Mon, 14 Sep 2009 15:19:24 The following, which has some very useful information on considerations for the choice of glasses, is extracted from an email exchange between Michael Minot, Seung-Woo Lee, and Henry.         Can I ask about soda lime glass vs borosilicate? I got the impression that soda lime wasn't as good for vacuum sealed tubes as borosilicate- what are the considerations? - The differences between those two glass substrates which mattered in the Flat panel display process were; (SL)   1. CTE : Since the thick film material in Plasma Displays were initially developed for Soda lime (CTE : 85-90), so they have been used Soda lime or PD200. Otherwise too much effort will be needed to modify material to match with Borosilicate(CTE 40-45). (SL)   2. Sodium content : Very low or almost no-sodium content was required to make TFT on the glass, otherwise the I-V characteristics will change over time due to Sodium migration. Therefore they started to use borosilicate such as Corning 7059, Corning 1737, Schott AF45, Asahi An635 and Corning Eagle. Because these glass materials were developed for LCDs, they are only available in thin thicknesses ( 0.5mm-1.1mm ). (SL)   3. Surface quality : Due to the thin film process and very fine geometry, LCDs require good quality of surface. Regular window glass(Soda lime) do not provide that. Therefore they had to create new glass or polish the surface of the glass. This is the part of reason they went to Borosilicate which were especially fabricated for LCDs. (SL) What are most vacuum tubes and PMT's made of?   1. CRTs and VFDs(Vacuum fluorescent Display) use Sodalime, FEDs used both Sodalime and Borosilicate ( Not Pyrex (Corning 7740), but AN635, Borofloat 40 or Corning 1737). There was a concern that the sodium migration may cause frit seal leakage at higher Electric field applications (10KV above) but it is not verified. (SL)   2. For the sake of frit seal, both of glass material should be OK as long as you use right material for the glass. (SL) There is also the question of chemical compatibility with the photocathode. Lastly, I had thought there was an advantage to having all components have the same CTE, and since the MCP's are borosilicate, was trying to keep the anode and spacers, and perhaps the window, the same. (this raises an interesting question- is the CTE for a glass capillary array the same as for the bulk glass?)         1. This is the reason I prefer to use discrete spacers rather than spacer sheets. CTE mismatch can be a problem only if seal area is big enough to affect thermal expansion of the medium. If we keep it small area and not connected to the others, it shouldn't be a problem. (SL)       2. For the same reason, I suggested to have MCPs float between the spacers not sealed to the glass plates. (SL)        3.  Regarding the CTE of the MCP's, we currently use a moderate expansion Borosilicate glass with at CTE = 55.  At this particular juncture in the development, using alternative glass materials is not particularly timely.  Ultimately however we can replicate our process with any of these glasses that we can form into tubes and that can be drawn into capillary.  I am confident that we can match the MCP with the glass selected for the Window, side walls and anode back plate.  (mjm)