How to design an equilibrium SAXS experiment

SAXS has emerged as a standard biophysical tool deployed routinely for characterizing macromolecules of biomedical interest. The relative logistical simplicity embodied in a technique that provides easy access to the size and low-resolution shape of macromolecules makes it an essential part of the biophysicists’ tool kit. With the introduction of Size-Exclusion Chromatography (SEC)-SAXS at BioCAT, which ensures monodispersity of even the most biochemically challenging molecules, structural parameters such as the Radius of Gyration (Rg), Maximum Dimension (Dmax), Volume and Molecular weight estimates can be determined with a high degree of success for a large variety of samples.

There are four equilibrium data acquisition strategies available at BioCAT: SEC-SAXS, SEC-MALS-SAXS, IEC-SAXS and batch mode SAXS.

Below we give some general guidelines for designing your SAXS experiment. If you have questions, or are a new user, please contact a beamline scientist.

What technique should I use?

BioCAT strongly encourages all users to use either SEC-SAXS or SEC-MALS-SAXS. If multiple components in solution cannot be separated by size, IEC-SAXS may be the appropriate choice. There are also some rare cases where sample concentration and volume are inadequate for SEC-SAXS, in which case you will use batch mode. Batch mode is also useful if you need to measure a large number of buffer conditions, such as in a titration experiment.

SEC-MALS-SAXS allows highly accurate quantification of molecular weight, making it generally superior to SEC-SAXS. However, the equilibration times for the SEC-MALS system are quite long (at least 6 hours), which limits the number of buffer changes. Because of the sensitivity of the system, the requirements on the sample quality are much higher than for SEC-SAXS.

SEC-SAXS is the right choice if …

  • Your system has a single well defined peak or several well resolved peaks (not including large aggregates that show up in the void)
  • You will need to make several buffer changes during your experiment

SEC-MALS-SAXS is the right choice if …

  • You have a complicated elution with several overlapping or poorly resolved peaks
  • You are measuring a complex or oligomer with unknown stoichiometry
  • You need at most one buffer change
  • There is a small amount of elution in the void

IEC-SAXS is the right choice if …

  • You have multiple components in solution that cannot be well separated by size
  • You can design an IEC gradient that separates your components

Batch mode SAXS is the right choice if …

  • You can’t meet the concentration and volume requirements for SEC-SAXS (see below)
  • You need a large number (>~5) buffer changes due to the nature of your experiment (e.g. a titration experiment)
  • Your system does not survive the a column, for example some weakly associating complexes

What sample concentration and volume do I need?


As a rule of thumb, you will get good SAXS data if you prepare a protein concentration in mg/ml that is:

  • 240/MW in kDa for SEC-SAXS or SEC-MALS-SAXS
  • 60/MW in kDa for batch mode SAXS

For example, using SEC-SAXS for a 20 kDa protein you would want a concentration of ~240/20 = 12 mg/ml whereas for a 150 kDa protein you would want a concentration of ~240/150 = 1.6 mg/ml. RNA and DNA samples scatter ~2.5 times more strongly than protein, so you can use a concentration of ~96/MW and 24/MW for SEC-SAXS and batch mode SAXS respectively.

Note that the MW calculation is done for the expected oligomeric MW. So if a protein is a 50 kDa monomer, but you expect it to be dimeric, you would prepare a concentration of ~240/100 = 2.4 mg/ml.

Note that in SEC-SAXS and SEC-MALS-SAXS experiments the sample is diluted by the column, and may elute in several peaks. This should be accounted for. The recommendation above adds a column dilution factor of ~4x for your SEC-SAXS sample preps. If a significant amount of your sample elutes in other peaks, further increase your concentration to compensate.

Also, for SEC-SAXS, if your sample has concentration dependent aggregation that prevents you from achieving the high concentrations needed, you may be able to compensate by increasing your load volume.


Typical SEC-SAXS and SEC-MALS-SAXS injection volumes are ~250-350 µl.

Batch mode experiments can be done routinely with 50 µl per measurement. Note that typical batch mode experiments require measuring several concentrations, and so can actually end up requiring more volume than a SEC-SAXS experiment.

We recommend that you bring enough sample volume for two measurements at each condition, that way if something happens during a measurement (such as the beam going down) we can repeat and obtain the desired data.

How many samples should I bring?

When considering how many samples to bring, you need to think both about the experiment time and the equilibration.

Experiment time

At BioCAT, users typically use the GE Superdex 200 Increase 10/300 GL column. This has a column volume of 24 and a flow rates of ~0.7 ml. That means that a 1.5 column volume (CV) experiment for SEC-SAXS or SEC-MALS-SAXS takes ~50 minutes. If you know that nothing elutes after 1 CV (including small molecules) you can further reduce this to ~34 minutes. So you should expect to run ~1-2 samples an hour.

