What Is Continuous Chromatography?
Continuous chromatography is a multicolumn purification strategy where feed is loaded without interruption by coordinating two or more columns through staggered load, wash, elute, and regeneration cycles. Unlike batch chromatography, where a single column sits idle during non-loading steps, continuous chromatography keeps every column productive at all times, dramatically improving throughput per unit of installed resin.
The concept originated in the petrochemical industry with simulated moving bed (SMB) technology in the 1960s, but its application to protein purification only gained traction in the 2000s as biopharmaceutical titers rose and the cost of chromatography resins, especially Protein A, became a dominant factor in downstream processing economics.
Three families of continuous chromatography now dominate bioprocessing:
- Periodic countercurrent chromatography (PCC) — 2–4 columns in staggered bind-and-elute cycles for capture steps
- CaptureSMB — a twin-column variant of PCC with integrated breakthrough control for Protein A capture
- MCSGP (Multicolumn Countercurrent Solvent Gradient Purification) — twin-column gradient elution with internal side-fraction recycling for polishing
Why Move from Batch to Continuous?
The economic case for continuous chromatography rests on three drivers: resin cost, buffer consumption, and facility footprint. A single Protein A column for a 2,000 L bioreactor can cost $500,000–$2,000,000 in resin alone. Continuous operation lets you achieve the same throughput with 40–60% less resin by loading columns closer to their dynamic binding capacity (DBC) rather than stopping at the conservative 80% breakthrough used in batch.
Buffer savings are equally significant. Batch Protein A capture typically uses 15–25 column volumes (CV) of buffer per cycle. Continuous processes reduce this to 8–15 CV because smaller columns require proportionally less wash and equilibration volume. At manufacturing scale, this translates to hundreds of thousands of liters of buffer per campaign saved, plus the WFI and storage infrastructure to support it.
The facility footprint benefit compounds with the other two. Smaller columns mean smaller skids, smaller buffer hold tanks, and less clean-in-place (CIP) infrastructure. For new greenfield facilities, continuous chromatography enables a meaningfully smaller building.
| Parameter | Batch (1 column) | 3-Column PCC | Twin-Col CaptureSMB |
|---|---|---|---|
| Resin utilization (% of DBC) | 60–80% | 85–95% | 85–92% |
| Productivity (g/L resin/h) | 5–15 | 15–40 | 12–30 |
| Buffer consumption (CV/cycle) | 15–25 | 8–15 | 10–16 |
| Resin volume needed (relative) | 1.0× | 0.4–0.6× | 0.5–0.7× |
| Product concentration in eluate | 5–15 g/L | 8–25 g/L | 10–25 g/L |
| Process complexity | Low | Medium | Medium |
| Number of columns | 1 | 3–4 | 2 |
Continuous Chromatography Modes Compared
Not all continuous chromatography is the same. The choice of mode depends on whether you need bind-and-elute capture or gradient polishing, the number of product-related impurities, and the required separation resolution. The table below summarizes the four main modes used in bioprocessing.
| Mode | Columns | Elution | Best For | Separation Type |
|---|---|---|---|---|
| SMB (Simulated Moving Bed) | 4–8 | Isocratic | Binary separations (2 components) | Flow-through or weak partitioning |
| PCC (Periodic Countercurrent) | 2–4 | Step gradient | Protein A/G capture, CEX capture | Bind-and-elute |
| CaptureSMB | 2 | Step gradient | Protein A capture with breakthrough control | Bind-and-elute |
| MCSGP | 2–3 | Linear gradient | Polishing (charge variants, PEGylation, aggregates) | Gradient with side-fraction recycle |
Traditional SMB was designed for petrochemical separations and is limited to isocratic elution, making it poorly suited to the step or gradient elution needed for most protein purifications. The bioprocess-adapted variants (PCC, CaptureSMB, MCSGP) solve this limitation through different engineering approaches.
Continuous Capture: PCC and CaptureSMB
Continuous capture processes work by overlapping the loading phase of one column with the non-loading phases (wash, elute, regeneration) of the others. The key innovation is deliberate overloading: in batch chromatography, you stop loading well before breakthrough to avoid product loss. In continuous capture, you intentionally load past the breakthrough point because product that breaks through the first column is captured by a second column connected in series during part of the cycle.
