Mixed-Mode Chromatography: When One Interaction Is Not Enough

1. What Mixed-Mode Chromatography Means

A mixed-mode or multimodal stationary phase creates two or more meaningful interaction types under the same operating conditions. A ligand may combine ionic and hydrophobic groups, hydrogen-bond donors or acceptors, aromatic surfaces, or metal/coordination behavior. The interactions are spatially coupled and may be cooperative, competitive or condition-dependent.

This is not merely a sequence of two columns, nor any resin that shows minor nonspecific adsorption. The medium is designed so multiple interactions contribute to selectivity. Because one condition can strengthen one mechanism while weakening another, retention often cannot be predicted from pI, net charge or hydrophobicity alone.

2. Ligand and Matrix Architecture

Functional-group charge, pKa, hydrophobic surface, spacer, ligand density and orientation shape the local environment. The base matrix determines pores, hydration, particle rigidity and pressure. A nominally similar ligand on a different matrix can produce different capacity and selectivity.

Some media expose multimodal ligand throughout the bead for bind-and-elute capture or polishing. Restricted-access core-shell media place adsorptive functionality inside a size-selective barrier so large targets flow through while smaller impurities enter and bind. Ceramic hydroxyapatite presents calcium and phosphate crystal surfaces rather than a conventional coupled organic ligand. Keep these architectures distinct during selection.

3. When Mixed Mode Adds Value

Mixed mode is not automatically more selective. Added interactions can broaden peaks, complicate elution or increase sensitivity to feed composition. Use it when the unresolved impurity and process context justify the added development burden.

4. pH, Salt and Additives Act Together

pH changes target charge and, for weak ionic groups, ligand ionization. Salt screens electrostatic forces, competes for charged sites and can strengthen or weaken hydrophobic association depending on identity and level. Multivalent ions and buffers may interact specifically. Equal conductivity from different formulations does not imply equal retention.

Map at least a two-dimensional pH–conductivity or pH–salt matrix within target stability. Add a salt-identity comparison when hydrophobic and ionic mechanisms are coupled. Test additives only for a defined problem and check matrix compatibility. A small change can shift the dominant mechanism, creating useful selectivity or an unstable process edge.

5. Bind-and-Elute, Flow-Through and Weak Partitioning

In bind-and-elute mode, product retention provides concentration and fractionation, but elution must reversibly disrupt enough interactions. A sequential change—salt followed by pH or vice versa—can identify the dominant mechanism. In flow-through mode, impurities bind while product passes; load is controlled by impurity capacity and product loss, not product mass alone.

Weak partitioning can provide high-load polishing when product and impurity interact differently without complete target binding. It may be sensitive to pH, conductivity and residence time. Define collection boundaries and breakthrough with representative feed rather than treating the mode as ordinary flow-through.

6. A Mechanism-Based Screening Workflow

1. Define the critical impurity and why IEX, HIC or SEC is insufficient.
2. Map target stability across candidate pH, salt, additive and concentration conditions.
3. Select media with deliberately different interaction architectures.
4. Run batch binding to eliminate inactive regions, with no-resin controls.
5. Confirm candidates in packed columns at matched load and residence time.
6. Collect all streams and use salt, pH or combined gradients to map release.
7. Test bind-and-elute and flow-through where both are plausible.
8. Challenge leading conditions with feed variability and operating-range edges.

High-throughput binding data do not predict pressure, breakthrough or gradient efficiency. Use batch formats to reduce the search space, then perform column confirmation before assigning capacity or scale.

7. Analytics and Mass Balance

Measure target in load, flow-through, washes, eluate, strip and precipitate. Pair total protein with target-specific identity and activity. Analyze the critical impurity directly—aggregate, fragment, DNA, host protein, leached ligand or variant—because overall purity can hide enrichment of a low-level high-risk species.

