Analytical Centrifuge SOP

To download a PDF of this SOP, click here.

Deciding which type of run you want to do
There are two types of run you can do:  sedimentation velocity (SV) or sedimentation equilibrium (SE).
            SV:      uses long solution column and high speed
                        gives you:  s = sedimentation coefficient
                                          D = diffusion coefficient
                                          S/D calculates buoyant mass (Mb)
                                          can use for mixtures if species have discrete boundaries

            SE:      uses short solution column, lower speeds
                        gives you: solution molar mass
                                         association constants
                                         association stoichiometries
                                         nonideality coefficients
                                         shape of molecular complex (if you have s from an SV run)
                                         ligand binding information
                                         ligand-mediated conformational change information

There are two types of detection you can use:  absorbance or interference.
            absorbance:  need sample that absorbs in the UV and buffer that doesn't
                                 has greater selectivity and sensitivity
                                 linear below 1.5 OD concentration
                                 wavelength range is 190 – 800 nm
            interference:  measures difference in refractive index between sample & reference
                                  no need for chromophores
                                  greater linearity, range, and accuracy but no selectivity
                                  sensitivity same as A280 for proteins


Guidelines for preparing your experiment
For an SV run, you need enough protein to be seen with absorbance (0.2—1.2 A units for linear response).  For interference optics, the lowest concentration is 0.1 mg/ml.  For OD>3, use the SE centerpieces.  To decrease nonideality, keep protein concentration <1 mg/ml.

Use buffer of  >50 mM ionic strength to suppress the nonideality stemming from charge effects.  If running absorbance optics, use a non-absorbing buffer.

A must!  Dialyze your protein against its buffer overnight (with 3 changes) so that the chemical potential of the macromolecule is defined under conditions of constant chemical potential of solvent.  Save some of the dialysis buffer to put into the reference sectors.

Volume needed
For a velocity cell, you need 200-450 μl each of sample and buffer, depending on the resolution you want to obtain.  For equilibrium cells, you need 40-120 μl of sample and buffer for each pair of sectors.  Larger samples make for better resolution.

Choose a temperature between 4º and 20º where your protein will be stable for many hours.  For an absorbance run at 4°, chill both the fuge and the rotor overnight first.  Interference optics don't work at 4° (water viscosity problem) so try 8 or 10° for samples which need to keep cool. 

Rotor speed
For SV runs, 42K is good for small globular proteins.  For large proteins, 30K will do.  For SE runs, start slowly and get to equilibrium there, then increase two steps, waiting for equilibrium each time.  There is a table in the Beckman manual (p. 4-26) to help you select rotor speeds for equilibrium runs, based on your monomer’s MW.  To change rotor speed during a run, see the Changing rotor speed section of this protocol.

For details:
Current Protocols in Immunology, Unit 18.8:  Measuring Protein-Protein Interactions by Equilibrium Sedimentation, 2004. John Wiley & Sons, Inc.

Protein-Ligand Interactions:  hydrodynamics and calorimetry, ed. Stephen E. Harding and Babur Z. Chowdry, 2001. Oxford University Press.  (chapters 4 and 5)
Choice of window type
Use quartz windows for absorbance runs and sapphire windows for interference runs.  If you need to run both at once, use sapphire windows but be aware that they block light below 240 nm. Quartz windows weigh 3.8 g and sapphire windows weigh 5.7 g.  The sapphire ones have an "X" etched into their sides.

Pre-chilling the rotor

Before assembling cells, turn on power to the centrifuge with the black switch on the right-hand side of the instrument. Open the fuge door and lower in the rotor you'll be using. Insert the monochromator arm and tighten it down. Be sure it doesn't wobble and that the rotor can spin beneath it.