Periodic Countercurrent Chromatography (PCC)
PCC uses 3 or 4 columns in a staggered cycle. At any moment, one column is in the loading zone (receiving feed), one is being washed and eluted, and one is being regenerated and equilibrated. The "periodic countercurrent" name refers to the fact that, during part of the loading phase, two columns are connected in series so that breakthrough from the first flows into the second. This interconnection phase is what allows loading to 85–95% of dynamic binding capacity, versus the 60–80% typical of batch.
CaptureSMB
CaptureSMB achieves similar performance with only two columns by using an inline UV detector to monitor breakthrough in real time. When the UV signal at the outlet of the first column indicates product is breaking through, the system automatically switches to an interconnected configuration where the second column captures the overflow. This real-time feedback (called AutomAb control) eliminates the need for conservative loading estimates and adapts to changes in feed titer automatically.
In a head-to-head comparison, Baur et al. (2016) found that all multicolumn processes (CaptureSMB, 3-column, and 4-column PCC) achieved similar maximum capacity utilization of 85–92% and significantly outperformed batch. The productivity improvement ranged from 2–5 fold depending on feed titer, with the greatest advantage at lower titers (1–3 g/L) where batch column idle time is proportionally largest.
Worked Example: Resin Savings from Continuous Capture
Scenario: mAb process, 2,000 L bioreactor, 5 g/L titer, total harvest = 10 kg mAb
Batch approach:
- Protein A resin DBC = 35 g/L resin at 4 min residence time
- Batch loading to 70% DBC = 24.5 g/L effective capacity
- Resin volume needed = 10,000 g / 24.5 g/L = 408 L resin
- At $12,000/L, resin cost = $4,896,000
3-column PCC approach:
- Same resin, but loading to 90% DBC = 31.5 g/L effective capacity
- Three smaller columns cycle multiple times per harvest (each ~60 L, running ~2 cycles each)
- Total installed resin = 3 × 60 L = 180 L (56% reduction vs batch)
- Each column processes 60 L × 31.5 g/L = 1,890 g per cycle; 3 columns × ~2 cycles each ≈ 11.3 kg capacity (margin for yield losses)
- Resin cost = 180 L × $12,000/L = $2,160,000
Savings: $2.7 M in resin per campaign, plus 30–40% buffer reduction.
Chromatography Calculator
Calculate column dimensions, linear velocity, residence time, and buffer volumes for your chromatography step.
Continuous Polishing: MCSGP
MCSGP solves a fundamentally different problem than PCC or CaptureSMB. In polishing chromatography (typically ion exchange), the target product co-elutes with closely related impurities like charge variants, aggregates, or differentially PEGylated species. In batch chromatography, you must choose between high purity (narrow cut, low yield) or high yield (wide cut, low purity). This yield–purity trade-off is an inherent limitation of single-column gradient elution.
MCSGP eliminates this trade-off by operating two columns in alternating interconnected and batch phases. During the batch phase, each column runs a standard gradient elution. The key innovation happens in the interconnected phase: impure side fractions (the leading and trailing edges of the product peak that contain co-eluting impurities) are not discarded but are instead loaded onto the second column for re-chromatography. This internal recycling concentrates the product while allowing impurities to be washed out over multiple cycles.
The results are striking. For a therapeutic peptide separation where batch chromatography achieved only 19% yield at the required purity, MCSGP achieved 94% yield at the same purity (De Luca et al., 2020). For mAb charge variant separation, Muller-Spath et al. (2008) demonstrated 93% yield and 93% purity simultaneously, a combination impossible with batch gradient elution.
| Product | Separation | Batch Yield | MCSGP Yield | Purity | Source |
|---|---|---|---|---|---|
| Therapeutic peptide | CEX gradient | 19% | 94% | ~90% | De Luca et al. 2020 |
| mAb charge variants | CEX gradient | ~70% | 93% | 93% | Muller-Spath et al. 2008 |
| Oligonucleotide | AEX gradient | 56% | 91% | 92% | De Luca et al. 2020 |
| Cannabidiol | RP gradient | 52% | 94% | >99.5% | De Luca et al. 2020 |
| mAb active variant | CEX gradient | ~60% | 90% | >95% | Muller-Spath et al. 2010 |
| PEGylated protein | CEX gradient | ~40% | 85% | >90% | Steinebach et al. 2016 |
Commercial Systems and Selection
Five commercial platforms cover the full range of continuous chromatography applications, from process development screening through GMP manufacturing. Selection depends on the chromatography mode needed (capture vs. polishing), the number of columns, and scale requirements.
