The Vacuum Way Back Machine:

A Look at How It Was in the Early Days In the 1960s, when I first became interested in vacuum technology, the only books available to me were two volumes in the Portland (Maine) Public Library. The first was John Strong’s Procedures in Experimental Physics [1], published in 1938. The second was John Yarwood’s High Vacuum Technique [2], published in 1945. I guess that vacuum was not exactly a hot topic in Portland.

Strong’s book covered many topics besides vacuum practice. Other chapters discussed optics, Geiger-Mueller tubes and counters, electrometers, optics, materials for physics, glassblowing and mechanical design. The book became a classic in its treatment of laboratory practice and has more recently been emulated by Building Scientific Apparatus [3] by Moore, Davis and Coplan where many of Strong’s topics are presented in updated form.

Today we are blessed with UHV compatible materials, advanced fabrication methods, standardized fittings and a plethora of pumps and gauges, almost all of which are available for purchase via the internet. We forget about the days when obtaining a decent vacuum involved a certain resolve and ability to make do with adapted components and various concoctions of sticky goops.

In this article we’ll take a look at some of the examples in Strong’s handbook and compare them to what we have now.

Strong’s RepresentativeVacuum Systems

Strong divided vacuum systems into two classifications: static and kinetic. Static system were of the type that would be used for achieving high levels of clean vacuum for applications such as evacuating x-ray tubes and surface studies. Kinetic systems were of the type that could be opened frequently for process type work and where small leaks and higher outgassing rates could be tolerated. While we don’t use those terms today, the classifications still apply in many respects.

A Static Vacuum System

A drawing of Strong’s static system is shown in Figure 1. The entire high vacuum side is made of glass with no joints. (Yes, glassblowing was a desirable skill for the vacuum worker.) The high vacuum pump is a mercury diffusion pump and the inlet side’s pressure is monitored with a McLeod gauge. A liquid air cold trap isolates the item to be pumped (in this case an x-ray tube). Cold traps are very effective in condensing mercury and water vapor, the main residuals in this system. The item being evacuated can be baked out using the oven. Bake out, of course, remains the standard method for cleaning up a UHV system.

McLeod gauges work on the principle of Boyle’s law. Gas at a low pressure is compressed by a known amount when the mercury reservoir is raised. This compression is sufficient to produce a visible reading on the gauge’s scale. McLeod gauges are gas-type insensitive and can produce accurate readings down to 10-6 Torr.

Figure 1. Static vacuum system per Strong [1]. Reprinted with the kind permission of Lindsay Publications. One item that appears to be missing is a high vacuum gauge. Triode ionization gauges were available at this time but they suffered from the x-ray effect. This limited the range to about 10-7 Torr, not much better than the McLeod gauge. The Bayard-Alpert configuration lay a dozen years in the future.

A Penning (Philips) cold cathode gauge would be another candidate. However Penning only developed the gauge a year before the publication of Strong’s book. That said, the original Penning gauges were only reliable down to about 10-5 to 10-6 Torr.

Kinetic Vacuum System

Strong’s depiction of a kinetic system is shown in Figure 2. This represents a lab-scale system for metallizing mirrors. This is a logical application for Strong as he was a leader in evaporative films for optics. At the time of the publication of his book he was producing aluminum films on telescope mirrors to 36 inches, and a decade later he coated the 200 inch Hale telescope mirror (first light in 1949). Needless to say, vacuum evaporation of aluminum was a tremendous advance over the chemical silver process as the silver film had to be renewed periodically – a messy and hazardous chore for big optics.

Figure 2. Kinetic vacuum system per Strong [1]. Reprinted with the kind permission of Lindsay Publications. The general layout will be familiar to anyone who has worked with a conventional evaporator although the construction and gauging are very much old school. The primary materials used in the pumping system are rubber tubing (fore and roughing lines), brass piping, nickel plated brass and copper for the diffusion pump bodies and chimneys, a mild steel baseplate and glass bell jar.

The foreline and roughing valves are very simple – pinch clamps. The high vacuum valve uses a packing on the shaft and was obviously made in the lab’s shop along with the rest of the connecting components. All of the permanent metal to metal connections were brazed. Inert gas orbital welding of stainless steel components had to wait for the nuclear and space industries in the early 1960s.

The separable connectors included rubber seals on the diffusion pump inlets and vacuum wax for the chamber seal.