With the coflow cell, BioCAT users now have the ability to run samples on the GE 5/150 columns without radiation damage. These columns provide significantly less separation, and so should only be used on a system with very well resolved peaks (ideally just one peak, or a peak plus elution at the void volume). However, if your sample is appropriate, the volume requirements and run times are much lower. With these columns, typical load volumes are ~100 µL and run times are ~10 minutes.

Note: If you bring your own column, run times will depend on the flow rate and volume for that column.

Batch mode samples are much faster, typically only 3-30 s of exposure. Throughput is limited by sample loading. Realistically, expect to do a sample every 3-5 minutes.


You will have to equilibrate the column at the start of your SEC-SAXS or SEC-MALS-SAXS experiment, and every time thereafter that you want to change buffers. For SEC-SAXS, we recommend a 2 CV equilibration, which for a GE Superdex 200 Increase 10/300 GL column will take ~1.25 hours. For SEC-MALS-SAXS, equilibration requires at least 6 hours, and is ideally done overnight.

What column should I use?

BioCAT provides a number of columns for users. Typically, users will use one of these:

  • Superdex 200 Increase, both 10/300 and 5/150 (MW ~10-600 kDa)
  • Superdex 75 Increase, both 10/300 and 5/150 (MW ~3-70 kDa)
  • Superose 6 Increase, both 10/300 and 5/150 (MW ~5-5,000 kDa)

Generally speaking, pick the column with the narrowest MW range that can accommodate your samples. But remember that the MW range is for globular proteins, extended proteins run as if they are higher MW! BioCAT recommends running a test separation in your lab, to ensure you can resolve your species. The default column at BioCAT for all experiments is the Superdex 200 Increase 10/300.

A full list of columns and the corresponding MW ranges is available for both SEC-SAXS and SEC-MALS-SAXS.

Users may also provide their own columns if desired. However, due to the dilution factor, we recommend that you only use analytical grade columns, not the larger prep columns like the Cytiva HiPrep or HiLoad columns.

How much buffer should I bring?

The following are intended as guidelines for users when planning their experiments. However, as most buffers do not contain precious components, we recommend bringing more buffer than you think you’ll need, for example taking the below numbers and adding 50%. You never know when you might want to change buffers and do one more run with a given sample, and have to equilibrate the column again.

If you have precious components in your buffer, there are ways to reduce your buffer usage. Please contact a beamline scientist to discuss those situations.

Given the large volume of buffer required for experiments, many of BioCAT’s users find it convenient to bring 10x concentrated stocks of buffer and then dilute on-site.


For SEC-SAXS experiments, you can calculate the amount of buffer you need as:

Buffer volume = 5*(column volume)*(number of samples + 1) + 250 mL

This accounts for both the per-sample use and the equilibration. Please note that the system cannot use all the buffer in a bottle, as you cannot risk drawing air into the system. There is also a fixed volume used for pump washing. This is the 250 mL offset in the above formula.

For example, if you are using the GE Superdex 200 Increase 10/300 GL column, it has a column volume of 24 mL. If you’re planning to run 5 samples in a particular buffer you should bring:

Buffer volume = 5*(24 mL)*(5+1) + 250 mL ~ 1 L

For these experiments, you should always bring at least 0.5 L of any buffer you are using.


For SEC-MALS-SAXS experiments equilibration needs significantly more buffer than SEC-SAXS experiments. Additionally, you cannot stop the buffer flow between experiments. In this case it is more useful to calculate buffer requirements by running time. Equilibration is done at the same flow rate as experiments. The coflow sheath also requires buffer. Given that, you can calculate the buffer you need as follows:

Buffer volume = 4*(experiment time)*(flow rate) + (equilibration time)*(flow rate)

For example, if you are using the Superdex 200 Increase 10/300 which has a flow rate of 0.6 mL/min, and you plan on 12 hours (720 minutes) of equilibration (overnight) and 8 hours (480 minutes) of experiments in a given buffer, you should bring:

Buffer volume = 4*(480 min)*(0.6 mL/min) + (720 min)*(0.6 mL/min) ~ 1.6 L

For these experiments, you should always bring at least 1.5 L of any buffer you are using.

Batch Mode

Batch mode experiments require a basic running buffer with ~ 1 L of volume. For buffers with precious or limited components, a basic running buffer need not contain that component. The same is true if you have lots of buffer changes (e.g. titration of a ligand or salt concentration).

Besides the basic running buffer, you need additional aliquots of a perfectly matched buffer for each sample. You nominally need just 200 µl of each matched buffer per sample (where each different concentration of the same system counts as a distinct sample). However, we never recommend bringing less than ~5 mL of each buffer, just in case. If you are in a situation where this is too much, please contact a beamline scientist to discuss how much buffer you need.