image
Home Journals Progress in Chemistry
Progress in Chemistry

Abbreviation (ISO4): Prog Chem      Editor in chief: Jincai ZHAO

About  /  Aim & scope  /  Editorial board  /  Indexed  /  Contact  / 
Online first  /  Current issue  /  All issues  /  Top access  /  RSS

Progress in Chemistry >

2026 , Vol. 38 >Issue 3: 502 - 531

DOI: https://doi.org/10.7536/PC20251208

Review

Metal-Organic Frameworks as Next-Generation Chemiresistive Sensors： From Fundamental Design to Advanced Applications

  • Shihzad Shakil 1 ,
  • Fan Wang , 2, * ,
  • Pengyu Sun 1 ,
  • Zhihan Zhou 1 ,
  • Xiaojing Lu 1 ,
  • Jun Du , 1, * ,
  • Jiarui Huang , 1, *
Expand
  • 1 College of Chemistry and Materials Science， Anhui Normal University， Wuhu 241002， China
  • 2 School of Materials Science and Engineering， Tongling University， Tongling 244000， China
* （Fan Wang）；
（Jun Du）；
（Jiarui Huang）

Received date: 2025-12-11

  Revised date: 2026-01-14

  Online published: 2026-03-21

Supported by

Scientific Research Project of Anhui Higher Education Institution(2024AH051843)

Scientific Research Project of Anhui Higher Education Institution(2024AH051839)

Fold

Abstract

Accurate and real-time sensing is fundamental to advancements in health diagnostics， environmental monitoring， and industrial safety. However， conventional sensing materials such as metal oxides， conducting polymers， and carbon-based composites are constrained by intrinsic trade-offs between sensitivity， selectivity， and operational stability. To address these limitations， metal-organic frameworks （MOFs） have emerged as a transformative class of materials， offering unparalleled structural tunability， ultrahigh surface areas， and programmable pore chemistry. This comprehensive review provides an in-depth analysis of MOF-based chemiresistive sensors， moving beyond a simple catalog of examples to establish a mechanistic understanding of how molecular-level design dictates sensing performance. We systematically deconstruct the evolution from often-insulating pristine MOFs to advanced composites where MOFs synergize with conductive fillers like graphene， carbon nanotubes， and polymers and to MOF-derived porous carbons and metal oxides. Each category is critically examined to highlight strategies for overcoming inherent challenges in electrical conductivity， response kinetics， and long-term stability. The review is structured to guide the researcher in the field from fundamental design principles and charge transport mechanisms to performance benchmarking against key metrics such as sensitivity， limit of detection， selectivity， and response/recovery times. A significant focus is placed on the integration of MOFs into next-generation applications， including flexible and wearable electronics， multi-parameter sensor arrays， and intelligent systems that leverage artificial intelligence for pattern recognition and drift compensation. Furthermore， we critically address the pivotal challenges hindering practical deployment， such as hydrothermal/chemical stability， mechanical robustness for wearable formats， and the urgent need for standardized testing protocols. By synthesizing insights from fundamental research and cutting-edge applications， this review serves as a rational design guide and a forward-looking perspective， outlining a concrete roadmap for harnessing the full potential of MOFs in the development of intelligent， reliable， and commercially viable next-generation chemiresistive sensing technologies.

Contents

1 Introduction

2 Design principles of MOFs

3 MOFs in chemiresistive sensing

3.1 Basic principles

3.2 Pristine MOF-based sensors

3.3 MOF composites and hybrids

3.4 MOF-derived materials

3.5 Challenges in stability and practical deployment

4 Future outlook & emerging applications

4.1 Integration with novel transduction mechanisms and flexible electronics

4.2 The path to intelligent and cognitive sensing systems

4.3 Multi-modal and extreme-performance sensing

4.4 The path to commercialization

5 Conclusion and future perspectives

Key words： chemiresistive sensors; metal-organic frameworks （MOFs）; MOF composites; MOF-derived materials; selectivity; wearable sensors; artificial intelligence

Cite this article

Shihzad Shakil , Fan Wang , Pengyu Sun , Zhihan Zhou , Xiaojing Lu , Jun Du , Jiarui Huang . Metal-Organic Frameworks as Next-Generation Chemiresistive Sensors： From Fundamental Design to Advanced Applications[J]. Progress in Chemistry, 2026 , 38(3) : 502 -531 . DOI: 10.7536/PC20251208

1 Introduction

Accurate， real-time sensing is critical for advancements in health diagnostics， environmental monitoring， and industrial safety. Despite decades of innovation， the performance of conventional sensing materials such as metal oxides， conducting polymers， and carbon-based composites is constrained by intrinsic trade-offs between sensitivity， selectivity， and operational stability[1]. For instance， metal oxides often require elevated temperatures and lack chemical specificity， while polymeric and carbon materials， despite better processability， can suffer from signal drift， fouling， and poor reproducibility.
This perspective is reinforced by a recent comprehensive review， which underscores that while traditional semiconductors and two-dimensional materials have been foundational for chemiresistive gas sensors， they remain fundamentally limited by poor selectivity and high operational temperatures[2]. The need for high-performance sensing is further amplified by emerging demands for ultratrace detection， portable and wearable formats， and operation in complex chemical environments. These requirements necessitate a class of materials that combines molecular precision in analyte recognition with robust structural integrity and a capacity for rational design criteria that traditional materials rarely satisfy simultaneously.
In this context， metal-organic frameworks （MOFs） have emerged as a transformative platform. MOFs are crystalline porous materials formed by the coordination of metal ions or clusters with organic linkers. This hybrid composition endows them with exceptional structural diversity， ultrahigh surface area， and intrinsic functionality[3]. A defining feature of MOFs is their programmable modularity， which enables precise control over pore size， shape， and chemical environment at the molecular level[4]. This tunability facilitates the design of frameworks with targeted host-guest interactions for specific analytes. Their immense porosity allows for the pre-concentration of analyte molecules， thereby dramatically enhancing sensor sensitivity[5].
Furthermore， coordinatively unsaturated metal centers， or open metal sites （OMS）， provide specific， high-affinity binding sites that directly address selectivity challenges. The inherent catalytic and redox activity of many metal nodes can also amplify sensing signals by catalyzing reactions of target analytes[6]. Collectively， these attributes position MOFs as premier next-generation sensing materials capable of bridging the long-standing gap between structural design and sensing performance[7] （Fig. 1）.
图1 MOF基化学电阻传感器的路线图：展示了从MOF基本结构单元（金属节点和有机连接体）到各种传感材料（原始MOF、复合材料、衍生材料）的构建过程，及其在可穿戴设备和人工智能电子鼻等先进智能应用中的集成

Fig.1 Roadmap of MOF-based chemiresistive sensing. This schematic illustrates the journey from fundamental MOF structural components （metal nodes and organic linkers） to the creation of various sensing materials （pristine MOFs， composites， derived materials） and their integration into advanced， intelligent applications such as wearable devices and AI-enabled E-noses

Despite the exponential growth of literature on MOFs， many reviews offer a broad overview of their applications in catalysis[8]， gas storage[9]， and sensing[10] without delving deeply into the specific mechanisms and material design rules for a particular sensing modality[11]. To address this gap， the present review provides a focused， in-depth analysis of MOF-based chemiresistive sensors. We concentrate specifically on the electrical resistance/conductance transduction mechanism， which is paramount for developing low-cost， portable， and miniaturizable electronic devices. Moving beyond a simple catalog of examples， this work critically analyzes how fundamental MOF design principles， including the choice of metal secondary building units， organic linker functionality， and framework topology， dictate chemiresistive behavior.
We systematically deconstruct charge transport pathways in both pristine and composite MOFs， exploring strategies to overcome their inherent insulating nature. The discussion is structured to guide the reader from fundamental principles to advanced applications， covering the performance of pristine MOFs， the enhanced functionality of MOF composites and hybrids， the utility of MOF-derived materials， and a forward-looking perspective on the integration of MOF sensors with advanced device architectures and artificial intelligence. By integrating mechanistic schemes， performance benchmarking tables， and a discussion of future challenges， this review serves as a rational design guide for harnessing the full potential of MOFs in next-generation chemiresistive sensing technologies.

2 Design principles of MOFs

Metal-organic frameworks （MOFs） are crystalline porous materials constructed from the coordination of metal-containing node ions or clusters known as Secondary Building Units （SBUs） with organic multitopic linkers （Fig.2）. Their evolution from serendipitous assemblies to rationally engineered materials has been propelled by pivotal conceptual breakthroughs[12]. The initial paradigm shift moved beyond a simplistic node-and-spacer model to the geometric principle， which exploits the predictable coordination geometries of polynuclear SBUs （e.g.， squares， octahedra） as structural anchors， providing a foundational blueprint for architectures with inherent mechanical robustness and permanent porosity. Building directly upon this predictability， the isoreticular principle established a powerful methodology for the systematic expansion of pore size and volume through the strategic elongation of organic linkers while rigorously maintaining the underlying network topology[13].
图2 次级结构单元（SBU）的关键作用：金属簇（SBU）的选择直接决定了所得MOF的几何结构、孔道化学性质及其在气体储存、分离和传感等应用中的适用性[17]

Fig.2 The pivotal role of secondary building units （SBUs）. The choice of metal cluster （SBU） directly dictates the resulting MOF’s geometry， pore chemistry， and its suitability for applications such as gas storage， separation， and sensing[17]. Redesigned by the authors， based on concepts from ref. [17]. Copyright 2018 American Association for the Advancement of Science

To imbue these pre-formed scaffolds with advanced functionality， post-synthetic modification （PSM） was developed as a versatile synthetic tool， allowing for the precise installation of diverse functional groups such as catalytic sites or binding moieties directly onto the pore walls[14]. Finally， the innovative concept of multivariate MOFs （MTV-MOFs） transcended single-linker systems by enabling the incorporation of multiple， functionally distinct linkers within a single crystalline phase， allowing for the deliberate design of complex， heterogeneous pore environments with synergistic properties[15]. This modular design strategy has enabled the targeted pursuit of specific material properties. A critical application has been the achievement of ultrahigh porosity[16]. A major milestone was MOF-5， which， with its robust cubic framework， demonstrated a BET surface area exceeding 2500 m2/g，significantly surpassing traditional porous materials like zeolites[18]. The quest for higher surface areas led to frameworks like MOF-177， which utilized a judiciously selected trigonal linker to achieve a non-interpenetrated network with a BET surface area over 4500 m2/g[19]. The principle of isoreticular chemistry has since been extended to create frameworks with exceptionally large pore apertures， exemplified by the IRMOF-74 series， which features pores large enough to accommodate proteins. Metal substitution in SBUs， as demonstrated by the MOF-74 series， provides a complementary route for tuning chemical properties without altering the overall structure[20].
For practical sensing applications， operational stability under demanding conditions is paramount. Frameworks like zeolitic imidazolate framework-8 （ZIF-8） and zirconium-based University of Oslo-66 （UiO-66） exhibit exceptional chemical resilience， maintaining their structure in boiling water， organic solvents， and highly acidic or basic environments[21]. The rich structural and chemical versatility of MOFs， as outlined above， can be harnessed through various sensing mechanisms. The following sections will focus specifically on their application in chemiresistive sensing， utilizing both the intrinsic properties of pristine MOFs and the enhanced functionality of composites and derivatives.

