How to Prepare and Perform Solution SAXS Experiments

Q Range

We typically do Small Angle X-ray Scattering (SAXS) experiments at an x-ray photon energy of 12 keV. The Pilatus 1M detector will be used for SAXS. In the case of simultaneously SAXS and WAXS measurements a second Pilatus 100 K detector will be used at a short distance from the sample.We have a dedicated SAXS camera with  a sample-to-detector distance of 3.5 m yielding a Q range 0.004--0.33 Å-1 which is suitable for protein molecules ranging from several kDa to a few MDa.

For our WAXS instrument, please contact the beamline scientist to discuss your needs.

Sample Quantity

To minimize radiation damage, typically 10 - 20 frames of short exposures (0.1 - 1s) are taken on a flowing sample in a capillary cell. The data are averaged over the multiple frames to improve the signal/noise ratio after confirming the absence of radiation damage. A minimum sample volume of 120 μl is recomended for Flow-Thru mode during the X-ray exposure. However, in special cases even a sample volume of 20 μl may also be able to yield reasonable SAXS data.

In the case of online chromatography SAXS, a sample volume of 150-500 μl of a few mg/mL is desirable to load the sample to GE Akta pure. Concerning the sample quantity and requirements for time-resolved experiments using the stopped-flow or microfluidic mixers, please contact the beamline scientist for details.

Sample Concentration


For proteins of a size comparable to lysozyme (14.3 kDa) or cytochrome c (~ 12 kDa), a concentration of 2 mg/ml can give reasonably good data quality. If the protein has twice the size of lysozyme or cytochrome c, the concentration can be reduced by a factor of two. Higher concentrations can be used to give better data quality if the protein does not suffer from aggregation. We can measure samples with a concentration of 0.5 mg/ml at long exposures and have good data quality in the measured q region.

DNA and RNA scatter x-rays more strongly than proteins, so the required concentration can be about 5 fold lower than proteins.

Matched Buffer

Scattering data taken on a protein solution contains signals from both the protein and the buffer. Scattering measurements will be done on both the protein solution and the buffer and then the  buffer scattering is then subtracted from the solution scattering. The buffer-subtracted data corresponds to the scattering signal of the protein alone. Therefore, the buffer measurement is as essential as the protein solution measurement. The exact chemical composition in the buffer and in the solution must be matched in order to obtain a proper buffer subtraction.

One way to get matched buffers is to dialyze the protein solution in a buffer and then bring both the protein solution and the buffer for SAXS measurement. Since the buffer may also be used to flush the capillary flow-cell before measurement, a volume of 10 ml of buffer is required for each protein solution measurement. It is recommended to bring 50-100 ml of buffer with an additional 10 ml for each protein sample. SEC SAXS requires more buffer to equilibrate the size-exclusion column which consumes 50 mL for each buffer condition. So a rule of thumb is always bring excess buffer solution. It is also possible to make buffers in our wetlab onsite. In such cases, please contact the beamline scientist in advance.

Performing Size Exclusion Chromatography SAXS

The default data collection strategy at BioCAT is Size-Exclusion Chromatography SAXS (SEC-SAXS), which involves having a liquid chromatography unit in line with the SAXS camera which facilitates purification of samples using an appropriate gel-filtration column immediately before exposure to the x-ray beam (Menhart 2004). This ensures optimal sample quality and separates the molecule of interest from potential aggregates, oligomers, or breakdown products which are often seen in samples that are purified days to weeks before data acquisition and can adversely affect data quality and interpretability. The continuously flowing sample very effectively counters the risk of radiation damage. This strategy is also particularly elegant because buffer matching is relatively simple and there is a built-in concentration series which samples a much more extensive number of concentrations with minimal effort compared to the traditional data collection strategies.


The FPLC unit being used for SEC-SAXS is called the AKTA-pure (GE HealthSciences), and it comes equipped with a UV-monitor, capable of measuring UV absorbance at 3 separate wavelengths (200nm to 700nm) simultaneously, a loop valve and a column valve which facilitate optimal efficiency. We are able to pre-load up to 5 samples at a time and also keep the system connected with up to 5 columns simultaneously which could be pre-equilibrated and therefore reduce the time consumed in transitioning to multiple samples and between different columns.

While we recommend that users bring their own columns for the sake of consistency and to avoid cross-contamination issues, we do provide a variety of columns in two different sizes, viz., superdex-200, superdex-75, and superpose-6 columns in 10/300 and 5/150 sizes (column volumes of 24mls and 3mls respectively). The default flow rate used for SEC-SAXS experiments is 0.7mls/min and the typical experiment takes ~40mins. For more detailed technical information on the AKTA-pure, please refer to the manufacturer's manual and for more specific queries about our setup contact the beamline scientists.

