Application note

AUC vs SEC: Is the 1% on Your Report Really 1% in Your Sample?

September 28, 2026 · Approx. 8 min read

SEC is the workhorse of aggregate testing — but it only sees the fraction of your sample that passes through the column unchanged. Analytical ultracentrifugation (AUC) uses no column and no standards, separating molecules from first principles in their own formulation buffer. That is exactly why regulators and industry use it as the orthogonal check on SEC.

1. When SEC Says "No Problem", There May Still Be One

SEC (size-exclusion chromatography) is the routine method for aggregate testing: fast, high-throughput and reproducible, sensitive to aggregates down to the 0.1% level — the mainstay of development and QC^[4]. But its results inherently carry four biases:

Four biases, hidden in routine operation

Dilution: Dilution before and inside the column can dissociate reversibly associated dimers^[9].
Mobile phase: To obtain good peak shapes, SEC often uses high-ionic-strength mobile phases that differ from the formulation, shifting the distribution of species^[4],[9].
Stationary-phase interaction: Non-specific interactions between sample and resin can both induce new aggregation and retain pre-existing aggregates on the column^[5],[9].
Separation range: Conventional SEC columns cover a limited size range.

The last bias has a hard boundary: SEC is essentially a filtration step, and aggregates larger than roughly 40–60 nm usually cannot pass the column bed^[4],[5]. "1% on the SEC report" does not mean "1% in the sample" — it may only describe the fraction that this column let through. SEC results should therefore always be checked against sample recovery^[4].

Terminology　What SEC "detects" is constrained by the separation range of the column bed; aggregates beyond the upper pore limit are filtered out before the detector, so "not detected" ≠ "not present".

Figure 1. No single method covers the full size range. SEC only sees the slice the column lets through; SV-AUC has no column bed and captures nanoscale aggregates in the same distribution (the lighter segment marks larger species that require a rotor-speed adjustment). Schematic; size classes follow the Narhi et al. classification and refs [4], [9]; method boundaries are typical values.

Orthogonality is not bureaucracy — it is physics

The US FDA guidance on immunogenicity assessment for therapeutic protein products recommends methods that, alone or in combination, enhance aggregate detection^[1]; ICH Q6B places molecular weight, size and aggregates under physicochemical properties and requires their separation and quantitation^[2]; the Chinese Pharmacopoeia 2020 (Volume III, general chapter on recombinant monoclonal antibodies for human use) likewise lists size variants as a mandatory test^[3].

The industry consensus: aggregate quantitation should be size-classified and rely on at least two orthogonal methods — because no aggregate reference standard exists, and any single method covers only a limited size window; the concentration in each window must be verified by a method based on a different measurement principle^[4],[6].

There is also published methodological precedent: the National Institutes for Food and Drug Control (NIFDC, China) has published the development and validation of an SV-AUC method for mAb size variants, optimizing rotor speed, temperature and data-processing parameters and systematically examining specificity, repeatability, precision and linearity^[7]. Full method-validation data are available on request.

2. AUC: A Race Without a Track

The physics of AUC is minimal: apply a centrifugal field, watch macromolecules redistribute in space in real time, and analyze the data quantitatively from first principles. No labeling, no chemical modification, no contact with any matrix or stationary phase — which makes it exceptionally gentle on perturbation-sensitive systems^[9],[11].

Figure 2. Analytical ultracentrifugation: in-solution separation without a column or labels. The optical system scans the cell repeatedly during centrifugation, recording concentration versus radius and time; species separate naturally by sedimentation coefficient (a function of mass, shape and density). Schematic, not to scale.

Two modes: how fast it sinks, or where it settles

Sedimentation velocity (SV-AUC) tracks the moving boundary at high rotor speed and derives a continuous sedimentation-coefficient distribution c(s) by direct fitting of the Lamm equation^[8], resolving species by s value — it is both a composition profile and a source of low-resolution shape information (frictional ratio). Sedimentation equilibrium (SE-AUC) waits at lower speed until sedimentation and diffusion balance, yielding absolute molecular weight and the association constant K_D directly from the equilibrium concentration profile^[9],[11].

Dimension	Sedimentation velocity (SV-AUC)	Sedimentation equilibrium (SE-AUC)
Setup	High speed; follow the sedimentation process	Lower speed; wait for the equilibrium distribution
Observable	Movement of the sedimentation boundary over time	Concentration distribution at equilibrium
Primary output	Sedimentation-coefficient distribution, species fractions	Absolute molecular weight, binding constants, stoichiometry
Best answers	What is in the sample, and how much of each	What is the molecular weight, how strong is the binding
Typical uses	Aggregate quantitation, oligomeric state, orthogonal check on SEC	MW confirmation, interaction thermodynamics

Table 1 | The two modes are independent in both design and analysis: decide first whether you need a "distribution" or an "affinity", then choose the mode.