3 MOFs in chemiresistive sensing

3.1 Basic principles

The fundamental charge transfer and host-guest interaction mechanism in MOF-based chemiresistors is illustrated in Fig.3. Chemiresistive sensors transduce chemical information into an electrical readout through a measurable change in a material’s electrical resistance upon analyte adsorption. Their appeal lies in a straightforward operating principle coupled with practical advantages such as low-cost fabrication， facile miniaturization， and simple integration into electronic platforms[22-24]. The specific mechanism of resistance modulation， however， depends on the material class. In metal-organic frameworks， multiple charge transfer pathways can be engineered. Adsorbed analytes can act as electron donors or acceptors， directly modulating the charge carrier concentration within the framework. The metal nodes and functional organic linkers serve as effective adsorption sites， while redox-active metal centers can undergo reactions that substantially influence overall conductivity. Furthermore， framework volume changes upon gas adsorption can alter inter-component electron hopping distances， leading to additional resistance variations[25]. The intrinsic insulating nature of many MOFs has been addressed through strategic material design. In intrinsically conductive MOFs （particularly 2D variants）， extended π-conjugation or through-bond charge transport allows analyte interaction to directly alter the band structure or charge hopping barriers.
图3 MOF中的传感机制示意图：（A） 金属节点处的主客体相互作用，展示其作为Lewis和Brønsted酸位点用于分析物结合的作用；（B） 分析物、金属节点和有机连接体之间的氧化还原平衡和电荷转移示意图，展示了分析物结合如何转化为电导率变化；（C） Cu3（HHTP）2与各种气体（N2、H2O、NH3） 相互作用的DFT优化结构及相应的态密度 （DOS）图，突出了对化学电阻响应有贡献的电子结构和层间距的相关变化[30]

Fig.3 Sensing mechanism in MOFs. （A） Proposed host-guest interactions at metal nodes， illustrating their role as Lewis and Brønsted acid sites for analyte binding. （B） Schematic of the redox equilibrium and charge transfer between the analyte， metal node， and organic linker， demonstrating how analyte binding is transduced into a conductivity change. （C） DFT-optimized structures and corresponding density of states （DOS） plots showing the interaction of Cu3（HHTP）2 with various gases （N2， H2O， NH3）， highlighting the associated changes in electronic structure and interlayer spacing that contribute to the chemiresistive response[30]. Copyright 2013 Royal Society of Chemistry

In a prevalent composite strategy， conductive fillers （e.g.， graphene， CNTs） form a percolation network while the MOF acts as a selective sieve， funneling analytes to the interface for electron transfer and a resultant resistance change. This synergy combines the supreme selectivity of the MOF with the excellent electrical properties of the filler[26-27]. MOF-derived metal oxides utilize surface reactions where target gases modulate space-charge layers formed by chemisorbed oxygen. Chemiresistive sensor performance is benchmarked against four essential parameters： response/sensitivity， response/recovery kinetics， selectivity， and operational stability. The sensor response， defined as ΔR/R0， and its derivative sensitivity， which determines the limit of detection （LOD = 3σ/Sensitivity）， are key quantitative metrics[27-30].
Sensor kinetics are characterized by the response time （tres）， the duration to reach 90% of the maximum resistance change， and the recovery time （trec）， the time needed to return to 10% of that change after analyte removal； optimizing these requires reducing activation barriers for surface reactions and enhancing analyte diffusion. Selectivity， the ability to distinguish a target analyte among interferents， originates from unique surface reactions and can be enhanced by tuning operating temperature， material composition， and the strategic incorporation of selective catalysts or molecular recognition elements. Finally， operational stability encompassing long-term reliability and reproducibility is a prerequisite for practical deployment， with a key challenge being performance degradation from environmental factors like water vapor adsorption； material design strategies that mitigate such deactivation pathways are therefore essential for developing commercially viable sensing platforms[31].
It is essential to distinguish between the charge transport mechanisms in intrinsically conductive MOFs （cMOFs） and those in frameworks with large electronic band gaps （>3 eV）. cMOFs， particularly low-dimensional variants with extended π-conjugation， facilitate charge transport via band-like （through-bond） or through-space conduction， where analyte interaction directly perturbs a delocalized band structure[32-33]. In contrast， for insulating or weakly conductive 3D MOFs such as many ZIFs and Zr-based frameworks， measurable chemiresistive responses typically originate from two primary alternative mechanisms： proton hopping and redox-hopping （electron hopping）.
Proton hopping， often following a Grotthuss-type mechanism， becomes significant in frameworks featuring hydrogen-bonded networks formed by protic defects （e.g.， μ3-OH groups）， coordinated water molecules， or guest acids within the pores[34]. Here， analyte adsorption can dramatically alter proton mobility by disrupting this network， leading to a change in electrical resistance. Redox-hopping， on the other hand， is operative in MOFs with redox-active components （e.g.， Co2+/Co3+ in zeolite imidazole framework （ZIF-67））. In this mechanism， charge propagates via thermally activated electron transfer between localized metal or ligand sites in mixed oxidation states[35-36]. This process is distinct from band conduction and exhibits a maximum conductivity when the populations of reduced and oxidized sites are equal[35]. Elevated operating temperatures are frequently employed to provide the activation energy necessary for this electron transfer， thereby enabling a detectable sensor response in otherwise insulating frameworks.

3.2 Pristine MOF-based sensors

3.2.1 Three-dimensional （3D） MOFs

Metal-organic frameworks have garnered significant attention as next-generation gas-sensing materials due to their record-breaking surface areas， which can reach approximately 8000 m2/g[37]. This exceptional porosity makes them particularly suitable for surface-mediated reactions. The pioneering work by Chen et al.[38] first demonstrated the chemiresistive sensing capabilities of pure MOFs using a cobalt-based zeolite imidazole framework （ZIF-67） for formaldehyde detection （Fig. 4A）[38]. Constructed from cobalt ions and methylimidazole linkers， this framework addressed the need to detect formaldehyde at low ppm concentrations， a known cause of sick building syndrome[39].
图4 基于3D MOF的传感器及其传感性能：（A） ZIF-67的晶体结构[38]；（B） ZIF-67在150 ℃下对5~500 ppm甲醛的动态电阻变化[38]；（C） [Co（IM）2]n的晶体结构[41]；（D） 在75 ℃下响应与三甲胺浓度的关系（插图为对2~50 ppm的响应[41]

Fig.4 3D MOF-based sensors and their sensing properties. （A） Crystal structure of ZIF-67[38]. （B） Dynamic resistance changes of ZIF-67 to 5~500 ppm of formaldehyde at 150 ℃[38]. （C） Crystal structure of [Co（IM）2]n[41]. （D） Response versus trimethylamine concentration （inset is response to 2~50 ppm） at 75 ℃[41]. Copyright 2012 Royal Society of Chemistry， 2025 American Chemical Society

To overcome ZIF-67’s intrinsic electronic band gap （~1.98 eV） and poor orbital overlap that rendered it non-conductive at room temperature， the sensors were operated at 150 ℃[40]. Despite this limitation， the material exhibited promising sensing characteristics with a detection limit of 5 ppm and a response factor of 1.8 （Fig. 4B）， performance attributes linked to its substantial surface area of ~1800 m2/g and consistent performance across varying humidity levels. Building on this， the same group explored cobalt-imidazole frameworks （Co[（im）2]n） for trimethylamine detection （Fig. 4C）[41]. These sensors achieved a response factor （Rgas/Rair） of 2 at 2 ppm when operated at 75 ℃ （Fig. 4D） and also maintained stable performance across diverse humidity ranges. The chemiresistive response is attributed to a redox-hopping mechanism facilitated by the Co2+/Co3+ couples， with elevated temperature activating the electron transfer between localized metal centers. Further expanding the scope， amine-functionalized Zr-based MOFs （NH2-UiO-66） were employed for sulfur dioxide sensing[42]. Operating at 150 ℃ in an argon atmosphere， these frameworks leveraged the high acidity of SO2 to facilitate charge transfer with the amine groups， resulting in a substantial resistance decrease （|ΔR/R0| = 21.6% to 10 ppm SO2） in the NH2-UiO-66 framework， which possesses a band gap of 2.75 eV[43]. In this case， the acidic SO2 analyte likely interacts with the amine-functionalized framework， modulating a proton-hopping conduction pathway along the hydrogen-bonded network within the pores.
Moving beyond common carboxylate- and imidazolate-based frameworks， metalloporphyrin-based MOFs offer a unique platform for enhancing host-guest interactions. A seminal study by Zhang et al. on a series of isostructural porphyrinic MOFs， PCN-222-M （M = Cu， Ni， Co， Fe）， for room-temperature NO2 sensing underscores the profound impact of molecular-level design[44]. They demonstrated that the central metal atom in the porphyrin linker plays a decisive role in tuning the electronic structure and sensing characteristics. PCN-222（Cu） exhibited an unprecedented sensitivity of 2209% ppm-1 and a detection limit of 0.93 ppb for NO2， ranking among the best performances for any room-temperature chemiresistive sensor. In contrast， PCN-222（Ni） showed the most rapid recovery. Combined with DFT calculations， they confirmed the critical role of metalloporphyrin units and their interaction with the Zr6 cluster in modulating the Fermi level and electron transfer with NO2.
This work highlights that the strategic choice of the organic linker’s metal center is a powerful and often overlooked tool for designing high-performance pristine MOF sensors. While these seminal studies established the fundamental feasibility of MOF-based chemiresistive sensors， they also revealed significant limitations： the underlying origin of selectivity often remains inadequately explained， the frequent necessity for elevated operating temperatures introduces system complexity， and substantial improvements in sensitivity are imperative for sub-ppm detection.

3.2.2 One-dimensional （1D） MOFs

Metal-organic frameworks with one-dimensional （1D） conductive structures represent an emerging and distinct class of sensing materials. Unlike their 2D and 3D counterparts that facilitate charge transport through layered or framework-wide pathways， 1D conductive MOFs typically consist of linear chains where charge transport occurs along a single axis. This anisotropic structure can yield unique electronic properties and highly exposed active sites along the chain， which are beneficial for specific sensing interactions.
A landmark study by Liu et al.[45] powerfully illustrated the potential of this class by developing a novel 1D conductive MOF， Cu2DADHA （where DADHA = 1，5-diamino-4，8-dihydroxyanthraquinone）， to address the critical safety challenge of real-time carbon monoxide （CO） detection from battery thermal runaway. The authors synthesized the framework in both powder and thin-film forms. Chemiresistive sensors based on Cu2DADHA films exhibited a remarkable response of 93.2% to 100 ppm CO at room temperature under anhydrous and oxygen-free conditions， with an ultralow detection limit of 235 ppb. The sensing mechanism was attributed to coordination-driven charge transfer， where CO molecules directly interact with the copper sites along the 1D chains. To underscore practical utility， the material was integrated into a wireless sensor module that provided real-time CO monitoring of a simulated battery cell， successfully transmitting data to a mobile device via Bluetooth， thereby demonstrating a complete sensor-to-system pipeline.
Building on this foundation， the same research group subsequently demonstrated a sophisticated multi-functional sensing platform[46]. They developed a new family of 1D conductive MOFs， termed MBTA （M = Cu， Ni）， and integrated them into a dual-parameter micro-electro-mechanical system （MEMS） sensor for thermal runaway monitoring. The device was engineered such that in-situ grown MBTA films could simultaneously detect carbon monoxide concentration and temperature on different channels of a single MEMS chip， with polydimethylsiloxane （PDMS） encapsulation used to decouple the signals. This work highlights the unique potential of 1D cMOFs to be engineered into complex， multi-modal sensor systems that provide both gas concentration data and crucial environmental context， thereby significantly enhancing reliability for critical early-warning applications. The synthesis of high-quality 1D conductive MOFs with robust electrical properties is a critical step toward their application in electronics. Shang et al.[47] contributed significantly to this foundation by reporting a one-dimensional conductive MOF， DDA-Cu， which features extended π-d conjugated nanoribbon layers （Fig.5）.
图5 MOF的电荷传输路径及DDA-Cu MOF的合成过程：（A） 扩展共轭平面；（B） 空间电荷传输；（C） “金属”指金属的d轨道，“配体”指配体的p轨道，“π核”表示配体的共轭π平面；（D） DDA-Cu的合成示意图；（E） DDA-Cu的晶体结构[47]

Fig.5 Charge transport pathways of MOFs and the synthesis process of DDA-Cu MOF. （A） Extended conjugation planes. （B） Charge transport through-space. （C） The “metal” means the d orbitals of metals， “ligand” is the p orbitals of ligand and “π core” represents the conjugated π plane of ligand. （D） Synthetic schematic diagram of DDA-Cu. （E） Crystal structure of DDA-Cu[47]. Copyright 2022 Springer Nature

The authors demonstrated that this structure facilitates efficient charge transport， yielding a high conductivity of ~9.4 S/m， orders of magnitude greater than that of many conventional 1D MOFs. Although demonstrated in energy storage and optoelectronic devices， the material’s high crystallinity， environmental stability， and favorable electronic band structure （~0.49 eV） position DDA-Cu as an exceptionally promising candidate for the future development of sensitive and stable chemiresistive sensors.

