Three material families dominate preparative monolithic chromatography: polymethacrylate, styrenic or methacrylic polyHIPE foams, and cellulose. The backbone decides two different things, and they are worth keeping apart. It sets the surface your molecule touches — hydrophilic from the start, or hydrophilic only after modification. And through the manufacturing route it sets the topology: how much void the structure can carry before it collapses, and how much pressure it wastes. MonoCore™ is a cross-linked cellulose monolith with a 15 µm channel size and a porosity around 80 %, supplied in 1.8 mL and 5.4 mL capsules with UNF 10-32 connectors.
The cauliflower: where the standard morphology comes from
The established polymethacrylate monolith is made by precipitation polymerisation. A mixture of monomer (usually glycidyl methacrylate), crosslinker (ethylene dimethacrylate), initiator and a porogenic solvent is polymerised; the growing polymer becomes insoluble in the porogen and precipitates. Nuclei grow into microglobules, microglobules stick together into clusters. What we call pores are simply the irregular gaps left over: macropores between clusters, mesopores between globules, and micropores inside them.
The porogen controls when the phase separation happens, and therefore how large the globules and the gaps become. In poly(GMA-co-EDMA) monoliths, raising the dodecanol content from 0 to 15 % shifts the mean pore size from 0.2 µm to 2.6 µm (J. Appl. Polym. Sci. 2012, DOI 10.1002/app.37514).
That structure has two consequences that rarely appear in product literature:
It has a mechanical ceiling on porosity. Modulus scales with porosity as E = E₀(1−ε)ⁿ, and the exponent depends on topology. For agglomerated-globule morphology the reported exponent is n = 3.61, which puts the maximum mechanically viable porosity at roughly 0.61. Foam-like topologies reach n = 1.86 and stay viable up to about 0.91 (J. Sep. Sci., DOI 10.1002/jssc.202300767). A cauliflower structure cannot be pushed to high void fraction; it fails first.
It is hydraulically wasteful. The dimensionless form factor φ = d²ε/K measures how much pressure a structure squanders at a given pore size. An empty tube would be 32. Silica monoliths sit at 22 to 64, polyHIPE foams at 66 to 100, and methacrylate monoliths at about 130. In the same published comparison, the measured permeability of a commercial methacrylate monolith disc is 5.74·10⁻¹⁵ m² — the lowest of the media compared, below that of a packed Protein A column (Ind. Eng. Chem. Res., DOI 10.1021/acs.iecr.5c01179).
That last number is worth pausing on, because it corrects a widespread belief. The advantage of a polymethacrylate monolith is the transport mechanism — convection instead of diffusion, and resolution that holds as the flow rises. It is not permeability.
Topology is a consequence of the process, not of the monomer
This is the part where an honest article has to contradict the convenient story. The cauliflower is the signature of the precipitation route, not of the methacrylate. Change the route and the same chemistry gives a different structure.
The clearest current demonstration comes from a competitor. In a 2026 paper, an inverted-flow morphology is described in which hollow spheres are packed into an interconnected network, reaching a void volume of up to 80 % and roughly one third of the pressure drop of conventional thermally polymerised methacrylate discs (Wheelwright et al., Sep. Purif. Technol. 394, 137559, 2026, DOI 10.1016/j.seppur.2026.137559). The material remains methacrylic. What changed is where the solid sits.
So the field has two axes, and they move independently:
Topology — how the void is arranged, how much of it the structure carries, how much pressure it costs.
Surface — what the molecule touches when it arrives.
Addressing one does not address the other. A monolith with an excellent void structure still has the surface chemistry of its backbone.
PolyHIPE: the best topology, with its own bill
Foams made from a high internal phase emulsion — more than 74 % internal phase — give the most favourable mechanics-to-porosity combination of the three families, with open cells connected by windows, porosities of 72 to 83 %, and hypercrosslinked variants above 95 %.
The bill comes in two parts, both documented. The emulsion requires 5 to 50 % surfactant relative to the external phase, and complete removal at scale is described in the literature as difficult, which matters for applications with demanding purity requirements. And the materials tend towards brittleness, a property directly attributed to the surfactant load.
Cellulose: hydrophilic at the start, but only if it is cross-linked
Cellulose brings what a synthetic backbone has to be given: abundant surface hydroxyl groups, straightforward derivatisation chemistry and weak non-specific adsorption (J. Chromatogr. A 2023, DOI 10.1016/j.chroma.2023.464202; review, PMID 35809519). The same argument is visible one unit operation upstream, where regenerated-cellulose filtration membranes are positioned against polyethersulfone and PVDF precisely because hydrophobic surfaces foul.
Methacrylate monoliths reach a comparable surface only through modification — hydrolysis of the epoxide to the diol, PEG-methacrylate grafting, zwitterionic coatings or polyethylenimine. These are routine and they work. The difference is where each material starts.
Two qualifications belong in the same breath, because they are the ones a careful reader will raise:
Cellulose is not alkali-stable as such. Unmodified cellulose hydrate media shrink, swell and ultimately decompose in caustic — that is the reason the patent literature gives for cross-linking them in the first place. Cross-linked cellulose is alkali-stable; the cross-linking step is a precondition of the claim, not a footnote. MonoCore's matrix is cross-linked cellulose for exactly this reason.
Cellulose is mechanically weak without it, too. Uncross-linked cellulose media compact under flow. The same step addresses both.
