This small molecule has good signals on 1H-15N HMBC:

However, upon incorporation into a polymer, the region expected to show correlation peaks is completely noise:

Why?

If you look at the 1D proton signals, plotted on top of the 2D spectra, the aromatic protons in the small molecule have sharp peaks, while those in the polymer have much broader peaks. The broadening is due to short T2. A simple though not the most rigorous way to estimate T2 is T2 = 1/(pi*Delta), while Delta is proton peak width in Hz.

Some 2D pulse sequences keep magnetizations on the xy plane for a long time (a few ms to > 100 ms). For protons with short T2, unfortunately, their signals cannot survive these pulse sequences. HMBC has the harshest requirements for T2.

While the no-show on 2D spectra is a bit disappointing, a short T2 does tell you important information about the dynamics of your molecules – it often indicates that they interact poorly with the solvent and are aggregating.