| System | Vendor | Columns | Mode | Scale Range | Single-Use |
|---|---|---|---|---|---|
| AKTA PCC 75 | Cytiva | 3 | PCC (capture) | Lab to process (up to 75 cm ID) | No (reusable flow path) |
| Cadence BioSMB PD | Sartorius (Pall) | 3–16 | PCC/SMB (capture) | PD screening (1–5 mL cols) | Yes |
| Cadence BioSMB 350 | Sartorius (Pall) | 3–16 | PCC/SMB (capture) | Clinical/commercial | Yes (disposable flow path) |
| Contichrom CUBE | YMC (ChromaCon) | 2 | CaptureSMB + MCSGP | Lab/PD (0.5–20 mL cols) | No |
| Contichrom TWIN LPLC | YMC (ChromaCon) | 2 | CaptureSMB + MCSGP | Pilot/GMP | No |
| BioSC | Novasep | 2–12 | SMCC (sequential) | Pilot to commercial | No |
| Octave | Semba Biosciences | 1–8 | CMCC (flexible) | Lab to pilot | No |
For mAb Protein A capture at clinical or commercial scale, the choice typically narrows to AKTA PCC (if the facility is already a Cytiva shop) or Cadence BioSMB (if single-use flow paths are preferred). For MCSGP polishing, the Contichrom platform from YMC/ChromaCon is the only commercial system with validated AutoPeak control algorithms.
Resin Lifetime Calculator
Track resin cycle count, dynamic binding capacity decay, and replacement scheduling for your chromatography columns.
Implementation and Regulatory Considerations
Implementing continuous chromatography requires addressing process development, validation, and operational considerations that differ from batch. The regulatory path is clear. FDA and EMA both encourage continuous manufacturing, and multiple approved biologics use PCC-based Protein A capture. The key regulatory requirement is not batch equivalence, but demonstration that the process reaches and maintains a cyclic steady state with consistent product quality.
Process Development Workflow
- Batch baseline: Establish single-column performance (DBC, step yield, purity) using standard resin screening.
- Breakthrough characterization: Run breakthrough curves at the intended residence time. This data drives PCC cycle timing and CaptureSMB AutomAb parameters.
- Cycle optimization: Use a lab-scale continuous system (Contichrom CUBE or BioSMB PD) to optimize interconnection time, load volume, and wash stringency.
- Steady-state verification: Run 20–30 cycles and demonstrate that yield, purity, and impurity clearance (HCP, DNA, leached Protein A) are consistent from cycle 5 onward.
- Scale-up: Scale by column dimensions (maintain residence time and linear velocity), not by cycle count.
Validation Strategy
Continuous chromatography validation follows the same ICH Q8–Q11 framework as batch, with two additions:
- Cyclic steady state definition: Define the number of start-up cycles before product collection begins (typically 3–5 cycles) and demonstrate that CQAs are within specification from that point onward.
- Resin lifetime under continuous use: Continuous operation exposes resin to more frequent CIP cycles. Resin lifetime studies must account for this accelerated cycling, typically validating 100–300 cycles rather than the 50–100 cycles common in batch.
Worked Example: Cycle Time Calculation for 3-Column PCC
Given:
- Column volume = 5 L (each of 3 columns)
- Residence time for loading = 4 min
- Resin DBC at 4 min RT = 35 g/L
- Target loading = 90% DBC = 31.5 g/L
- Feed titer = 5 g/L
- Flow rate = column volume / residence time = 5 L / 4 min = 1.25 L/min
Step 1: Load volume per column per cycle
Load volume = (31.5 g/L × 5 L) / 5 g/L = 31.5 L
Step 2: Load time per column
Load time = 31.5 L / 1.25 L/min = 25.2 min
Step 3: Non-load time (wash + elute + CIP + re-equil)
Wash: 3 CV × 4 min/CV = 12 min
Elute: 5 CV × 4 min/CV = 20 min
CIP: 3 CV × 4 min/CV = 12 min
Equil: 3 CV × 4 min/CV = 12 min
Total non-load = 56 min
Step 4: Check feasibility
With 3 columns, you have 2 columns × 25.2 min = 50.4 min of non-load time available while the other columns are loading. This is slightly less than the 56 min needed, so you would either increase residence time slightly (to 4.5 min) or reduce CIP to 2 CV. At 4.5 min RT: load time = 28.4 min, non-load window = 56.7 min. Feasible.
Frequently Asked Questions
What is continuous chromatography in bioprocessing?
