Canberra found success in Ge(Li) detector manufacturing by making refinements to the basic processes that Fiedler brought with him. Although we had talented technicians who became very proficient at production and Orren was able to keep cryostat design at a competitive level, we had no one on the staff doing device development. Bob Lothrop, our Si(Li) man, was encouraged to develop processes for making HPGe detectors, and he even made a couple of planars using borrowed Si(Li) technology. He really didn’t have his heart in it, however, and left the company. We found ourselves in the mid-1970s with a dying product line and with no replacement in sight.
Unlike Ge(Li) detectors which were widely made in university and research laboratories, the HPGe detector was developed mostly by commercial firms. There were some published works on device structures and manufacturing techniques, but most involved exotic processes which could not be reproduced easily. We talked to several individuals who had reported some success in making HPGe detectors but few had a convincing story to tell when the right questions were posed. We had hoped to hire someone with a proven track record but no such person was available, there being only a few individuals in the world who could claim any knowledge about HPGe detector fabrication.
In our search for help we interviewed Dr. Andy Zolnay, a young Ph.D from Ohio State University who had worked on a detector project financed by Picker Nuclear. Several OSU grad students had been involved in this Ge gamma camera development project which ran for several years. Andy did not join us at first but rather took a job at EDAX where he worked on Si(Li) detectors. Later we had further discussions with Andy, who now professed to know the secret of making good contacts on Ge. For personal reasons Andy did not join Canberra at that time but he put us in contact with another grad student at OSU, Mike Yocum, who had also worked on the Picker project. Mike joined Canberra in 1977 and, after some time, was joined by Zolnay who had become available again. Mike and Andy worked together for several months. It soon became clear that that Mike was terrific at experimental work, and theories were not so much in demand after several failures. Eventually Andy sought employment elsewhere with our blessing.
Mike was experienced in vacuum evaporation techniques and soon we abandoned all the exotic processes and settled on evaporated gold (P+) and diffused Lithium (N+) contacts. After making a good device, however, it was a challenge to keep it good through temperature cycles. With Ge(Li) detectors, it was not necessary to have really clean cryostats because the molecular sieves pumped well and never warmed up to outgas and contaminate the detector element. With HPGe detectors, however, a new level of cryostat cleanliness had to be achieved and some kind of passivation coating was found to be useful, if not essential, in maintaining detector stability.
We attacked all these problems and used one fix after another in the early days- again looking for the good passivating coating thought to be necessary. It was de rigueur in those days to do five temperature cycles on each detector to prove stability, and even though we followed this rule, our confidence in our claim of indefinite cyclability was not very high. Fortunately, most customers cooled down detectors and kept them cold so we were never plagued by device failures at any time.
Mike was a whiz at device and process development. He would carry out a dozen experiments in a day if necessary in order to devise, debug, or refine a process. Once we had basic processes in place for the fabrication of planar and coaxial Ge detectors, these device were turned over to production but Mike’s attention was required to right problems that came up routinely. Eventually, as the training of production workers became more complete, we were able to expend more effort in broadening the product line. Reverse-Electrode (N-Type) detectors came easy even if, unlike ORTEC who had an ion-implanter (they had developed an ion-implanter and sold a few after all), we had to use an outside implantation service for some time. Larry Darken told me much later that ORTEC believed their capability in implantation would be a huge barrier for their competitors.
Ge Well Detectors were problematic for Canberra, in part because they were made in low quantities, had high capacitance, and tended to be microphonic because of this and the tricky core contact. Canberra made a huge breakthrough in Well detectors in 2013, with the development of the Small-Anode Ge Well Detector (SAGe-Well), which evolved from the development of core-less coaxial detectors (SAGe Detectors) that Canberra began making back in the 1980s and which became popular as tools for Underground Science much later, thanks to pioneering work by Dr. Juan Collar at the University of Chicago.
The Low-Energy (LEGe) detector, an aptly named product (the resolution is not great at high energies), was developed by Canberra as a replacement for Planars, which had relatively high capacitance with attendant noise and microphonism. This device became very popular in Lung Burden systems as well as a general purpose X-Ray detector. For many years, it was believed that Ge detectors were virtually unusable below 3 keV because of poor peak shape. Some (LBL included) wrote that this was due to “hot electron escape” from the interaction of photons very close to the entrance window. Mike Yocum developed a proprietary entrance window for the LEGe detector one and for all demonstrating that good low-energy performance could be had with Ge detectors. We called this detector the Ultra-LEGe and it became very popular in X-ray applications once dominated by Si(Li) detectors. Because the energy required to produce a hole-electron pair in Ge (~3 eV) is less than that required for Si (3.5eV) the energy is resolution fundamentally better simply because of the statistics of charge production.
The Extended-Range Ge (XtRa) Ge detector was developed because it was difficult to obtain N-type Ge crystals with the great charge collection properties associated with P-type crystals. We realized that many users were buying n-types simply for their low-energy response and not for radiation damage resistance so Mike Yocum perfected a proprietary entrance window contact that extended the energy range of p-types down to below 10 keV.
The Broad-Energy Ge (BEGe) Detector was developed because of the market need for larger and larger LEGe detectors especially in applications such a lung burden. The resolution of LEGe detectors gets progressively worse with increasing energy and size. For about a week we called this detector the Wide-Energy Ge (WEGe) Detector. Orren introduced this product at a Canberra User’s Group Meeting and suddenly realized that no one who had ever been a schoolboy in the US would pay money for a WEGE, so we promptly change the name. In recent years this detector structure has become popular choice for underground science.
The SAGe aka Spot SEGe Detector evolved from detectors we began making for a customer at Idaho Falls in the 1980s. They needed a coaxial detector with good low energy resolution so we set about to reduce the core dimensions to reduce capacitance, continuing until there was virtually no core left, just a surface electrode or “spot” on the rear face. Because the charge collection in p-type germanium was so good the performance was good over a wide range of energies. It occurred to Orren and Mike Yocum during that time that this structure could be used to make troublesome Ge Well Detectors but it would be decades later that this came about.
Around 2004 I got a call from Dr. Juan Collier at the University of Chicago asking if we could make a coaxial detector with extremely low noise for Dark Matter experiments. I suggested the “Spot SEGe” concept and he gave us an order for one. Interestingly earlier efforts to make low capacitance detectors by Paul Luke at LBNL focused on n-type material and the charge collection was poor at best. We had only a vague recollection of this work including the name that Paul had used which was “Point Contact”, after the point contact transistor I suppose. Anyway Juan was pleased with the performance of this first detector and he found that, by means of pulse shape analysis, he could identify the location of photon interactions within the detector element. After these results were published scientists realized that highly segmented detectors might not be required for the various underground double beta decay experiments in the planning stages.
Around 2014, Canberra resurrected the idea of making Ge Well detectors with the Small-Anode structure and soon introduced the SAGe-Well Detector which greatly improved the performance of these devices and made it practical to build detectors with larger well sizes while maintaining good low-energy resolution.