be employed in not only electronics gadgets but also electronic vehicles and so on. In line with this tread, layered MOFs and MONs have been employed as active electrode materials for LIBs [51][52][53][54][55][56], taking advantage of their porous and layered structures. One of the examples was contributed by Sakaushi and Nishihara [52], where a conductive bis(diimino)nickel framework (NiDI) was utilized (Fig. 21a). As shown in Fig. 21b, the neutral form of a mononuclear model complex features reversible redox reactions in both oxidation and reduction directions, with which the authors anticipated enhanced LIB capacity by counter ion uptake (Fig. 21a). Fig. 21c displayed a close-up TEM image of NiDI, with a hexagonal periodicity consistent with the expected 2D lattice shown in Fig. 21a. A typical crystalline domain size was 30 nm in diameter. Cyclic voltammetry was performed to confirm the redox activity of NiDI. The redox couple associated with anodic and cathodic peaks at 3.73 and 3.56 V vs Li + /Li, respectively, was assignable to a transition between the neutral and positively charged (oxidized) states of NiDI accompanied by the insertion/desertion of PF 6 -. On the other hand, a cathodic peak at 3.21 V coupled with a broad anodic peak was ascribed to a transition between the neutral and negatively charged (reduced) state of NiDI, which was followed by Li + -insertion/desertion. The specific capacity was calculated from the voltammograms; the capacity with a negative cut-off potential of 2.0 V showed more than 100 mAh g À1 in both positive and negative scans, while it decreased to ca. 40 mAh g À1 with a potential window of 3.0-4.5 V. Electrochemical impedance spectroscopy quantified the charge transfer resistance and ion diffusion constant. A coin cell was fabricated, and the performance of the NiDI cathode was further investigated by charging/discharging with current densities of 10-500 mA g À1 (Fig. 21e). During the cell test, two sets of plateau regions were found in the potential range of 2.0-4.3 V, which should stem from the insertion/desertion of Li + or PF 6 -, as mentioned in cyclic voltammetry. The specific capacity relied on charging/discharging current density, but at 10 mA g À1 the cell achieved a specific capacity of 155 mAh g À1 , corresponding to a specific energy density of 434 Wh kg À1 . The cell featured one of the highest specific capacity among MOF-based cathode materials, and comparable to commercially available LiCoO 2 and LiFePO 4 . The cell showed good durability up to 300 cycles at 250 mA g À1 , and the coulombic efficiency was mostly more than 99 % (Fig. 21f).

Supercapacitors
Supercapacitors, or electric double-layer capacitors (EDLCs), exploit charging to, and discharging from the electric doublelayer formed close vicinity to an electrode surface. Short charging/discharging duration derived from the intrinsic fast process and low internal resistance is advantageous for EDLCs over LIBs, while the realization of EDLCs with greater capacities is yet to be solved. Electrode materials for EDLCs are required to possess rich porosity and good conductivity, and there are several attempts to adopt layered MOF and MONs to such materials [57][58][59].
Dincȃ and coworkers demonstrated that Ni 3 (HITP) 2 served as an electrode material for an EDLC (Fig. 22a) [57]. This material features high conductivity of 5000 S m À1 , and one-dimensional cylindrical channels of $1.5 nm diameter, which penetrate the 2D lattice and afford a BET surface area of 630 m 2 g À1 (Fig. 22b). This series of characteristics is desirable for EDLCs; in fact, Et 4 N + , Et 4 N + -Á7CH 3 CN, BF 4 -, and BF 4 -Á9CH 3 CN derived from Et 4 NBF 4 and CH 3 CN used in EDLCs may be incorporated into the channels (Fig. 22b), considering their ionic diameters, 0.68 nm, 1.3 nm, 0.46 nm and 1.16 nm, respectively. Ni 3 (HITP) 2 powder pressed into a Pt-mesh working electrode was subject to cyclic voltammetry in 1 M Et 4 -NBF 4 /CH 3 CN (Fig. 22c). No faradaic process was found in a potential range from À0.6 V to 0.5 V, thereby quantifying a working potential window of ca. 1.0 V for Ni 3 (HITP) 2 . Then, a twoelectrode symmetrical supercapacitor cell was fabricated, where Ni 3 (HITP) 2 powder was pelletized at a pressure of 100 kg-force cm À2 . Cyclic voltammetry and galvanostatic charge and discharge of the symmetrical cell exhibited, respectively, rectangular-and triangular-shaped charts, which are typical of capacitive behavior (Fig. 22d,e). The specific capacitance of the Ni 3 (HITP) 2 EDLC was calculated from galvanostatic discharge curves, and at a low discharge rate of 0.05 A g À1 the capacitance reached 111 F g À1 , which was in the same range as activated carbons. The capacitance was deteriorated with greater discharge rate, which the authors attributed to decreased ion accessibility to the pore surface. After 10,000-cycle charging/discharging between 0 V and 1 V at 2 A g À1 , the EDLC device experienced a capacitance reduction by only 10 %.

