1H 1D-NOE Difference Spectra and Spin-Saturation Transfer Experiments on the GN500

Transcription

1 UGN526 VVM-21JUN88CD VVM-31OCT91UD 1H 1D-NOE Difference Spectra and Spin-Saturation Transfer Experiments on the GN500 Double-resonance experiments are techniques which use a second irradiating field (B 2 ). A wide variety of experiments are available, which differ largely in the magnitude of the applied power of the second irradiating field. Commonly used double-resonance techniques include proton-broadband-decoupled carbon spectra, homonuclear decoupled spectra, and presaturationsolvent-suppression. 1D-NOE and spin-saturation-transfer experiments are two more double resonance techniques; this handout describes the practical implementation of these two experiments using GEM software (see NOTE 1). THE NOE EXPERIMENT 1D-NOE difference spectra are generated by subtracting two spectra from one another. The first spectrum is a proton spectrum having a low-power decoupler irradiation at a resonance of interest, while the second spectrum ideally has that same irradiation located at some remote, signalfree region of the spectrum. To acquire these spectra: 1. The best signal to noise is achieved with the 1 H-only probe (S/N 400:1 using 0.1% ETB) which is scheduled by request (see MSL staff.) If your NOE is small, the 1 H/ 13 C probe (S/N 180:1 using 0.1% ETB) may not work well enough for you. 2. If possible, the probe should be tuned by MSL staff to optimize the NOE (this is essential when running VT). If you require many NOE experiments, you might want to learn to tune the probe yourself. 3. Run a macro to set default parameters (#1 or #2 for proton on the GN500), then set additional parameters as needed to get a good 1PULS spectrum (GN, LB, PE, SF, SW,...). Check the 90 pulse and T1 of the sample. 4. Acquire a 1H spectrum with a 60 flip angle and 5*T1=D5 + AT. 5. Set CB = 64K, N. To get a good subtraction, a minimum of 32K is required. 6. Adjust the spectral window to observe only the region of interest with the Z0 command. While still in Zoom mode, type W. This sets the spectral window to the desired region. Type ^F. -- Note the acquisition time (usually 2-5 sec.) Acquire the spectrum again and see if it corresponds to the zoom window set in 6 and save it. Make sure that foldover (aliasing) of peaks is not a problem. 7. Set the decoupler, first directly on the resonance to be irradiated and then on a signal-free region. It should be close to the peak of interest - within 2 ppm if possible. The best results are obtained when only two regions are irradiated in one experiment: the signal-free region and the region of interest. To set the decoupler: Type EF, ^R, use knob A to move the red cursor to the desired irradiation site, type D, {note the resulting number and later make sure it is the first number in the CD list}, then move the cursor to a signal-free region and type D again {note the resulting number} <return>. You can enter up to 32 different decoupler offset frequencies, where only one offset needs to be in a signal-free region, but realize that your GS number (used later) must equal the number of different decoupler offsets. If the chemical shift difference of the irradiated peaks is too great, you might need two or three reference spectra, one for each region of interest. If you are irradiating a multiplet, all the peaks in the multiplet should be set individually. See note 3.

2 8. Set: EX = PRESAT P2 = tip angle: usually a 60 tip angle (approx 5 µsec on the 1 H only probe) works well. A 90 or the default tip angle will also work. Use a composite 90 if selective population transfer (SPT) occurs (see the RESULTS section)) D5 = interpulse delay msec If acquisition time is short (i.e. 2 s) then increase D5 to 1-3 seconds (the total time between acquisitions should be about 10 sec., i.e., D5 + D10 + AT = 10 seconds This will depend upon the T1 value.) D10 = saturation time - According to Derome, this value should be three times the longest T1 in the molecule. Usually, a value of 2 to 3 seconds is adequate for most samples. D12 = 10 µsec (a switching time for the electronics) L1 = decoupler power - 30 db usually works well. To optimize this value, you must observe the decoupled spectrum. If the peak is not decoupled, increase L1 by 5dB until the region of interest is from 70-90% decoupled. This gives the best results. If the peak is negative or has an odd shape, reduce L1 by 5dB. CD = + (or CD "Decoupler offset list on?" - turn ON). The values should match those selected with EF, or you can type them in. Note that the PRESAT pulse sequence controls the decoupler settings. You do not need to set any DC parameters except those prompted for by the EX command. 9. Turn off the spinner - AF. This will result in fewer subtraction errors. The lock level may need to be adjusted at this point if the sample is not well shimmed. 10. Set WD=Ø& NA = 4* (use some multiple of four) GS = x*n (see below) DATASET A = filename.001 <return> COMMENT (enter comment) <return> AUTO REPLACE? Y <return> Good results are often obtained when NA = 16 and GS = 2*4 if at least 10 milligrams of sample are used. If less material is available, increase n, e.g. GS = 2*8. This starts the interleaved acquisition where (for GS = x*n): x = number of offsets in the CD list (the number of different datablocks), and; n = an integer which determines how many times NA scans are acquired. If NA = 1, then GS = 2*64 puts 64 scans in each of two files. If NA = 16, then GS = 2*4 puts 4 sets of sixteen scans (64 semi-interleaved scans) into 2 separate files. The DECOUPLER PULSE light on the Decoupler Status Board should pulse on/off during this experiment. If it doesn't, going into monitor, MO, typing GEM, taking a scan ZG, going back into MO, typing GEM16, EX = PRESAT, and reentering values will solve this problem. Reloading software might help, too. Contact MSL staff if necessary. Now generate the difference spectrum. The method for this is given below.

