Introduction by Terry Quinn FRS
What follows is a short note by John Martin on his career, which began as a technician at the National Physical Laboratory (NPL). He joined the NPL in 1960 as the most junior of junior technicians and retired in 2000 as a highly respected senior member of the NPL staff with a worldwide reputation and more than 50 scientific papers to his name, either as co-author or principal author. What is perhaps unusual is that having arrived at the NPL with his only academic qualifications being A-levels in physics and mathematics he took no further formal academic courses but absorbed the necessary science from those around him, a long apprenticeship in the best meaning of the term!
The NPL changed greatly between 1960 and 2000. It was founded in 1900 with the clear objective of providing for British industry what the German Physikalisch-Technische Reichsanstalt (PTR) had been doing for German industry since 1887. However, during most of the twentieth century the NPL suffered many vicissitudes at the hands of the British government, which, from very early on, seemed to have forgotten what it was for. The periodic reviews of how it should be managed, notably in 1919, when the Royal Society lost the battle with the Department of Scientific and Industrial Research (DSIR) as to which of them would have prime responsibility for the scientific policy and programme of the NPL, and then from the 1960s, with successive changes in name and policy of the departments responsible for the NPL, left it rather battered and unsure of itself. This is in contrast with the PTR (which became the Physikalisch-Technische Bundesanstalt (PTB) after World War II) and the National Bureau of Standards (NBS) in Washington, founded in 1901, which both had and continue to have clear missions wholly supported by their respective governments. In the case of the NBS, now the National Institute of Standards and Technology (NIST), the US government had always provided very substantial funding, at present nearly US $500 million and recently set to reach US $1 billion by 2016. Three current members of staff have won Nobel Prizes for their work at NIST, a fact perhaps not wholly unrelated to the continuous strong support of the US government.
John Martin's career, however, was relatively untouched by the successive and changing roles of the NPL because he worked for most of his time in the area of basic physical standards, which always remained its core function. His principal activity, as you will see, was in the development with me of the cryogenic radiometer. This was a project I began thinking about when I spent a year at the NBS in 1967–68 and there came across work on a cryogenic heat-flow calorimeter to be used for the measurement of the thermodynamic temperature of the boiling point of water by measuring the total thermal radiation from a black body. Our project really began in 1971 after I returned from a Temperature Conference in Washington where I had heard that the NBS project had been abandoned as a result of intractable problems related to diffraction and scattering. I thought I knew how to overcome these problems and drew up a sketch of how it could be done. I sent this off to Ralph Hudson, then head of the NBS Heat Division and with whom I had worked when I was there, for his comments. A short while later the reply came back that although my ideas looked interesting, his view and that of his colleagues was that my proposals would certainly not overcome the basic problems that they had encountered. I am not sure that I showed this to my then boss, Cecil Barber; because the project already had his approval I had every intention to embark on it and I probably thought it would simply worry him unnecessarily! This was long before the Rothschild customer/contractor principle had taken hold of the NPL and things were simpler. In 1972 John joined me and, together with John Compton, a low-temperature physicist who unfortunately left the NPL soon afterwards, we set to work to build the instrument and to demonstrate its capabilities. We planned to measure the Stefan–Boltzmann constant directly by using thermal radiation from a black body at the temperature of the triple point of water and then the thermodynamic temperature of the boiling point of water. All of this was successful, although it took some time, and was published in 1985 in a 100-page paper jointly authored by John and myself in Philosophical Transactions of the Royal Society series A. John went on to design and build the first commercial prototype, then manufactured by Oxford Instruments and later by an American company (in Cambridge, MA), and it is now used by all national metrology institutes around the world as the basic reference for the measurement of light and thermal radiation. He was later closely involved in the design of a cryogenic radiometer for use in space for measurements of the total solar irradiance, a project that, sadly, has yet to be funded despite its increasing importance in the provision of baseline data for climate change. In its latest incarnation, however, under the acronym TRUTHS (Traceable Radiometry Underpinning Terrestrial- and Helio- Studies), funding is looking increasingly likely.
In one sense, John's career is the story of his key role in the development and subsequent use of the cryogenic radiometer.
I got A-levels in physics, pure mathematics and applied mathematics at school, and was accepted for a ‘thick sandwich’ course at Rugby Engineering College in 1958 for a degree in mechanical engineering. I only survived one year there; having been declared unfit for National Service, I saw an advert in the paper for an Experimental Officer (EO) at the NPL. This was part of the Civil Service, and the advert offered the prospect of a salary of £1000 a year in 10–15 years' time. I started in 1960 as ‘Assistant Scientific’ at about £300 a year and worked with Linc Smith for the next 12 years. Smith came from the Clarendon Laboratory at Oxford as a Senior Scientific Officer (SSO), although his interest in science was on the wane by then. He ended up as a training officer, and his main hobby was bricklaying as his father had been a stonemason at Portland Bill. I was initially ‘unestablished’ but after a year I became ‘established’, which meant that after passing a medical I had a permanent post and was entitled to a pension. An ‘Establishment Officer’ also checked your security background!
I originally worked on specific heats at low temperatures in Applied Physics near Building 56. In those days the NPL was much more concerned with standards, and only about 650 people worked there. At that time the quantum Hall and Josephson effects had not yet been discovered. The building had changed little in the years before I arrived although there were no longer ‘boys’ whose role was only to take down notes and record numbers.
I was part of a team of four that ran the Collins helium liquefier under Smith from 1960 to 1964; this included George Crockett and Joyce Nicholls. Joyce had a degree and was a Scientific Officer (SO), and George was a Senior Experimental Officer (SEO). He was a brilliant glassblower and made glass dewar flasks at the National Chemical Laboratory; he made glass animals in his spare time. However, by the early 1960s, glassblowing had gone out of fashion with the advent of stainless steel dewars, and it was no longer an important part of the workshop's operations. He was also highly skilled in the hard and soft soldering techniques that were essential in low-temperature work, and he taught me his skills. This early experience with the liquefier gave me my enthusiasm for low-temperature work.