Photocatalytic hydrogen evolution
Photocatalytic water splitting using particulate semiconductor photocatalysts is considered a promising way to produce sustainable hydrogen from solar energy, and to break away from dependence on fossil fuels. This process comprises both HER and OER reactions, and the loading of a cocatalyst on the surface of semiconductor photocatalysts is frequently used to promote HER. As HER cocatalysts, metals and metal oxides such as Pt and RuO x have been used typically, while molecular cocatalysts are seldom employed due to their low stability against redox reactions and light illumination. An overall photocatalytic water splitting was demonstrated by Sakamoto and Abe with the help of conductive  layered MOF, NiBHT (Fig. 23a) as a cocatalyst of HER [60]. NiBHT/ CoO x /SrTiO 3 :Al (CoO x : cobalt oxide as an OER cocatalyst; SrTiO 3 :Al: aluminum-doped SrTiO 3 ) composite underwent photocatalytic water splitting with a long term stability (no degradation after 120 h) as well as an apparent quantum efficiency (AQE) of 6.5 % at 350 nm (Fig. 23b). In contrast, the control experiment with Pt as a HER cocatalyst (Pt/CoO x /SrTiO 3 :Al) showed the gas production stopped apparently as H 2 and O 2 were accumulated, due to the coexistence of the unfavorable backward reaction (H 2 O H 2 + 1/2-O 2 ). According to the theoretical investigation, NiBHT turned out to suppress the backward reaction, which afforded the sustainable gas evolution.

Others
Sakamoto and Nishihara fabricated a MON featuring the bis (dipyrrinato)zinc complex motif [36]. The complex motif has intense absorption in the visible region, which allowed the nanosheet to absorb visible light (450-550 nm), leading to a photoelectric conversion function, just as 1D counterparts performed [61][62][63]. Flexible structural design is one of the greatest virtues in molecule-based nanosheets, which also allows us to reinforce their functionality. Taking advantage of the strong point, the authors customized the dipyrrin MON to enhance the photoelectric conversion ability, materializing a chessboardlatticed MON comprising a porphyrin-dipyrrin hybrid ligand [64]. The MON exhibited a broader photoelectric conversion response covering the whole visible region (400-650 nm), with a maximum quantum efficiency more than twice as much as that of the prototype. 6. Sensors