3.2.3 Two dimensional （2D） conductive MOFs

Recent advances have yielded 2D metal-organic frameworks with remarkable electrical conductivity， opening new frontiers for electronic applications. Charge transport in these materials is enhanced through two primary strategies： through-bond conduction， which improves electron orbital overlap between metal nodes and organic linkers， and through-space conduction， which creates electrical pathways through electroactive fragments within the porous architecture[48-51].
The pioneering work by the Dincă group demonstrated the first 2D conductive MOF （cMOF）- based chemiresistive sensor for room-temperature ammonia detection[23]. The framework Cu3（HITP）2 achieved an electrical conductivity of 0.2 S/cm and detected NH3 from 0.5 to 10 ppm （ΔG/G0 = 0.1% to 2.5%）. Further advancing this technology， sensor arrays comprising Cu3（HHTP）2， Cu3（HITP）2， and Ni3（HITP）2 were developed to identify and categorize 16 different volatile organic compounds into distinct classes using principal component analysis[52]. The underlying sensing mechanism involves charge transfer at the metal sites， as evidenced by the contrasting responses of Cu- and Ni-based cMOFs （Fig. 5C） and directly confirmed by DFT and IR studies showing NH3 coordination to Cu（II） sites， which induces unit cell expansion and an increased band gap. Complex behaviors， such as the concentration-dependent response of Cu3（HITP）2 to n-butylamine， suggest additional contributions from hydrogen bonding with organic ligands. The utility of cMOFs for complex environments is powerfully demonstrated by their integration into sensor arrays paired with machine learning. Benedetto， Simon， Mirica et al.[53] developed a chemiresistive array of M3（HHTP）2 （M = Ni， Cu， Zn） where variations in the metal node generated distinct response patterns for ppm levels of CO， NH3， SO2， H2S， NO， and their binary mixtures at room temperature. Using principal component analysis and random forest classification， the array could accurately identify gas composition， creating an effective “electronic nose” to address selectivity challenges.
Beyond chemical design， synthetic and morphological control is crucial. Huang et al.[54] reported a general template method to synthesize nanostructured 2D cMOFs with hollow and flower-like morphologies via insulating MOFs-to-cMOFs transformations. The resulting Cu-HHB nanoflowers exhibited a sensor response over 250% greater than the bulk material， highlighting the importance of nanostructuring. Similarly， Huang， Dong， Feng et al.[55] created hierarchical cMOF films with a nanoporous shell and hollow voids （Fig.6A~G）， increasing gas permeability by more than 8-fold and yielding an NH3 sensor with a response speed ten times faster than its non-hierarchical counterpart.
图6 合成策略与结构表征[55]

Fig.6 Synthetic strategy and structural characterization[55].Copyright 2023 Springer Nature

To address fabrication challenges and improve practicality， innovative integration methods have been developed. The Mirica group demonstrated direct self-assembly of Cu3（HHTP）2 and Ni3（HHTP）2 on shrinkable polymer films with hand-drawn graphite electrodes， creating chemiresistors for NH3， NO， and H2S[56]. They further grew cMOFs directly on woven cotton fabrics， creating flexible， wash-stable textile sensors （Fig.7A） that achieved high responses （e.g.， ΔG/G0 = 98% for H2S） and low theoretical detection limits （e.g.， 0.23 ppm for H2S） （Fig.7B）[57]. In parallel， Xu et al. used a layer-by-layer spray-coating technique to produce crystalline Cu3（HHTP）2 nanofilms with precise thickness control （Fig.7C）. A 20 nm-thick film exhibited a high response （|ΔR/R0| = 45% to 10 ppm NH3） and a low detection limit （0.5 ppm）， maintaining stability for three months （Fig.7D）[58].
图7 基于2D导电MOF的传感器件：（A） MOF组装棉织物的制备过程及其照片和SEM图像[57]；（B） 基于MOF-棉花复合材料的传感器传感性能[57]；（C） 采用逐层喷涂法制备Cu3（HHTP）2薄膜气体传感器的示意图[58]；（D） Cu3（HHTP）2纳米薄膜的NH3传感性能[58]

Fig.7 2D Conductive MOF-based sensing devices. （A） Fabrication process of MOF-assembled cotton fabrics and their photos and SEM images[57]. （B） Sensing properties of MOF-cotton composites-based sensors[57]. （C） Schematic illustration of synthesis of Cu3（HHTP）2 thin film gas sensors by using the layer-by-layer spray-coating method[58]. （D） NH3-sensing properties of Cu3（HHTP）2 nanofilms[58]. Copyright 2017 American Chemical Society， 2017 Wiley-VCH

The rational design of the electronic structure is fundamental. Shan et al.[59] demonstrated this by modulating cMOF conductivity over five orders of magnitude through ligand design and topology control， which directly correlated with an outstanding and tunable response to NH3. Integration into advanced device architectures further enhances functionality. Lu， Mao et al.[60] constructed flexible field-effect transistor （FET） sensors based on Ni3（HHTP）2 films on diverse substrates， achieving a high-sensitivity NO2 detection limit of 56 ppb. More et al.[61] showed that a Zn-HHTP-based Chem-FET platform unlocked superior CO sensing characteristics compared to a chemiresistor， leveraging the electrostatic gating effect. Lin et al.[62] grew highly oriented Cu-HHTP films on a 3D nanowire array， creating a 3D MOF thin film that synergized efficient electrical pathways with rapid gas diffusion， resulting in a top-performing room-temperature ammonia sensor. Despite these advances， challenges remain， including limited response to a narrow range of gases， insufficient response magnitude and detection limit compared to alternative materials， and needs for improved stability and reproducibility （Table 1）. While the development of 3D conductive MOFs[63-64] and the potential for catalyst functionalization[65] offer future avenues， current strategies focus on enhancing intrinsic properties. Lin et al.[66] engineered hydrogen bonds into a π-d conjugated MOF， FeCo3（DDA）2， achieving exceptional structural stability for over 2000 hours， a design principle applicable to sensing. Lee， Park， et al.[1] developed a groundbreaking photoactivated platform by integrating cMOF films onto a micro-LED （μLED） array （Fig.8A~D）. This system used light to generate charge carriers， enhancing sensitivity and enabling full reversibility at room temperature with ultra-low power （587 µW）. When paired with a deep learning algorithm， it achieved 99.8% classification accuracy for multiple analytes. Furthermore， the cMOF platform’s versatility is exemplified by the incorporation of lanthanides， as shown by Jiang， Yang et al.[67] with Lu-HHTP， which functioned as a high-performance humidity sensor.
表1 纯MOF用于化学电阻传感器的传感性能总结

Table 1 The summary of the sensing properties in recent studies on pure MOFs for chemiresistive sensors

Gases Materials Operating temperature Response Detection limit Ref
Trimethylamine [Co（imidazole）2]n 150 ℃ Rgas/Rair = 15 at 100 ppm 2 ppm 41
Formaldehyde ZIF-67 150 ℃ Rgas/Rair = 15 at 100 ppm 5 ppm 38
Sulfur dioxide NH2-UiO-66 RT |ΔR/R0| = 22% at 10 ppm 1 ppm 42
Cu-BTC/graphene ΔRgas/Rair = 4% at 100 ppm 100 ppm 68
Ammonia Cu3（HITP）2 RT ΔG/G0 = 2.5% at 10 ppm 500 ppb 23
Cu3（HHTP）2 nanofilms （thickness = 20 nm） |ΔR/R0| = 45% at 10 ppm 500 ppb 58
Methanol Cu3（HHTP）2 RT ΔG/G0 = 9% at 200 ppm N/A 52
Cu3（HITP）2 ΔG/G0 = 4% at 200 ppm N/A 52
Ethanol Ni3（HITP）2 RT ΔG/G0 = 4% at 200 ppm N/A 52
Cu3（HHTP）2 nanorods |ΔI/I0|= 1.8% at 80 ppm 2.5 ppm 56
Co3（HHTP）2/graphite ΔG/G0 = 10% at 80 ppm 5 ppm 69
Nitrogen monoxide Fe3（HHTP）2/graphite RT ΔG/G0 = 10% at 80 ppm 5 ppm 69
Hydrogen sulfide Ni3（HITP）2 on cotton RT ΔG/G0 = 81% at 80 ppm 160 ppb 57
Ni3（HHTP）2 nanorods ΔI/I0 = 4.2% at 80 ppm 2.5 ppm 56
Ni3（HHTP）2 on cotton ΔG/G0 = 98% at 80 ppm 230 ppb 57
Ni3（HITP）2 on cotton ΔG/G0 = 97% at 80 ppm 520 ppb 57
图8 本研究示意图，展示了在超低功耗微发光二极管（μLED）气体传感器顶部使用2D导电金属有机框架（cMOF）层，该装置既用作气体传感器又用作电子鼻：（A） 用于气体传感的优化2D cMOF薄膜示意图，显示了薄膜构型（密度和厚度）和组成（导电层和催化覆盖层）的变化；（B） μLED光波长和强度优化过程示意图，以最大化cMOF的气敏能力（电极宽度和间距：5 μm）；（C） 由不同传感器（不同的μLED类型、光强度和cMOF类型）组成的cMOF阵列；（D） 集成深度学习电子鼻系统的光激活cMOF化学电阻气体传感器阵列[1]

Fig.8 Schematic overview of this study， illustrating the use of a 2D conductive metal-organic frameworks （cMOF） layer atop an ultra-low power micro light-emitting diode （μLED） gas sensor， which is employed as both a gas sensor and an e-nose. （A） An illustration of the optimized 2D cMOF films for gas sensing， showing variations in film configuration （density and thickness） and composition （conduction layers and catalytic overlayers）. （B） An illustration of the optimization process of the light wavelength and intensity of μLED to maximize the gas-sensing capability of the cMOFs （width and spacing of electrodes： 5 μm）. （C） An array of cMOFs composed of different sensors （varying μLED types， light intensities， and cMOF types）. （D） Light-activated cMOF chemiresistive gas sensor array integrated with a deep learning-enabled e-nose system[1]. Copyright 2025 Springer Nature

Finally， the intrinsic properties of MOFs that make them excellent for gas separation， such as abundant active sites， tunable pore architectures， and molecular sieving capabilities[70-71] can be directly leveraged to address the selectivity challenge in chemiresistive sensors through the integration of MOF membranes[72].