Start a "fake spin" to maximally chill the rotor. Set the rpm to zero, the time to Hold and the temperature you want. Then hit the Start button. (not the vacuum button)
Assembling cells (Inspect all parts for cleanliness and/or damage as you go.)
           Velocity (two-sector centerpieces) (see next page for diagram)

  1. Prepare two window assemblies. From the bottom up:
    1. window holder – with key notch at 12 noon position
    2. window gasket – thin clear one, use a new one each time
    3. window liner – bent into horseshoe shape with gap at 6:00 position, not frayed
    4. window – with notch mark at 12 noon (triangle on side pointing down), put in at an angle, press flat with lens paper
  2. Into a velocity cell housing (with collar at the bottom), put one window assembly (window side up), a two-sector centerpiece (has no sidedness), then another window assembly (window side down). These will all align correctly when the keyways slide down the rib in the housing. Be sure all components have slid down as far as they can. Push gently with the cell aligning tool if necessary.
  3. Apply a thin film of Spinkote onto a screw ring washer (brown) and into the threads of a screw ring. Place the washer into the housing on top of the window holder. Screw in the screw ring with the word "OUT" on the outside. Hand tighten, then put cell in the torque stand and tighten to 120 inch-pounds.
  4. Fill the cell by laying it on its side in a white drawer liner with the screw ring facing you and the filling holes on the top. With a gel loading tip, add 440 ul buffer to the left hand sector. Then load 440 ul sample into the right hand sector. Avoid bubbles when loading by sacrificing volume rather than adding them.
  5. Plug the filling holes with 1 or 2 red plug gaskets, using the gasket rod to tamp each one down evenly. Tighten the housing plugs to hand tight but don't use excessive force as that can distort the entire cell housing. If the housing plugs are sticking out too far, remove one of the gaskets.
  6. Weigh the cell. It should weigh within 0.5 g of the cell spinning opposite it. If the opposite cell is the counterbalance (CB) cell, the sample cell should weigh more than the CB cell (in case of leakage). The weight of the CB cell can be adjusted with the different screw weights provided.

Equilibrium (six-sector centerpiece)

  1. Prepare two window assemblies as in step 1 above.
  2. Take an equilibrium cell housing (open at both ends) and turn it so the part number on the outside is upside down. Put the six-sector centerpiece into the housing beveled side down, aligning the keyway with the housing's rib. Next put in one window assembly, window side down.

Follow step 3 above but tighten the screw ring to only 60 inch-pounds.





  1. Turn the housing right side up and orient it with the housing plug holes and part number to the left. Using a gel-loading tip, fill the bottom three sectors with 120 ul buffer and the top three sectors with 120 ul sample, placing the highest concentration sample in the leftmost sector.
  2. Put the second window assembly in, window side down. Follow step 3 above but this time tighten the screw ring to 120 inch-pounds. Now turn the cell over and tighten the other screw ring all the way to 120 inch-pounds. No housing plugs are needed.
  3. Weigh the cell as in step 6 above.

Loading the rotor
To stop the "fake spin", first press the red Stop button. Then press Vacuum to release the vacuum and wait for the hissing to stop. (about 45 sec)

The CB cell always spins in hole 4 of the An-60 Ti rotor and hole 8 of the An-50 Ti. The arrow on the top points away from the center of the rotor and the hatch marks at the bottom must align. After aligning the CB cell, gently tighten it in place with the small hex nut on top using the Allen wrench.

Both velocity and equilibrium cells are loaded with the part numbers right side up and the housing plugs facing toward the center of the rotor. For velocity cells, the hatch marks on the bottom must align with the hatch marks on the rotor. For equilibrium cells, align the key notch in the housing with the hatch mark on the rotor. A loose cell can be held in place with a piece of thread or a human hair if necessary.

If you don't have enough samples to fill the rotor, leaving rotor holes empty is fine.

Setting up rotor and monochromator in chamber
If the centrifuge power is not already on, turn it on with the large black toggle switch on the outer right hand wall of the fuge.

Lower the rotor carefully onto the spindle so as not to damage the overspeed disk. Give it a little spin by hand to confirm that it's seated properly. You'll hear a click.

Align the monochromator arm so that it plugs into its socket properly. (There is only one orientation for this.) Hand tighten the screw at the bottom then check that the monochromator is seated firmly by trying to make it wobble. If everything is set correctly, close the chamber door.

Set the speed to 0 and hit Start to start both the vacuum and diffusion pumps. Watch to see that the vacuum descends steadily. If it stalls out or rises 50-100 µ, you have a leaking cell. The fuge can't start the spin until the vacuum is below 50 µ.

