Ask someone to describe how graphene is made and you will usually hear about chemical vapor deposition — hot furnaces, copper foils, atomically perfect films. That is how the highest-quality graphene is made. It is not how most graphene is made.
By weight, the overwhelming majority of graphene sold commercially is produced by liquid-phase exfoliation (LPE) and its variants. If you have bought graphene nanoplatelets by the kilogram, you have almost certainly bought an LPE product. Understanding the method explains a great deal about why commercial graphene behaves the way it does.
The Basic Idea
Graphite is a stack of graphene layers held together by van der Waals forces. Within each layer, carbon atoms are bound covalently and extremely strongly. Between layers, the attraction is comparatively weak. Exfoliation means supplying enough energy to overcome the interlayer attraction without destroying the layers themselves.
Liquid-phase exfoliation does this by suspending graphite in a liquid and applying mechanical energy — usually ultrasonication or high-shear mixing. The liquid does two jobs. It transmits the energy that separates the sheets, and it prevents the freshly separated sheets from immediately restacking.
That second job is the harder one, and it is where solvent choice becomes critical.
Why Solvent Selection Dominates the Process
Separated graphene sheets are only stable in suspension if the energy cost of creating a graphene-liquid interface is roughly equal to the cost of creating a graphene-graphene interface. In practice this means matching the surface energy of the solvent to that of graphene, which sits at roughly 70 mJ/m².
A small number of organic solvents — N-methyl-2-pyrrolidone (NMP) and dimethylformamide (DMF) being the classic examples — match well and produce good suspensions. Both are also high-boiling, difficult to remove completely, and carry meaningful health and handling concerns. NMP in particular is subject to restrictions in several jurisdictions.
This creates the central tension in LPE. The solvents that work best are the ones you least want in your final product. Manufacturers respond in three ways:
- Solvent exchange. Exfoliate in NMP, then transfer the graphene into a more benign carrier. Effective but adds cost and process steps.
- Aqueous surfactant systems. Water plus a stabilizer such as sodium cholate or a polymeric dispersant. Safer and cheaper, but the surfactant stays in the product and may interfere with downstream chemistry.
- Polymer-assisted exfoliation. Exfoliate directly into the polymer or resin that will host the graphene, so the “contaminant” is a component you wanted anyway. Elegant when it fits the application.
If you are buying dispersions rather than powders, ask which of these routes was used. The answer determines what else is in the bottle.
Sonication Versus Shear
Ultrasonication is the laboratory workhorse. Cavitation bubbles collapse near graphite particles and generate the local forces that pry sheets apart. It works, it is easy to set up, and it scales badly. Energy delivery drops off sharply away from the sonicator tip, so scaling up means more time or more tips rather than a bigger vessel.
High-shear mixing scales much better. A rotor-stator mixer generates turbulent shear throughout the vessel, and the process behaves more like conventional industrial mixing. A Trinity College Dublin group made the point memorably by demonstrating usable shear exfoliation in a household kitchen blender — not a manufacturing recommendation, but an effective illustration that the physics does not require exotic equipment.
Most industrial LPE production today uses shear-based or related mechanical routes precisely because they translate into standard process engineering.
What LPE Actually Produces
This is the part that matters most for buyers. LPE does not produce monolayer graphene at scale. It produces a distribution.
A typical LPE product contains flakes ranging from a few layers to several tens of layers, with lateral dimensions from a few hundred nanometers to several microns. Yield of true monolayer material is low, often single-digit percentages by number and lower still by mass. The commercial term for the resulting material is graphene nanoplatelets, and the terminology matters — a nanoplatelet is a legitimate and useful material, but it is not monolayer graphene and should not be priced or specified as if it were.
The upside is that LPE is comparatively gentle. Because it is a mechanical process rather than a chemical one, the basal plane of each flake retains most of its structure. Defect density is low compared with oxidation-reduction routes, and the material typically shows better electrical and thermal performance than reduced graphene oxide at similar loadings.
The downside is variability. Flake thickness, lateral size, and the ratio between them all shift with feedstock graphite, energy input, processing time, and solvent. Two batches from the same supplier can differ measurably, which is why anyone using graphene in a qualified product needs incoming inspection rather than trust.
Where LPE Sits Against the Alternatives
Against CVD: LPE is orders of magnitude cheaper per gram and produces bulk powder or dispersion rather than continuous film. If your application needs a conductive additive dispersed in something, LPE wins. If it needs a continuous transparent conductive layer, LPE cannot compete at all. See our manufacturing coverage for the full comparison.
Against oxidation routes: LPE material has far fewer basal-plane defects and no residual oxygen functionality, which means better intrinsic properties. But graphene oxide disperses in water without help and offers reactive sites for functionalization. LPE flakes are chemically inert and correspondingly difficult to disperse — a tradeoff explored in our article on the dispersion problem.
Against electrochemical exfoliation: several commercial producers use electrochemical routes that intercalate ions between graphite layers before separation. These sit between LPE and oxidation on the defect spectrum and have proven scalable, as the production capacities of the larger producers in our company profiles demonstrate.
What to Ask a Supplier
If a supplier tells you their material is made by liquid-phase exfoliation, useful follow-up questions include:
- What is the layer-number distribution, and how was it measured?
- What solvent or surfactant system was used, and what residual level remains?
- What graphite feedstock is used, and does its source vary between batches?
- What is the batch-to-batch variation in specific surface area?
- Is the product supplied as dry powder, wet cake, or dispersion — and if dispersion, at what solids loading and in what carrier?
The answers tell you more about whether the material will work in your system than any single headline specification. LPE is a good, cheap, scalable process. It is also a process whose output is defined by a distribution rather than a number, and buyers who treat it as a single-value commodity are the ones who get surprised.
This article is part of our Manufacturing series. For cost context, see Graphene Production Costs. For the contrasting high-quality route, see our coverage of chemical vapor deposition.