3.3 MOF composites and hybrids

MOF composites synergistically combine the unique properties of frameworks with other functional materials to overcome limitations of individual components. A classic example is the use of MOFs as molecular sieves for enhancing sensor selectivity. Palladium is a well-established hydrogen-sensing material， where resistance changes upon the formation of palladium hydride （PdHx）. A significant challenge， however， is competitive oxygen adsorption， which poisons the Pd surface and degrades performance[73-74].
To address this， Koo et al. utilized a ZIF-8 layer as a molecular sieve on Pd nanowires （Pd NWs@ZIF-8） （Fig.9A， B）[75]. The precise pore size of ZIF-8 （0.34 nm） permits hydrogen （0.289 nm） to diffuse through while excluding larger oxygen molecules （0.346 nm） （Fig.9C）. This strategic integration dramatically accelerated sensor kinetics， reducing the response and recovery times for 1% H2 from 164 s to 7 s and 229 s to 10 s， respectively， despite a modest reduction in response magnitude （Fig.9D~F）.
图9 MOF-Pd复合材料及相应的H2传感性能：（A） ZIF-8包覆Pd纳米线的合成示意图；（B） Pd NWs@ZIF-8的SEM图像；（C）有/无ZIF-8层的Pd NWs传感模型示意图，传感器在室温空气中的H2传感特性：（D） 响应，（E） 响应时间，（F） 恢复时间[75]

Fig.9 MOF-Pd composites and corresponding H2-sensing performance. （A） Schematic illustration of the synthesis of Pd NWs covered by ZIF-8. （B） SEM image of Pd NWs@ZIF-8. （C） Schematic illustration of the sensing model of the Pd NWs with and without the ZIF-8 layer. H2-sensing properties of the sensors at room temperature in air： （D） response， （E） response time， and （F） recovery time[75]. Copyright 2017 American Chemical Society

It is important to note that the molecular sieving action of ZIF-8 is influenced by its intrinsic structural flexibility. The ZIF-8 framework can undergo a “gate-opening” phenomenon， where the rotation of the 2-methylimidazole linkers transiently enlarges the pore aperture beyond its crystallographic size of ~3.4 Å. This flexibility allows ZIF-8 to admit certain molecules， such as N2 （3.64 Å）， under specific conditions[76-77].
The selectivity for H2 over O2， therefore， is not based on a perfectly static and rigid sieve but on a kinetic differentiation. The small kinetic diameter and high diffusivity of H2 enable facile passage， while the energy barrier required for linker rotation to accommodate O2 （3.46 Å） is less readily overcome at room temperature， especially in a thin-film configuration. This results in the effective exclusion of O2 and protection of the Pd surface， as demonstrated by the dramatically improved sensor kinetics. For applications demanding the highest selectivity， strategies to enhance framework rigidity and suppress gate-opening are an active area of research. These include synthesizing mixed-ligand ZIF-8 frameworks[78]， introducing heterometallic nodes to strengthen coordination bonds[79]， and applying postsynthetic thermal treatments under controlled conditions[77]. Such methods can finely tune the flexibility of MOF membranes， offering a pathway to engineer even more precise and stable molecular sieves for next-generation sensor designs.
The sophistication of composite design extends to heterostructures with controlled energy alignment. Chen， Xu et al.[80] engineered a core-shell UiO-66@TDCOF heterojunction， where strategic functionalization enabled a light-induced transition from a Type-I to a Type-II configuration. This controlled modulation enhanced charge separation， resulting in a NO2 sensitivity that ranks among the highest for heterojunctions. Similarly， Ding et al.[81] designed a 0D-1D heterostructure by anchoring ZIF-8 nanoparticles onto carbon nanotubes （CNTs）， which prevented aggregation and provided excellent charge transport， yielding an ultrahigh response （57.87） to 40 ppm isopropanol at room temperature.
The ultimate expression of this rational design may lie in MOF-on-MOF heterostructures， as demonstrated by Yao et al.[82] with van der Waals-integrated Cu-TCPP-on-Cu-HHTP films， which combined molecular sieving and conductivity for unparalleled benzene sensing. A major application of MOF composites is addressing the poor selectivity of metal oxide sensors （e.g.， ZnO， SnO2， WO3）[83]. Coating metal oxides with MOF membranes leverages their molecular sieving capabilities. Yao et al.[84] developed ZIF-CoZn-coated ZnO nanowires （ZnO@ZIF-CoZn） for selective acetone detection （Fig.10A）. The hydrophobic ZIF-CoZn layer also provided humidity-independent operation at 260 ℃ （Fig.10B）. Tian et al.[85] used a ZIF-8 shell on ZnO nanowires （Fig.10C） to selectively detect formaldehyde （0.243 nm） while excluding larger interferents like ethanol （0.453 nm） and toluene （0.525 nm）， drastically improving selectivity （Fig.10D， E）. Dmello et al.[86] encapsulated SnO2 nanoparticles in ZIF-67 （SnO2@ZIF-67）， leveraging the framework’s high CO2 affinity to achieve a two-fold higher response compared to pristine SnO2. These and other examples （Table 2） underscore the potential of MOF membranes to impart selectivity when uniformly deposited on various sensing materials[87].
图10 MOF-金属氧化物复合材料及相应的气敏性能：（A） ZIF-CoZn包覆ZnO纳米线的合成示意图[84]；（B） ZIF-CoZn包覆ZnO在260 ℃下的丙酮传感性能[84]；（C） ZIF-8包覆ZnO纳米线的分子筛效应示意图[85]；（D） ZIF-8包覆ZnO纳米线在300 ℃下的选择性特性[85]；（E） ZnO纳米线在300 ℃下的选择性特性[85]

Fig.10 MOF-metal oxide composites and corresponding gas-sensing performance. （A） Schematic illustration for synthesis of ZIF-CoZn coated ZnO nanowires[84]. （B） Acetone-sensing properties of ZIF-CoZn coated ZnO at 260 ℃[84]. （C） Schematic illustration of the molecular sieving effect of ZIF-8-coated ZnO nanowires[85]. （D） Selectivity property of ZIF-8 coated ZnO nanowires at 300 ℃[85]. （E） Selectivity property of ZnO nanowires at 300 ℃[85]. Copyright 2016 Wiley-VCH， 2016 American Chemical Society

表2 近年来报道的用于化学电阻传感器的MOF膜的传感性能总结

Table 2 The summary of the sensing properties reported in recent literature of MOF membranes for chemiresistive sensors

Gases Materials Operating temperature Sensing property Ref
Acetone ZnO@ZIF-CoZn （T = 5 nm） 250 ℃ ΔR/Rgas = 27.5 at 10 ppm independent to 0%~90% RH 84
Formaldehyde ZnO@ZIF-8 nanowire 300 ℃ Rgas/Rair = 12.5 at 100 ppmSratio： improved from 0.74 to 2 85
Carbon dioxide SnO2@ZIF-67 205 ℃ ΔR/R0 = 16.5% at 0.5%
2-fold higher than pristine SnO2		 86
Hydrogen Pd nanowires@ZIF-8 （T = 160 nm） RT tres = 7 s & trec = 8 s at 1%20-fold faster than pristine Pd 75
ZnO@ZIF-8 nanowire （T = ∼150 nm） 300 ℃ Rgas/Rair = 1.44 at 50 ppm， no response to some gases 88
ZnO@ZIF-8 nanorods （T = ∼ 85 nm） 250 ℃ ΔI/I0 = 80% at 50 ppm， Sratio： improved from 0.45 to 2.0 89
ZnO@Pd@ZIF-8 nanowires 200 ℃ Rgas/Rair = 6.7 at 50 ppm， no response to some gases 90
The functionality of MOF-metal oxide composites can be further enhanced. Ling et al. developed a UV-activated sensor using Au-decorated ZnO@ZIF-8 core-shell nanorods for room-temperature NO2 detection. In this tripartite system， ZIF-8 preconcentrated the analyte， ZnO acted as the semiconductor， and Au nanoparticles enhanced charge separation， enabling remarkable responses up to 34700% and full recovery[91]. Min et al.[92] created an N-carbon-doped ZnO/ZIF-8 nanophase via controlled oxidation， which evolved an interconnected conductive network upon NO2 exposure， yielding an extreme sensitivity of ≈130 ppm-1 and a sub-ppb detection limit （0.63 ppb）.
Beyond oxides， composites with carbon materials and polymers offer distinct advantages. Roh et al. created robust composites between conductive MOFs （cMOFs） and conductive polymers （cPs）. The interfacial energy band alignment promoted hole-enrichment in the cMOF， selectively enhancing analyte desorption kinetics and enabling improved recovery and long-term stability at room temperature across multiple cMOF systems[93]. For environmental stability， Russell et al.[94] developed a nano-sandwich structure confining MOF nanoparticles between graphene oxide （GO） nanosheets （Fig.11A~G）. This architecture preserved crystallinity， provided a protective barrier against harsh environments （even for water-sensitive MOFs）， and significantly boosted electrical conductivity， presenting a universal platform for robust sensors.
图11 （A） 通过受限冷冻组装制备纳米三明治的过程；（B） 随机组装的MOF-GO复合材料的SEM图像；（C） 用于TEM研究的ZIF-8-GO超薄切片；（D， E） ZIF-8-GO纳米三明治的电子断层扫描和SEM图像；（F） 4个MOF-GO纳米三明治的数码照片，验证该策略的普适性和放大潜力；（G） 不同组装程度的有序纳米颗粒[94]

Fig.11 （A） The fabrication process of nano-sandwiches by confined freeze assembly. （B） SEM images of the randomly assembled MOF-GO composites. （C） Ultramicrotomy of ZIF-8-GO for TEM study. （D， E） Electron tomography and SEM images of the ZIF-8-GO nano-sandwich. （F） Digital photographs of four MOF-GO nano-sandwiches to verify the generality of the strategy and potential for upscaling. （G） Ordered nanoparticles with different degrees of assembly[94].Copyright 2025 Springer Nature

3.4 MOF-derived materials

MOF-derived materials， obtained through the thermal treatment of MOF precursors， inherit advantageous properties such as high porosity， large surface area， and compositional tunability， making them highly promising for sensing[95-97].

3.4.1 MOF-derived carbon composites

Pyrolysis of MOFs yields porous carbon composites with excellent chemical/thermal stability. The cavities of the MOF precursor can be used to confine active materials like transition metal dichalcogenides （TMDs）. Koo et al.[98] used ZIF-67 as a template to create a few-layered WS2 confined within a Co， N-doped carbon matrix （Fig.12A）. The confinement during pyrolysis suppressed WS2 growth， resulting in abundant exposed edge sites （Fig.12B~D）， which have a high affinity for NO2 （adsorption energy： -1.4 eV vs. -0.4 eV for basal planes）[99]. The composite exhibited a high NO2 response （ΔR/Rgas > 48.2% to 5 ppm） with excellent selectivity and stability at room temperature （Fig.12E， F）， and a 10-fold increase in reaction kinetics compared to the pristine derived carbon[99]. For complex sensing challenges， MOF-carbon composites can be paired with machine learning. Losic et al. developed an extrusion-printed sensor from an NU-1000/graphene composite. By integrating principal component analysis （PCA）， the sensor could detect methanol at ppb levels amidst a high background of its structural analog， ethanol， demonstrating a viable path for non-invasive health monitoring[100].
图12 MOF衍生碳复合材料及相应的传感性能：（A） MOF衍生碳/WS2复合材料的合成示意图；（B） TEM图像；（C） 放大TEM图像；（D） WS2纳米片功能化的MOF衍生碳复合材料的HRTEM图像；（E） 传感器在室温下对1~5 ppm NO2的动态响应变化；（F） 传感器对各种分析物的响应[98]

Fig. 12 MOF-derived carbon composites and corresponding sensing performances. （A） Illustration of the synthesis of MOF-derived carbon/WS2 composites， （B） its TEM image， （C） magnified TEM image， and （D） HRTEM image of MOF-derived carbon composites functionalized with WS2 nanoplates. （E） Dynamic response transitions of the sensors toward 1~5 ppm of NO2 at room temperature. （F） Response of the sensors to various analytes[98]. Copyright 2018 Wiley-VCH

3.4.2 MOF-derived metal oxides

Calcination of MOFs produces porous metal oxide architectures that retain the morphology of their templates[95,101-102]. Lu et al. first demonstrated this by calcining ZIF-67 to form porous Co3O4 concave nanocubes （Fig. 13A， B）[103]. These exhibited high ethanol selectivity at 300 ℃ （Fig. 13C）， with the calcination temperature critically balancing crystallinity and porosity； the sample calcined at 300 ℃ showed the highest response （Rair/Rgas = 1.8 to 10 ppm） （Fig.13D）. The composition can be tailored by the MOF precursor， yielding porous In2O3， ZnO， and Fe2O3 from Materials Institute Lavoisier-68 （MIL-68）， MOF-5， and Materials Institute Lavoisier-88A （MIL-88A）， respectively， for detecting formaldehyde and acetone[104-106].
图13 （A） ZIF-67衍生Co3O4凹面纳米立方体的合成示意图，（B） 其SEM图像，（C） 其在300 ℃下对4种气体的响应，（D）其在300 ℃下对不同浓度乙醇的响应[103]