Track actual pH and conductivity of fractions. Retention can change within the band as buffer components exchange. Pool from impurity and target data rather than UV. If recovery is low, distinguish irreversible adsorption, precipitation, incomplete elution and analytical nonrecovery before changing the ligand.

8. Scale-Up, Cleaning and Lifecycle

Determine the parameter that controls performance: product dynamic capacity, impurity capacity, residence time, gradient slope or flow-through breakthrough. Preserve bed height and relevant flow basis while confirming pressure, packing, distributors, system delay, buffer mixing and collection. Feed composition may be more influential than geometry.

Mixed interactions can retain complex foulants strongly. Develop cleaning using representative feed and track capacity, selectivity, recovery and pressure. Ensure that cleaning restores both interaction modes rather than only gross capacity. Define storage, sanitization, cycle history and replacement criteria. Custom resin development and method development may help when the required selectivity is not available from standard media.

9. Common Pitfalls

• Calling any resin with nonspecific adsorption “mixed mode.”
• Predicting retention from protein pI without considering ligand ionization and hydrophobic patches.
• Changing pH and salt simultaneously without learning which mechanism controls separation.
• Comparing media under one borrowed condition that favors only one architecture.
• Using product mass to set flow-through load while ignoring impurity breakthrough.
• Assuming higher binding is better when elution and cleaning become difficult.
• Reporting total protein purity without measuring the critical impurity.

10. Selection Matrix and Method Handoff

Document ligand or material architecture, matrix, particle and pore properties, column, packing, buffers, feed conditioning, load, residence time, wash, elution or flow-through collection, pool rules, strip, cleaning and storage. Include a mechanistic rationale and conditions that failed.

Compare the selected mixed-mode step with the best single-mode alternative using usable yield, critical-impurity clearance, pool volume, buffer consumption, time, lifecycle and downstream transition. Mixed mode earns its place when it improves the whole process, not when it merely produces a different chromatogram.

Set reevaluation triggers for target or feed changes. A modification that shifts charge distribution, hydrophobic exposure, glycosylation, assembly or impurity load can alter several mixed-mode interactions at once. Confirm the operating window after meaningful upstream changes.

11. Building a Mechanistic Design Space

Start with a small factorial map rather than a one-variable sequence. Choose pH levels that alter target or ligand ionization, conductivity levels that change electrostatic screening, and at least two salt identities when hydrophobic effects may contribute. Include a low normalized load initially. Measure target and critical impurity in unbound, wash and eluted material.

Use the pattern of response to infer—not assume—the dominant interaction. Loss of retention with rising salt supports an electrostatic contribution; increased retention with a salting-out salt supports hydrophobic contribution; phosphate- or chelator-specific changes can indicate coordination. Because variables also change protein conformation and solubility, pair every interpretation with recovery and stability controls.

After locating a selective region, challenge it with realistic feed variation. Spike or select lots that represent high critical-impurity load, high conductivity, upper product concentration and longer feed hold. For flow-through operation, collect sequential fractions to identify breakthrough. For bind-and-elute, test the load and wash boundaries that control target loss.

Define a failure response based on mechanism. A high-conductivity feed might be diluted or diafiltered; early impurity breakthrough may require lower load or more resin; difficult elution may require a different pH–salt sequence or a less interactive medium. Avoid correcting all failures by stronger cleaning or harsher elution.

The method package should include the experimental condition map, not only the chosen point. State acceptable ranges, interaction hypothesis, evidence, unresolved uncertainty and monitoring plan. This information is essential because a small upstream change can alter more than one retention mechanism at the same time.

References

1. Zhang K, Liu X. Mixed-mode chromatography in pharmaceutical and biopharmaceutical applications. J Pharm Biomed Anal. 2016;128:73–88.
2. Carta G, Jungbauer A. Protein Chromatography: Process Development and Scale-Up. 2nd ed. Wiley-VCH; 2020.
3. Gagnon P. Technology trends in antibody purification. J Chromatogr A. 2012;1221:57–70.