Back to a list of spectrometers

Complex NMR experiments: 2D, selective, etc.

Contents

• Introduction to 2D NMR experiment: gradient selected COSY and DQF COSY

  ---

• 2D NOESY

  ---

• 2D 1H-13C HSQC

  ---

• 2D 1H-13C HMBC

  ---

• DOSY

  ---

• Organic Structure Elucidation (separate document)

Back to Top

Introduction to 2D NMR experiment: gradient selected COSY and DQF COSY

Basic 2D NMR experiment setup. Gradient selected and DQF COSY.

Prerequisite: You MUST have mastered performing 1D 1H NMR on manual Bruker instruments (NOT with ICON-NMR) before you may set up a 2D experiment.

The tutorial covers a basic gradient-selected COSY (more sensitive) and a double-quantum filtered COSY (less sensitive but it removes all singlets and gives a cleaner 2D plane).

https://youtu.be/djtE96oh6Ak

Back to Top

2D NOESY

Prerequisite: You MUST practice the COSY before you can set up NOESY.

---

 

Setup of the experiment

To set up two-dimensional NOESY, please, read and follow step-by-step guidelines from the Topspin Guide Book: Advanced NMR experiments:   2D_NOESY.pdf. This book is available in your Topspin installation in Manuals section.

---

 

Optimization of parameters

• Mixing time, D8
  (optimally, should be around average T1 of a molecule):
  
  
  • 2-3 sec for Mw < 250 Da
    
    
  • 1 sec for 250 < Mw < 400 Da
    
    
  • 0.5 sec for Mw >400 Da

  NOTE: In practice, one will perform several NOESY experiments with different D8. For example, for a larger molecule (> 400 Da), I would set up experiments with D8 set to 0.1 sec, 0.25 sec, 0.5 sec, 0.75 sec, and 1 sec thus probing a range of mixing times around the optimal value of 0.5. Protons that are very close in space will develop NOE much sooner than more distant protons. As D8 increases, long-distance NOEs add to the picture while the cross-peaks between the closer-spaced protons remain too. The maximum mixing time is expected to show cross peaks for all protons in the molecule that are within 5 Angstroms. 
  
  KEEP IN MIND: the longer the D8, the lower the overall amount of signal (sensitivity) in the NOESY spectrume. You may need a larger NS to observe NOE cross peaks at long D8 values!

  ---

  
  
  

• Recycle delay, D1
  

Mw, Da	If Mw is unknown, use # protons **)	D1, seconds
< 275 Da	< 30 protons	2 sec
275-350 Da	30-40 protons	1.5 sec
> 350 Da	> 40 protons	1 sec

---





• Direct dimension points, TD F2
• 2048

  ---

  
  
• Increments in the indirect dimension, TD F1
  • 128 for a quick survey of a signal
  • 256 or 512 for higher resolution (if 128 appeared not adequate)

    ---

    
    
    
• Number of scans, NS
  • 4  - only for quick survey of the overall signal
  • 16  (required by the phase cycle)

    ---

Processing

Topspin ProcPars Tab

• SI F1 = 1024
• SI F2 = 2048
• SR F1 and SR F2 = 0 Hz

Phasing

Phase 2D to make diagonal peaks positive.

---

 

 

Interpretation

Below are quick notes on properties of NOE cross peaks. For comprehensive discussion, see
Burns and Reynolds, "MInimizing risk of deducing woring natural product structures from NMR data", Magn Reson Chem, 2021, 59: 500-533

and

High-Resolution NMR Techniques in Organic Chemistry, 3rd Edition, by Timothy D.W. Claridge. Elsevier Science (May 27, 2016), ISBN-10 ‏ : ‎ 0080999867, ISBN-13 ‏ : ‎ 978-0080999869

---

 

Small to medium sized molecules, approx. < 600 Da, in a low-field spectrometer (400-500 MHz)

• NOE cross peaks appear with an opposite sign relatively to diagonal peaks
  
  
• EXSY, exchange peaks appear with the same sign as diagonal peaks.
  • Exchange with water is observed as exchange cross-peaks at water frequency
  • A problem: exchange peaks may overlap with NOE cross-peaks cancelling each other.
    