Fig.13 （A） Schematic illustration for the synthesis of ZIF-67 derived Co3O4 concave nanocubes， （B） its SEM image， （C） its response to four gas species at 300 ℃， and （D） its response to various concentrations of ethanol at 300 ℃[103]. Copyright 2014 American Chemical Society

To enhance performance， hierarchical structures and heterojunctions can be engineered. Jang et al.[107] created hierarchical MOFs from ZIF-67 and Zn ions， which， upon calcination， yielded n-p heterojunctions of ZnO sheets and Co3O4 rods （Fig. 14A~C）. This structure showed a significantly enhanced acetone response （Rair/Rgas = 29.0 to 5 ppm） compared to pure Co3O4 （Rair/Rgas = 1.05） （Fig. 14D）.
图14 分级和异质MOF衍生金属氧化物及相应的传感性能：（A） 分级MOF及其衍生物的合成示意图；（B） 其TEM图像；（C） MOF衍生的ZnO-Co3O4的TEM图像；（D） MOF衍生的ZnO-Co3O4在450 ℃下的丙酮传感性能[107]

Fig.14 Hierarchical and heterogeneous MOF-derived metal oxides and corresponding sensing performance. （A） Illustration of the synthesis of hierarchical MOFs and their derivatives， （B） its TEM image， （C） TEM image of MOF-derived ZnO-Co3O4， and （D） acetone-sensing properties of MOF-derived ZnO-Co3O4 at 450 ℃[107]. Copyright 2018 American Chemical Society

3.4.3 Catalyst-functionalized MOF-derived oxides

The porosity of MOFs allows for the encapsulation of ultrasmall， uniform metal nanoparticles （e.g.， via ship-in-bottle or bottle-around-ship methods[119]）， which can be transformed into catalytic sites upon calcination. Koo et al.[116] used Pd@ZIF-67 to create PdO-functionalized Co3O4 hollow nanocubes[120]. The PdO nanoparticles act as electronic sensitizers， enhancing the acetone response （Rair/Rgas = 2.51 at 5 ppm） at 350 ℃. Since many MOF-derived oxides are p-type， combining them with n-type metal oxides like SnO2 can create more responsive heterostructures[115]. Jang et al.[121] used a galvanic replacement reaction to convert PdO-Co3O4 nanocubes into n-type PdO-SnO2-Co3O4， drastically improving the acetone response to Rair/Rgas = 22.8 at 5 ppm[118]. Similarly， Koo et al.[114] decorated WO3 nanofibers with catalysts derived from Pd@ZIF-8， significantly promoting their gas-sensing properties.
Beyond noble metals， doping with high-valence cations is an effective strategy. Yang et al.[122] synthesized Mo6+-doped Co3O4 microrods from an MOF template. The doping increased oxygen defects and active Co3+ sites， enhancing the xylene response （Rgas/Rair = 29.8 to 100 ppm） and lowering the operating temperature to 140 ℃， with the sensing mechanism elucidated by in-situ DRIFTS. The sensing properties of various MOF-derived materials are summarized in Table 3， showing performance comparable to state-of-the-art sensors. Finally， the principle of using a structured template extends beyond MOFs. Xie et al.[123] used polymer cubosomes to synthesize a library of ordered nanoporous metal oxides， such as WO3， which exhibited superior H2S sensing performance due to enhanced mass transport.
表3 近年来MOF衍生物用于化学电阻传感器的传感性能

Table 3 The sensing properties in the recent papers on MOF derivatives for chemiresistive sensors

Gases Materials Operating temperature Response Detection limit Ref
Nitrogen dioxide ZIF-67-derived carbon composites-loaded MWCNTs RT |ΔR/R0| = 1% at 5 ppm 100 ppb 108
ZIF-67-derived WS2 functionalized carbon composites RT ΔR/Rgas = 18% at 1 ppm 100 ppb 98
SWCNTs loaded PdO-Co3O4 derived from Pd@ZIF-67 RT ΔR/Rair = 27.3% at 20 ppm 10 ppm 109
ZIF-67 templated Co3O4 concave nanocubes 300 ℃ Rgas/Rair = 3.25 at 200 ppm 10 ppm 103
Ethanol HKUST-1 templated Cu2O/CuO cages 150 ℃ Rgas/Rair = 6.5 at 200 ppm - 110
In/Ga-MOF-68 derived In/Ga oxides 235 ℃ Rair/Rgas = 80 at 200 ppm 2 ppm 111
MOF-5 templated ZnO nanocages 300 ℃ ΔR/Rair = 1.28 at 0.1 ppm 100 PPB 106
Benzene ZIF-8 templated hollow ZnO 450 ℃ ΔR/Rgas = 1.9 at 5 ppm - 112
Formaldehyde MOF templated mesoporous In2O3 nanorod 210 ℃ ΔR/Rgas = 7.5 at 10 ppm - 104
p-Xylene ZIF-67 derived hollow Co3O4 nanocages 225 ℃ Rgas/Rair = 78.6 at 5 ppm 250 ppb 113
ZIF-67 derived hollow Co3O4 nanocages 225 ℃ Rgas/Rair = 43.8 at 5 ppm - 113
Toluene Pd@ZIF-8 templated Pd@ZnO-WO3 nanofibers 350 ℃ Rair/Rgas = 22.22 at 1 ppm 0.1 ppm 114
Pd@ZIF-ZnCo derived PdO@ZnO/ZnCo2O4 hollow spheres 250 ℃ ΔR/Rair = 69% at 5 ppm 400 ppb 115
Pd@ZIF-67 derived PdO@Co3O4 nanocubes 350 ℃ Rgas/Rair = 2.51 at 5 ppm 100 ppb 116
Acetone Pd@ZIF-8 templated PdO@ZnO-SnO2 nanotubes 400 ℃ Rair/Rgas = 5.06 at 1 ppm 100 ppb 117
Pd@ZIF-67 templated Co3O4-PdO loaded SnO2 nanocubes 450 ℃ Rair/Rgas = 22.8 at 5 ppm 5 ppb 118
Hierarchical MOF-derived ZnO-Co3O4 450 ℃ Rair/Rgas = 29 at 5 ppm 1 ppm 107

3.5 Challenges in stability and practical deployment

The preceding sections have detailed the remarkable progress in designing MOF-based chemiresistive sensors with high sensitivity， selectivity， and novel functionalities. However， the transition from laboratory demonstrations to commercially viable and reliable devices is predicated on overcoming significant challenges in operational stability and practical deployment. This section critically examines these barriers， which often remain underemphasized in fundamental research but are paramount for real-world application （see Table 4 for a summary）.
表4 MOF基化学电阻传感器的主要挑战与缓解策略概述

Table 4 Summary of key challenges and mitigation strategies for MOF-based chemiresistive sensors

Challenges category Specific challenge Proposed mitigation strategies Ref
Stability Hydrolytic degradation： water molecules attack metal-linker bonds， leading to framework collapse， especially in humid environments.
Chemical/redox degradation： framework degradation or irreversible adsorption by aggressive （NO2， SO2） or redox-active analytes		 Use metals forming high-energy bonds： employ Zr（IV）， Cr（III）， or Fe（III） clusters to form robust M―O bonds resistant to hydrolysis.
Hydrophobic pore engineering： Post-synthetically graft hydrophobic groups （e.g.， ―CF3， ―CH3） or use hydrophobic linkers to create water-repellent pores.
Apply impermeable protective coatings： Encapsulate MOF crystals or films with stable matrices like graphene oxide （GO） or atomic layer deposition （ALD） oxides to create a physical barrier.
Employ redox-inert or stable metal nodes： Utilize Zr（IV）， Al（III）， or Ti（IV） centers less prone to redox reactions.
Design robust linkers： Use chemically stable organic struts （e.g.， perfluorinated or aromatic carboxylates） resistant to oxidation.		 128-129
Mechanical strength & processability Intrinsic brittleness： MOF powders or crystalline films are fragile， prone to cracking/delamination， and difficult to process into devices Fabricate composite architectures： Embed MOF particles into flexible， stress-dissipating polymer matrices （e.g.， PDMS， polyurethane）.
In-situ growth on flexible substrates： Directly grow MOF films on textiles， polymers， or paper to form interpenetrated， adhesive interfaces.
Construct nano-sandwich or core-shell structures： Confine MOF nanoparticles between GO sheets or within polymer shells to enhance cohesion and flexibility.		 128
Scalability & reproducibility Batch-to-batch variability： Inconsistent sensor performance due to difficulties in controlling MOF synthesis （crystallinity， defect density， particle size） at scale
Device integration hurdles： Challenges in forming low-resistance， stable electrical contacts with MOF active layers and integrating them into full sensor systems.		 Develop standardized， scalable fabrication protocols： Implement techniques like continuous flow synthesis， spray coating， or chemical vapor deposition for uniform thin films.
Adopt machine learning （ML）-guided synthesis： Use ML models to optimize synthesis parameters and predict outcomes for consistent material quality.
Engineer dedicated interfacial layers： Introduce conductive adhesion layers or employ direct photolithographic patterning on MOF films to improve contact.
Engineer dedicated interfacial layers： Introduce conductive adhesion layers or employ direct photolithographic patterning on MOF films to improve contact.		 129
Standardization & benchmarking Lack of unified testing protocols： Performance data collected under idealized lab conditions （dry air， single analyte） prevents fair comparison and obscures real-world viability. Establish benchmark testing under realistic conditions： Advocate for tests at standard temperature/humidity levels
（e.g.， 25 ℃， 50% RH） and in complex gas mixtures.
Report long-term stability metrics： Mandate reporting of continuous baseline drift， response decay over weeks/months， and cycle-to-cycle reproducibility.
Define common interferent sets for selectivity： Propose benchmark gas mixtures （e.g.， for a benzene sensor， test against toluene， xylene， CO， NO2） to quantify selectivity.		 129
Signal integrity & data management Signal drift & environmental interference： Sensor baseline and sensitivity shift over time due to aging， humidity fluctuations， or temperature changes.
Data security in networked systems： Risk of data interception or manipulation when sensor networks are connected to the cloud for IoT applications		 Integrate on-sensor compensation： Co-fabricate humidity/temperature sensors or use built-in micro-heaters for active compensation.
Implement machine learning for drift correction： Apply AI algorithms （e.g.， adaptive filters， neural networks） to raw sensor data to differentiate drift from true analyte signal.
Employ robust encryption protocols： Implement end-to-end encryption for all data transmitted from sensor nodes to central hubs.		 129

3.5.1 Hydrothermal and chemical stability

A primary concern is the chemical instability of many MOFs， particularly via hydrolysis， where humidity attacks metal-linker bonds， causing structural degradation， loss of porosity， and irreversible performance decline[124]. While zirconium-based MOFs （e.g.， UiO-66） and zeolitic imidazolate frameworks （e.g.， ZIF-8） exhibit superior stability， many promising conductive MOFs， particularly those with first-row transition metals， are susceptible.
For instance， the performance of some 2D cMOFs can drift significantly in humid environments， complicating data interpretation[125]. Strategies to enhance hydrothermal stability focus on strengthening the framework. These include node engineering with high-valence metals （e.g.， Zr， Cr， Ti， Fe） for stronger coordination bonds， and hydrophobic functionalization of linkers with groups like ―CF3 to create water-repellent pores[126-127]. When intrinsic material stability is insufficient， protective coatings offer an external solution by encapsulating MOF crystals or films within inert， stable matrices； a prime example is the nano-sandwich structure with graphene oxide （GO） reported by Lu et al.[94]， which effectively shielded even water-sensitive MOFs from harsh environments. Aggressive analytes like NO2， SO2， or H2S can permanently oxidize or reduce metal nodes， degrading active sites and causing baseline drift. This necessitates using redox-inactive or robust metal centers for long-term stability.

