In DNA hybridization, stability of the probe-target interaction is achieved when the probe results in the formation of at least 15 bases. For as log as the complementary bases are present in the target, it should be possible to form a stable capture regardless of how long the target is. Also, you need to choose a unique probe that is complementary to recurring sequences in the target to make sure that the chances of interaction is higher. There are electrochemical DNA sensors that ahd been used for whole virus or whole organisms that had been shown to be stable even after an extended period of time from the target capture. This publication might be of help.

http://pubs.acs.org/doi/abs/10.1021/ac051450i (Z. P. Aguilar, “Rapid Small Volume Microelectrochemical DNA-hybridization Detection of Plasmodium falciparum,” Analytical Chemistry, 2006,. 78(4), 1122-1129.)

I'm not sure that I understand what you are trying to achieve, or indeed how!

One problem may be that if you are trying to capture single copy sequences, you may need to wait for quite some time at the appropriate annealing temperature (too low will capture lots of other non-target sequences).  You can increase the chances of capture by using a large amount of DNA, but this may increase viscosity (unless you increase the volume ( in which case the contact time with the sensor will be reduced) and bear in mind that still only a ~millionth of this will be what you want.  Sonicating that DNA will reduce viscosity, but that might decrease sensitivity if your readout is based on the amount (= length) of DNA captured.