Fullerene separation and purification — five dedicated COSMOSIL columns (Buckyprep, Buckyprep-M, PBB, PYE, NPE) covering analysis and preparative purification of C60/C70 and metallofullerenes.
Fullerenes (C60, C70 and higher fullerenes) dissolve poorly and recover badly on conventional columns, which makes them difficult to separate. The COSMOSIL Buckyprep family, developed specifically for fullerenes, achieves high-loading, tail-free separation and preparative purification in toluene and other good solvents.
The Buckyprep column (pyrenylpropyl phase) is the industry reference for fullerene separation. With toluene as the mobile phase, C60 and C70 give sharp, tail-free peaks and the load can be raised considerably relative to a C18 column, making it the first choice for preparative fullerene purification.
Buckyprep-M is optimized for endohedral metallofullerenes and higher fullerenes, offering selectivity different from Buckyprep and suiting metallofullerenes of similar properties that conventional columns cannot distinguish.
The PBB column (pentabromobenzyl phase) gives stronger retention and selectivity than Buckyprep, suiting high-purity preparation of C60/C70 and separation of higher fullerenes (C76, C78, C84 and others) where more resolution is needed.
The PYE column (pyrenylethyl phase) provides strong π–π interaction and excels at separating structural isomers and derivatives; the NPE column offers complementary selectivity, suited to fine separation of fullerene derivatives in organic photovoltaic materials such as PCBM.
Cage and principal derivative structures, rendered by RDKit as orthogonal projections of the three-dimensional cages. Cages are drawn as skeletal representations without marking 6,6-bond positions; addition sites on fullerenols and fluorinated fullerenes are indicative and n is a representative value — the actual number and distribution of additions depend on the reaction conditions. Swipe horizontally on mobile.
| Dedicated column | Bonded phase | Best suited to | Characteristics |
|---|---|---|---|
| Buckyprep | Pyrenylpropyl | C60 / C70 Standard Separation | The industry reference; high-load preparative work |
| Buckyprep-M | Phenothiazine | Metallofullerenes / higher fullerenes | Selectivity complementary to Buckyprep |
| PBB | Pentabromobenzyl | High-purity preparation / higher fullerenes | Stronger retention and resolution |
| PYE | Pyrenylethyl | Structural isomers / derivatives | Strong π–π selectivity |
| NPE | Nitrophenylethyl | Fullerene derivatives / PCBM | Complementary selectivity |
| Note: toluene is the preferred mobile phase for fullerenes, giving high solubility and good recovery; all the columns above are optimized for toluene systems. | |||
Data from the COSMOSIL fullerene column series; swipe horizontally on mobile for the full comparison.
| Packing | Buckyprep | Buckyprep-M | PBB | PYE | NPE |
|---|---|---|---|---|---|
| Silica substrate | High-purity porous spherical silica | ||||
| Mean particle size | 5 μm | ||||
| Mean pore size | About 120 Å | ||||
| Surface area | About 300 m²/g | ||||
| Stationary phase structure | |||||
| Bonded group | Pyrenylpropyl | Phenothiazine | Pentabromobenzyl | Pyrenylethyl | Nitrophenylethyl |
| Bonding type | Monomeric | ||||
| Endcapping | Nearly complete | None | Nearly complete | — | — |
| Carbon load | About 17% | About 13% | About 8% | About 18% | About 9% |
| Characteristics | Standard fullerene separation column | Metallofullerene separation | Preparative separation of C60/C70 | Fullerenes and structural isomers | Fullerene derivatives |
The core combination from analytical selection to industrial preparative purification
Extraction of fullerenes is a relatively mature step; the constraints on cost and efficiency sit in separation and refining. The table below compares the main routes reported in public sources by principle, achievable purity and demonstrated scale. Purity and scale are listed as publicly disclosed and have not been verified by us.
| Technology | Principle | Achievable purity | Demonstrated scale | Applicability and limitations |
|---|---|---|---|---|
| Soxhlet / maceration extraction | Fullerenes dissolve in toluene, o-xylene and CS₂, separating from amorphous carbon | Crude mixture | Industrial scale | Mature and a small share of cost; solvent recovery and residue control govern downstream quality |
| Alumina / activated carbon column chromatography | Differences in adsorption strength; C₆₀ elutes before C₇₀ | About 95%–99% | Pilot to industrial | Low cost and scalable; limited purity ceiling, with yield lost to irreversible adsorption on the stationary phase |
| Chlorobenzene elution + activated carbon | Exploits the higher solubility of C₆₀ in chlorobenzene | Reported > 99% | Batch production feasible | Public sources describe it as scalable; chlorobenzene is a controlled solvent with demanding environmental and residue requirements |
| Preparative HPLC | Fullerene-specific stationary phases such as pyrenyl, phenothiazine and pentabromobenzyl | 99.9%–99.99% | Kilogram / annual scale | Good purity and cage-type resolution; low throughput, high solvent consumption and a large share of cost in the stationary phase |
| Fractional crystallization | Solubility differences between C₆₀ and C₇₀ in CS₂ or o-xylene | Mainly enrichment | Gram scale in the literature | No stationary phase losses; requires prolonged stirring and sometimes heating, and scale-up efficiency remains to be demonstrated |
| Selective complexation / host–guest precipitation | γ-cyclodextrin, calixarenes, porphyrin cages, triptycene–bowl molecules, AgNO₃ molecular cages | Reported up to 99.5% in the literature | Laboratory scale | Avoids the irreversible adsorption losses of chromatography; host molecule cost and recyclability are prerequisites for industrialization |
| Simulated moving bed chromatography | Multi-column switching gives continuous countercurrent separation | Depends on the stationary phase | No commercial fullerene installation publicly disclosed | Mature in peptide and chiral separation, and a plausible route to lowering the cost of fine fullerene separation |
| Vacuum sublimation | Removes residual solvent and low-volatility impurities | Sublimation grade, 99.9%+ | Kilogram to pilot scale | The grade dividing line for vacuum deposition applications; energy consumption and yield must be balanced |
| Note: the table is compiled from public sources, and purity and scale figures for the same route differ between sources. "Demonstrated scale" refers to the largest implementation publicly disclosed and does not indicate current production capacity. | ||||
Public sources are consistent on this point: extraction is mature, while separation is costly and inefficient, and is the main reason high-purity fullerene prices have stayed high. Competition in industrialization tends to centre on fine separation and sublimation refining rather than on synthesis.