3.5.2 Mechanical stability for flexible and wearable devices

The future of sensing lies in flexible and wearable electronics， which impose stringent requirements on mechanical robustness. Most MOF crystals are inherently brittle， and thin films can crack， delaminate， or lose electrical connectivity upon repeated bending， stretching， or compression. This is a significant hurdle for integrating MOFs into smart textiles， epidermal patches， or flexible medical devices. Promising approaches to address the challenge of mechanical robustness are being actively developed， focusing on integrating MOFs into resilient architectures.
Strategies for mechanical robustness include： （1） forming composites by embedding MOFs in stress-absorbing polymers （e.g.， PDMS， PU）； （2） in-situ growth on flexible substrates （e.g.， textiles） to create interpenetrated， robust films[57]； and （3） designing intrinsically flexible MOFs with hinge-like linkers to withstand deformation.

3.5.3 The critical need for standardization and benchmarking

A lack of standardized testing， particularly for humidity at defined levels （e.g.， 30%， 50%， 80%）， impedes objective material comparison and must be addressed for commercialization. Equally critical is the adoption of long-term and drift analysis， providing data on continuous operation over extended periods of weeks or months， rather than a few cycles， to credibly prove reliability. Furthermore， the establishment of selectivity benchmarks testing against a common set of structurally similar and commonly co-existing interferents， such as evaluating a benzene sensor against toluene， xylene， CO， and NO2， would allow for a true and comparable assessment of selectivity. Finally， the consistent quantification and reporting of stability metrics， such as baseline drift and response decay over time， are essential for evaluating practical viability. In conclusion， while the structural and functional tunability of MOFs presents unparalleled opportunities for chemiresistive sensing， their journey from the lab bench to the market is inextricably linked to solving these challenges of stability and reproducibility. Future research must pivot towards a co-design philosophy， where materials are engineered not just for high performance， but also for environmental resilience， mechanical integrity， and scalable integration， underpinned by this rigorous and standardized evaluation.

4 Future outlook & emerging applications

The future of MOF-based sensing lies in their evolution from simple analytical devices into intelligent， connected systems that integrate advanced materials， novel device architectures， and artificial intelligence. This convergence is unlocking new paradigms in sensing technology.

4.1 Integration with novel transduction mechanisms and flexible electronics

The utility of conductive MOFs （cMOFs） extends beyond chemiresistance. For instance， Cui et al.[130] integrated Cu3（HHTP）2 onto a surface acoustic wave （SAW） sensor， leveraging the strong H2S-Cu（II） interaction to induce mass and acoustoelectric changes. This device achieved an exceptional detection limit of 6 ppb at room temperature with remarkable long-term stability， showcasing cMOFs in highly sensitive， alternative transduction platforms. The merger with wearable electronics is equally promising. Kim et al.[131] constructed a flexible healthcare platform using a 3D conductive network of Cu3（HHTP）2 on hollow MXene spheres （Fig.15A， B）. This bionic， dual-mode sensor detected both NO2 gas and physical pressure. Integrated with a mobile app and machine learning， it assessed asthmatic risk factors with 97.6% accuracy， demonstrating potential for transformative telemedicine. Similarly， Wang et al.[132] developed a flexible piezoresistive sensor for continuous intraocular pressure （IOP） monitoring by assembling Cu3（HHTP）2 on polymer fibers， showcasing its potential in minimally invasive biomedical devices.
图15 本研究的概念设计：（A， B） 受生物触觉感官结构和神经形态系统的启发，开发了一种具有多功能仿生传感器的柔性智能可穿戴警报系统，用于哮喘患者生理信号的无线监测和区分，并辅以基于一维CNN的机器学习算法[131]

Fig.15 Conceptual design of this study. （A-B） Inspired by the biological tactile sensory structure and neuromorphic system， a flexible smart wearable alarming system with a multifunctional bionic sensor is developed for wireless monitoring and differentiating for the asthma patients’ physiological signals， as assisted by a machine learning algorithm based on 1-D CNNs[131]. Redesigned by the authors， based on concepts from ref. [131]

4.2 The path to intelligent and cognitive sensing systems

The convergence of artificial intelligence with gas sensing is pivotal for advancing MOF-based sensors beyond basic detection， transforming them into adaptive and intelligent platforms. Liu， Wang et al.[133] have extensively reviewed this paradigm shift， highlighting how AI interventions， including deep learning for feature extraction， drift compensation， and hardware-software convergence through olfactory chips and neuromorphic processors， are providing the core technical support to overcome the long-standing limitations of artificial olfaction. Building on this foundation， specific implementations showcase this power. Gustafson and Wilmer employed a genetic algorithm to computationally identify optimal MOF arrays for methane sensing， maximizing the array's ability to probabilistically identify gas mixtures and moving beyond trial-and-error[134]. Providing a theoretical framework for such arrays， Sousa and Simon framed gas sensing as an inverse problem， introducing a method to evaluate the 'fitness' of a MOF combination based on its resilience to measurement noise， enabling robust quantitative analysis[135].
Beyond data analysis， AI can guide the sensor design process itself. Nurhuda et al.[136] developed a high-throughput computational protocol using density-of-states and binding energy calculations to screen MOF-metal oxide composites for specific disease biomarkers in breath， establishing a powerful 'sensing informatics’ framework for predictive medicine. The ultimate expression of intelligent sensing is the creation of cognitive and neuromorphic systems that mimic biological functions. Wang et al.[137] demonstrated a photonic neuromorphic autonomous perceptual decision system using a heterostructure of ZnO nanorods and Cu3 （HHTP）2 MOFs， which enabled a robotic dog to perform autonomous navigation decisions through multi-signal visual evidence accumulation and processing， bridging neuromorphic materials with bioinspired cognitive architectures.
In a parallel breakthrough for visual intelligence， Kim et al.[138] developed an information-filterable artificial retina， where the pore diameter and functionality of isoreticular Zr-based MOFs were used to modulate synaptic plasticity， mimicking the brain's short-term and long-term memory functions to enable attention-based visual signal processing （Fig.16A~I）. This work， alongside the artificial olfactory memory system using a conductive Ce-HHTP framework demonstrated by Yin， Wang et al.[139]， which exhibits learnable responses to different vapors， lays the foundation for a new generation of bionic sensing systems that can perceive， process， and remember information.
图16 带有MOF层的信息可过滤人工视网膜系统示意图：（A） 人类信息处理；（B） 神经元信号总和；（C， D） 突触信号（短/长期）；（E） 编码模式（注意/刺激）；（F） 系统图像；（G） 信号流示意图；（H） MOF层及离子渗透；（I， J） LTP后的保留[138]

Fig.16 Schematic of the information-filterable artificial retina system with MOF layers. （A） Human information processing. （B） Neuronal signal summation. （C， D） Synaptic signals （short/long-term）. （E） Encoding modes （attentional/stimulus）. （F） System images. （G） Signal flow schematics. （H） MOF layers & ion penetration. （I， J） Retention post-LTP[138].Copyright 2025 Springer Nature

4.3 Multi-modal and extreme-performance sensing

MOFs enable hybrid sensors that combine multiple sensing principles to achieve enhanced functionality and real-world applicability. Ouyang et al. demonstrated a humidity-resistant colorimetric sensing platform by constructing a COF-on-MOF heterostructure （Dye@ZIF-8@COF） （Fig.17A~E）， which leveraged the MOF core for VOC preconcentration and a hydrophobic COF shell for environmental stability， enabling direct visualization of VOC sensing processes with AI-assisted classification of matcha drying stages with 95.74% accuracy[140].
图17 （A） 核壳ZIF-8@COF复合材料的制备示意图；（B） ZIF-8@COF材料的XRD图谱；（C） 复合材料的FT-IR光谱；（D） N2吸附-脱附等温线；（E） 复合材料的孔径分布[140]

Fig.17 （A） Schematic illustration for the preparation of the core-shell ZIF-8@COF composites. （B） XRD patterns of ZIF-8@COF materials. （C） FT-IR spectra of the composite materials. （D） N2 adsorption-desorption isotherms， and （E） pore-size distribution of the composite materials[140]. Copyright 2025 Wiley-VCH

This approach exemplifies the power of multi-modal sensing combined with intelligent data analysis. Similarly， Jaishi， Xian et al.[141] developed a piezoelectric-colorimetric sensor by coating a micro quartz tuning fork with a MOF-thymol blue hybrid， allowing simultaneous mass detection （frequency shift） and visual color change for monitoring plant stress volatiles. Pushing the limits of sensitivity， Deng et al.[142] reported a core-sheath structure （TiO2 core/NH2-MIL-125 sheath） for non-contact nitro-explosive vapor detection.
The MOF sheath preconcentrated the analyte by a factor of 101 while optimal band alignment enabled efficient visible-light harvesting. This synergy resulted in a self-promoting sensing mechanism and a detection limit of ~0.8 ppq for hexogen， a sensitivity that surpasses sniffer dogs and represents the highest reported without pre-concentration. As highlighted by Hu et al.[143]， the future of this field lies in hardware-software co-design， where the synergy between tunable materials like MOFs and intelligent algorithms facilitates a transition from standalone sensors to fully responsive， ecosystem-level monitoring networks.

4.4 The path to commercialization

The transition of MOF-based sensors from laboratory prototypes to commercially viable products hinges on addressing key challenges of scalability， power management， multifunctional integration， and regulatory compliance. Recent advancements demonstrate promising pathways toward this goal across multiple high-value application domains. In healthcare diagnostics， significant progress is evident. Huang et al. developed a deep-learning-enhanced MOF e-skin that integrates biomolecular sensing （glucose， lactic acid）， motion detection， and electrocardiography in a single device， utilizing transformer neural networks for precise signal differentiation and micro-expression recognition， representing a significant advancement toward clinical-grade wearable health monitoring[144]. Yu， Wang et al.[145] developed a groundbreaking photoresponsive nanozyme sensor array using ZnTCPP-based MOFs for neurological disease diagnosis， achieving rapid， multiplexed detection of neurotransmitters in complex biological samples （serum and cerebrospinal fluid） through light-driven catalysis and machine learning pattern recognition. For point-of-care testing， Liu et al.[146] advanced the practical application by developing a 3D Zn-MOF for fluorescence detection of biomarkers and antibiotics， successfully integrating it into test strips and mixed matrix membranes and validating its performance in simulated urine and actual water systems with excellent recovery rates （93.5%~100.5%）.
In the realm of wearable technology， Ren et al.[147] developed a commercially viable piezoresistive sensor using ZIF-8 and reduced graphene oxide in a polyurethane sponge matrix， achieving exceptional performance metrics including ultrahigh sensitivity （243.24 kPa-1， a broad detection range （0~200 kPa）， fast response/recovery （70/80 ms）， outstanding cyclic stability （5000 cycles）， and crucial bacteriostatic properties for long-term wearable comfort. For power management， Lin et al.[148] developed a flexible piezoelectric sensor using a MOF@g-C3N4 heterojunction that achieved self-powering capability through mechanical energy harvesting with wireless data transmission via bluetooth， while the photoactivated μLED platform reported by Lee， Park et al.[1] achieved ultra-low power consumption （587 µW）.
For industrial and environmental safety， Zhong， Mirica et al.[149] made significant strides by developing a scalable templated fabrication method for Cu3（HHTP）2 MOF on polyester textiles， creating large-area （100 cm2） devices capable of simultaneous SO2 sensing at sub-ppm levels， filtration meeting OSHA permissible exposure limits， and catalytic detoxification. In food safety， Liu et al.[150] developed a monitoring platform using MOF-derived amorphous Co3O4 nanozymes fabricated via scalable dielectric barrier discharge plasma， creating colorimetric sensor arrays integrated with smartphone imaging and machine learning for intelligent identification of phenolic compounds in real food samples. For manufacturing scalability， techniques such as layer-by-layer spray-coating[58]， direct self-assembly on textiles[57]， and extrusion printing[100] demonstrate the potential for large-area， cost-effective fabrication. However， significant hurdles remain， including the development of standardized encapsulation methods to ensure long-term stability， the reduction of high-purity ligand costs for mass production， and the establishment of rigorous validation protocols under realistic operating conditions. The future commercialization of MOF sensors will likely be driven by these specialized， high-value applications where their superior selectivity and functionality justify initial costs， gradually expanding to broader consumer markets as manufacturing efficiencies improve.