    
• COSY artifacts: split (dispersive) cross-peaks with positive and negative components. 
  • Helpful trick: Using maximum mixing time maximizes NOE but not the COSY, therefore the COSY artifacts may be identified. 
    
    
• Pulse program artifacts: peak patterns looking as a mirror image of a diagonal, or pairs of peaks running parallel to the diagonal. They are unrelated to the NOE effect, - to be ignored.
  

  ---

Larger molecules, > 600 Da, in a low-field spectrometer (400-500 MHz)

• NOE effect vanishes beyond 600 Da at low field magnets and changes sign to positive as molecular weight increases further (1-2 kDa)
  
  
• NOE cross peaks in a higher molecular weight region will appear with  the same sign as diagonal peaks
  
  
• EXSY, exchange peaks also appear with a positive sign: you cannot distinguish NOE cross-peak from exchange cross-peaks in this molecular weight range!
  
  
• COSY artifacts: split (dispersive) cross-peaks with positive and negative components. 
  • Helpful trick: Using maximum mixing time maximizes NOE but not the COSY, therefore the COSY artifacts may be identified. 
    
    
• Pulse program artifacts: peaks looking as a mirror image of a diagonal, or pairs of peaks running along the diagonal. They are unrelated to the NOE effect, - to be ignored.
  

  ---

  
  
  

NOTE: The 600 Da boundary for NOE is approximate. The change of sign of NOE is controlled not by molecular weight or shape but by the rotational diffusion coefficient of the molecule in the current solvent as it compares to the spectrometer field strength. Greater sovent viscosity, lower temperature, and stronger magentic field all shift this boundary to smaller molecular weights. Example: 650 Da molecule exhibits positive NOE in DMSO at 800 MHz that is the "zero NOE" boundary shifted to lower Mw at 800 MHz in DMSO. In practice, it is advisable to record both NOESY and ROESY and compare results for your sample/solvent/temperature/magnet combination.

---

 

Back to Top

2D 1H-13C HSQC

Prerequisite: You MUST have practiced the COSY before you can set up HSQC.

Important notes

HSQC experiment is NOT very sensitive! You should aim for at least 50 mM sample concentration to be able to record HSQC spectrum in a reasonable time (30-60 min).

There are two kinds of HSQC you may perform: regular and multiplicity-edited:

• Regular experiment gives you best sensitivity but is not "broad band" in carbon. This means that you cannot simultaneously have a spectrum of all your aliphatics (around 10-40ppm) and aromatics (around 120 ppm). You will have to place O1 in one region and record then shift O1 and record the second experiment.

  ---

• The multiplicity edited experiment is used in all assignment work. The resulting 2D plan contains CH and CH3 peaks in the same sense (usually, phased to positive) and CH2 peaks appearing negative. This experiment is lower sensitivity than regular HSQC but is broad-banded, that is will give you aliphatics and aromatics in the same spectrum

  ---

To set up HSQC on a Bruker instrument

1. Use a parameter set ND_HSQC for regular experiment or ND_HSQC_MULT for the multiplicity-edited one

   ---

2. Probe must be tuned to proton AND carbon: issue atma while in the experiment that you just made. Wait for the tuning routine to finish (you may use the trick with the 'wobb' described in the basic training).

   ---

3. Parameters to adjust:
   
   
   1. Proton parameters: O2 and SW2 - same as in COSY
      
      
   2. Carbon parameters: O1 and SW1 (center frequency and spectral width in indirect dimension).
      
      
   3. NS mus be an integral number of phase cycles (check PulseProg tab for phase cycle used in the program). If sensitivity in your experiment is not enough you will increase NS in the phase cycle steps.
      
      
   4. Use TD F1 = 128 for your initial experiment (gives 64 increments in indirect dimension). If you will need more resolution you will reacquire later.