Yield losses on the chromatographic route come mainly from irreversible adsorption on the stationary phase, which is the argument behind the host–guest route. Using fullerene-specific bonded phases with a toluene system can improve loading and recovery while retaining resolution.
Different applications differ considerably in their requirements for purity, residual solvent and batch consistency, which in turn determines how far the separation step must go. The table below sets out consumption characteristics, industrialization status and demand drivers; the assessments are inferences from public information.
| Application | Consumption and specification requirements | Industrialization status | Assessment of demand pull |
|---|---|---|---|
| Electron transport layers in perovskite and tandem cells | Vacuum-deposited C₆₀ thin films, requiring sublimation-grade purity and low residual solvent; consumption per GW of line capacity, calculated from film thickness and utilization, is well above current research and cosmetics demand | Several GW-scale lines in China came on stream from 2025 and larger-scale manufacturing began in 2026 (per industry reports); because C₆₀ costs are high, some manufacturers use a C₆₀/SnO₂ composite electron transport layer | On publicly available information, this appears the direction most likely to pull ton-scale fullerene capacity. The constraints are raw material price (public sources cite around RMB 200/g) and deposition material utilization |
| Organic photovoltaics and organic photodetectors | PC₆₁BM / PC₇₁BM or deposited C₆₀; demanding on batch consistency | OPV products are commercially available; organic photodetectors for full-screen fingerprint sensing are described by manufacturers as close to commercialization | Consumption per unit area is small, but it generates steady orders for high-purity derivatives and is the main outlet for Japanese combustion-process capacity |
| Lubricant and lubricating material additives | Manufacturers report that around 100 ppm markedly reduces wear on metal parts; mixed fullerenes are acceptable | Commercial products exist; early applications included satellite gyroscope lubrication, and Chinese manufacturers have launched additives for trucks | Purity requirements are relatively relaxed, so intermediates outside the fine separation stream can be absorbed, which helps overall line economics |
| Cosmetics | Fullerenes and fullerenols as antioxidant ingredients, with residual solvent and metal impurities to be controlled | Serums, masks and concentrates are established categories; several Chinese raw material suppliers also sell finished cosmetics | High unit price, small total volume. Market research firms' estimates of the size and growth of this segment differ widely and are of limited comparability. Public reporting from 2024 also records administrative penalties in the sector over advertising language, so claim compliance warrants attention |
| Biopharmaceuticals | Metallofullerenes such as Gd@C₈₂ for MRI contrast; fullerenols in antioxidant and antitumour research | Companies state that kilogram-scale raw material lines are built; as of the search date, no public record of an approved IND for a metallofullerene drug in China was found | The highest unit price segment, but constrained by drug review timelines, so large-scale raw material demand is unlikely in the near term |
| Metal matrix composites | Introducing fullerenes into aluminium and copper matrices | Reports state that fullerene-laminated aluminium reaches an ultimate bending strength comparable to 45 steel at about one third the density, and that copper composites show roughly one third the friction coefficient of pure copper | If validated in rail transit, bearings and similar applications, consumption per unit would be substantial; publicly reported volume applications are currently limited |
| Lithium battery electrolyte additives | Fluorinated fullerenes as entropy-tuning additives in lithium metal batteries | Published jointly by companies, universities and research institutes in 2026; at the research stage | Consumption would be considerable if industrialized; no public information on pilot or volume production at present |
| Other | Resin anti-ageing, laser protection, photocatalytic formaldehyde removal, agricultural and veterinary protection, catalysis and adsorption | Mostly at trial or small-batch stage | Dispersed demand, limited in aggregate in the near term |
Deposition-grade C₆₀ and PCBM for photovoltaics place the strictest demands on purity, residual solvent and batch consistency, and normally require fine separation on dedicated columns together with sublimation refining; lubricant and some industrial uses can be relaxed to column-chromatography-grade intermediates.
If your application is unusual, contact our technical team directly — selection consulting is free of charge.