5 Conclusions and future perspectives

This review has systematically charted the remarkable journey of metal-organic frameworks from novel porous materials to sophisticated， high-performance components in chemiresistive sensing. We have elucidated how the intrinsic properties of MOFs， their unparalleled structural tunability， ultrahigh surface area， and rich host-guest chemistry provide a fundamental design platform that directly addresses the critical limitations of sensitivity and selectivity in traditional sensing materials. The evolution from pioneering studies on often-insulating pristine 3D MOFs to the development of intrinsically conductive 2D and 1D frameworks represents a paradigm shift， enabling room-temperature operation and direct electronic readout of chemical interactions.
The strategic formation of MOF composites and hybrids has further empowered this technology， synergistically combining the molecular sieving and preconcentration capabilities of MOFs with the excellent electrical properties of conductive fillers like carbon nanotubes， graphene， and polymers. Similarly， MOF-derived porous carbons and metal oxides inherit advantageous morphological features from their precursors， offering a powerful route to create highly sensitive and stable sensing architectures with tailored active sites. The integration of these materials into flexible textiles， wearable devices， and multi-modal sensor arrays underscores their practical versatility for real-world applications in health， safety， and environmental monitoring.
Looking forward， the trajectory of MOF-based chemiresistive sensors is unequivocally pointed towards intelligent and cognitive systems. The convergence of tunable MOF materials with artificial intelligence and machine learning， as demonstrated in advanced e-nose platforms， is effectively overcoming the perennial challenge of selectivity in complex environments. The emergence of neuromorphic sensing， where MOFs contribute to artificial olfactory memory and decision-making， heralds a new era of bio-inspired devices that can learn and adapt. Furthermore， the push for commercial viability will hinge on a concerted focus on overcoming remaining challenges： enhancing hydrothermal and mechanical stability through robust material design， developing scalable and cost-effective fabrication techniques like spray-coating and printing， and establishing standardized benchmarking protocols to rigorously evaluate performance under realistic conditions.
In conclusion， the “designer” nature of MOFs positions them not merely as incremental improvements but as transformative enablers for next-generation sensing. By continuing to intertwine rational material synthesis with advanced device engineering and data science， the field is poised to deliver a new class of intelligent， connected， and reliable sensors that will permeate every facet of our lives， from personalized healthcare and environmental protection to industrial safety and beyond.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Lee K, Jo Y M, Sohn M S, Jeon M, Kim C. Nat. Commun., 2025, 16: 9612.

[2]
Feng S, Lu Y, Wang C, Torabi M, Huang X. Nat. Commun., 2025, 16: 9473.

[3]
Chen C X, Cui X L, Xiong Y Y, Lan P C, Tan K, Wei Z W, Niu Z, Shan C, Yang L F, Ma S Q, Su C Y. Nat. Synth, 2026, 5(1): 129.

[4]
Shakil S, Yuan D J, Li M G. J. Electrochem. Soc., 2024, 171(6): 067512.

[5]
Zhang P C, Zhang Y X, Wu F, Xiao W X, Hua W W, Tang Z W, Liu W, Chen S W, Wang Y X, Wu W S, Pan D Q. Nat. Commun., 2025, 16: 2361.

[6]
Akagi K, Naito H, Saikawa T, Kotani M, Yoshikawa H. Sci. Rep., 2024, 14: 12021.

[7]
Peng Y, Shakil S, Yuan D J, Li M G. Anal. Chim. Acta, 2024, 1318: 342921.

[8]
Xie H, Khoshooei M A, Mandal M, Vornholt S M, Hofmann J. Nat. Chem., 2025, 17: 1514.

[9]
Sabat M, Hassan H, Roukos R, Abou Dargham S, Romanos J. Sci. Rep., 2025, 15: 33054.

[10]
Xu S, Su Z, Jiang R, Wu X, Wang J. Sci. Rep., 2025, 15: 5235.

[11]
Shakeel M, Bari A, Shakil S, Hamid A, Habib A. Microchem. J., 2025, 217: 115014.

[12]
Cleeton C, Sarkisov L. Nat. Commun., 2025, 16: 4806.

[13]
O’Shaughnessy M, Glover J, Hafizi R, Barhi M, Clowes R, Chong S Y, Argent S P, Day G M, Cooper A I. Nature, 2024, 630(8015): 102.

[14]
Bari A, Aslam S, Khan H U, Shakil S, Yaseen M, Shahid S, Yusaf A, Afshan N, Shafqat S S, Zafar M N. Plasmonics, 2025, 20(10): 8319.

[15]
Ji Z, Mukherjee S, Andreo J, Sinelshchikova A, Peccati F, Jiménez-Osés G, Wuttke S, Boxer S G. Nat. Chem., 2025, 17(12): 1920.

[16]
Lv Z, Lin R, Yang Y, Lan K, Hung C T. Nat. Chem., 2025, 17: 177.

[17]
Kalmutzki M J, Hanikel N, Yaghi O M. Sci. Adv., 2018, 4(10): eaat9180.

[18]
Jimenez Lopez L, Morales-Ospino R, Castro-Gutiérrez J, Sumbhaniya H O, Sdanghi G, García Dalí S, Celzard A, Fierro V. Int. J. Hydrog. Energy, 2025, 173: 151272.

[19]
Gaikwad S, Kim Y, Gaikwad R, Han S. J. Environ. Chem. Eng., 2021, 9(4): 105523.

[20]
Fan X Y, Wang S, Pan M Y, Pang H, Xu H B. ACS Energy Lett., 2024, 9(6): 2840.

[21]
Shakil S, Akmal M, Zahid R, Azeem M, Bari A, Asghar R. J. Electrochem. Soc., 2025, 172(1): 017514.

[22]
Koo W T, Jang J S, Kim I D. Chem, 2019, 5(8): 1938.

[23]
Campbell M G, Sheberla D, Liu S F, Swager T M, Dincă M. Angew. Chem. Int. Ed., 2015, 54(14): 4349.

[24]
Aubrey M L, Wiers B M, Andrews S C, Sakurai T, Reyes-Lillo S E. Nat. Mater., 2018, 17: 625.

[25]
Rubio-Giménez V, Almora-Barrios N, Escorcia-Ariza G, Galbiati M, Sessolo M, Tatay S, Martí-Gastaldo C. Angew. Chem., 2018, 130(46): 15306.

[26]
Liu Y X, Dong X C, Chen P. Chem. Soc. Rev., 2012, 41(6): 2283.

[27]
Schroeder V, Savagatrup S, He M, Lin S B, Swager T M. Chem. Rev., 2019, 119(1): 599.

[28]
Franke M, Koplin T, Simon U. Small, 2006, 2(1): 36.

[29]
Ammu S, Dua V, Agnihotra S R, Surwade S P, Phulgirkar A, Patel S, Manohar S K. J. Am. Chem. Soc., 2012, 134(10): 4553.

[30]
Park C, Baek J W, Shin E, Kim I D. ACS Nanosci. Au, 2023, 3(5): 353.

[31]
Torsi L, Magliulo M, Manoli K, Palazzo G. Chem. Soc. Rev., 2013, 42(22): 8612.

[32]
Lu S, Jia H X, Hummel M, Wu Y N, Wang K L, Qi X Q, Gu Z R. RSC Adv., 2021, 11(8): 4472.

[33]
Le T A, Hai N D, Tran T T N, Trinh K T L, Tran N Q. Chem. Commun., 2025, 61(72): 13543.

[34]
Chen J, Mei Q, Chen Y, Marsh C, An B. J. Am. Chem. Soc., 2022, 144: 11969.

[35]
Li J G, Kumar A, Johnson B A, Ott S. Nat. Commun., 2023, 14: 4388.

[36]
Yan M L, Bowman Z, Knepp Z J, Peterson A, Fredin L A, Morris A J. J. Phys. Chem. Lett., 2024, 15(48): 11919.

[37]
Furukawa H, Cordova K E, O’Keeffe M, Yaghi O M. Science, 2013, 341(6149): 1230444.

[38]
Chen E X, Yang H, Zhang J. Inorg. Chem., 2014, 53(11): 5411.

[39]
Wargocki P, Wyon D P, Baik Y K, Clausen G, Fanger P O. Indoor Air, 1999, 9(3): 165.

[40]
Yang H, He X W, Wang F, Kang Y, Zhang J. J. Mater. Chem., 2012, 22(41): 21849.

[41]
Chen E X, Fu H R, Lin R, Tan Y X, Zhang J. ACS Appl. Mater. Interfaces, 2014, 6(24): 22871.

[42]
DMello M E, Sundaram N G, Singh A, Singh A K, Kalidindi S B. Chem. Commun., 2019, 55(3): 349.

[43]
Long J L, Wang S B, Ding Z X, Wang S C, Zhou Y G, Huang L, Wang X X. Chem. Commun., 2012, 48(95): 11656.

[44]
Zhang Y M, Liu Q, Sun Q Q, Li H, Shen J S, Liu H Y, Chen W M, Zhang Y X, Chen Y L. ACS Sens., 2023, 8(11): 4353.

[45]
Liu X, Wu J, Fan C, Zhang Y W, Quan W J, Li J Z, Hu N T, Yang J H, Zeng M, Yang Z. ACS Sens., 2025, 10(9): 6464.

[46]
Liu X, Wu J, Li J, Wang Q, Wang Y. Adv. Funct. Mater., 2025, 36: e11152.

[47]
Shang S C, Du C S, Liu Y X, Liu M H, Wang X Y, Gao W Q, Zou Y, Dong J C, Liu Y Q, Chen J Y. Nat. Commun., 2022, 13: 7599.

[48]
Sun L, Campbell M G, Dincă M. Angew. Chem. Int. Ed., 2016, 55(11): 3566.

[49]
Feng D, Lei T, Lukatskaya M R, Park J, Huang Z. Nat. Energy, 2018, 3: 30.

[50]
Sheberla D, Bachman J C, Elias J S, Sun C J, Shao-Horn Y, Dincă M. Nat. Mater., 2017, 16(2): 220.

[51]
Talin A A, Centrone A, Ford A C. Science, 2014, 343: 66.

[52]
Campbell M G, Liu S F, Swager T M, Dincă M. J. Am. Chem. Soc., 2015, 137(43): 13780.

[53]
Benedetto G, Damacet P, Shehayeb E O, Fabusola G, Simon C M, Mirica K A. ACS Sens., 2025, 10(10): 7787.

[54]
Huang C, Sun W, Jin Y. Angew. Chem. Int. Ed., 2024, 63: e202313591.

[55]
Huang C, Shang X, Zhou X. Nat. Commun., 2023, 14: 3850.

[56]
Smith M K, Jensen K E, Pivak P A, Mirica K A. Chem. Mater., 2016, 28(15): 5264.

[57]
Smith M K, Mirica K A. J. Am. Chem. Soc., 2017, 139(46): 16759.

[58]
Yao M S, Lv X J, Fu Z H, Li W H, Deng W H, Wu G D, Xu G. Angew. Chem. Int. Ed., 2017, 56(52): 16510.

[59]
Shan Z, Xiao J Z, Wu M. Angew. Chem., 2024, 136: e202401679.

[60]
Lu G R, Zong B Y, Tao T, Yang Y H, Li Q J, Mao S. ACS Sens., 2024, 9(4): 1916.

[61]
More M S, Bodkhe G A, Singh F, Dole B N, Tsai M L, Hianik T, Shirsat M D. Synth. Met., 2023, 296: 117357.

