Rainer,
QFN benefit from the leadframe molded into package - thermal  
conductivity of the copper slug type of ground is more efficient for  
heat transfer.  the glass format  would not have such a benefit,  
even  it is  formed using thin film technology with wettable contact.
you better use the real - or dummy made similar to the real part -  
use X-ray, 3D if you have to to study the paste spread.  cross  
section to validate if you have metallurgical  joints  or not. Glass  
chip is great for flip chip, but not for QFN (correct me if I am  
wrong please). IMO.
        jk
On Jul 21, 2014, at 10:53 AM, Rainer Taube wrote:

> Thank you all for the comments. I´m not looking for the wetting
> characteristics but for the paste spreading in a capillary as we  
> have under
> BTCs like QFNs, especially the parts with pitches <= 500µm. I have not
> seen any research on this issue.
>
> From videos with little glass slides on the paste deposites with  
> nearly the
> size and weight of a qfn we saw that the spreading of the paste  
> sometimes
> is very critical. This may cause shorts and solderballs under the
> components. To verify this observation it would be nice to have glass
> components with wettable terminations, because the wetting of glass  
> slides
> with terminations may be more realistic compared to a slide without
> terminations, but in general, there may be a risk with the  
> behaviour of solder
> past under BTCs especially if the paste hight/stencil thickness is  
> very low.
>
> Regards
>
> Rainer
>
>
> Am 21 Jul 2014 um 10:13 hat Steven Creswick geschrieben:
>
>> Rainer,
>>
>> I understand what you are looking for, as we did it ourselves.
>>
>> You need to find a thin film house that can deposit NiC [or Ti], Ni
>> [or Cu], Au [flash] in the pattern you desire onto the glass
>> substrate.
>>
>> My experience has been that this works well for cleaning evaluations
>> and underfill flow evaluations [but the wetting characteristics are
>> not exactly the same as a laminate...], but I would not put any
>> merit in mechanical test results [such as temp cycle, etc.] as the
>> pads are much more likely to lift off the glass than your solder is
>> to fail.
>>
>> Once the array of patterns is deposited on the glass, just dice them
>> apart and pick & place as if they were normal parts.
>>
>> When I get back later today, I will see if I can dig up any possible
>> sources for you.  In the mean time, I suggest looking for those that
>> have thin film capability, or those that sputter [or evaporate] on
>> glass routinely, such as optical filter manufacturers, etc.
>>
>>
>> Steve Creswick
>> Sr Associate - Balanced Enterprise Solutions
>> http://www.linkedin.com/in/stevencreswick
>>                          616 834 1883
>>
>>
>>
>>
>> ____________________________________________________________________
>> __
>> This email has been scanned by the Symantec Email Security.cloud
>> service.
>> For more information please contact helpdesk at x2960 or
>> [log in to unmask]
>> ____________________________________________________________________
>> __
>>
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud  
> service.
> For more information please contact helpdesk at x2960 or  
> [log in to unmask]
> ______________________________________________________________________

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________