The ge-2D HMBC experiment is the gradient-enhanced version of the conventional 2D HMBC experiment in which coherence selection is achieved by means of PFG. Thus, clean 2D HMBC spectra can be recorded in a single scan per t₁ increment without need for phase cycle when sample concentration is high. Other advantages are the optimal dynamic range, improved water and artefact suppression, and reduced t₁ noise in the minimally required experiment time.  The HMBC experiment allows to trace out longa-range (typically two- and three-bonds away) ¹H-X pairs via the small ⁿJ_HX coupling constant.
   

Easy implementation on any AVANCE spectrometer equipped with pulsed field gradients (PFGs) and inverse probehead.

The incorporation of PFGs in the HMBC pulse train affords interesting advantages. The basic pulse sequence ( 91JMR648-91 and 93MRC287 ) is quite similar to ge-2D HMQC experiment in which the following modifications are made:

• An optional low-pass J-filter (d2-90°(X) where d2=1/2*(¹JXH) can be optionally inserted after the first pulse to minimize direct responses.
• The evolution period is optimized to an 1/2*(ⁿJXH).
• The last refocusing delay is omitted.
• X-broadband decoupling during acquisition is not applied.
• PFGs are usually applied for coherence selection and only one of two desired coherence-transfer pathways are selected, thereby producing magnitude-mode spectra. Defocusing gradients are usually applied during the variable evolution period and the refocusing gradient is applied just prior to acquisition. Qualitative heteronuclear long-range connectivities, mainly on quaternary carbons or through heteronuclei can be extracted.

Several modified pulse sequences have been proposed:

• Analog experiments using the HSQC pulse train has been also reported ( 97JMR503-124 , 99MRC413 , 00MRC265 )
• Modified pulse sequences to suppress undesired direct responses:
  A purged TANGO scheme is incorporated as a preparation period followed by the classical low-pass J-filter. Under these conditions, milder gradient strengths can be used and improved t1-noise reduction is achieved. ( 95JMRA241-112 ).
  Use of a double low-pass filter ( 00MRC981 , 00MRC963 )
  use of a BIRD sandwich ( 00MRC963 )
• Sensitivity of the HMBC experiment can be increased by using semi-selective pulses which suppress homonuclear ¹H-¹H J modulation and allows correlating poorly resolved proton multiplets ( 96JMRA134-119 ).
• A decoupled-HMBC version has also been proposed to observe small cross-peaks ( 95TL2817 ).
• On the other hand, a constant-time version (CT-HMBC) has been reported ( 98TL7337 ) to remove the undesired J(HH) modulation during the t₁ period.
• The ACCORD-HMBC experiment based on the Accordion spectroscopy ( 98MRCS44 , 99MRC517 , and 00MRC452 ). Variants are IMPEACH-MBC ( 99JMR274-140 ), CIGAR-HMBC ( 00MRC143 ) or a pulse scheme to differentiate two-bond from three-bond correlation peaks ( 00JMR232-146 ).
• Co-addition of several HMBC experiments optimized for different delays (( 00MRC981 ))
• Improved resolution in the F₁ dimension can be achieved by applying a band-selective carbon pulse ((see Semi-selective HMBC experiment)).
• In the J-resolved HMBC pulse scheme ( 95MRC632 , 99TL6271 , 00TET2935 ), long range ¹H-¹³C, ¹H-15N, and ¹H-31P coupling constants have been measured by stepwise incrementation of the coupling evolution time in a constant-time manner. A 3D version of this experiment can be easily designed (ge-3D J-Resolved HMBC experiment). Application on ¹¹⁹Sn has been reported ( 00PROG271 ).
• Modified sequences have been also proposed to extract nJ(CH) from phase-sensitive HMBC spectra ( 96JMRA120-119 and 98JMR296-130 ).
• The basic pulse sequence have been also used to obtain ¹H-¹⁵N HMBC spectra at natural abundance on several natural products ( 96TL1447 , 95JHC1665 , and 95JHC1759 , 98JNP969 , and 00JNP543 ) and to differentiate regioisomers ( 98MRC35 ). The application of ¹H-¹⁵N ACCORD-HMBC and ¹H-¹⁵N IMPEACH-MBC experiments that use accordion spectroscopy has been also reported ( 00MRC251 ). In addition, ⁿJ(HN) coupling constants could be measured with accuracy from the corresponding anti-phase cross-peaks in its phase-sensitive HMBC version (with the echo-antiecho approach) as demonstrated for the indoloquinoline alkaloid cryptolepine ( 96JNP2 ). Practical applications on 31P ( 97MRC227 ), and 119Sn ( 97JMR218-124 and 96ORG1920 ) have also been published. An analog long-range optimized 31P-¹H HSQC experiment has also been published ( 96MRC33 ).

The ge-2D HMBC experiment is usually recorded in routine/automation modes and minor changes are required if a predefined parameter set is available. Magnitude-mode spectrum is normally recorded because the experiment is usually analyzed in qualitative terms. Important parameters to consider:

The user need to define the strength, the duration, the shape of the gradients and the recovery delay.

Offset and spectral widths in each dimension.

Optimization of the evolution delay as a function of 1/2(ⁿJCH) (50-80 ms).

More details on practical implementation of ge-2D HMBC experiments on AVANCE spectrometers can be found in

Tutorials: 2D inverse experiments

Tutorials: 2D gradient-based inverse experiments

The HMBC spectrum correlates chemical shifts of heteronucleus X (F₁ dimension) and protons (F₂ dimension) via the long-range heteronuclear coupling,  ⁿJ(XH) whre n>1. The effective suppression of unwanted ¹H-¹²C or ¹H-14N magnetization by means of PFGs allows to obtain ultra-clean 2D spectra from which clear analysis of even tiny cross-peaks can be done. Direct correlations can be observed as a large double resonances due to ¹J(CH). An optional low-pass filter is usually included in the pulse sequence to minimize them.

Related experiments:

2D Inverse experiments

2D Inverse gradient-enhanced experiments