[62]
Lin Y, Li W H, Wen Y Y, Wang G E, Ye X L, Xu G. Angew. Chem. Int. Ed., 2021, 60(49): 25758.

[63]
Pedersen K S, Perlepe P, Aubrey M L. Nat. Chem., 2018, 10: 1056.

[64]
Kung C W, Platero-Prats A E, Drout R J. ACS Appl. Mater. Interfaces, 2018, 10: 30532.

[65]
Schroeder V, Savagatrup S, He M, Lin S B, Swager T M. Chem. Rev., 2019, 119(1): 599.

[66]
Mourdikoudis S, Dutta S, Kamal S. Adv. Mater., 2025, 37: 2415399.

[67]
Jiang S, Mo X, Zhao X. Anal. Sens., 2024, 4: e202400024.

[68]
Travlou N A, Singh K, Rodríguez-Castellón E, Bandosz T J. J. Mater. Chem. A, 2015, 3(21): 11417.

[69]
Ko M, Aykanat A, Smith M, Mirica K. Sensors, 2017, 17(10): 2192.

[70]
Li H, Wang K C, Sun Y J, Lollar C T, Li J L, Zhou H C. Mater. Today, 2018, 21(2): 108.

[71]
Han X, Godfrey H G W, Briggs L. Nat. Mater., 2018, 17: 691.

[72]
Liu G, Chernikova V, Liu Y. Nat. Mater., 2018, 17: 283.

[73]
Penner R M. Acc. Chem. Res., 2017, 50(8): 1902.

[74]
Yang F, Kung S C, Cheng M, Hemminger J C, Penner R M. ACS Nano, 2010, 4(9): 5233.

[75]
Koo W T, Qiao S, Ogata A F. ACS Nano, 2017, 11: 9276.

[76]
Hobday C L, Woodall C H, Lennox M J. Nat. Commun., 2018, 9: 1429.

[77]
Hao J, Babu D J, Liu Q. Chem. Mater., 2021, 33: 4035.

[78]
Kim Y, Kim T, Kim J, Lee N, Yoo S H, Kwon H T, Eum K. J. Membr. Sci., 2025, 723: 123963.

[79]
Li Y M, Cai X T, Chen C J, Li L B, Zhao Y L, Wu W F, Wei Y Y. J. Membr. Sci., 2026, 741: 125069.

[80]
Chen Y J, Xu Y M, Ye X L. Nat. Commun., 2025, 16: 6634.

[81]
Ding J, Qiao J, Zheng Z. Talanta, 2025, 285: 127444.

[82]
Yao M S, Xiu J W, Huang Q Q. Angew. Chem., 2019, 131: 15057.

[83]
Sun Y F, Liu S B, Meng F L, Liu J Y, Jin Z, Kong L T, Liu J H. Sensors, 2012, 12(3): 2610.

[84]
Yao M S, Tang W X, Wang G E, Nath B, Xu G. Adv. Mater., 2016, 28(26): 5229.

[85]
Tian H, Fan H, Li M. ACS Sens., 2016, 1: 243.

[86]
DMello M E, Sundaram N G, Kalidindi S B. Chem., 2018, 24(37): 9220.

[87]
Rodenas T, Luz I, Prieto G. Nat. Mater., 2015, 14: 48.

[88]
Drobek M, Kim J H, Bechelany M, Vallicari C, Julbe A, Kim S S. ACS Appl. Mater. Interfaces, 2016, 8(13): 8323.

[89]
Zhou T T, Sang Y T, Wang X X, Wu C Y, Zeng D W, Xie C S. Sens. Actuat. B Chem., 2018, 258: 1099.

[90]
Weber M, Kim J H, Lee J H, Kim J Y, Iatsunskyi I, Coy E, Drobek M, Julbe A, Bechelany M, Kim S S. ACS Appl. Mater. Interfaces, 2018, 10(40): 34765.

[91]
Murugesan T, Kumar R R, Ranjan A, Lu M Y, Lin H N. Sens. Actuat. B Chem., 2024, 402: 135106.

[92]
Min H, Kwon O, Lee J. Adv. Mater., 2024, 36: 2309041.

[93]
Roh H, Kim D H, Cho Y. Adv. Mater., 2024, 36: 2312382.

[94]
Lu Y, Fang Y G, Chen Y. Nat. Commun., 2025, 16: 3397.

[95]
Xia W, Mahmood A, Zou R Q, Xu Q. Energy Environ. Sci., 2015, 8(7): 1837.

[96]
Li X, Zheng S, Jin L. Adv. Energy Mater., 2018, 8: 1800716.

[97]
Kim C R, Uemura T, Kitagawa S. Chem. Soc. Rev., 2016, 45(14): 3828.

[98]
Koo W T, Cha J H, Jung J W. Adv. Funct. Mater., 2018, 28: 1802575.

[99]
Ko K Y, Song J G, Kim Y. ACS Nano, 2016, 10: 9287.

[100]
Hassan K, Trong Tran A T, Jalil M A, Tung T T, Nine M J, Losic D. ACS Sens., 2025, 10(6): 3931.

[101]
Indra A, Song T, Paik U. Adv. Mater., 2018, 30(39): 1705146.

[102]
Wang X F, Song X Z, Sun K M, Cheng L, Ma W. Polyhedron, 2018, 152: 155.

[103]
Lü Y Y, Zhan W W, He Y, Wang Y T, Kong X J, Kuang Q, Xie Z X, Zheng L S. ACS Appl. Mater. Interfaces, 2014, 6(6): 4186.

[104]
Wang J L, Zhai Q G, Li S N. Inorg. Chem. Commun., 2016, 63: 48.

[105]
Gao P Y, Liu R, Huang H H, Jia X, Pan H B. RSC Adv., 2016, 6(97): 94699.

[106]
Li W, Wu X, Han N. Sens. Actuators B, 2016, 225: 158.

[107]
Jang J S, Koo W T, Kim D H, Kim I D. ACS Cent. Sci., 2018, 4(7): 929.

[108]
Rui K, Wang X S, Du M, Zhang Y, Wang Q Q, Ma Z Y, Zhang Q, Li D S, Huang X, Sun G Z, Zhu J X, Huang W. ACS Appl. Mater. Interfaces, 2018, 10(3): 2837.

[109]
Choi S J, Choi H J, Koo W T, Huh D, Lee H, Kim I D. ACS Appl. Mater. Interfaces, 2017, 9(46): 40593.

[110]
Wang Y T, Lü Y Y, Zhan W W, Xie Z X, Kuang Q, Zheng L S. J. Mater. Chem. A, 2015, 3(24): 12796.

[111]
Cui Y F, Jiang W, Liang S, Zhu L F, Yao Y W. J. Mater. Chem. A, 2018, 6(30): 14930.

[112]
Li W, Wu X, Liu H. New J. Chem., 2015, 39: 7060.

[113]
Jo Y M, Kim T H, Lee C S, Lim K, Na C W, Abdel-Hady F, Wazzan A A, Lee J H. ACS Appl. Mater. Interfaces, 2018, 10(10): 8860.

[114]
Koo W T, Choi S J, Kim S J, Jang J S, Tuller H L, Kim I D. J. Am. Chem. Soc., 2016, 138(40): 13431.

[115]
Koo W T, Choi S J, Jang J S, Kim I D. Sci. Rep., 2017, 7: 45074.

[116]
Koo W T, Yu S, Choi S J, Jang J S, Cheong J Y, Kim I D. ACS Appl. Mater. Interfaces, 2017, 9(9): 8201.

[117]
Koo W T, Jang J S, Choi S J, Cho H J, Kim I D. ACS Appl. Mater. Interfaces, 2017, 9(21): 18069.

[118]
Jang J S, Koo W T, Choi S J, Kim I D. J. Am. Chem. Soc., 2017, 139(34): 11868.

[119]
Chen L Y, Luque R, Li Y W. Chem. Soc. Rev., 2017, 46(15): 4614.

[120]
Yang D J, Kamienchick I, Youn D Y, Rothschild A, Kim I D. Adv. Funct. Mater., 2010, 20(24): 4258.

[121]
Oh M H, Yu T, Yu S H. Science, 2013, 340: 964.

[122]
Yang W, Fang B J, Zhang Y H, Ma G M, Meng H, Liu S T. Sens. Actuat. B Chem., 2025, 422: 136658.

[123]
Xie W, Huang X, Zhu C. Adv. Mater., 2024, 36: 2313920.

[124]
Gui W, Du M, Zhang X. J. Alloys Compd., 2025, 1020: 179341.

[125]
Cao H, Chen L, Huang Z X, Ding X E, Liu Y Y, Yu J S, Xu F. Inorg. Chem., 2025, 64(23): 11600.

[126]
Chen Z, Hua Y N, Chen B, Feng J W, Liu G, Jiang B. Chem. Eng. J., 2025, 506: 160004.

[127]
Liu B T, Huang Y C, Dong Z F, Wu Y H, Huang H J, Bing L H, Zhu Z G, Wei J F, Wang R. Polym. Degrad. Stab., 2025, 241: 111611.

[128]
Liao J H, Li B F, Ran L F, Han J, Yang C, Li Z J, Jiao T F. J. Environ. Chem. Eng., 2025, 13(6): 119324.

[129]
Kidanemariam A, Cho S. Sensors, 2025, 25(21): 6539.

[130]
Cui B, Lang X, Ren Z. Sens. Actuators B, 2025, 423: 136817.

[131]
Zhou Q, Ding Q, Geng Z. Nano-Micro Lett., 2024, 17: 50.

[132]
Zha X, Zhou N, Hu M. ACS Sens., 2025, 10: 6656.

[133]
Liu M H, Ren R R, Zhou X Y, Zhu S, Wang T. Chem. Commun., 2025, 61(55): 9996.

[134]
Gustafson J A, Wilmer C E. ACS Sens., 2019, 4(6): 1586.

[135]
Sousa R, Simon C M. ACS Sens., 2020, 5(12): 4035.

[136]
Nurhuda M, Otake K I, Kitagawa S, Packwood D M. Adv. Theory Simul., 2025, 8(5): 2401404.

[137]
Wang Y, Li S X, Yao Y X. Adv. Funct. Mater., 2025, 36: e09600.

[138]
Kim S, Kwon O, Kim S. Nat. Commun., 2025, 16: 162.

[139]
Yin X, Zhang H, Qiao X. Nat. Commun., 2024, 15: 8409.

[140]
Ouyang Q, Rong Y N, Xia G F, Chen Q S, Ma Y J, Liu Z H. Adv. Sci., 2025, 12(12): 2411621.

[141]
Jaishi L R, Ding W, Kittelson R A, Tsow F, Xian X J. ACS Sens., 2024, 9(12): 6553.

[142]
Deng W H, Yao M S, Zhang M Y. Natl. Sci. Rev., 2022, 9: nwac143.

[143]
Hu C C, Zhang W, Yang J Q, Pei Y H, Tan X Y, Dong B, Song H W, Xu L. Chem. Soc. Rev., 2025, 54(23): 11302.

[144]
Ke X Y, Duan Y, Duan Y. Device, 2025, 3: 100650.

[145]
Yu K, Lu S, Qiu K. Adv. Sci., 2025, 12: e05333.

[146]
Liu T, Ji M, Hao H. Appl. Organomet. Chem., 2025, 39: e70079.

[147]
Ren M C, Huo W J, Liu T, Cao M H. ACS Appl. Electron. Mater., 2025, 7(5): 2048.

[148]
Lin Q B, Fu J, Peng Q, Chen S X, Wu Y B, Shi Q S. Nano Mater. Sci., 2025, doi: 10.1016/j.nanoms.2025.09.008.

[149]
Zhong Z R, Damacet P, Sánchez-González E. Chem, 2025, 11: 102580.

[150]
Liu C, Liu N, Xia Q, Cao Y, Huang Q. Food Chem., 2025, 490: 145102.

Options
Outlines

/

〈 〉