Metal–organic frameworks can fall apart under the electron beam used to study them. Yet advanced transmission electron microscopy is turning that apparent contradiction into a way to see their atoms, defects, growth, and transformations.

Metal–organic frameworks are porous crystalline materials composed of inorganic metal nodes and organic ligands. Their tunable topologies, large specific surface areas, and adjustable chemical compositions support applications from catalysis and gas separation to energy storage and chemical sensing.

Structure–activity relationships can guide the rational design and applications of flexible and functional MOFs, but atomic-scale determination of the crystal structures is a well-founded prerequisite for understanding those relationships. Surfaces, interfaces, defects, and host–guest interactions are the main microstructures that directly affect MOF properties.

Understanding growth, transformation mechanisms, and evolution pathways under working conditions is also crucial for improving applications. Using TEM for MOF characterization is challenging because MOFs are extremely sensitive to electron-beam irradiation.

That fundamental hindrance originates from their organic components and the coordination bonds linking organic parts to metals. Electron irradiation can cause structural decomposition before detection is complete, or remove intrinsic characteristics during acquisition, limiting the feasibility and utility of TEM research on MOFs.

To make TEM suitable for beam-sensitive materials, researchers developed low electron dose and low temperature strategies, along with three-dimensional electron diffraction, direct-detection electron-counting cameras, and integrated differential phase-contrast scanning transmission electron microscopy.

The main radiation damage mechanisms under a high-energy incident electron beam include radiolysis, knock-on damage, and thermal effects. In practical situations, identifying the predominant mechanism can help determine the proper method to minimize damage.

Reducing the electron dose is a general solution for MOFs because all these electron-beam-induced irradiation damages are dose dependent. A preliminary assessment is required to determine whether low-dose conditions can maintain crystallinity stability. Electron diffraction is an effective and feasible way to determine the electron dose that a MOF can withstand: its pattern changes as dose increases, indicating structural changes and the appearance of disordered phases.

The maximum electron dose depends on the material and TEM operating conditions. Under a three-hundred-kilovolt accelerated electron beam, MIL-101 chromium can withstand about sixteen electrons per square ångström, UiO-66 zirconium begins damage from ten to twenty electrons per square ångström, and ZIF-8 zinc about twenty-five electrons per square ångström.

Because in situ TEM requires longer irradiation than static characterization, it needs an electron dose well below the damage threshold to preserve the region of interest. Cryo-TEM preserves MOF stability over prolonged irradiation times because cryogenic temperatures diminish radiolysis and compensate for thermal effects to a certain extent.

At liquid nitrogen temperature, high-resolution TEM imaged the complete pore structure and crystal lattice periodicity of MOF-5 zinc nanocrystals. Cryogenic temperatures also allowed researchers to elucidate the ordered internal architecture of large-area conductive two-dimensional copper two TCPP MOF films by high-resolution TEM and electron diffraction, while electron diffraction determined highly porous CAU-7 bismuth structures at one hundred twenty kelvin.

Three-dimensional electron diffraction allows effective electron-diffraction data collection, followed by ab initio structure determination and analysis. A single-crystal dataset is a sequence of diffraction patterns recorded at different tilt angles of the TEM goniometer.

The corresponding diffraction peaks are produced from that sequence, and three-dimensional electron diffraction has evolved from stepwise strategies to faster continuous data collection. Continuous crystal rotation speeds data acquisition while still obtaining relatively accurate and complete electron-diffraction intensities as a movie.

By limiting goniometer tilt, the strategy is particularly suitable for crystals with low symmetry. Continuous rotation electron diffraction has resolved the structures of several types of MOFs and has supplemented relatively ambiguous X-ray diffraction data in reported work.

For example, UU-100 cobalt was solved as a tetrahedral unit cell with lattice parameters of twenty-seven point three ångström, twenty-seven point three ångström, and nineteen point six ångström, a possible P four over m b m space group, and rectangular channels with elliptical pores.

