First, needless to say, everything must be scrupulously clean before making the seal. I have found it relatively unnecessary to put exactly the right angle on the window, but you do need to have the correct angle on a carbon rod (preferably POCO CZR-1 or better grade rod - email me if you need the source). The rod should be substantially smaller (like 50%) than the ID of the tube (OD of the window).
The easiest way I've found to grind the OD of the window is to cut the correct angle on a scrap of tubing that is really close to (or slightly smaller) than the ID of the tube you'll be sealing the window into (in other words, use a scrap of tubing that slips nicely into the tube you'll be sealing the window onto - heat a piece that's too large and stretch it down if necessary). Then, attach the window to the tube with green lapidary doping wax or something similar (again, contact me for a source of supply - doping wax has a strong grip to glass and cleans easily with acetone). Using the side of a diamond saw blade grind the window down to the desired size (until it's even with the OD of the holder), or use a SiC blade to within a few thousandths and then dress the window down to size with diamond 'star pads' (Wale Apparatus and others). SiC will imbed particles into the quartz whereas the diamond is non-contaminating.
The graphite rod should be solid and not drilled to accommodate vacuum. The contacting end of the rod should be wiped clean with paper towel, then coated with AquaDag thinned with DI or DW to a paint consistency and allowed to dry at room temperature - applying heat will result in it spalling off later. When dry (about an hour or two) polish the AquaDag to a high sheen with a piece of notebook paper.
Clean the glass pieces with soap and water (I definitely prefer Micro to Alconox), then rinse profusely with water and a final rinse of DI or DW. Do _not_ use HF (it will damage the glass) or a solvent (most will leave a residue), just use DW.
Slowly preheat the carbon rod with a bushy flame to drive off residual moisture in the rod - don't carried away - red hot is completely unnecessary.
Chuck the outer jacket tube into the headstock of your lathe. Chuck the graphite rod into the tailstock. Rotate the lathe to check for concentricity/alignment. Handling the window with a tissue to avoid fingerprints, slip it into the open end of the tube. Rotate the lathe slowly to enable the lens to tumble into the tube slightly, and then introduce the graphite rod into the tube. Blow into the tube (a non-contaminating means can be described upon request if concerned that your breath will contaminate the glass) to 'paste' the window against the rod. Position the rod and window where you want them - the lens should be about an inch or so inside the open end of the tube. Stop the lathe such that the lens is nicely supported where you want it, by the rod.
Heat the outside of the jacket at a point directly over the side of the window (not over an extending tip - I don't know quite how else to describe this in words except to say I'm referring to the minor and not the major diameter of the ellipse) and when the glass is plastic, and while blowing like hell into the blow hose to keep the window pasted to the end of the carbon rod, touch the tube with the tip of a pointed graphite tool (again, CZR-1 or better) to 'tack' the tube to the window. Rotate the work piece 180 degrees and repeat.
Remove the graphite rod - you are finished with it. Chuck a matching piece of quartz tubing into the tailstock (with a second blowhose attached) and seal it onto your work piece.
Returning to where you made a tack, reheat and make the tack point better (larger and stronger). Repeat to reinforce the other tack point. Finish by working the seal, nursing the contact point between the tube and window all the way around. You may have to suck slightly to pull the tube down onto the window, or blow to keep from forming a cusp.
If you are working with H2 and O2 this should be fairly fast and bloom buildup should be minimal. If working with city gas and O2, It can help if you coat the faces (not the edges) of the window with thinned AquaDag and let it air dry before proceeding (again, heating AquaDag quickly is asking for trouble - in a worst-case scenario you will trap a bit of spalled-off AquaDag where you need to make the seal, and this can be a potential leak path). The AquaDag is non- contaminating and will protect the window from bloom (a little something I've discovered over the years and am, I think, revealing here for the first time). The AquaDag that is too close to the hot zone simply vaporizes. When done, the remaining AquaDag washes off easily with soap and water (easiest in an ultrasonic bath if the piece is robust). In general, any residual bloom can be etched off with 10% HF (but only for an instant and only if you really need to).
The key here is to use the graphite rod as little as possible, otherwise it will impart contamination to the window.
All trademarks used in this report are the exclusive property of their respective owners.
Last edited 02-18-03