I don't understand the rationale for sputum dessication. Sputum is difficult to obtain, there are always caveats like contamination with saliva, undersampling, sampling for areas of the lung that are more healthy. Sputum is a nasty tissue and it's role in research and patient care is not impressive, possibly TB diagnosis is the most important one. In general, sputum seems a difficult sample to spend so much effort. - The ideal situation is for some small amount of sputum to be useful for diagnosis or decision-making in a clinical context (change medications, consider invasive therapies, etc). So I think the highest yield that can be obtained from working with sputum is to connect sample findings with clinical outcome. - Sputum has fluid, cells, debris, flora and other. By manipulating the sample through dessication, some of these might be seriously altered (fluid is the one you want to shrink to later expand, cells will die, flora will increase, debris will augment). Therefore, through manipulation you might decrease the value of your sample by worsening its metabolite or cell profile. People spend a lot of time playing with DTT concentrations and volumes just to bring the sputum to a fluid state so it can be homogenized, aliquoted and frozen. I think it's best to use as little sputum as possible for important clinical decision making. So it's best to see what the patient has, and how else the sputum can be analyzed to get the proper information.