      ---

4. Parameters to NOT touch: TD F2, AQ, and D1. 
   
   This experiment uses heteronuclear decoupling, which may damage the probe. These parameters are set to safe values by default.

   ---

5. Issue 'pulsecal' to calibrate a proton pulse

   ---

6. Issue 'rga' to adjust Gain

   ---

7. Issue 'expt' to see experimental time

   ---

   
   

Back to Top

2D 1H-13C HMBC

Prerequisite: You should have acquired 1D carbon and 2D HSQC on your sample prior to setting up the HMBC.

To set up HMBC on a Bruker instrument

1. Use a parameter set CMCse_HMBC

   ---

2. Probe must be tuned to proton AND carbon: issue atma while in the experiment that you just made. Wait for the tuning routine to finish (you may use the trick with the 'wobb' described in the basic training).

   ---

3. Parameters to adjust:
   
   
   1. Proton parameters: O2 and SW2 - same as in COSY or HSQC

      ---

      
      
   2. Carbon parameters: O1 and SW1 - same as in 1D carbon.
      
      IMPORTANT: SW1 must be larger than that of HSQC because HSQC only detects protonated carbons while HMBC is capable of detecting all carbons.

      ---

      
      
   3. NS must be an integral number of phase cycles (check PulseProg tab for phase cycle used in the program). If sensitivity in your experiment is not enough you will increase NS in the phase cycle steps.
      
      IMPORTANT: HMBC is less sensitive than HSQC. You should double the number of scans relatively to HSQC on the same sample.

      ---

      
      
   4. Use TD F1 = 128 for your initial experiment (gives 64 increments in indirect dimension). This will be coarse but reasonably fast. If you will need more resolution you will reacquire later.

      ---

      
      
   5. Set AQ according to your molecular weight:
      
      
      • AQ=0.3 sec for Mw <= 400 Da
        
        
      • AQ=0.4 sec for Mw > 400 Da

        ---

        
        
   6. D1 must be set considering the total recycle delay: Recycle delay = D1 + AQ
      

Mw, Da	If Mw is unknown, use # protons **)	Recycle delay, seconds
< 275 Da	< 30 protons	2 sec
275-350 Da	30-40 protons	1.5 sec
> 350 Da	> 40 protons	1 sec

For example, for the compound of 500 Da, I will use AQ=0.4 sec. The recycle delay must be 1 sec, therefore, D1 = Recycle delay - AQ = 1 - 0.4 = 0.6 sec.

---




Multiple-bond J(CH) setting CNST13



• common: CNST13 = 8 Hz - will be sensitive for J couplings of 3-4 Hz but will miss smaller couplings;
  
  
• optional:  CNST13 = 4 Hz - will detect more peaks for small couplings but may miss stronger ones.
  NOTE: Remember that the experiment becomes less sensitive with this setting: you should double the NS.
  
  

Issue getprosol to set probe parameters

---

Issue pulsecal to calibrate a proton pulse

---

Issue rga to adjust Gain

---

Issue expt to see experimental time

---

Processing HMBC in Topspin

• Set in ProcPars:
  
  
• SI F1 = 2048
  
  
• SI F2 = TD F2

  ---

  
  
• Enter on a command line:
  
  
• proc2d y
  
  
• xf2m

  ---

Processing HMBC in MNova

• Use default settings

 

 

Back to Top

DOSY

• DOSY on Bruker

  ---

• DOSY on Varian 600

  ---

DOSY on Bruker

SAMPLE CONCENTRATION REQUIREMENTS

DOSY is low-sensitivity experiment because of long waiting times in the pulse program to allow for diffusion to occur. For 1H detection, the approximate concentration requirements are as follows:

• Bruker 800 MHz with TXI probe: about 1-2 mM with 1 hour per experiment and 10-15 hours for the complete diffusion series. If you can make a 10 mM sample, the full diffusion series will take a couple of hours.

