TWIN-COLUMN POLISHING

MCSGP Continuous Polishing Chromatography

Recover what batch chromatography leaves behind. MCSGP is the twin-column preparative HPLC process that breaks the purity–yield trade-off in reversed-phase and ion-exchange polishing — same resin, same buffers, no chemistry change. With AutoPeak® dynamic process control, MCSGP delivers manufacturing-grade reproducibility on the Contichrom® platform, validated at GMP scale in peptide and oligonucleotide manufacturing.

Validated across 4 PPQ runs: 99.2% Purity | −96% In-process controls | 235 h unattended operation | +9% Yield gain

Source: Eisenhuth & Müller-Späth, Processes 2025 (Bachem Bivalirudin PPQ campaign, Open Access)

Request a Webinar

Download the PPQ Paper

Request a Webinar

Download the PPQ Paper

The Purity-Yield Trade-off

Every gradient polishing step faces the same constraint. A tight centre-cut achieves target purity — but discards product in the early and late fractions flanking the peak. Widening the cut recovers that product — but fails the purity specification. Storing side fractions and reprocessing them with additional batch runs only improves yield at a massive cost: it multiplies handling steps, analytical burden, buffer consumption, and campaign time.

This is not a method development failure. It is an inherent limitation of single-column batch chromatography. When your product and its impurities are closely-eluting, no fractionation strategy can simultaneously maximize both purity and yield on a single column.

MCSGP Changes the Constraint

MCSGP (Multi-column Counter-current Solvent Gradient Purification) does not optimise the cut — it eliminates the trade-off by continuously recycling the impure side-fractions rather than discarding them.

Two identical columns operate in a coordinated cyclic process. While one column elutes, the other regenerates and loads fresh feed in parallel. The early-eluting and late-eluting impure fractions — the side-cuts that batch processes collect, pool, store, and reprocess — are instead transferred directly from the first column to the second column with in-line dilution, ready to re-adsorb and be re-separated in the next switch. Product never leaves the system unless it is pure. The result is that high yield and high purity are achieved simultaneously, in a continuous automated process.

Because MCSGP operates with the same gradient as your existing batch polishing step, there is no change to resin, buffers, or eluent composition. Your batch method is the starting point.

The performance gains are quantified across multiple validated industrial applications. Bachem AG and Bristol Myers Squibb have both published results across peptide and protein applications.

10–50%

Absolute yield increase at target purity

75% ↓

Solvent reduction vs. batch process

Zero

Re-chromatography runs, internal recycling replaces pooling

5 days → 24h

Campaign time (peptide API), Bachem, GMP production scale

Bachem AG peptide API manufacturing, Eisenhuth, R.; Müller-Späth, T. Processes, 2025, (Open Access), Weldon, R. et al. J. Chromatogr. A, 2022 (Open Access). Weldon, R. et al. Org. Process Res. Dev., 2025 (Open Access).

How MCSGP Works

Each MCSGP cycle consists of two switches, with one elution from each column per cycle. Within a switch, four phases run in sequence:

Phase	Upstream column	Downstream column
P1 – Elution start	Gradient started; weakly adsorbing impurity (W) removed	Regeneration completed; preparation for loading
P2 – Weak recycling	Columns interconnected; early low-purity fractions (W/P) transferred to downstream column with in-line dilution for re-adsorption	Receives recycled W/P fraction + binds product
P3 – Product collection	Pure product (P) center-cut collected	Fresh feed loaded
P4 – Strong recycling	Late low-purity fractions (P/S) transferred to downstream column with in-line dilution for re-adsorption	Receives recycled P/S fraction + binds product

At the end of the switch, the two columns exchange positions by valve switching — no physical movement — and the cycle repeats. The process reaches a cyclic steady state within 1–2 cycles, at which point product quality and yield are consistent from cycle to cycle.

In-line dilution during the recycling phases (P2 and P4) lowers the eluotropic strength of the transferred fraction, ensuring the recycled product re-adsorbs onto the downstream column rather than passing through. The dilution factor is calculated automatically by the MCSGP Wizard from the batch chromatogram.

The MCSGP Process Principle (.pptx)

Method Design with the MCSGP Wizard

MCSGP methods are designed in the MCSGP Wizard in ChromIQ® 9 — a tab-based interface that takes a batch chromatogram from the Contichrom® CUBE and guides you through to a complete, ready-to-run method.

Step 1 — Load the batch chromatogram and enter fraction data

Load the single-column batch design run from the CUBE directly into the wizard. Select the UV channel to use for phase boundary setting. Enter purity and concentration data from offline fraction analysis of the batch run (optional but strongly recommended). The wizard displays predicted batch yield, purity, and product loss based on the loaded chromatogram as a reference.