Think of it this way: SV sends swimmers of different builds into the water together and watches who sinks fastest; SE waits until everyone stops moving and reads the depth at which each one floats. One measures how fast things sediment, the other where they finally rest.

Figure 3. Raw data shapes of the two modes. A: in SV the boundary moves toward the cell bottom over time — its velocity gives the sedimentation coefficient, its broadening the diffusion — and c(s) fitting converts this into a species distribution. B: in SE, sedimentation and diffusion reach equilibrium and concentration rises exponentially along the radius; the shape depends only on molecular weight and interactions, not on molecular shape.

3. Parameters, Use Cases and Three Common Misconceptions

Reusable parameters (backed by published method validation)

Parameter	Value	Basis
Rotor speed	50,000 rpm	Speed from method optimization [7]
Temperature	20 °C	Temperature from method optimization [7]
Optics / wavelength	UV / 280 nm	Multi-wavelength acquisition can separate protein and nucleic-acid signals [7]
c(s) grid resolution	150	Resolution = 150 [7]
s-value range	0.1 – 20 S	Upper bound s_max = 20 S [7]

Table 2 | Parameters from a published, validated SV-AUC method [7]; a reference starting point for similar methods. Sugar/polyol formulations and high-concentration samples need separate assessment [9], [10].

Behind these conditions sits a complete validation package: repeatability RSD 0.91% over 6 runs, overall precision RSD 0.78%, and good linearity across 0.2–0.8 mg/mL^[7].

SV-AUC delivers specificity, repeatability, precision and linearity data — it is not a research-only "look-and-see" tool but an analytical method that can be written into regulatory submissions and withstand review scrutiny; published best practices already organize SV-AUC validation within the ICH Q2 framework^[10].

Figure 4. SV-AUC c(s) distribution of a mAb sample (illustrative). A single run reads fragments, monomer, dimer and higher-order aggregates on one axis; integrating each peak gives its relative fraction, and minor peaks require significance testing (e.g., F-statistics) [9].

Questions you can hand to AUC

Decision questions: Verifying the credibility of SEC results; confirming oligomeric state and stoichiometry; interactions and binding constants K_D; aggregation-propensity assessment during formulation screening; orthogonal data for release consistency, changes and originator comparability.
Sample types: mAbs / bispecifics / Fc-fusions; peptides (GLP-1 receptor agonists); viral vectors (AAV / VLP — multi-wavelength optics resolve empty, partial and full capsids and estimate nucleic-acid content); ADCs; non-antibody proteins such as collagen.

Sample type	Questions it answers
mAb / bispecific / Fc-fusion	Aggregates, fragments, batch-to-batch consistency, formulation buffer screening, verification of SEC results
Peptides (GLP-1 receptor agonists)	Purity heterogeneity, aggregation-propensity assessment, method validation for US filings
Viral vectors (AAV / VLP)	Multi-wavelength resolution of empty / partial / full particles with nucleic-acid content estimation; aggregate control
ADC	Size variants and aggregation behavior after hydrophobic-payload conjugation; orthogonal confirmation of conjugation-related attributes
Collagen and other non-antibody proteins	Aggregate and assembly-state analysis under near-native conditions
Protein interactions / nucleic-acid complexes	Antigen–antibody ratios, ligand–receptor K_D, protein–nucleic-acid complexes

Table 3 | The scenario list summarizes our service scope; suitability for a specific sample requires a method assessment.

Three common misconceptions

Misconception 1: "SEC already measured aggregates — AUC is duplicate work."

The separation mechanisms and the physical state of the sample differ completely: AUC has no stationary phase, measures in the formulation buffer, and sees species the SEC column cannot pass. AUC is a cross-check of SEC, not a repetition^[1],[9],[10].

Misconception 2: "AUC is just for molecular weight."

SV focuses on species distributions; SE focuses on molecular weight and thermodynamic parameters — entirely different experimental designs and data paths^[11].

Misconception 3: "If the two methods disagree, the experiment was done badly."

Quite the opposite — the discrepancy itself is the point. Especially in heat-, stirring- or shaking-stressed samples, SEC and SV-AUC frequently disagree, and that signals species the SEC run failed to recover: a diagnostic clue that must be investigated^[4],[9].

