There is not a hard and fast rule to answer your question, but I can give you some guidelines and what others have done.

 When you do a plasma cleaning and surface preparation, you are doing several things:  
 

1. Removing trace contaminants.  That effect is permanent (as long as the substrate is not re-exposed to the contaminant).
2. Dehydration.  Plasma is extremely efficient at removing moisture from surfaces.  When the substrates are removed from the chamber, they start to re-hydrate.  After initial reaction which forms a mono-layer of water, further hydration takes several days.
3. Removal of thin oxides.  Aluminum bond pads for instance, have about 60Å of Aluminum oxide.  Exposing to plasma can remove the oxide and leave a fresh layer of aluminum.  The re-formation of Aluminum oxide takes several days.
4. Bond activation.  Surface becomes very hydrophilic.  This allows liquids to flow easily and form thin strong bond lines.  The effect on a surface varies with the material.  Soft plastics tend to rearrange themselves and "swallow" the effect.  Harder crystalline materials tend to last longer.

 A large component of how long the effects last depends on storage conditions.  If left out in the atmosphere, the surface will rehydrate and the surfaces will revert to pre-plasma condition.  If they are stored under N2, the effects will last longer.  Depending on the sensitivity of the process, we have customers that have placed as short as 4 hour limits between plasma cleaning and subsequent process.  A more usual timeframe is 48 hours. 

 One other factor to consider is that if there is an unplanned delay between plasma and coating, simply repeating the plasma is acceptable.  There is so little material being removed, that you would need to run many, many cycles to damage the parts.