Starting an absorbance run

  1. Open the Proteomelab software. Wait for the window with blue lightning bolts to close. Choose File, New File. The scan parameters window will appear. Choose whether you have the 4 hole or 8 hole rotor. Check the XL Settings box to see that the temperature is correct. Leave the speed at 3000 rpm for the radial calibration. Click the Velocity or Equilibrium button and then the Absorbance box for each cell. The Rmin, Rmax and Wavelength 1 boxes will become active. For SV runs, set R max to 7.15. For SE, set it to 7.2. If you want all cells to be identical, check the box All Settings Identical to Cell 1 but be sure to follow step 2 below.
  2. Enter a descriptive phrase for each sample in the Comment field so you can tell which cell holds which sample. Don't select the counterbalance cell or the fuge will take data for it.
  3. For additional scan parameters for a velocity run, click the Detail button for each cell. The absorbance portion of the Velocity Detail window gives you options for 2nd or 3rd wavelengths, to decide what radial step size you want (this governs how long your run will be), and whether you want replicate readings taken. Leave the Mode on Continuous and be sure the 2 channel centerpiece is selected. For Data directory name, leave the default. If you're doing an equilibrium run, the Equilibrium Detail window will open. You have the same choices as for a velocity run, but you want to choose Step mode and click on the 6 channel centerpiece. Do this for every cell individually.
  4. Now back on the scan parameters window, click Method. The Delayed Start option does not work. We time the standard hour of temperature equilibration time ourselves. Time between scans is in hr:min:sec. Scanning every hour is sufficient for equilibrium runs. For velocity runs, scan every 0-5 min. depending on the number of samples you have. You can ask for 999 scans and the instrument will keep going until you come and stop it. Bear in mind that some data analysis software can handle only 50 data points.
  5. Now back on the scan parameters window, click Options. Choose Radial calibration before first scan. We do a low-speed spin at 3000 rpm for this calibration. Check off whether you want the fuge to stop after the last scan and whether you want the last 3 scans overlayed on the screen display.
  6. Go to File, Save File As to name your method and save it.
  7. ┬áIf the vacuum is <50µ, start the run by clicking Start Method Scan. The Scan Monitor window (and the XL Monitor window for equilibrium runs) will open so you can see what's going on. The radial calibration takes about 8 minutes, then the scans will start. Check one scan for each cell to be certain there have been no leaks by looking at the position of the menisci.
  8. Now stop the scans by pulling down from Scans to Stop Scan. When the software says they've stopped, stop the spinning by going to XL and Stop XL. When the rotor has stopped, the fuge will beep six times. If significant sedimentation occurred during the calibration, turn off the vacuum, open the fuge, remove the monochromator arm and tilt the rotor to restore your samples to homogeneity.
  9. Now return the rotor and monochromator arm. Get the vacuum pulling down again with the speed set to 0 rpm. While that's pulling down, change your method to have the speed you actually want to spin at and be sure to remove the calibration step under Options. Now save your method again. (overwrite)
  10. When the vacuum is down and the temperature has returned to the set point, wait one hour for temperature equilibration before clicking Start Method Scan.
  11. If you want to look at your data but leave the fuge running, stop the scans and move your data onto a flash drive. Go to the DLS computer and work on it there. If you need more scans, start the scans again.
  12. To stop the fuge, see Stopping the centrifuge on page 9.
  13. Please always sign the logbook during your run so that we know who to contact should something go wrong. This also tells people which rotor was used last so they know what to expect in terms of the radial calibration.

