Safety information for you and for equipment in the 800 lab

  1. The inner lab has stray fields in excess of 5 gauss. Some individuals should never enter this area, including cardiac pacemaker wearers. Handle any ferromagnetic objects with great care in this area. The bigger the object, the farther away from the magnet you need to worry about it.
  2. A quench of the 800 would be a serious event. If one occurs, evacuate the lab area immediately for an area with fresh air. If the Oxigraf oxygen sensor is showing or sounding an alarm, do not enter the inner lab until you are sure it is safe to do so.
  3. The lab area is normally locked and Omnilock codes are issued only to people who need them to work on the instrument. If you admit someone to the lab, you assume responsibility for their safety and the consequences of what they do.
  4. If you see signs of trouble from the BMPC (such as the flashing yellow light comes on, notify me immediately.
  5. Probe damage will result if the sample is inserted too deep in the probe. The sample depth guage is superglued at a safe level of 19 mm. The standard shim sets are also made for this sample position. Do not exceed a safe depth or use a sample tube that is not gripped tightly by the spinner turbine.
  6. Active temperature control should be on all the time (with edte) with an airflow of 670 liters/hour.
  7. Make sure there is normal airflow before letting go of a sample in the magnet. You hear and feel the air at the top of the magnet, or check the pressure of the house air behind the console (normally 80-85 psi).
  8. If you use the BCU-05 precooling bath, note that excessive humidity in the compressed air can cause this bath to ice up. Interruption in the air flow will lead to an emergency warmup of the cryoprobe. There is a visual humidity indicator on the compressed air line in room 104. If the air is too humid, we will have Facilities check the compressed air dryer.
  9. Setting a sample temperature below 14 C or above 55 C will lead to an emergency cryoprobe warmup.
  10. The sample would not ordinarily be spinning for nD and water suppression experiments anyway, but do not use sample spinning with the cryoprobe.
  11. The “power check” feature is enabled in TopSpin to cap the transmitter power at a safe level. Do not attempt to defeat this mechanism. The great majority of the standard experiments will work safely with default parameters and the values loaded by getprosol.
  12. Some conditions are not explicitly checked by power check. These include decoupling duty cycle, decoupling of carbon and nitrogen simultaneously, and the power level and pulse spacing of nitrogen CPMG pulse chains. In all parameter sets, d1 should be 1 second or more and aq with X-nucleus decoupling should be less than 100 msec. For the handful of experiments with double decoupling (typically they have “sm” in the name”) carbon and nitrogen decoupling powers should both be reduced by 3 dB and the decoupling pulse width should be lengthened by 40%. The document “Cp800.pdf” contains the complete list of power restrictions on the cryoprobe.
  13. You can create or modify parameter sets and pulse programs any way you want as long as they are safe for the hardware. However, do not modify (i.e. change something and then wpar) the standard Bruker parameter sets. Save your modifications under a different name by appending your initials to the name of the file. As of TopSpin 2.0, pulse programs are in a database file and you must save them under a new name in order to edit them in any case.