Continuous chromatography uses two or more columns operating in a coordinated cycle so that feed is loaded without interruption while other columns undergo wash, elution, and regeneration. This eliminates the idle time inherent to batch chromatography, increasing resin utilization by 40–60% and reducing buffer consumption by 30–50% compared to single-column batch operation.
What is the difference between SMB and PCC chromatography?
Traditional simulated moving bed (SMB) chromatography uses 4–8 columns with simulated countercurrent flow for isocratic binary separations. Periodic countercurrent chromatography (PCC) uses 2–4 columns in a simpler load-wash-elute cycle with staggered timing, making it better suited for bind-and-elute protein capture. PCC is the dominant format for Protein A capture of monoclonal antibodies.
How does MCSGP differ from CaptureSMB?
CaptureSMB is a twin-column process designed for bind-and-elute capture steps (e.g., Protein A), where breakthrough from the first column is captured by the second. MCSGP is designed for polishing separations with gradient elution, automatically recycling impure side fractions to break the yield–purity trade-off. CaptureSMB maximizes resin utilization; MCSGP maximizes recovery of closely eluting variants.
What productivity improvement can continuous chromatography achieve?
Continuous capture processes typically achieve 2–5 fold productivity increases over batch, measured in grams of product per liter of resin per hour. CaptureSMB and 3-column PCC show 40–60% higher capacity utilization, while MCSGP polishing can increase yield by 20–75 percentage points at equivalent purity. Buffer savings of 30–50% are consistently reported across platforms.
Which commercial systems are available for continuous chromatography?
Major commercial platforms include AKTA PCC (Cytiva, 3-column PCC), Cadence BioSMB (Sartorius/Pall, 3–16 columns), Contichrom CUBE/TWIN (YMC/ChromaCon, twin-column CaptureSMB and MCSGP), and BioSC (Novasep, sequential multicolumn). All offer lab-to-GMP scale options, with the Cadence BioSMB 350 and AKTA PCC 75 being the most widely adopted for clinical and commercial manufacturing.
Is continuous chromatography accepted by regulators for GMP manufacturing?
Yes. FDA and EMA have approved biologics manufactured using continuous chromatography, and both agencies actively encourage continuous manufacturing through regulatory guidance. Multiple clinical-stage and commercial mAb processes use PCC-based Protein A capture. The key regulatory requirement is demonstrating process understanding and consistent product quality across the cyclic steady state.
Related Tools
- Chromatography Calculator — Column dimensions, linear velocity, residence time, and buffer volume calculations.
- Resin Lifetime Calculator — Track DBC decay, cycle count, and resin replacement scheduling.
- TFF/Filtration Calculator — Membrane sizing and diafiltration volume calculations for the UF/DF step following chromatography.
References
- Muller-Spath T, Aumann L, Melter L, Strohlein G, Morbidelli M. Chromatographic separation of three monoclonal antibody variants using multicolumn countercurrent solvent gradient purification (MCSGP). Biotechnol Bioeng. 2008;100(6):1166–1177. doi:10.1002/bit.21843
- Muller-Spath T, Krattli M, Aumann L, Strohlein G, Morbidelli M. Increasing the activity of monoclonal antibody therapeutics by continuous chromatography (MCSGP). Biotechnol Bioeng. 2010;107(4):652–662. doi:10.1002/bit.22843
- Angarita M, Muller-Spath T, Baur D, Lievrouw R, Lissens G, Morbidelli M. Twin-column CaptureSMB: A novel cyclic process for protein A affinity chromatography. J Chromatogr A. 2015;1389:85–95. doi:10.1016/j.chroma.2015.02.046
- Baur D, Angarita M, Muller-Spath T, Steinebach F, Morbidelli M. Comparison of batch and continuous multi-column protein A capture processes by optimal design. Biotechnol J. 2016;11(7):920–931. doi:10.1002/biot.201500481
- Steinebach F, Muller-Spath T, Morbidelli M. Continuous counter-current chromatography for capture and polishing steps in biopharmaceutical production. Biotechnol J. 2016;11(9):1126–1141. doi:10.1002/biot.201500354
- De Luca C, Felletti S, Lievore G, et al. Modern trends in downstream processing of biotherapeutics through continuous chromatography: The potential of Multicolumn Countercurrent Solvent Gradient Purification. TrAC Trends Anal Chem. 2020;132:116051. doi:10.1016/j.trac.2020.116051