The gauging is interesting and would be entirely foreign to most of today’s vacuum users. A simple discharge tube is used between the small booster pump and the main diffusion pump. This rather lowly device can give an experienced operator quite a bit of information about the system. For example, the length of the cathode dark space will provide a rough indication of pressure. If the system is pretty much leak free, the discharge color will transform from the purplish glow of nitrogen to the whitish glow of water vapor, the dominant residual gas in a leak free system. Finally, the discharge will “go dark” as the pressure heads toward a few milliTorr. So, for the price of a piece of glass tubing with a couple of electrodes you get a semi-quantitative pressure gauge and RGA all rolled into one visually appealing device.

The other gauge is a Knudsen gauge in the improved version developed by Dumond and Pickels (1935). I won’t go into a lot of detail, but this gauge consists of a vane that is held by a tight tungsten fiber and where each end of the vane is in close proximity (less than one mean free path) to a heated filament. The higher velocity molecules from the filaments bombard the vane and produce a deflection of the vane. Mounted with the vane is a mirror. A point source of light is reflected off of the mirror and shows a spot on a scale external to the gauge’s housing.

Without going into the math, the deflection is proportional to pressure and independent of the molecular weight of the gas. The Knudsen gauge will work down to pressures on the order of 10-6 Torr.

Concerning the fixturing in the evaporation chamber, what is shown is very simple. For larger work (or multiple small pieces), Strong covers the use of multiple sources, rotating fixtures and shutters. All of these are standard in today’s vacuum evaporators.

Greases and Waxes

The early vacuum worker had access to a wide variety of sealing goops. These ranged from homemade preparations like beeswax/rosin, shellac, Khotinsky cement (a mix of shellac and pitch) to much more advanced sealants such as Apiezon W wax and Picein wax. Most of these materials have room temperature vapor pressures on the order of 0.1 to 1 milliTorr. Apiezon W wax was an exception – its room temperature vapor pressure is about 10-8 Torr. Apiezon W was so successful that it is still available, along with a number of other Apiezon brand vacuum compatible sealing waxes and greases. It’s usable to temperatures to about 70 °C and is applied at 100 °C. W wax is perfect for affixing glass windows to metal apparatus as well as other sealing chores where components can’t be subjected to the heat associated with welding or brazing. Unlike epoxies, a W wax joint also easy to take apart. Simply warm the joint, separate the parts and then clean the residue with an organic solvent. I bought my first stick of Apiezon W in the late 1960s and have always kept a supply on hand.

Going back to the ‘30s, these materials were used everywhere. For example, on large steel evaporation chambers, Strong recommended coating the entire inside with Apiezon W wax and the outer surface with red Glyptal (another handy material that’s still available). As a side note, I last observed the use of Glyptal when examining a pharmaceutical freeze dryer about a dozen years ago. The pump line was coated with it. Obviously there had been a leak and, rather than taking the time to pinpoint the leak, the operator just kept applying Glyptal until the leak was stopped up. Not exactly a good example of troubleshooting technique.

Valves

Valves are an important component in today’s vacuum systems and the situation was no different in the 1930s. However, for the most part you couldn’t go to your favorite vacuum supplier and get a ready-to-go valve. You adapted something – more than likely a brass water or steam valve.

For a basic modification you had to modify the packing on the valve stem. If the valve didn’t have to be adjusted except infrequently, you could melt Apiezon W wax on the valve stem to seal it. When you needed to adjust the valve, just warm the wax with a torch, make the adjustment and let the wax cool again.

A proper valve would be one that replaced the packing with a bellows, and this is what we use today in the form of poppet or bellows valves. Weston Fulton (1871–1946) developed the seamless bellows just after 1900. He was a meteorologist and prolific inventor. The process consisted of spinning a metallic cylinder on a multi-part, removable mandrel. Fulton called his invention a sylphon, named after the fairy sylphs of legend. One of the first applications was in temperature measuring and control devices where the bellows enclosed a volume of liquid.Figure 3 shows a recording thermograph developed by Fulton [4].

Figure 3. The recording thermograph of Weston Fulton. A sylphon is employed as the temperature sensing element. Illustration from [4]. The sylphon quickly found use in industrial valves. A good example is shown in Figure 4.

Figure 4. Sundstrom’s packless valve from 1933 [5]. Strong, being a very hands-on experimental physicist, made many of his own valves from commercial valves. (Albert Ingalls, the author of Amateur Telescope Making, commented that Strong spent so much time with Russell Porter, the “patron saint” of amateur telescope makers and the designer of much of the mechanical structure of the 200 inch Hale telescope, “that the amateur's outlook has rubbed off on him.”) Figure 5 shows one of Strong’s adapted valves.

Figure 5. Industrial valve adapted for vacuum use [1]. Reprinted with the kind permission of Lindsay Publications.