Starting an interference run

  1. Open the Proteomelab software. Wait for the window with blue lightning bolts to close. Under XL Settings, enter your run temperature and a speed of 3000 rpm. (You'll do the laser setup at this low speed so there won't be any rotor stretch involved.) Under XL, click Start XL to begin the run.
  2. Go to the Interference button and pull down to Laser Setup. The fringe display at the top of the screen should have parallel black and white lines showing. If there are no fringes to the left of the meniscus, check the right-left orientation of the fringe window. If the lines are all grey, check that the monochromator is not loose and that the condenser lens cover is clean. See Debby if the problem persists.
  3. Now choose the counterbalance cell (either cell 4 or 8 depending on your rotor) and set the laser delay. You won't see fringes at both ends of the window so move to the R end of the fringe window. You can click on the Auto Adjust Laser Delay button and watch where the fringe lines are the most distinct. Remember the number where this happened and go back to that region to set the laser delay manually. To do this, move the Laser Delay box to the left of where the fringes are visible. Now start moving the box to the right, counting the mouse clicks until you've moved to the right of where the fringes are visible. Move left halfway back and set the laser delay there by clicking on OK. The Laser Duration should be left at 0.2.
  4. Now set the radial settings for the CB cell by going to Interference and pulling down to Radial Calibration. To be sure you're still in the correct cell, click on Laser Setup. Then click on OK to get back to the Radial Calibration window. Choose Inside, then move the cursor to the rightmost edge of the fringes on the left. Set that to be 5.85. (This may involve mirror alignment.) Hit Set Radius. Now choose Outside and put the cursor on the leftmost edge of the fringes on the right. Set that to be 7.15. Hit Set Radius. ( The left pixel column should be in the range of 135-148 for good alignment. If it isn't, see Debby for mirror alignment.)
  5. To set up your Method, go to File and New File. The scan parameters window opens. Choose whether you have the 4 hole or 8 hole rotor. Check the XL Settings box to see that the temperature is correct. Check the Velocity or Equilibrium button first, then the Interference box.
  6. Click on Detail for each cell to be set up. The Velocity Detail window will appear. Select the correct centerpiece. Click the Laser Setup button and set the laser delay as in step 3 above. Set radii for each cell using the Radii section of the Detail window. For a velocity centerpiece, set the left (inside) limit at the top of the sector and the right (outside) at the bottom of the sector. For an equilibrium centerpiece, set the left and right limits for each of the three sectors. (Click Inside radius, click on the left margin, then click the Set button. Click Outside radius, click on the right margin, then click the Set button. The order matters.) In this window, the Pixels per fringe value should be ~21.75—22 and set the same for each cell.
  7. Click on Method. The Delayed Start option does not work. Enter your desired Time Between Scans and Number of Scans. Scans take only 7 sec. so you should leave enough time between scans so that all 999 don't happen before your samples pellet. Go to File, Save File As, name your method and save it.
  8. Now stop the rotor (Stop XL under XL). The fuge will beep 6 times when spinning has stopped. If significant sedimentation occurred during calibration, release the vacuum, open the chamber and remove the monochromator. Pick up the rotor and tilt it to mix up your samples to homogeneity again. Replace rotor and monochromator. Close the chamber door and turn on vacuum pump. Set the speed to 0, hit Start to get the diffusion pump working. Once set temp. has been reached, leave the fuge to equilibrate for one hour.
  9. You might want to look over all the parameters you've chosen one time to be sure they're correct. It's likely that you'll need to change the rotor speed to something higher than 3000 rpm. Now save your method again. If the vacuum is <50µ, you can start your run by clicking Start Method Scan.
  10. To see if you've reached equilibrium at a given speed, stop the scans and put your data onto a flash drive and look at it with WinMatch. (See step 11 under Starting an absorbance run, page7.) To change speeds during a run, see the Changing rotor speeds section of this protocol.
  11. Please always sign the logbook during your run so that we know who to contact should something go wrong. This also tells people which rotor was used last so they know what to expect in terms of the radial calibration.

Changing rotor speeds during a run
To change rotor speeds during a run, go to Scan and choose Stop Scan. Collect 5-10 new scans if you've reached equilibrium. Then in the Method window, change the speed manually. Click on Start Method Scan again and the computer will make a new folder for the next data set.

Starting a wavelength scan

  1. To see where your protein is absorbing and whether there's enough of it, you can do a wavelength scan at 3K rpm but you'll need to use quartz windows to see below 240 nm. Either centerpiece type can be used. Open the XL-I software and go to File, New File. On the scan parameters window choose Wavelength and Absorbance. Set the minimum and maximum wavelengths you want for the scan. Rmin refers to the center of the sector to be read so for a velocity centerpiece just accept the default of 6.5. For an equilibrium centerpiece, it will read the center sector unless you specify Rmin = 5.95 for the inner sector, Rmin = 6.5 for the center sector and Rmin = 7.1 for the outer sector.
  2. Enter a descriptive phrase in the Comment field then click on the Detail button. In the Wavelength Detail window, be sure the correct centerpiece is selected. You should select 5-10 replicates to get a good spectrum. You can send the data to your own folder.
  3. Go back to the scan parameters window and click Options. Choose to have radial calibration done automatically and choose the number of scans you want to see overlayed on the screen. To begin the run, click Start Single Scan. To save the parameters to a file, go to File, Save As.
  4. When the computer displays your graph, you can change the tick labeling by clicking on an axis which will highlight both axes in red. Then double-click on the axis you want to change. A window will pop up to take your choices.