Bruker 800 Sample Setup for Biomolecular Experiments with TopSpin 2.0

  1. Double click the TopSpin icon on the Linux desktop to start the software.
  2. Set the sample depth (19 mm for a full length sample, centered around 0 for a shorter sample) in the spinner turbine. Push the LIFT ON-OFF button to start the air and insert the sample. Push the LIFT ON-OFF button again to turn the air off.
  3. Type edte to adjust the sample temperature set point. Normally the air bath and the probe heater are on at all times for active temperature regulation.
  4. Type wobb to start probe tuning. A 2 MHz window is sufficient to see all of the tuning dip that matters. The tuning is done underneath the magnet. The tuning adjustments are color coded with yellow being proton, blue being carbon, and red being nitrogen. Each adjustment is labelled either T for tune or M for match. All active channels are tuned starting from the lowest frequency (nitrogen if active) to the highest (proton). From under the magnet, push the f1 button on the top of the preamp to activate the display; the Chn up and Chn down buttons can be used to change the channel. Normally the nitrogen and carbon changes little if any between samples; the proton will change significantly. The behavior of the tune adjustment is always the same and you can predict if you need to go clockwise or counterclockwise depending on whether you are off to the left or off to the right. You can predict the direction of the match adjustment IF you know how the ionic strength of your sample compares with the last one: clockwise for less salty/less polar samples, counterclockwise for more salty/more polar. If you start with tune and need to go clockwise with that, then the match will need to go counterclockwise, and vice versa. A good tune is reached when the center bar and no more than one other bar are lit on the horizontal row of LED's; a good match when there are 2-3 bars lit in the vertical row. When tuning is finished, click the stop icon and the return icon in the TopSpin window to stop the wobble.
  5. Click on the lock window icon to bring up a new lock window. Click on the lock sample icon in that window and click on the solvent (normally D2O or H2O/D2O). Lock should be found automatically; if not, try reading (with rsh) default shims for the probe from the file called CPTCI, or from a shim file you have saved yourself. Note that z1 is rising over time and that an old shim file may need a manual adjustment of z1 (watch the beat pattern in the lock window with the lock turned off) to be valid.
  6. Shimming: it may be helpful to adjust z1, z2, z3, x, y, xz, and yz manually to maximize the lock level before performing gradient shimming. Type topshim gui. I recommend doing 1D shimming routinely, deselecting the z6 option, and taking all other defaults. It may be beneficial to do 3D shimming on some samples. (To see the topshim manual for a complete description of options, type topshim? on the command line.) After the shimming is completed, it may be helpful to click the ONAXIS button on the BSMS panel, select z1, and then click AUTO SHIM. This will activate background z1 autoshimming during the experiment.
  7. Proton pulse calibration: click the new document icon or type new to create a new data set. Put in a sample name and load the parameters onepulsecal, and type zg. Process the spectrum with ef and mc. Measure the splitting in Hertz between the tallest points of the doublet. Invert this number, and divide it by 80 to get the proton pulse width. The pulse width will be quite dependent on the ionic strength and length of the sample; the higher the ionic strength, the longer the 90 will be. The highest allowable proton power in the cryoprobe is -1.5, which gives a 90 degree proton pulse of 8 microseconds on an organic sample, longer on salty samples. The carbon and nitrogen pulses can generally be left at the default values. If desired, you can check them by loading the parameter sets HSQCETGPSISP2.NULL for carbon or HSQCETFPF3GPSI.NULL for nitrogen. These load pulse sequences which are 1D versions of their 2D counterparts and the X pulse is automatically doubled. When the carbon pulse p3 reached 90 degrees, the signal will be nulled out; the same goes for p21 in the nitrogen experiment. If you wish, you can use the command “popt” to automate this procedure.
  8. Probe file update for proton 90: with the experimentally determined pulse width and power, you can update the probe file without actually editing it by typing

recalcProsol all 9.5 -1.5 1H.F1.A2

in the case where the experimental 90 is 9.5 at power level -1.5. This should be done on every new sample. If you wish to check/update pulses for carbon and nitrogen, use the syntax

recalcProsol all 16.5 -1 13C.F2.A1 for carbon (assuming 16.5 us 90) or

recalcProsol all 39.5 -2.5 15N.F3.A3 for nitrogen (assuming 39.5 us 90)

For carbon-detected experiments, the channel assignments of proton and carbon are reversed, in which case the syntax is 1H.F2.A2 for proton and 13C.F1.A1 for carbon.

  1. Acquisition of water suppressed spectrum: create a new experiment number, and choose the parameters ZGPR (for presat), P3919GP (for watergate with hard rf pulses), ZGPGWG (for watergate with shaped pulses), or ZGW5 (for W5 suppression). Type getprosol to update pulse widths and powers. Type rga to set the receiver gain. It may be helpful to type gs and go in to “go setup” mode to iterate various parameters that can affect the quality of the suppression. The goal is to get the smallest signal intensity of the received signal indicating maximum removal of water.The parameters to adjust might include o1, the carrier position; in presat, pl9, the power level used for the presaturation; in hard pulse watergate, the length of the last pulse, pl27; in soft pulse watergate, the power level of the soft pulse (the length is normally fixed). (For more complex sequences incorporating a flipback pulse, it is a good idea to try fine tuning the flipback pulse power, sp1.) When finished with gs, save any changes if desired, then click stop and return. Set the number of scans for the final spectrum (normally 4 dummy scans, DS, are also taken) and type zg to acquire the data.
  2. With shims and pulse widths/powers set, if the water suppression is good, any desired experiments can be set up by typing new, choosing parameters, and then typing getprosol. Update o1p as desired for optimal water suppression, or type gs for iterative adjustment of any desired parameters. Change sweep widths and point counts as desired in any dimensions; check the experiment time by typing expt or clicking the clock icon. Type rga to set the receiver gain. Type zg to take data in a single experiment. If you want to queue multiple experiments, set them up as sequentially numbered data sets. Return to the lowest numbered data set and type multizg; enter the number of experiments to be queued.
  3. There are numerous variations on the basic hsqc experiments. The ones I recommend for labeled proteins are HSQCETFPF3GPSI for nitrogen and HSQCCTETGPSISP for carbon.
  4. Like the hsqc experiments, many of the 3D sequences will work with as few as 2 scans per block. This puts the experiment time at about 4 hours. The more sensitive 3D experiments will give useful results that way. Many parameter sets are supplied with 16 scans per block as the default which puts the experiment time at 32 hours, whether the experiment is sensitive or not. In many cases, you do not need this much time to get good results. We have some projection reconstruction parameter sets available that can give useful results in as little as 1 hour in favorable cases.

