Seeing the Unseeable

How Graphene Liquid Cells Revolutionize Our View of Nanoscale Reactions

The Quest for Atomic Clarity

For decades, scientists studying chemical reactions at the nanoscale faced a fundamental dilemma: transmission electron microscopes (TEMs) require high vacuum to function, yet most transformative reactions occur in liquids. Early liquid cell designs used silicon nitride windows, but their thickness (20-50 nm) scattered electrons severely, limiting resolution to ~10 nm—like viewing cells through frosted glass.

"Watching real-time chemical reactions in liquids at the atomic-scale is a dream for chemists and physicists" 8

This changed in 2012 when researchers at Lawrence Berkeley National Lab pioneered a breakthrough: sealing liquids between atom-thin graphene sheets 8 . Overnight, resolution leaped to the atomic scale, enabling real-time observation of nanoparticles forming, transforming, and interacting in their native liquid environments.

Resolution Comparison
Key Advantages
  • Atomic-scale resolution
  • Native liquid environment
  • Real-time observation
  • Reduced beam damage

Why Graphene? The Physics of a Perfect Window

Graphene structure
Graphene's hexagonal carbon lattice enables its remarkable properties
Atomic Thinness

At just 0.34 nm thick—a single carbon atom layer—graphene is nearly invisible to electron beams. This minimizes scattering, allowing >90% of electrons to pass through versus <50% for silicon nitride windows 3 9 .

Strength and Impermeability

Despite its thinness, graphene withstands pressure differences exceeding 1 atm. Its covalent carbon lattice blocks even small molecules like water vapor, preventing leakage in TEM vacuum 1 3 .

Radical Mitigation

When electron beams strike water, they generate reactive radicals (e.g., •OH, H•) that damage samples. Graphene acts as a "radical scavenger," absorbing these species 7 4 .

Thermal Conductivity

Graphene efficiently dissipates heat from beam exposure (5,000 W/mK vs. 150 W/mK for silicon nitride), reducing sample drift and boiling artifacts 3 .

Graphene: 5000 W/mK
SiN: 150 W/mK
Electron Transmission

Capturing Chemistry in Action: The Platinum Growth Experiment

Methodology: Sealing a Reaction in Atomic Pockets

In the landmark 2012 study 8 , researchers captured platinum nanocrystal growth in real time:

  1. Grid Preparation: Copper foil coated with multilayer graphene (3–5 layers) was etched using sodium persulfate, transferring graphene onto gold TEM grids with holey carbon supports.
  2. Liquid Encapsulation: A droplet of platinum salt solution (H₂PtCl₆ in water) was pipetted onto two graphene-coated grids.
  3. In Situ TEM Imaging: Using the TEAM I microscope (resolution: 0.5 Å), movies were recorded at 1–5 frames/second.
TEM microscope
Transmission Electron Microscope used for GLC experiments

Revelations: Birth and Transformation of Nanocrystals

The movies revealed two growth pathways with atomic precision:

  • Classical Growth: Pt²⁺ ions slowly reduced, adding atoms to existing crystals.
  • Oriented Coalescence: Colliding nanoparticles fused along identical crystallographic planes ({111} facets), then reshaped into single crystals.
Coalescence Events in Platinum Nanoparticle Growth
Nanoparticle Size (nm) Coalescence Rate (events/min) Preferred Attachment Plane
2–3 nm 3.2 ± 0.8 {111}
5–6 nm 1.1 ± 0.3 {111}

Data revealed >80% of collisions occurred along identical crystallographic planes, enabling defect-free fusion 8 .

How GLCs Transformed Resolution Limits
Liquid Cell Type Window Thickness Resolution
Silicon nitride windows 20–50 nm ~5 nm
Graphene liquid cells 0.34 nm 0.14 nm

Thinner windows and liquid layers enable atomic-scale imaging previously impossible 3 8 9 .