  ---

• Bruker 500 with BBFO is three times less sensitive, therefore, for the same acquisition times you need 3-6 mM (15 hour diffusion series) or 30-60 mM (to be done within a couple of hours)

  ---

• Bruker 400 is further less sensitive, so will only work well with samples in 30+ mM range.

  ---

• If you want to do 31P or 19F diffusion, the sensitivity is further lower, therefore, acquisition times will need to increase quadratically.

  ---

BRUKER DOCS

Bruker guidelines on DOSY expts (PDF)

---

REFERENCE DATA

D of residual H2O in pure D2O at 298K is 1.902e-9 m2/s (in Claridge, p317 in 2nd and p397 in 3rd) => log D = -8.72

---

  DOSY measurement

1. Have VT off for enough time - 15 min
   
   
2. Create a proton 1D
   
   
3. Verify that D1+AQ >= 3 T1
   
   
4. Run a proton 1D
   
   
5. Turn off spinning: 'ro off'
   
   
6. Create DOSY experiment:
   
   
   1. For Proton DOSY: 
      1. Create new experiment with "DOSY" parameter set
      2. run pulsecal
         
         
   2. For other nucleus (31P, 13C, 113Cd, etc.): Open the Setup_DOSY template for this nucleus. Create new experiment with Start:Create Dataset:Use current parameters. If you want to use a nucleus that have never been used for DOSY, we have to create an experiment for this nucleus before you can proceed.
      
      

   
   

7. Verify that
   1. Verify that D1+AQ >= 3 T1
      
   2. p30 < 3ms
   3. p30/(D1+AQ) < 0.05
      
      
8. Run test of first and last spectral intensities: issue xau dosy 2 95 2 l y y 
   NOTE: symbol "l" is a lowercase "l" as in "lower".

   ---

   
   
   
9. OPTIMIZATION
   Extract first and second fids and overlay - they must have intensities 10:1 or 20:1
   Considerations (from Dosy an Diffusion by NMR, p.9): The smallest signal to be detected (i.e. at highest gradient strength) has to be above the noise. If the signal intensity is already totally gone, reduce the gradient strength (gpz6). If the signal is still to big, you have to increase either the diffusion time ∆ (d20) or the gradient length δ (p30 - no more than to 3 ms!!!). Increasing δ is favorable, because it results in a bigger effect. δ2 is determining the signal attenuation, while ∆ is only affecting the exponential decay function linearly (see chapter 1). If you change ∆, you have to take the relaxation into account (T1 relaxation for all STE type sequences).
   
   To perform extraction follow these steps:
   
   1. Extract first trace (the smallest gradient)
      1. issue 'rser 1 X01', where X stands for the EXPNO number of your DOSY experiment, like if DOSY is #3, then use 301
      2. Process: ef; apk;

         ---

   2. Extract 2nd trace (the strongest gradient)
      1. Display DOSY experiment again
         
      2. Issue 'rser 2 X02'
      3. ef; apk
      4. If you see no signal, go to AcqPars/ square wave and reduce D20

         ---

   3. Enter Multiple Display mode and overlay the first trace (X01)
   4. Assess relative intensities: should be 10:1 or 20:1 for THE PEAKS OF INTEREST
   5. Adjust D20 further if needed

      ---

      
      
10. After you made adjustments, check if
    1. p30 < 3ms
    2. p30/(d1+aq) < 0.05
       
       
       
11. Run full experiment:
    
    1. set NS to phase cycle, or 1/2 of it. DS to 8
    2. issue xau dosy 2 95 N l y y where N is number of gradient steps (10-15)

---

DOSY Processing

You should always perform initial processing in Topspin to verify that experiments worked! Topspin gives you default DOSY plot. MNova is good at peak-by-peak fits and reconstructing a simulated spectrum.

• Processing in Topspin (DOSY plot)
  
  
• Processing in MNova (peak-by-peak fitting)
  

 

 

Back to DOSY

---

 

DOSY on Varian 600

1H Varian protocol (PDF)

31P Varian protocol (PDF)

 

Back to DOSY

---

Back to Top