Eleven “NMR Common Sense” recommendations for optimizing data quality on the 800

  1. A properly filled Shigemi tube will give you better water suppression than a regular tube.
  2. Make sure the lock phase is adjusted for maximum lock level. The lock phase will change if the probe temperature is changed. An incorrect lock phase will lead to errors with hand shimming and will also create lineshape problems in the transformed spectrum that cannot be removed through later phase adjustment (“lock pull”).
  3. Small missettings of probe tuning and matching have no negative effects. Be more concerned with the tune than the match. Smashing the tuning and matching capacitors through overuse leads to very expensive probe repairs.
  4. Watch what the gradient shimming does and what shim values it leads to. Normal values for z1 are 37,000-38,000. All other z and low order shims have much smaller values on normal samples, typically with an absolute value of less than 5000. If you get different values, either your sample geometry is not optimal, or you are not getting a good shim. In either case, the lineshape and water suppression will be worse. If you are working in a deuterated solvent and use 2H gradient shimming, watch the profiles extra carefully to make sure they are all valid. Once in a while, you will get a bad profile that can throw the shims way out of normal values, and the program may not find its way back in the remaining iterations.
  5. Turn on z1 autoshim during any long experiments. Z1 drifts slowly upward a few units a day and this change should be tracked and compensated for.
  6. Optimize the power of any shaped proton pulses used for flipback and watergate using the “gs” command. These pulses are 1 msec sinc pulses and their power is predicted from the value of the hard 90 degree pulse. This prediction tends to be inaccurate because of radiation damping. A typical power value is 33-35 dB. In many sequences, the optimal value will be very nearly 1 dB higher attenuation (=1 dB lower power).
  7. If there is a watergate sequence available in the standard Bruker pulse programs, it probably means that you will benefit by using it. For those sequences without a watergate option, you should be able to get good water suppression without it.
  8. Compensate for probe heating effects during heteronuclear decoupling by adjusting z1 during the dummy scan period (may typically go down 100 units).
  9. Pay attention to the receiver gain value that is returned by rga. The desired signal is typically small and rga reflects how much residual signal there is from water, buffer, etc. With a properly prepared sample, good shimming, and optimized water suppression, rga should return a gain of 1024 for an hsqc and 2048 or more for most 3D sequences. You may choose to proceed with an experiment if you can't get those values, but there will be a penalty in terms of signal to noise if the gain is low. Bruker says that rga is conservative, and a higher value can often be successfully used, but evaluate it first in a 1D and a 2D plane to make sure there is no baseline distortion.
  10. It is chancy to run a 3D without doing at least a 1D to verify there is appropriate signal. It is better to do 2D planes, at least an f1f3 plane. Some sequences that are constant time or semi constant time in t2 don't give great results in the f2f3 plane, but still come out well in the full 3D. However the final 3D won't look better than the first f1f3 plane. If the signals are reasonably narrow and uniform in intensity, but weak because of low protein concentration, you can compensate with more scans. But if you have variations in intensity and linewidth, the full 3D will also be that way. If you are sure you have the best conditions you can find, you may choose to do the experiment anyway; but if the experiment really isn't going to work, it is better to decide this before you invest the time (and money).
  11. Processed Bruker data files can be opened directly in Sparky. You do not have to reprocess the data in NMRPipe unless you want to do so. If you do use NMRPipe, you may save time if you write down the values of the phase corrections from TopSpin. The numbers will be the same as those in NMRPipe, except that the signs will be the opposite of what they are in TopSpin.
  12. Running this instrument is like driving a luxury car! Enjoy it!