Conventionally, electron diffraction reveals average spatial information, such as periodic crystal structures or abnormal macroscopic features. Its inherent deficiency in probing local structures makes TEM and STEM imaging indispensable for resolving MOF structures.

Scanning electron diffraction based on four-dimensional scanning transmission electron microscopy was proposed to overcome those limitations. Two-dimensional diffraction patterns scanned across the sample reflected defect-nanodomain size, morphology, local orientation, and spatial distribution in single-crystal UiO-66 hafnium particles.

A focused electron probe achieved two-to-five-nanometer spatial resolution with a convergence angle of about one milliradian, revealing interfaces between reo and fcu domains on the two-one-one plane. Figure 2 presents two advanced electron-diffraction strategies.

In panel a, three-dimensional electron diffraction collects a sequence of patterns as a single crystal is tilted through different angles; panel b illustrates scanning electron diffraction, recording a transmission pattern at each probe position. Panel c shows a defect-engineered UiO-66(Hf) particle under cryogenic, gas, liquid, vacuum, and heating conditions, pairing images with diffraction patterns and structural models.

Together, the figure shows how electron diffraction can connect local structure, defects, and changing environments. Advances in spherical-aberration correctors and contrast-transfer-function correction have greatly enhanced TEM imaging of atomic-scale information.

Highly sensitive scintillators and optical fibers can also promote a high signal-to-noise ratio in detecting signal electrons. The problem with conventional cameras is that photons generated by electron–scintillator interactions must be converted into an electrical signal, and that conversion consumes a significant portion of the electrons essential for imaging.

As a result, conventional cameras lack the sensitivity needed to collect high-quality data at a very low electron-beam dose. The coexistence of ordered missing-linker and missing-cluster defects in UiO-66 zirconium was discovered at sub-ångström resolution using low-dose TEM and electronic crystallography.

The missing-linker defects were first identified as a topology of bcu net because of defect-terminating formate ligands. High-quality HRTEM images were essential to unambiguously resolve all structural components by three-dimensional reconstruction. The reconstruction revealed an eight-connected network with zirconium six oxygen eight clusters, BDC linkers, and terminating formate groups.

In TEM mode, a coherent electron beam causes contrast reversion with defocus, so a series of defocused images and crystallographic analysis is required to determine true structural information. STEM mode uses a convergent electron beam and incoherent phase-scattering imaging, giving it higher resolution.

High-angle annular dark-field imaging produces Z-contrast, with contrast proportional to the square of the atomic number. Despite the difficulty of imaging low atomic number elements, STEM images can directly identify elements with relatively obvious contrast differences.

STEM therefore has advantages for imaging non-periodic local structures such as defects, surfaces, interfaces, and deformations, and for analyzing chemical elements. To achieve a lower dose, higher signal-to-noise ratio, and better contrast, iDPC-STEM offers high efficiency in collecting electron signals.

It uses four-quadrant segmented detectors in the new generation of Cs-STEM without additional commercial equipment. This direct electronic phase-imaging mode produces contrast approximately linear with atomic number, reflecting electrostatic potential information in the lattice projection.

Light and heavy elements can be distinguished simultaneously at sub-ångström resolution. Sufficient electron utilization and filtering of non-integrable noise allow iDPC-STEM to achieve high resolution and signal-to-noise ratio under an extremely low electron dose.

Figure five traces iDPC-STEM from measurement to structural image: the electron beam is deflected by the sample’s potential field, and four detector segments record separate signals that are combined into the DPC image. A two-dimensional integration then produces the iDPC image of MIL-101(Cr), whose magnified view matches the structural model.

The corresponding FFT shows information transfer up to one point eight angstroms, illustrating how this phase-imaging mode reveals the material’s cage structure. In a practical comparison of HAADF-STEM and iDPC-STEM, MIL-101 chromium cages were identified with the same resolution of four point seven ångström, using a beam current of two picoamps and a total dose of fifty-four electrons per square ångström.