Stopping the centrifuge

  1. Go to Scan and choose Stop Scan. There may be a delay but wait until you're asked to confirm the stopping by clicking Yes. Check on the Scan Monitor window to see that the scan has stopped.
  2. Go to XL and choose Stop XL. The fuge will beep 6 times when it's finished spinning. Turn off the vacuum. You may get the screen message "Holding for DP cooling" before you hear the hiss of releasing vacuum. After the hissing stops (about 45 sec.), open the chamber door and remove the monochromator arm and rotor. If no one else will be using the fuge for a while, close the software, then turn fuge power off with the toggle switch on the right side wall of the machine.
  3. If the computer has crashed, you can still stop the fuge using its control panel. Just push the red STOP button.
  4. You can recover your samples or just clean the cells. (See cell cleaning protocol.) Recover samples from velocity cells through the filling holes before dismantling cells. Recover equilibrium samples by loosening the top screw ring (cell oriented with part number rightside up) and carefully removing the top window assembly.
  5. Please enter into the logbook the revolution counter number at the end of your run.

Recovering your data

  1. Move your data onto a flash drive only if the scans have stopped.
  2. Programs for processing the data can be found on the computer next to the DLS in a folder named AUC Programs. In it you will find shortcuts to six useful programs: SedFit, SEDNTERP, WinMatch, WinReed, Winnl106, and UltraScan. You're welcome to take these programs with you.

Cleaning cell parts  (also see AUC quick clean-up instructions)

  1. Loosen the screw ring with the torque wrench to take apart the cells. Remove the housing plugs and gaskets from the side of the cell for easier removal of the internal parts. Screw rings and their gaskets get rinsed with ethanol to remove Spinkote, then water and ethanol again. All other metal and plastic parts get rinsed with copious amounts of water then with ethanol before drying. Windows and centerpieces should be handled with plastic forceps and are sonicated in a detergent bath (0.5 ml of RBS 35 detergent in ~125 ml water in a plastic beaker) for 10 minutes. Be sure the edges of the windows don't contact other windows. This is how windows get scratched. After sonication, windows and centerpieces are rinsed with copious amounts of water. They are rinsed with ethanol before drying.
  2. All parts may be dried with the compressed air lines except the windows which must be dried by hand with lens paper. (Compressed air cans put a residue on the windows.) Wrap the windows individually in lens paper and be sure to put them in the correct drawer. (Weighing windows disambiguates this.) All other parts get returned to their proper drawers as well.

Modes for the AUC  (slit movement) (For interference only, Mode is not relevant.)

Step:  only mode available for wavelength scans (abs)
           for equilibrium scans (abs or both)
Continuous:  for velocity scans (abs or both)                                                                                    

To run absorbance and interference optics at the same time (SV)

  • Use sapphire windows. (but they don't transmit light below 240 nm)
  • Mode is Continuous for SV. (Use Step for SE and see special SE protocol.)
  1. Check off absorbance and interference for each cell.
  2. Ask for many scans and radial calibration before the first scan. Start method scan at 3K rpm.
  3. After the radial calibration has ended (~8 min), stop the scans.
  4. Now set the laser delay and interference radial calibration for the CB cell while the rotor still spins at 3K. Then set laser delay and radii for each cell using the Detail window.
  5. After all interference calibration is complete, change the method to remove "radial calibration before first scan".
  6. Also change the method to put in your higher speed.
  7. Save the method. (Save File As...)
  8. Stop the fuge. Shake and tilt the rotor to resuspend everything. Wait an hour after the temp. reaches your desired temperature again.
  9. Start Method Scan.

Updated: 5/22/2017

Note:  This protocol is a practical hands-on guide to using the instrument.
Reading this is not a substitute for reading the manufacturer's manual, something we strongly recommend that you do.