Size exclusion chromatography is a gentle polishing step, but resolution depends on choices made before the run: pore size, bead size, flow rate, and sample volume. This guide explains how each variable affects separation, reports the flow velocities and sample loads tested on WorkBeads 40 SEC…
Size exclusion chromatography SEC separates molecules by size and geometry, not by affinity. A molecule larger than the resin's exclusion limit cannot enter the pores and travels through the packed bed in the void volume. Smaller molecules diffuse into the pores, take longer to elute, and eventually co-elute at roughly one column volume once they are small enough to access nearly all of the pore volume. Because the mobile phase is simply the buffer the sample is already in, SEC is gentle: it suits biomolecules that are sensitive to pH, ionic strength, metal ions, and temperature, which is why it is usually placed as a polishing step at the end of a purification process rather than at the front.
The first decision is pore size, because the separation range of a resin is primarily determined by its pores. For the WorkBeads 40 SEC line, the vendor describes two useful brackets. A resin with a 150 kDa exclusion limit suits small proteins, peptides, and small oligonucleotides such as antisense oligonucleotides ASOs . Resins with exclusion limits in the 10,000 to 30,000 kDa range suit larger targets such as antibodies and extracellular vesicles. Molecules smaller than a resin's lower separation limit all elute at one column volume, and molecules larger than the exclusion limit all elute in the void volume, so in both regimes resolution between differently sized analytes is lost. The practical rule: choose the smallest pores that still let the target molecule in, so that the steep part of the selectivity curve covers the size difference you actually need to resolve.
One gap in the vendor's text is worth flagging. It gives these exclusion limits but does not state which named resin, 40/100, 40/1000, 40/10000, or Macro, matches which limit. That mapping sits in a table that was not reproduced. Before ordering material, a buyer should confirm the exclusion limit for the specific resin name rather than assume it from the name alone. Porosity drives selectivity, but the optimal resin depends on three variables together: porosity, bead size, and the application.
All four WorkBeads 40 SEC resins share an average bead size of 45 µm and differ only in pore size. Bead size sets a trade-off. Smaller beads give better resolution because diffusion distances into and out of the pores are shorter, but they generate higher backpressure. At preparative scale, the 45 µm bead is a compromise between resolution and the pressure limits of the equipment. Some applications need larger beads, above 100 µm , for example when sample viscosity is high; WorkBeads 200 SEC is the vendor's larger-bead option. The vendor gives no performance data for that resin in the study, so the viscosity argument is asserted rather than measured.
The resins are highly cross-linked, rigid agarose beads. Rigidity matters at scale because it keeps the bed stable at higher flow and permits harsh cleaning: the vendor states the material tolerates extreme pH, extreme ionic strength, and cleaning with 1 M NaOH , the standard regime for sanitization in biopharmaceutical production.
Pore size distribution PSD is as important as mean pore size. A narrow PSD means molecules of a given size experience similar diffusion paths, so they leave the column within a narrower elution band; a broad PSD stretches the band and thins the peak. More uniform mass transport in and out of the pores translates into narrower peaks and, in a purification, higher purity. The vendor measured PSD by inverse SEC iSEC , injecting dextrans of predefined sizes and detecting them with a refractive index detector. The method matters because it probes the pores in the hydrated beads, the state they are in during a real run, rather than in dried or derivatized material. The measurements cover four resins: WorkBeads 40/100 blue , 40/1000 green , 40/10000 red , and Macro 40/30000 grey . The comparison shows the four resins occupying different pore-size windows, with relatively tight distributions within each window.
Selectivity, how strongly elution volume changes with molecular weight, is best read from a selectivity curve : the distribution coefficient K d plotted against log molecular weight. The vendor generated curves using globular proteins, thyroglobulin, bovine serum albumin 66,300 Da , and ribonuclease A 13,700 Da , detected at 280 nm . A steeper curve means a given size difference translates into a larger elution volume difference, exactly what resolving close-sized molecules requires.
K d is also the workhorse diagnostic for column health. It expresses a molecule's elution volume relative to the void volume and the total column volume. K d = 0 means the molecule elutes at the void volume; K d = 1 means it elutes at one column volume. The operational range for separations is K d between 0.1 and 0.9 . If K d comes out above 1, suspect non-specific interactions between the sample and the matrix. If K d is negative, the packed bed contains microchannels that let sample bypass the resin, and the column needs repacking.
A caveat on the selectivity data: they were collected with globular proteins. Antibodies are roughly globular, but peptides, ASOs, and especially extracellular vesicles are not. Non-globular molecules have different effective radii at the same molecular weight, so published selectivity curves should be treated as a floor for planning, not a precise predictor.
