The DEPT experiment (Distortion-less Enhancement by Polarization Transfer) is a very powerful method for determining the multiplicity of coupled spins. It is most often used in ¹³C NMR. DEPT can be used for other nuclei as well. It has been successfully applied to the study of Rhodium, Silicon and Nitrogen. The experiment works on all of the facility's spectrometers.

The principal reason for DEPT to be such a popular experiment is that it relies on transferring polarization from protons to the X-nucleus . The enhancement equals the ratio of [[gamma]]. For ¹³C, this leads to an enhancement of about four. Since the proton magnetization is being transferred, the relaxation time of carbon is irrelevant. The repetition rate of the experiment is determined by proton T1 only. In most (but not all) compounds the proton relaxation is usually considerably faster than the carbon relaxation, therefore a faster repetition rate can be used. NOE is not observed, a great advantage for nuclei with negative NOE such as ¹⁵N, ²⁹Si etc.

Assuming equal phase correction for all spectra, the peaks will be positive, null or negative as a function of [[Theta]] and multiplicity:

From this it follows that with only two experiments one can assign all multiplicities. All non-protonated carbons will be suppressed. When a compound is measured for the first time it might well be reasonable to use a 135° DEPT to obtain some information as quickly as possible followed by a 90° experiment and a normal Broad Band Decoupled spectrum should quaternary carbons be needed. The main problem is to find the proper phase correction. The experiments normally used on our spectrometers have been modified such that the same phase correction is used for DEPT and a Broad Band Decoupled spectrum. If necessary, the ASTM standard sample can be used to obtain the phase parameters. Either before or after running a DEPT, take a couple of scans with the ASTM sample with or without decoupling. transform and phase. The DEPT spectra can then be phased to the same parameters.

In practice one should check the proton decoupler 90° pulse, especially if a long run will be done. It takes very little time and will verify proper instrument performance. The DEPT experiment itself can be used.

With a sample containing a CH₂ group, take one scan with a [[Theta]] -pulse estimated to be about 45°. Phase the spectrum. Increase the [[Theta]] -pulse until a null is obtained, corresponding to a 90° decoupler pulse. If a sample with more than one resonance is used, the approximate phase parameters can be established at the same time. The ASTM standard sample placed near the instruments contain 39% Dioxane. This allows determination of the 90° pulse but exact 1^st order phase parameters cannot be established.

If the pulse width differs much from the expected value (>10%), the probe tuning might be off. Check the probe.

The proton 90° and 180° pulses should be corrected if they differ by more than about 10% from the pre-determined value (P1, P2 on the AM spectrometers, P3, P4 on the AMX spectrometers).

CARBONx.400(500) is the parameter file for carbon. It contains approximate values which can be used for DEPT. Unless your sample is in CDCl3, use 'SO' to set the solvent shift. If desired, reduce the observe window using 'OW', or, in EP, with CTRL/O command to needed width. Type 'II' to initialize the interface.
DEPT90.AU and DEPT135.AU are customized versions of the generic DEPT.AU and generate the 90° and 135° spectra.
Use, e.g., "AS DEPT90.AU" and check pulses and delay times. (decoupler levels S3 = 0H, S2 = 13H {AM-500}, or = 17H {AM-400}; PW=RD=0.) Then type 'AU' to start the experiment.

The carib05.300 and carqn05.400 parameter sets contain reasonable starting parameters for DEPT. The only change needed is the appropriate pulse program.
We have added pulse programs with the .cnt extension, e.g., dept135.cnt, which store the FID to disk every NS scans. NS must be a multiple of 32 for the sequence to work! To select the pulse program, type `pulprog dept135.cnt' (or whichever program you want to use).
To check or change parameters, type `ased' to invoke the parameter editor.

• Values are approximate. They are given for:
• AM-400, AMX-400: 5 mm QNP-probe.
• AMX-300: 5 mm Inverse BB probe.
• AM-500: 5 mm Inverse C/H probe.

 AM-400        AM-500*      AMX-400      AMX-300       Function                
 D2   3.5 ms   D2   3.5 ms   D2  3.5 ms    D2   3.5 ms  1/2J                    
 P6  11.8 us   P6  15.6 us   P2  11.4 us   P2  13.4 us  Carbon  180°         
 P5   5.9 us   P5   7.8 us   P1  5.7 us    P1   6.7 us  Carbon  90°          
 P1   11.0us   P1  11.8 us   P3  8.3 us    P3   8.0 us  Proton  90°          
 P2  22.0 us   P2  23.6 us   P4  16.6 us   P4  16.0 us  Proton  180°         
 D1            D1            D1            D1           Relaxation delay        
 P3            P3            P0            P0           [[Theta]]  pulse        

The relaxation delay D1 should be set to 1.3 x proton T1 for best S/N.

* 3/13/96: The AM-500 90° Decoupler pulse should really be closer to 8 ms.

The experiment has some limitations. If only a few scans are taken, quaternary carbons will still be seen. Using two to four dummy scans eliminates the quaternary signals.Since the deuterated solvents are also suppressed the solvent cannot be used for chemical shift reference! Use TMS or collect a few scans with a regular 1-pulse experiment to set the offset on the solvent resonance. On the AM-400/500, use the 'SR' value as determined in a Broad Band decoupled experiment; on the AMX-300/400, use the `offset' parameter.
The sequence is somewhat sensitive to the choice of 1/2 J. Assuming the decoupler pulse is properly set, errors in the 1/2 J delay result in incomplete nulling of CH₂ and CH₃ resonances for [[Theta]] = 90° . Reasonable values are between 3.5 ms to 3.9 ms (J ~150Hz to 130Hz).

In practice, it is often not possible to completely null CH₃ groups. Nulling requires extremely good homogeneity of the decoupler rf field. On some of the probes this is known to be less than ideal.
For wide sweep widths (in Hz), the first order phase correction may change noticeably for different [[Theta]] pulse width. Reduce the sweep to reasonable width (often there are no protonated carbons beyond 150 ppm).

The most difficult part is the phase correction. The easiest method is to acquire a regular decoupled carbon spectrum (even if only the solvent is visible) with a few scans. After the FT and default phase correction (i.e., `efp'), further correct the 0-order phase only. On the AMX spectrometers, use only the middle Mouse button; on the AM-spectrometers, use only the C-Knob for phasing. If you do this before starting the DEPT experiments, the subsequent phase corrections (`efp') will work!
For plotting a decoupled carbon spectrum with the DEPT 90 and 135 spectra, use the xau program "BB-DEPT.PLOT" on the X32 or Sun workstation.

If a compound has no quaternary carbons of interest, a DEPT with a 45° pulse will give the same information as a BB spectrum in less the time.

If coupling constants are needed, the DEPT experiment without decoupling during acquisition can be used. Select the experiment from the library or reduce power setting (L2, S2) to minimum. The experiment is superior to Gated Decoupling since proton relaxation determines the optimal repetition rate (Acquisition time +D1). In addition, the 135° experiment may help to reduce overlap of multiplets.

The standard sequence is somewhat sensitive to the 1/2J delay vs. actual coupling constants. The multiplets may show some phase and intensity distortions. Several variations of DEPT have been designed to improve this. The sequences, DEPTPP.AU (AM), or deptppnd (AMX), bring about a considerable improvement. Note though that the results of a 45° and 135° experiment are reversed compared to a regular DEPT.

Last Update: 3/14/96.