Biosensors
Nanosheets capable of interacting with biomolecules may be utilized as key platforms of biosensing [65,66]. Xia and coworkers prepared lanthanide-based MONs, such as MOF-La (Fig. 24a), which was modified with single-strand DNAs (ssDNAs) [65]. Dyelabeled ssDNAs (TAMRA-P1, FAM-P1) as luminescent probes were adsorbed on MOF-La, and when a target ssDNA, T1, was added to TAMRA-P1+MOF-La and FAM-P1+MOF-La, the MON conjugates underwent luminescence spectral changes derived from the probe dyes (TAMRA-P1+MOF-La: increase at 580 nm; FAM-P1+MOF-La: decrease at 520 nm; Fig. 24b,c). The hybridization with T1 and resultant formation of double-stranded DNAs (dsDNAs) led to the liberation of TAMRA-P1 and FAM-P1 from the MOF-La surface and charge transfer luminescence quenching (Fig. 24d). This should provide the dye probes with luminescence enhancement, however, the negatively charged carboxyl and phenolic hydroxyl groups on FAM bound the molecule on the edge of MOF-La (Fig. 24d), leading to further quenching. The authors exploited the series of optical phenomena to produce a two-color intracellular adenosine imaging in living cells (Fig. 24e-g).

Chemiresistive sensors
Chemiresistive sensors are a kind of chemical sensors that detect analytes as changes in electrical resistances caused by interaction between the analytes and conductive materials. Conductive layered MOFs were applied as active materials for chemiresistive sensor systems for the detection of gases and volatile organic compounds by Dincȃ and coworkers [67][68][69]. Cu 3 (HITP) 2 ( Fig. 25a; HITP = 2,3,6,7,10,11-hexaiminotriphenylene) with a bulk conductivity of 0.2 S cm À1 could be incorporated into a chemiresistive sensing system for NH 3 gas (Fig. 25b) [67]. The sensor prepared by dropcasting Cu 3 (HITP) 2 dispersion detected sub-ppm levels of NH 3 gas, which was accompanied by a linear conductivity-gas concentration relationship in the 0.5 to 10 ppm region, being allowed for quantitative detection (Fig. 25c,d). Moreover, the sensor displayed a good response even under an ambient atmosphere and high humidity conditions. Cu 3 (HITP) 2 , Ni 3 (HITP) 2 , and Cu 3 (HHTP) 2 (an analogue of Cu 3 (HITP) 2 ; Fig. 25a; HHTP = 2,3,6,7,10,11-hexahy droxytriphenylene) turned out to be capable of sensing various volatile organic compounds such as alkanes, alcohols, ketones/ ethers, aromatics and amines (Fig. 25e) [68]. The conductive layered MOFs exhibited different sensing responses against the five  categories of organic compounds based on the functional groups, distinguishing them resultantly (Fig. 25f). A one-step fabrication method was developed for textile-supported devices based on Ni 3 (HHTP) 2 and Ni 3 (HITP) 2 [69]. The modified textile displayed electrical conductivity, enhanced porosity, flexibility and stability to washing. The chemiresistive device using the textile could detect ppm-level NO and H 2 S semiquantitatively with its function maintained under a condition of 18 % relative humidity (Fig. 25g,h). The theoretical limits of the NO and H 2 S detection were estimated to be, respectively, 0.16 and 0.23 ppm.
The selective detection of target neurochemicals in a multianalyte solution remains a challenge due to interference from other analytes. Mirica and coworkers reported the selective detection of neurochemicals, ascorbic acid (AA), dopamine (DA), uric  Fig. 27a) [71]. These chemicals play significant roles in human physiology, such as mental, metabolic, and nutritional health. The authors demonstrated the selective detection of DA and 5-HT in the presence of crucial interferents, such as AA and UA. In addition, the authors quantified DA and 5-HT by using differential pulse voltammetry (DPV) technique (Fig. 27b,c)), and achieved a high analyte sensitivity with the nanomolar detection limit (63 nM and 40 nM for DA and 5-HT, respectively) with a Ni 3 HHTP 2 modified electrode, which is much higher than an unmodified glassy carbon electrode.      [77,78]. They are half metals, in which all carrier electrons are spin-polarized perfectly in one orientation (Fig. 28b,c).