HAADF-STEM can serve as a reference for outlining structures because it is hardly affected by crystal-zone-axis deflection, although it has limited ability to image light atoms. Compared with traditional STEM imaging, iDPC-STEM offers potential for low-dose imaging of high-crystallinity, beam-sensitive materials such as MOFs.

Its local-structure performance is comparable to cryogenic DDEC cameras, and its images are highly interpretable because contrast directly correlates with atomic number. Figure seven uses iDPC-STEM, alongside CTF-corrected HRTEM, to visualize MIL-101(Cr) surfaces prepared with hydrofluoric acid, acetic acid, or no additive.

The images resolve super tetrahedrons and two distinct {111} surface terminations, modeled with complete twenty-nine and thirty-four angstrom cages. The final panels extend the approach to FeNi-BA-T, where highlighted regions and the structural model identify missing ligands, showing how imaging can reveal surface structures and linker defects directly.

The liquid phase is an important synthetic condition for MOFs, where chemical interactions typically generate many tiny intermediates that quickly reach equilibrium. In situ liquid-cell TEM provides insights into MOF synthesis in solution by directly visualizing nucleation, growth, and self-assembly dynamics.

Liquid-phase synthesis conditions and mechanisms can be investigated through contrast, morphology, and growth behavior, while concentration, temperature, and node-to-linker ratio can subsequently be verified by ex situ TEM. Because MOFs are beam sensitive, liquid synthesis studies must emphasize electron tolerance and methods to reduce damage, while controlling the surface chemistry of the liquid-cell viewing membrane together with electron-beam conditions.

Figure eight combines in situ liquid-cell electron microscopy with schematics to show two solution-phase formation pathways. For ZIF-8, the sequence moves from a homogeneous solution through phase separation and condensation into an amorphous cluster, before crystallizing into nanocubes.

For the two-dimensional Hf-MOF, the images and diagrams depict cluster formation, nucleation, and crystallization into an extended ordered network, revealing intermediate stages that would be hidden in an endpoint-only measurement. MOF periodic frameworks can undergo reversible lattice transformations called pore breathing when temperature or pressure fluctuates, or when guest molecules are adsorbed and desorbed.

In situ environmental TEM directly visualized MIL-53 chromium lattice changes caused by water adsorption and desorption during temperature cycles from heating to three hundred degrees Celsius and cooling to twenty-seven degrees Celsius, with a cumulative dose of about five electrons per square ångström.

The pore breathing was initially triggered by adsorption of the first single water molecule per unit cell at three hundred degrees Celsius. Those molecules were anchored by hydrogen bonds with bridging mu-OH groups and remained stable during later temperature-modulated adsorption and desorption.

Figure nine shows in situ electron-beam irradiation of MIL-101, with iDPC-STEM images before and after exposure. White and red circles mark the initial and final pore positions at the crystal edge, revealing pore evolution that depends on the crystal plane and local position.

This matters because it directly visualizes irradiation-driven crystal shrinkage and deformation at molecular-scale resolution, helping clarify how MOF structures respond during electron microscopy. MOFs are susceptible to electron-beam irradiation even as TEM provides powerful structural information.

Low electron dose and temperature are key experimental considerations for preserving structural integrity. Advanced TEM technologies, including three-dimensional electron diffraction, direct-detection electron-counting cameras, and iDPC-STEM, have enabled groundbreaking discoveries beyond traditional TEM-based methods.

Static structural characterization reveals surfaces, interfaces, defects, and host–guest interactions, while dynamic exploration provides insights into MOF formation, phase transitions, pore breathing, and on-demand structural modification under electron-beam irradiation.

The review’s central message is that low dose, low temperature, and advanced TEM methods make fragile MOFs accessible to atomic-scale and real-time study. That matters because both structure and structural change govern their function.