Sample and Buffer Requirements:

Most commonly used buffer systems (e.g. Tris, HEPES, Phosphate buffer, MES, etc) are compatible with our data-acquisition strategy. Salt concentrations up to 1M are tolerated but lower salt concentrations are recommended. Use of reducing agents such as DTT and TCEP, which are also known free radical scavengers is recommended to counter the possibility of radiation damage in concentrations of up to 1mM. Glycerol has been successfully used in the past at concentrations as high as 10%. Higher concentrations of Glycerol are discouraged as they would preclude the optimal flow-rate of 0.7ml/min on account of higher back pressures thus increasing the likelihood of radiation damage. Optimal sample concentrations are dependent on the size of the molecule. For relatively small molecules in the 10-25kDa range, we recommend preparing ~500ŵl of ~5mg/ml and for medium sized molecules in the 25-75kDa range, ~300ŵl of ~5mg/ml and for large molecules (75kDa and above), ~200ŵl of ~5mg/ml is sufficient for one SEC-SAXS run.

Samples with nucleic acids (DNA, RNA or protein-nucleic acid complexes) also show strong scattering and need relatively less sample than protein. We strongly suggest test-runs with analytical columns before arriving at the beam line, thus, facilitating a higher degree of predictability and easier assessment of results on site. Please find more helpful sample preparation guidelines in this presentation on sample preparation.

We also provide the more traditional equilibrium SAXS setup without in-line chromatography for which please find instructions below.

Performing Equilibrium SAXS Experiments

Loading Samples

SAXS sample loading system
Sample loading system with a Hamilton syringe pump.

The protein solution and buffer are contained in eppendorf tubes 1.5 ml or 0.5 ml before being loaded into the brass flow-cell. The sample is then flowed back and forth during SAXS/WAXS measurements with a Hamilton programmable syringe pump. We provide centrifuges for both eppendorf tube sizes. You will need to bring your own filters if you want to filter your samples. The protein solution is measured immediately following the measurement of its matched buffer. Before switching to the next buffer and protein solution, the flow cell must be flushed with the sequence: water, 20% bleach, water, 100% ethanol, water, acetone, and water to remove any possible protein deposits left on the flow-cell wall by interactions with the high-flux x-ray beam.

Data collection for equilibrium SAXS

Flow control of SAXS sample
The flow control interface for equilibrium SAXS.
The User Interface shown below controls the sample loading before the X-ray exposure and flow through the quartz capillary during the X-ray exposure. It also automates the washing process between measurements. When 'Wash cell' is clicked, the capillary flow cell is flushed subsequently by water, bleach, water, isopropyl alcohol, water and then is air dried. 'Load cell' will pick up the fluid in the Eppendorf tube and pull it into the capillary and 'Pull' and 'Push' can be used to fine tune the position of the sample segment inside the capillary and the tubing. In order to reduce the potential effects of radiation damage, the sample will flow through the X-ray beam during the X-ray irradiation. Immediately before starting the detector image collection, one needs to click 'Flow-thru' to initiate the flow. The 'Speed' and 'Step' can be set according to the sample volume and image collection protocol. After the measurements if the user would like to recover the sample and save for further analysis, it can be done by click 'Return sample'. Otherwise, one can click 'Wash cell' and it will be flushed to waste.

Incident beam intensity
Intensity mesurement of the incident and transmitted X-ray beam.

The X-ray intensity of the incident beam is measured with a National Instruments analog I/O card and a LabView program is developed to acquire the incident and transmitted beam intensity during the X-ray exposure time. The X-ray scattering is collected with a Pilatus 1M detector mounted at the end of the vacuum flight tube of a fixed camera length. The exposure time is set using the following EPICS interface which communicates with the Dectris camserver. The camserver controls and communicates directly with the detector hardware.

Xray data collection with Pilatus 1M
X-ray data colleciton with Pilatus 1M.

Data Reduction

The raw data for solution SAXS consists of the detector images and the corresponding incident and transimitted intensity values which are used to normalize the SAXS curves. The 2-D images are converted to SAXS curves after radial averaging and q-calibration. Automatic data reduction pipeline is available at the beamline computer. SAXS curves are available to the user for further processing in real time (only a few minutes after the images are saved).

Online data analysis for quality control

The Linux version of ATSAS is installed on the beamline computer 'dracula' and the program 'SAS Data Analysis' (primusqt) can be used to plot, average, and subtract SAXS curves immediately after it is reduced. Guinier analysis, p(r) calculation and ab initio modeling can also be done using the ATSAS software in order to check on the data quality.

Data Analysis and Modeling

For detailed data analysis and modeling we rely on the ATSAS software package developed at EMBL Hamubrg and the programs in CCP-SAS. Please refer to the respected website for more information. Our beamline scientists are experinced BioSAXS expert and if you would like to collaborate with us on data analysis and modeling, please contact them.