Step 2 — Define section boundaries by drag-and-drop

Drag five vertical boundary markers on the batch chromatogram to define the four MCSGP zones: weak recycling start (t2), product collection start (t3), product collection end (t4), and strong recycling end (t5). The wizard updates predicted yield, purity, and product loss in real time as you move the markers — no manual calculation required. In-line dilution requirements for the recycling phases are calculated automatically from the boundary positions.

Step 3 — Configure washing, regeneration, and AutoPeak®

Set column washing and regeneration steps in the Washing & Regeneration tab. Activate AutoPeak® process control in the “AutoPeak Process Control” tab — each of the four phase transitions (t2–t5) can be activated independently. For early development runs with well-defined conditions, fixed timing is sufficient. For scale-up and extended manufacturing campaigns, AutoPeak® is essential for maintaining consistent yield and purity as conditions drift. See Dynamic Process Control → for full configuration guidance.

Step 4 — Review performance prediction and generate

The Method Settings & Performance tab shows predicted process performance — productivity (g/L/h), solvent requirement (L/g), column loading (g/L), and pool concentration — before initiating the MCSGP run. Set the number of cycles for the main method and review the buffer volumes per pump inlet. Hit Save: the wizard generates a complete method package including Startup, Main, Shutdown, and Clean methods, plus a procedure file that loads them in the correct sequence. Open the procedure in ChromIQ® and run MCSGP.

01 / 04

Want to learn MCSGP?
Join our Continuous Chromatography Workshop.

AutoPeak® — Dynamic Process Control

Fixed-time phase boundaries work during early MCSGP development when conditions are well-defined and the process is closely supervised. For manufacturing campaigns — where column aging shifts retention times, buffer batches vary, and the process must run automatically for hundreds of cycles — AutoPeak® is essential.

AutoPeak® replaces fixed-time phase transitions with real-time UV-triggered boundaries. The UV signal at the outlet of the eluting upstream column is monitored continuously. When the signal crosses the defined threshold — an absolute mAU value on the ascending peak front, or a percentage of that switch’s actual peak maximum on the descending tail — a new process phases is started. Four transitions can be controlled independently: Weak Recycling Start (t2), Product Collection Start (t3), Product Collection End (t4), and Strong Recycle End (t5).

The critical detail is that the relative triggers (% of peak maximum) are recalculated fresh every switch from that switch’s actual UV maximum — not from a fixed reference. This means AutoPeak® adapts to changes in peak height as well as peak position. If feed concentration drops and the elution peak shrinks, the tail triggers still work at the correct relative point. Collection windows follow the peak wherever it moves, across the full column lifetime, without any operator input after initial configuration.

The trigger values are pre-populated from the section boundaries set on the batch chromatogram — no manual entry required.

See Dynamic Process Control → for the full mechanism and trigger type guidance

Verifying Results: The Evaluation Center

Every MCSGP run is captured in the Evaluation Center — ChromIQ®’s integrated data analysis module.

Cycle Overlay — Superimpose all cycles of a run in one view. Consecutive UV traces that are no longer changing indicate cyclic steady state has been reached — typically within 1–2 cycles for MCSGP. Visual confirmation of process consistency at a glance
Auto Integration — Quantitatively tracks product recovery, purity, and yield for each fraction across every cycle. Confirms when steady state is reached and documents per-cycle performance for process records
Superimpose — Overlay runs from separate files to compare different setpoints, resin lots, or scale-up experiments side by side
PDF and Excel export — Per-cycle chromatograms, integration tables, and process parameters compiled into a structured report for development records

The ChromIQ® logbook records every system event and AutoPeak® trigger with timestamp and triggering value — an unalterable audit trail of every decision the system made.

Platform Advantages at a Glance

Advantage	What it means in practice
10–50% absolute yield increase	Product recovered from side fractions that batch processes discard — directly reduces the crude material required per lot
Re-chromatography eliminated	No pooling, storage, or reprocessing of side fractions — QC burden and campaign time reduced proportionally
Up to 5× throughput gain	Parallel column operation and smaller columns eliminate batch idle time — more product from the same equipment footprint
Same resin and buffers	No change to chromatographic chemistry — your existing batch conditions are the starting point for MCSGP development
24/7 unattended operation	AutoPeak® maintains consistent output quality without operator supervision across extended campaigns
Direct scale-up path	Methods developed on the Contichrom® CUBE transfer directly to the Contichrom® PILOT 300X and TWIN HPLC

MCSGP Applications

MCSGP is applicable to any gradient polishing step where the product peak co-elutes with structurally similar impurities — the most common and most difficult scenario in downstream processing.