Decoding Nano-Reactions: Key Discoveries Enabled by GLCs

1. Nanoparticle Coalescence Dynamics

The platinum study showed that particles under 5 nm don't merge randomly. They:

  • Rotate to align crystal lattices before fusion
  • Fuse along low-energy planes ({111} for Pt)
  • Reshape into equilibrium structures within seconds

This overturned models assuming chaotic "sticking" and revealed strategies for defect-free nanostructures 8 9 .

2. Confined Liquids Behave Strangely

Water trapped in 10-nm GLC pockets exhibits bizarre properties:

  • Viscosity Surge: Diffusion coefficients for 5-nm gold nanoparticles drop to 0.12 nm²/s—10⁹× slower than in bulk water 2 .
  • Non-Gaussian Motion: Nanoparticles show "stop-and-go" diffusion due to molecular crowding 2 .
3. Etching and Growth Mechanisms

GLC-TEM visualized:

  • Site-Selective Etching: Iodide ions preferentially remove gold atoms from high-energy nanocrystal edges 4 .
  • Electrochemical Dynamics: Silicon nanoparticles expand by 300% during lithium-ion battery charging 3 .
Growth vs. Etching Mechanisms Revealed by GLCs
Process System Key Observation Implication
Nanoparticle growth Pt in aqueous solution Oriented attachment along {111} planes Defect-free crystal growth possible
Oxidative etching Au nanorods + I⁻/O₂ Corner dissolution creates concave facets Shape control via surface chemistry
Battery cycling Si nanoparticles + LiPF₆ Anisotropic swelling cracks SEI layers Explains battery degradation

Real-time imaging reveals kinetic pathways for designing functional nanomaterials 4 8 3 .

The Scientist's Toolkit: Essential Components for GLC Experiments

Key Reagents and Tools for GLC-TEM
Material/Reagent Function Technical Notes
Multilayer graphene on Cu Provides ultra-thin, impermeable cell windows CVD-grown; 3–5 layers optimize seal strength & resolution 4
Sodium persulfate (Na₂S₂O₈) Etches copper foil without damaging graphene Avoids HNO₃ (creates defects) 4
Gold TEM grids Supports graphene membranes; inert to chemicals Holey carbon film (e.g., Quantifoil) aids liquid trapping 5
Heavy water (D₂O) Reduces bubble formation during imaging Radiolysis slower vs. H₂O 1

Beyond Materials Science: Biology in Its Native Habitat

Amyloid fibrils
Amyloid Fibrils

GLC-TEM captured intermediate fibril states that rapidly disintegrate, revealing early-stage instability in Alzheimer's-related proteins 7 .

Ferritin
Ferritin

Iron-storing proteins were imaged at 2.7 nm resolution, showing iron-core dynamics in physiological buffers 9 3 .

Virus
Virus Assembly

Hepatitis B virus capsids self-assembled in GLCs while avoiding beam-induced denaturation 7 .

Challenges and Horizons: Where GLCs Go Next

Persistent Hurdles
  • Random Liquid Pockets: Blister formation is stochastic. Solution: Lithographic wells now pattern uniform cells (e.g., 100-nm wells in hBN spacers) .
  • Hydrophobicity: Graphene repels water. Fix: UV-ozone treatment increases wettability 5 .
  • Radiolysis Control: Beams still alter chemistry. Strategy: Scavengers like iodide mitigate radical effects 6 .
Next Frontiers
  • Cryo-GLCs: Freezing reactions mid-flow for cryo-ET of cellular processes.
  • Integrated Electrodes: Real-time voltammetry during TEM imaging .
  • AI-Driven Analysis: Automating particle tracking in terabyte-scale movie datasets 2 .

"This technique will become instrumental in answering questions regarding the synthesis of materials in liquids at the atomic scale" 8

Conclusion: A New Era of Atomic-Scale Storytelling

Graphene liquid cells transformed TEM from a static snapshot tool into a dynamic atomic-scale cinema. By sealing attoliter volumes behind graphene's invisible walls, scientists now scrutinize chemistry as it happens—from nanoparticles forging new bonds to proteins folding in water.

With each innovation in cell design and imaging, we move closer to a fundamental truth: in the nanorealm, seeing is not just believing—it's understanding.

References