4. Quickgene Analytical Services: AUC Offering and Deliverables

Our analytical ultracentrifugation platform covers method development, full method validation, routine testing and data fitting — every step already exercised in regulatory projects. Representative work includes:

US registration of peptide products: Full validation of SV-AUC methods for semaglutide, tirzepatide and others (specificity, precision, accuracy, limit of quantitation, linearity and range, robustness);
AAV release testing: Discrimination and quantitation of empty / partial / full capsids and aggregate control — full method validation delivered;
BLA filings for antibody drugs: Development and validation of SV-AUC methods for size variants in support of submission dossiers;
ADC characterization: Size-variant and aggregation assessment after hydrophobic-payload conjugation, with orthogonal confirmation of conjugation-related attributes alongside LC-MS;
Collagen and other recombinant proteins: Aggregate and assembly-state analysis under near-native conditions.

Standard deliverables: c(s) profiles, quantitative species fractions, molecular weight or binding constants (mode-dependent), method and results reports, and ICH-structured validation protocols and reports.

Explore SV-AUC aggregation and self-association analysis

Aggregation is a quality attribute no single column can fully reveal. Put the sample back into its own buffer and let it sediment once under first principles — that often supplies the part of the answer "SEC cannot explain".

If your program faces challenged SEC results, insufficient originator-comparability data, or peptide/AAV aggregation and empty-capsid questions needing orthogonal evidence, talk to us. For timelines and pricing, email [email protected] or request a quote online.

AUC in three sentences

It is an in-situ measurement: No stationary phase, no labels, no standards — you see molecules as they are in solution, giving SEC an independent second opinion.
Question first, mode second: SV answers "distribution and fractions"; SE answers "molecular weight and affinity". The experiment is dictated by the question, not the instrument.
It is a cross-check, not a replacement: Best for verifying critical samples and attributing disputed data; given run time and sample constraints, routine high-throughput testing still relies on SEC.

References

U.S. Food and Drug Administration. Guidance for Industry: Immunogenicity Assessment for Therapeutic Protein Products. CDER/CBER, August 2014. fda.gov/media/85017/download
ICH. Q6B: Specifications: Test Procedures and Acceptance Criteria for Biotechnological/Biological Products. Step 4, March 1999.
国家药典委员会. 中华人民共和国药典（2020 年版）三部：人用重组单克隆抗体制品总论. 北京：中国医药科技出版社, 2020. (Chinese Pharmacopoeia 2020, Vol. III — general chapter on recombinant monoclonal antibodies for human use.)
Uchiyama S. Aggregates Quantification of Biopharmaceuticals in a Wide Range of Sizes Using Orthogonal Methods. American Pharmaceutical Review, 2019. americanpharmaceuticalreview.com
Carpenter JF, Randolph TW, Jiskoot W, Crommelin DJ, Middaugh CR, Winter G. Potential inaccurate quantitation and sizing of protein aggregates by size exclusion chromatography: essential need to use orthogonal methods to assure the quality of therapeutic protein products. J Pharm Sci. 2010;99(5):2200–2208. doi:10.1002/jps.21989
den Engelsman J, Garidel P, Smulders R, et al. Strategies for the assessment of protein aggregates in pharmaceutical biotech product development. Pharm Res. 2011;28(4):920–933. doi:10.1007/s11095-010-0297-1
National Institutes for Food and Drug Control (NIFDC), China. Development and validation of a sedimentation-velocity analytical ultracentrifugation method for size variants of monoclonal antibodies (in Chinese). (Full bibliographic details being updated.)
Schuck P. Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and Lamm equation modeling. Biophys J. 2000;78(3):1606–1619. doi:10.1016/S0006-3495(00)76713-0
Uchiyama S, Noda M, Krayukhina E. Sedimentation velocity analytical ultracentrifugation for characterization of therapeutic antibodies. Biophys Rev. 2018;10(2):259–269. doi:10.1007/s12551-017-0374-3
Best Practices for Aggregate Quantitation of Antibody Therapeutics by Sedimentation Velocity Analytical Ultracentrifugation. J Pharm Sci. 2022;111(7):2121–2133. doi:10.1016/j.xphs.2021.12.023
Berkowitz SA. Role of analytical ultracentrifugation in assessing the aggregation of protein biopharmaceuticals. AAPS J. 2006;8(3):E590–E605. doi:10.1208/aapsj080368