Molecule class	Stationary phase	Typical separation
Monoclonal antibodies	Cation exchange (CEX)	Charge variant separation — acidic/basic isoforms from main species, aggregate (HMW) and fragment (LMW) separation
Bispecifics / antibody variants	CEX, mixed-mode	Complex charge variant profiles; half-antibody, parental antibody, and mis-paired species removal
Antibody-drug conjugates (ADCs)	Hydrophobic interaction (HIC), RP	DAR species separation; unconjugated antibody removal
Peptides	Reversed phase (RP)	Deletion sequences, oxidized variants, closely eluting impurity removal
Oligonucleotides	Anion exchange (AEX)	n−1 / n+1 shortmers, backbone variants, closely resolved sequence impurities
Recombinant proteins	CEX, AEX, HIC	PEGylation isoform separation; aggregate and charge variant removal
Impurity isolation	RP, AEX	Preparative isolation of low-abundance impurities for reference standard production

Recommended Systems for Continuous Polishing (MCSGP)

The Contichrom® platform offers both development and scale-up systems for MCSGP.

Systems Overview

Contichrom® CUBE

The essential platform for chromatography process development. This versatile 100-bar benchtop FPLC system runs batch, integrated batch, and continuous chromatography — including MCSGP and CaptureSMB — out of the box. Develop purification processes for mAbs, peptides, oligonucleotides, ADCs, and viral vectors, with verified scale-up at over 100× to pilot and production systems.

Contichrom® PILOT

Scale peptide and oligonucleotide purification to pilot and clinical GMP manufacturing. This 100-bar preparative HPLC system features MCSGP with AutoPeak® to eliminate the yield-purity trade-off — delivering +10–50% absolute yield increase at target purity and significantly less solvent than batch HPLC. ATEX Zone 2 / Class I Div 2 certified for organic solvents. Flow rates up to 300 mL/min.

Contichrom® TWIN HPLC

Produce peptides and oligonucleotides at commercial GMP scale. This 80-bar industrial HPLC system uses MCSGP with AutoPeak® to eliminate the yield-purity trade-off — increasing recovery to >90% while reducing solvent consumption compared to batch. ATEX Zone 2 / Class I Div 2 certified for organic solvents, and flow rates up to 36 L/min.

Contichrom® TWIN LPLC – Polishing

Bring continuous polishing to biopharmaceutical manufacturing at production scale. This sanitary GMP system uses MCSGP with AutoPeak® to eliminate the yield-purity trade-off achieving +10–50% absolute yield increase at target purity while reducing buffer consumption compared to batch. CIP-ready design for mAbs, viral vectors, ADCs, and recombinant proteins.

Systems Overview

Recover Everything. Reprocess Nothing.

Book a CUBE Rental Request a System Quote

Does MCSGP require different resins or buffers than my existing batch polishing step?

No. MCSGP uses the same stationary phase and mobile phase conditions as your batch gradient polishing method — your existing resin and buffer conditions are the starting point. The MCSGP Wizard derives all process parameters from a single-column batch run performed on the Contichrom® CUBE using those same conditions.

Why does the batch chromatogram have to come from the Contichrom® CUBE?

The MCSGP Wizard in ChromIQ® loads the batch run directly from the ChromIQ® measurement database. Chromatograms from other systems use different file formats and cannot be imported. The batch design run on the CUBE is typically a short single-column experiment — it is the starting point for MCSGP development, not a separate preparatory step.

How many MCSGP cycles are needed to reach steady state?

MCSGP typically reaches cyclic steady state within 1–2 cycles after the Startup phase. This is fast compared to other cyclic processes because the Startup method pre-loads Column 1 to approximate the steady-state loading level before the main cyclic method begins. The Evaluation Center Cycle Overlay makes steady state visible immediately — consecutive UV traces that are no longer changing confirm it has been reached.

When should I activate AutoPeak®, and which transitions do I need?

AutoPeak® is not needed during initial process development when conditions are stable and the process is supervised. It becomes essential for scale-up and manufacturing campaigns where column aging, buffer variability, and temperature shifts would otherwise cause collection windows to drift off-target. For most separations, activating t3 (Product Collection Start) and t4 (Product Collection End) is sufficient — these control the critical boundary around the product peak. Weak Recycling Start (t2) adds value for separations where the early side-cut boundary is also sensitive to drift. Strong Recycle End (t5) should only be activated for Anti-Langmuir peak profiles. All trigger values are pre-populated from the batch chromatogram boundaries — no manual calculation is required.

What is the scale-up path from CUBE development to manufacturing?

Methods developed on the Contichrom® CUBE transfer directly to the Contichrom® PILOT 300X (up to 300 mL/min, GMP-ready, 21 CFR Part 11, ATEX Zone 2) and from there to the Contichrom® TWIN HPLC production system. The same section boundary positions, and AutoPeak® trigger values apply at each scale — column diameter and bed height change, method parameters do not require re-derivation.