3 SPIN SATURATION TRANSFER EXPERIMENTS Set up the PRESAT experiment the same as above, except now both D5 and D10 need to be 5 times the T1 value. An L1 power level of 35 db seems to work fairly well. Note the temperature dependence of chemical exchange; you may want to run the experiment at higher temperatures to generate faster exchange rates. GENERATING THE DIFFERENCE SPECTRUM The difference spectrum can be generated in two ways: 1) To reduce subtraction errors, it is often best to subtract the FID's using the AS command: AS <return> DATA SET B = filename.øø2 <return> (calls the control spectrum) ADD FULL DATA SETS? Y DATA SET A = filename.øø1 <rtn> (calls the resonance-irradiated spectrum) <rtn> After the second <return>, the FID may then be saved with SA and processed as normal. e.g., LI = BCEM(ZF)FTPS <return> AU = 1 You must use PS to phase the difference spectrum. At this point, you have generated the difference spectrum and can proceed with analysis. However, you can't change individual add/subtract scales or offset parameters by this method. 2) To generate the difference spectrum, you can also first process each dataset (FID) to give a spectrum which is saved as a separate file. Each file should be processed with the same parameters (apodization, phasing, etc); this can easily be done with a link: GA = filename.øø1 <return> process (i.e. BC, EM, FT, PS) LI = GABCEMFTPSSB <return> INCREMENT NAMES? Y AU = # of total decoupler offsets you used. DATASET A = filename.øø1 <return> DATASET B = filename.øø1 <return> AUTO REPLACE? Y Now type: AS <return> DATASET B = filename.5øx <return> (calls the control spectrum) ADD FULL DATASETS? Y DATASET A = filename.5øy <return> (calls the first resonance-irradiated spectrum) The difference spectrum will be displayed in purple. Now knob A controls the offset, and knob B controls the scaling (the absolute value of k). The default equation is A - kb, but typing K = 1.0 <return> will generate the equation A + kb (K = -1.0 <return> will regenerate A - kb). When all parameters are set correctly, type <return>, and the difference spectrum will be retained in memory. You can now plot or save the new difference spectrum as any other spectrum, and proceed with analysis. By altering the scaling factor, you can get a rough approximation of the magnitude of the resulting NOE. This is the easiest method to determine the magnitude of the difference NOE.

4 Additional considerations: NOTE 1: If you want to use PRESAT with CHARM software, realize that CHARM's D6 is the irradiation time (=D10) and CHARM's D4 is the switching time (=D12). Note that decouplerpower levels are very different on the QE-300. Attempts to get NOE data on the QE-300 have met with failure. NOTE 2: Wider multiplets require either higher power or a frequency cycling method (see note 3). DO NOT, however, go above 55 db for NOE or spin-saturation-transfer studies. Too high a power level is indicated by: 1) Large glitches in the spectrum, 2) Decoupler power spilling over onto neighboring multiplets, which produces negative signals in the difference spectrum for near neighbors, or 3. Localized sample heating by the decoupler (hard to directly detect but undesirable in the spectrum). Note 3: If you have multiplets which are broad, especially if they are closely spaced, you may have trouble delivering enough irradiation power to the multiplet without it spilling over to neighboring signals. If so, it is best to cycle the irradiation frequencies among the lines of a multiplet (which is more efficient for a given power level). To use an "irradiation-cycling" scheme see MSL staff. RESULTS There are five results in the difference spectrum: 1. The irradiated resonance will give a very strong negative signal. 2. Resonances with NOEs typically will give small positive signals. 3. Resonances involved in spin-saturation-transfer will give small negative peaks. 4. Resonances which look like glitches or have dispersive-looking character are probably artifacts of the subtraction (caused by instrument instability). 5. Antiphase lines within a multiplet are due to selective population transfer (SPT). Realize that, with all of these results, incorrect interpretations may occur if the power levels were high enough to irradiate more than just the intended multiplet! Misc. Comments - The NOE is represented by h; this value is the enhancement, and is added to the intensity of the non-enhanced signal which equal 1.0). For proton homonuclear cases, the maximum possible NOE is between -1.0 and +0.5 (-100% to +50%), but typical measured values are several percent. - "Observation of an NOE between two protons does not, on its own, provide sufficient evidence that they are close"; from A.E.Derome's "Modern NMR Techniques for Chemistry Research", 1987, pp109ff. - To quantitate NOE's (a procedure fraught with possible errors), use the difference spectrum and assign the saturated resonance to -100% (assuming it is from a one-proton multiplet). A direct integration will yield the %NOE from the positive peaks. - the TT and TR commands do not correct for the GS=x*n format, and will give incorrect experiment times. - For optimal (quantitative) measurements, samples should be degassed by freeze, pump, thaw method and stored under argon or dry N 2. All paramagnetic material removed to ensure the presence of only dipolar relaxation. - To improve spectral subtraction:

5 1. Use the interleaved acquisition technique described above; 2. Increase the number of scans (better averaging); 3. Add more line broadening (LB and EM); up to 1-2 Hz for proton. 4. Use enough datapoints to fully describe your resonances. 64K is best. 5. Set lock to "lock slow" and properly phase the lock signal; 6. If highest resolution isn't required, don't spin the sample; 7. Use a low-viscosity solvent having a narrow lock resonance (i.e. acetone); 8. Stabilize the temperature (esp. for aqueous solutions). Realize that higher decoupler powers, especially over 50 db, can cause localized sample heating. 9. Use a frequency cycling method to lessen selective population transfer (SPT) and use lower power levels; 10. Use composite 90 pulses to eliminate SPT. REFERENCES Andrew E. Derome, "Modern NMR Techniques for Chemistry Research", Pergamon Press, 1987, Chapter 5 (The Nuclear Overhauser Effect). J.H.Noggle and R.E.Schirmer, "The Nuclear Overhauser Effect - Chemical Applications", Academic Press, 1972.