Resolution in SEC is governed by how much time a molecule has to diffuse into and out of the pores. Lower flow rates improve resolution, up to a point: at very low flow, longitudinal diffusion starts to broaden peaks, and small molecules, which diffuse fastest, are the first to show it. The vendor tested two model protein pairs on prepacked 16 mm × 600 mm columns. Set 1 paired myoglobin 17,600 Da with apo-transferrin 78,300 Da on WorkBeads 40/100 SEC. Set 2 paired ribonuclease A 13,700 Da with bovine serum albumin 66,300 Da on WorkBeads 40/1000 SEC. Resolution values were calculated with UNICORN 5.31 .
| Protein | Molecular weight Da | Protein set | Resin tested |
|---|---|---|---|
| Myoglobin | 17,600 | 1 | WorkBeads 40/100 SEC |
| Apo-transferrin | 78,300 | 1 | WorkBeads 40/100 SEC |
| Ribonuclease A | 13,700 | 2 | WorkBeads 40/1000 SEC |
| Bovine serum albumin | 66,300 | 2 | WorkBeads 40/1000 SEC |
In the vendor's flow-rate experiments Figures 3A to 3D , both protein sets reached baseline resolution, conventionally R s ≥ 1.5 , at flow velocities of 15 cm/h 0.5 mL/min and 30 cm/h 1 mL/min . As flow velocity increased, resolution decreased. At 60 cm/h 2 mL/min the two tested columns still separated the model proteins clearly, but not at baseline resolution.
| Flow velocity cm/h | Flow rate mL/min | Observed resolution |
|---|---|---|
| 15 | 0.5 | Baseline resolution R s ≥ 1.5 for both protein sets |
| 30 | 1 | Baseline resolution R s ≥ 1.5 for both protein sets |
| 60 | 2 | Acceptable separation on both columns, below baseline resolution |
The practical translation: for fractionation of analytes that elute close together, run at 15 to 30 cm/h. For group separations, where the target and the impurities elute in very different volume ranges, 60 cm/h may be acceptable because the peaks have a large elution-volume gap to spare.
Sample volume is the second operating lever. A large injection volume widens every peak by the width of the injected plug, which reduces resolution between close peaks. The vendor's recommendation for preparative SEC is a sample volume of 0.5 to 4% of column volume . Group separation and desalting can tolerate much larger loads, up to 30% of column volume , because target and impurity peaks are so far apart that broadening does not merge them. In the vendor's sample-volume experiments Figures 4A to 4D on WorkBeads 40/1000 SEC, resolution was not significantly affected by increasing sample volume across the tested range, which suggests that load capacity is not the binding constraint in a 16 mm × 600 mm format at these volumes. The caveat is "in the tested conditions": the data cover one column format and two protein pairs.
| Application | Recommended sample volume % of column volume |
|---|---|
| Preparative SEC fractionation | 0.5 to 4 |
| Group separation or desalting | up to 30 |
Sample concentration, by contrast, usually has little effect on resolution. SEC bands are dilute and the separation mechanism is not concentration-dependent, so operators do not usually need to control concentration tightly, within the limits of solubility and viscosity.
Group separations relax both constraints. A common example is virus purification: the goal is often to elute the virus in the void volume while retaining impurities inside the resin. Because the virus and the impurities elute at opposite ends of the chromatogram, the run tolerates higher flow and much larger sample volumes than a fine fractionation.
All of the quantitative operating data above come from the vendor's own experiments. They are internally consistent and methodologically straightforward, but they have not been independently replicated in the peer-reviewed literature, and the selectivity and resolution numbers were produced with globular model proteins on a single column format. Treat them as carefully measured starting points, not as guarantees for a particular feedstock.
The adjacent published evidence concerns the problem SEC is asked to solve, not the resins themselves. In CHO cells expressing IgG1 in 2 L bioreactors, extending culture from day 8 to day 17 increased the number of host cell proteins detected after protein A capture from 72 to 475 and decreased product monomer percentage from 98% to 95.5% PMID 34751518 . For a process that ends with SEC polishing, that means the polishing column can face a substantially heavier and more diverse impurity load, and a lower monomer fraction, depending on when the bioreactor is harvested. Population balance modelling of fed-batch CHO cultures, combined with orthogonal measurements, suggests that a significant fraction of extracellular host cell proteins is actively secreted by viable cells rather than released only by cell breakage PMID 35320326 . Actively secreted host cell proteins form a dynamic, size-diverse population, a reminder that pore selection should consider the size distribution of the impurities, not just the target.
Clinical and regulatory registries do not carry chromatographic operating conditions, which is why resin guidance of this kind lives in vendor technical documentation. For context on the product classes involved: a chikungunya virus vaccine candidate is being followed in a phase 3 study with 363 participants NCT04838444 , and a PD-L1 × 4-1BB bispecific antibody program was registered for phase 1 and phase 2 but withdrawn before enrolling participants NCT04841538 . These entries document that vaccine and antibody programs reach or approach the clinic, but they contain no information about resin choice, flow rate, or column load, and they neither confirm nor contradict the vendor's operating recommendations.
What remains unresolved: the exact exclusion limit and separation range for each named WorkBeads 40 resin; the quantitative relationship between sample concentration and resolution; whether the vendor's referenced case studies would survive independent scrutiny; how the 16 mm × 600 mm findings translate to production-scale columns; and the exclusion limits and separation ranges of WorkBeads 200 SEC.
Each of these rules is conditional on the evidence base described above. The mechanism of SEC is well understood; the specific resin numbers are vendor-generated and should be confirmed on the actual feedstock before a process is locked in.
Related reading: Hybrid Fragment Synthesis Expands Peptide Manufacturing Options, PEC Purification for GLP-1 Manufacturing: Liraglutide Case Study, Balancing Chemistry and Timelines in Complex Peptide Synthesis, Therapeutic Peptides: Classes, Applications and Synthesis Challenges.