Overview
Several techniques are available to successfully deliver your crystalline sample into the X-ray interaction point in a serial fashion.
The importance of successful sample delivery
In order to measure X-ray diffraction from crystals using the serial femtosecond crystallography method, we need to replenish the sample in the X-ray interaction point after each X-ray pulse. A stable sample delivery method is therefore critical for any LCLS experiment.
Next, to replenish the crystalline sample before the next X-ray pulse arrives, the sample delivery method should:
- Not cause any damage to the sample
- Be compatible with running in vacuum, helium, or air
- Minimize the background scattering from the carrier medium (for example, water molecules or the lipidic cubic phase)
- Operate reliably for hours
Sample delivery methods available at LCLS
A wide array of sample delivery methods is available for experiments at LCLS, each with its own advantages and disadvantages. Although an overview is given below, feel free to reach out to your point of contact for advice.
| Pros | Cons | Ideal For |
|---|---|---|
| Gas Dynamic Virtual Nozzle (GDVN) | • Reduced jet diameter: low background scattering • Compatible with a high X-ray repetition rate • Compatible with vacuum geometries | • High sample consumption (20-100 µL/min) • Smaller capillary which can clog more easily • Challenging to align in non-vacuum geometries |
| GDVN plus mixer | • Enables time-resolved mixing experiments | • Mixing/dilution time can be limited by geometry, typically milliseconds to seconds |
| Double Flow Focusing Nozzles (DFFN) | • Similar to the GDVN, the DFFN uses a second liquid for focusing the jet and reducing sample consumption • Enables time-resolved mixing experiments when combined with a mixer • Reduces sample consumption • Compatible with vacuum and non-vacuum geometries | • Typically, high total flow rate (sample and focusing liquid), above 100 µL/min • Smaller capillary which can clog more easily • Two fluids must be compatible (not cause precipitation) • Mixing time typically milliseconds to seconds |
| Rayleigh Jets | • Simple jetting system, easily operated and stable compared to other liquid jet systems • Compatible with vacuum or non-vacuum geometries • Larger capillary results in less clogging | • High sample consumption rates (>100 μL/min) • Jet thickness dozens of micrometers or more, which can cause higher background signal • Inverted driving pressure and liquid flow rates are directly related to the fourth power of the inner diameter of the tubing • High vacuum loads • Prohibitive fluid properties (not suitable for high viscosity fluids) |
| High Viscosity Extruders | • Suitable for viscous media, such as the lipidic cubic phase (for membrane protein crystals) • Low sample consumption • Compatible with vacuum or non-vacuum geometries | • Low speed incompatible with high X-ray repetition rate • High background due to thick jet • Clogging issues • High vacuum loads • Using full transmission of X-ray pulse can destabilize jet |
| Microfluidic Electrokinetic Sample Holder (MESH) | • Can accommodate larger crystals and mitigate clogging without increasing sample consumption | • Depending on capillary size, can get clogged • Liquid must be doped with cryoprotectant when used in vacuum • Not compatible in helium atmosphere |
| Concentratic Mesh (CoMesh) | • Can mix viscous media • Can accommodate larger crystals and mitigate clogging without increasing sample consumption | • Mixed fluids must be compatible (not cause precipitation) • Depending on capillary size, can get clogged • Liquid must be doped with cryoprotectant when used in vacuum • Not compatible in helium atmosphere |
| Droplet Injectors | • No wasted sample in between pulses if synchronized • Can accommodate larger crystals • Compatible with time-resolved experiments | • Extremely sensitive droplet generation • Complicated for in-vacuum operation • Large volume droplets • Deceptively high sample consumption rates despite being efficient (e.g., 80 μm droplets at 120 Hz consume ~2 μL/min) |
| Fixed Targets | • Minimal concerns with fluid/slurry properties (no clogging, high viscosity, etc.) • Compatible with some pump-probe time-resolved experiments • No wasted samples in between pulses if synchronized and sample locations are known • No clogging | • Difficult to mount sample onto substrate • Complicated in-vacuum operation • Can’t match faster pulse structures • Need to know sample locations on substrate • Substrate contributes to the background signal • Incompatible with other time-resolved experiments (mixing, complex/multiple pump probes) |
| Lens Stack | • Suitable for single particle imaging experiments • Can have high vacuum pressure loads | • Large amount of sample lost during transmission • Not used for crystalline samples • Sample may need buffer exchange into volatile media |
In addition, there are some methods not owned by SLAC that have been used at LCLS in the past and may be available through collaboration.
- Acoustic levitation
- Rowed runner
- Tape drive
Preparing for a serial crystallography experiment
We can advise LCLS users on which injection method is best suited for your project. In addition, we offer tools for extensive testing and characterization of the chosen sample delivery method before your beamtime takes place. For more information, please have a look at the laboratories we have available for our users.