And one honest note on hydrophobicity: it is not a defect. Hydrophobic interaction is a chemistry people deliberately buy. In one viroid concentration study, a butyl monolith recovered up to 60 % where the anion exchanger managed 30 %. The fair statement is not that hydrophobic surfaces are bad, but that a hydrophobic backbone is a default you have to modify away before you can put it back deliberately where you want it.
What the surface actually decides: getting the product off again
For viruses, viral vectors and fragile particles, the limiting property is usually not how much binds. It is how much comes off intact.
The best available evidence comes from lentiviral vector work on anion exchange membranes. Standard membranes lost 70 to 80 % of the vector at scale, through a time-dependent conformational change of the bound particle leading to irreversible multi-point attachment. Reducing the ligand density from 100 to 8 µmol/mL — a factor of 12.5 — raised particle recovery from about 30 % to roughly 90 %, with functional titre improving from 17–19 % to 50–73 % (Pamenter et al., Mol. Ther. Methods Clin. Dev. 33(3), 101533, 2025, DOI 10.1016/j.omtm.2025.101533).
Read that carefully, because it is not the result a media supplier would like to claim. The lever in that study was ligand density and surface architecture, not hydrophilicity. The general principle it supports is the one that matters: with large, fragile particles, the surface decides releasability, and releasability decides yield — while capacity, the number everyone compares, is the wrong figure of merit.
A hydrophilic backbone is therefore a starting point, not a result. It is an honest argument only in that form.
The price of a coarse channel
Nothing here is free, and the trade-off is geometric. Accessible surface area scales as a ≈ 4/d, permeability as K ∝ d². A 15 µm channel has roughly one seventh of the geometric surface area per millilitre of a 2 µm channel, and about fifty times the permeability. Coarse channels buy pressure headroom, throughput and tolerance for difficult feeds, and they pay for it linearly in surface area and therefore in capacity per millilitre.
Where that trade pays off is documented for large molecules. In a study on anion exchange monoliths at 2, 3 and 6 µm, the open-circular plasmid isoform was captured convectively in the narrower channels, causing yield loss and irreversible pressure build-up; channels larger than about 3.5 µm were required for plasmids up to 16 kbp (Electrophoresis, DOI 10.1002/elps.202300035). The common commercial standard is 2 µm. MonoCore runs 15 µm.
For a 150 kDa antibody, that same trade is a bad deal, and we say so: for capacity-driven protein capture a packed-bed resin remains the economical choice.
On sustainability, briefly and honestly
A cellulose backbone is renewable, and that is true but small. Life cycle assessments of biologics manufacturing put the dominant burdens on cleanroom energy and water; the mass of resin is negligible against the mass of water (Curr. Opin. Green Sustain. Chem. 2022, DOI 10.1016/j.cogsc.2022.100628). The LCA literature also warns explicitly that bio-based origin does not by itself reduce environmental impact.
The environmental argument that survives scrutiny is a process argument: fewer cycles, less buffer, less water, a smaller eluate to handle downstream. Those attack a documented hotspot. "Made from a renewable material" is a fair secondary note, not a headline — and no life cycle assessment compares chromatography media by backbone material at all.
What we cannot show you — yet
Being specific about the gaps is more useful than pretending they are closed:
No controlled comparison of backbones exists. Same ligand, same ligand density, same pore size, only the backbone varied, with recovery as the endpoint — that study is not in the literature. Anyone claiming a percentage advantage for one backbone over another is extrapolating.
No published caustic cycling study for a cross-linked cellulose monolith at 80 to 90 % porosity. Membrane and bead data do not transfer directly, because the load-bearing skeleton is different.
No volume-normalised ligand density comparison between cellulose and methacrylate monoliths. Surface areas quoted per gram are misleading at 90 % porosity.
We are working on the first and the second. If you have a molecule where the backbone question decides your process, we would rather run it than argue about it.
MonoCore™ in one paragraph
Cross-linked cellulose monolith, 15 µm standard channel size (other sizes on request, fixed within any one monolith), porosity around 80 %, hydrophilic backbone with low non-specific adsorption. Available as Q and S (strong and weak ion exchange), HIC (butyl and phenyl), Protein A and Protein G. Pre-assembled 1.8 mL and 5.4 mL capsules, UNF 10-32 coned ports, larger formats on request. Detailed performance data are available on request.
Frequently asked questions
Are all monoliths made of polymer?
No, and this is a common misconception. Silica monoliths are established in analytical chromatography, and cellulose monoliths exist in preparative bioprocessing. We are not aware of a second commercial cellulose monolith, but the category is broader than the usual definition of a polymer block suggests.
What is the cauliflower structure in a methacrylate monolith?
The morphology that results from precipitation polymerisation: microglobules growing from nuclei and sticking together in clusters, with the pores being the gaps between them. It is a consequence of the manufacturing route rather than of the monomer.
Does a hydrophilic backbone improve recovery?
It starts you in a better place. The published evidence for recovery losses with fragile particles points at ligand density and multi-point binding rather than at hydrophobicity as such, so the honest formulation is that a hydrophilic surface is a favourable starting condition, not a guaranteed result.
Is cellulose stable in sodium hydroxide?
Cross-linked cellulose is. Uncross-linked cellulose hydrate media shrink, swell and eventually decompose in caustic, which is precisely why they are cross-linked. Always check which of the two a product is.
Why do coarse channels have lower capacity?
Because accessible surface scales inversely with channel diameter (a ≈ 4/d). A wider channel means less area per millilitre. What you get in return is permeability, which scales with the square of the diameter, plus tolerance for feeds that would block a finer structure.