Topological phases
In the latest decades, topological phases have collected significant attention in the research field of condensed matter physics [79]. Topological insulators (TIs) occupy an important series of the topological phase, featuring insulating bulk parts but and spin-polarized metallic edge or surface states. Thus, TIs are expected to be key materials in next-generation spintronics.
The verification of the TI state has been demonstrated using inorganic materials exclusively, such as bulk Bi 2 Te 3 . On the other hand, 2D TIs are much rarer, for example, a quantum well composed of HgTe/CdTe and a Bi bilayer on Bi 2 Te 3 . Besides, organicbased TI systems have remained to be unexplored. Within this context, Liu and coworkers predicted virtual organometallic 2D TI lattices [80]. Then, single-layer NiDT (Fig. 3a) was reported to be the first existing candidate for organic 2D TIs by the authors [81]. As shown in Fig. 29a,b, first-principles calculations gave a semiinfinite single-layer of NiDT a flat band and a pair of Dirac bands lying above the Fermi level. Therefore, proper electron doping is required for NiDT to behave as a TI. Furthermore, the topological edge states was also confirmed (Fig. 29c,d). Band gaps D 1 (13.6 meV) and D 2 (5.8 meV) stemmed from the spin-orbit coupling of the nickel center, and an analogue of NiDT with the heavier Au center exhibited greater band gaps (22.7 and 9.5 meV). The greater band gap is useful for the preservation of the TI phase at higher temperatures. Nishihara and coworkers synthesized a Pt analogue of NiDT (PtDT), including at the same time theoretical calculation that predicted its potential 2D TI property [21].

Conclusions and outlook
This review article is devoted to assembling recent topics on layered MOFs and MONs, especially focusing on their functionalities and potential applications. Layered MOFs and MONs have collected prominent interest of researchers, which has been reflected in the steep increase in the number of relating publications. A part of layered MOFs and MONs enjoys physical properties such as electrical conductivity, memristive behaviour, proton conductivity, magnetism, redox, and luminescence: such basic properties allow layered MOFs and MONs to be key active materials for a wide series of applications. Electrocatalysis is one of the representative examples, where electrical conductivity is beneficial and concerts with the metal centres existing in layered MOFs and MONs trivially. Porosity is also a distinctive factor for layered MOFs and MONs, which plays important roles in the separation and filtration of gaseous or nano-sized matters. In this review article, other important applications are reported, including energy devices (LIBs, supercapacitors, photoelectric conversion), sensors (biosensors, chemiresistive sensors, electrochemical sensors). Theoretical approaches are also introduced, seeking for unveiled properties such as peculiar spin states and topological phases.
As mentioned and discussed above, we may recognize the bright future of layered MOFs and MONs as functional nanomaterials; however, still there is large room left for unexplored topics [82]. The ultimate applications for layered MOFs and MONs are those exploiting them as single-or few-layer forms. In contrast, most current applications utilized them as bulk forms, or with unidentified stacking patterns or layer numbers. Moreover, especially for single-or few-layer MONs, their quality (i.e. crystallinity, degrees of disordered or defect sites) has been seldom quantified: in fact, practical analytic methods to evaluate layered MOFs and MONs quality are not sufficient. This series of difficulties, however, does not deteriorate the allure of layered MOFs and MONs; instead, it will stimulate future challenges therein.
An unexplored stage that will enable the application of layered MOFs and MONs is the synthesis of high-quality samples (large single crystals and ultra-high purity). This is because historically, many innovations in materials science have occurred when extremely high-quality samples have been obtained. The realization of high-quality sample preparation of layered MOFs and MONs is directly related to the identification of intrinsic parameters of physical and chemical properties vital for basic science and the expression of high performance important for applications. Another challenge is to fabricate heterostructures (heterolayers and transverse heterojunctions) of different layered MOFs and MONs. Progress in this challenge using high-quality layered MOFs and MONs will expand the field of basic and applied research in materials science, chemistry, physics, and electronics.

Data availability
No data was used for the research described in the article.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.