Wayne,
baking at  above 65C in vacuum  chamber does make difference based on  
my experience.  absorption and desorption process take place as long  
as it has sufficient energy (assume required bake is not strongly  
bonded - reacted moisture, like corrosion product for extreme   
example).  You might want to put moisture indicator in the chamber  
with and without vacuum to see if it make any difference (assume you  
do have a good vacuum chamber).  Or you can do TGA figure out where  
is the desorption take place of your specimen (use 90% RH/65 C to  
saturated 1st and ramp down to room temp slowly to prepare your  
specimen for example).  IMHO.
baking above 105 C is great with or without vacuum, but many device  
can't go that high (or packaging material... for example, laser  
alignment might slightly off, if it not designed to operated at high  
temp).  low temp baking, a good vacuum helps the process in my  
experiences (long ago.. don't ask me for data or publication... ).
jk
On Apr 15, 2018, at 12:43 PM, Wayne Showers wrote:

> The problem with vacuum drying is not so much the boiling point of  
> water as in an unimpeded environment, those values are: At 25" Hg :  
> 56C, 27" Hg : 45C, 28" Hg :40C, and so on.  The problem is degree  
> of impediment caused by the actual plastic.  To build off what Yuan- 
> chia Joyce Koo has stated but with a different vector in mind, the  
> plastic material of moisture bags for instance is highly resistant  
> to diffusion.  Different polymers will follow accordingly.
>
> For information, many plastics must be dried immediately prior to  
> molding otherwise they suck up moisture like a sponge. Common  
> hygroscopic polymers are Nylon, ABS (Acetyl Butyl Styrene), PC  
> (Polycarbonate), and PMMA (Poly (methyl methacrylate)).  Nylon and  
> PMMA are very common materials in component shells.  ABS and PC  
> index* at temperatures not much higher than boiling water and are  
> not typically used.  PP (Polypropylene) and PE (Polyethylene) are  
> not so bad, hence their use in capacitors, but these index in the  
> 200C range making them a bear for SMT placement.
> My general rule is the higher the MSL number, i.e., the more likely  
> a part is to absorb moisture, the easier it is to dry.
>
> For normal bake (NO vacuum), here is the formula I derived several  
> years ago:
> Thickness (mm) X Thickness (mm) (Divided by) Bake Temp Factor
> D125oC= 0.4427 // D90oC= 0.1270 // D40oC= 0.0131, Result is time in  
> hours.
>
> I have never experimented with Vacuum Acceleration, but I will  
> query the Penn State Plastics Engineering group with what we are  
> trying to resolve and post those results to the J-STD-033 committee  
> as well as here.
> * Index is the temperature at which a plastic flows well enough to  
> be molded.  It is neither solid nor liquid, but soliquid.
>
> This all sounds like a job for Clumpy and Kloumpios when they are  
> next available.