Initially it was thought that we would need to liquefy hydrogen; the liquefier was housed in a building with a special safety roof, although in the event we never actually used hydrogen. When we worked on helium, companies such as BOC were not in a position to supply liquid helium, and the NPL itself was the only supplier. The liquefier had come to the NPL as part of the Marshall Plan, but this was on the basis that the NPL would supply university departments and research establishments over the UK, as far afield as Bangor. Pure helium gas came from the USA, and there was a large demand for liquid helium.
The liquefier operated on the principle of an expansion engine working on the Carnot cycle with a Joule–Thomson valve as the final stage. It was run by George Crockett, and I had the job of a sort of ordering clerk. Gradually I was shown how to run the liquefier, a job that included manipulating the Joule–Thompson valve to set the rate of liquefaction. It was extremely hard to ‘see’ the liquid, and to do this you had to put a dipstick into the liquefier to make the helium ‘ripple’. It liquefied using clean gas at a rate of 8 litres an hour, and up to 30 litres of liquid was stored in the liquefier. The liquid helium was then transferred into transport dewars by using a vacuum-insulated siphon—50% of the liquid was boiled off during the process. The dewars were made of copper with a narrow neck consisting of four chambers: one for the helium, one for liquid nitrogen and two vacuum chambers. Making them was quite an art, and they had to be specially made by a copper spinner working for Dupree-Swift, a company based in the East End of London. They had no knowledge of low-temperature engineering but were experts in copper spinning and in making vacuum-tight solder joints.
The dewars were transported around the country on the rail network system; sometimes they were dropped on railway platforms, causing the helium to boil off so that universities complained that there was no helium on arrival. At other times, the guards refused to let the dewars on the trains. Because helium was so expensive, users had to return the gas in a compressed form in cylinders, but the gas was often contaminated with oil vapour and had to be cleaned.
At about this time (1964) I was promoted to Assistant Experimental Officer, and I moved with Linc Smith to the high-pressure section in Building 17 (near Glazebrook Hall), whose focus was on producing artificial diamonds. In those days you had a choice of where you went if you were a scientist with a university degree, but in the EO class we had to go where we were told. I worked on the collapse of crystal structure under high pressure, and with Norman Owen, a SEO, we developed a miniature diamond anvil press that permitted the use of X-ray techniques to determine the structural changes in iii–v semiconductor compounds. At the same time I started to attend Birkbeck College to learn more about X-ray crystallography and the interpretation of powder diffraction photographs. The group did produce some diamonds but could not compete with private companies making the same product, so the section was closed down. I was co-author of scientific papers with Owen and Linc Smith on the results of our work.1
After this, in about 1965 or 1966, I went back with Smith to the newly built Basic Physics Laboratory in Building 95. The group was called Low Temperature Techniques, and we designed and constructed a helium-3 cryostat that went down to a base temperature of 0.3 K, which was a first for the NPL. In 1970 this group was absorbed into Bob Barber's temperature section, which Terry Quinn had joined in 1962. By then I was an EO, and I worked with Linc Smith on a 3He/4He dilution refrigerator. I designed the equipment and did the technical drawings, and the workshop manufactured the machine. Linc Smith left at this point to become a training officer, and I was left alone to commission the refrigerator. It reached its objective of a continuous low temperature of 25 mK and a one-shot temperature of 5 mK. During this period Bob Barber recruited several SOs and SSOs into the group, such as John Compton (nuclear orientation thermometry), Tony Colclough (acoustic thermometry), Keith Berry (gas thermometry) and Richard Rusby (who used the dilution refrigerator for his development of the rhodium–iron thermometer). Various EOs, such as Barry Jolliffe, Martin Ford and Terry Chandler, also provided assistance to the SOs.
In 1967–68, things changed at the NPL when a number of scientists came back from the USA. This was part of a major effort by the Civil Service to reverse the ‘brain drain’, and some top-class people were brought back. When I had finished the work on the dilution refrigerator, Bob Barber asked me to work on his pet project, which was an international comparison of platinum resistance thermometers. This work was important and contributed towards what became known as the Ward–Compton comparison. However, I turned the offer down, and he reluctantly agreed that I could work on the construction of a cryogenic radiometer with Terry Quinn and John Compton, because I had heard about it on the grapevine. The radiometer was to be used for measuring the thermal energy from a blackbody radiator in the temperature range −30 to 100 °C. From these measurements it would be possible to determine the Stefan–Boltzmann constant and the thermodynamic temperature.2
John was responsible for the cryogenic radiometer and Terry was in charge of the blackbody radiator—my function was to bring all their ideas together to design a workable instrument. The other major component was a radiation trap at a temperature of 4 K. This housed two apertures a set distance apart to define the solid angle of radiation from the black body as well as significantly reducing the amount of scattered radiation reaching the radiometer—this was part of Terry's responsibility. Those three components were assembled in a common vacuum chamber. Over the next two years I completed nearly 150 technical drawings from which our workshops could start to construct the instrument.
There were many problems to be solved during the design phase, such as how to build a copper chamber that could be filled on a daily basis with helium by pumping the pressure lowered to produce superfluid helium. One end of the apparatus was at 2 K, the other at 273 K, and there were big problems with mechanical stability and thermal contraction. We had to hang the equipment perfectly vertically on brick pillars with an anti-vibration mounting designed by Norman Owen. We needed to know the thermal properties of the copper used for the apertures and the radiation trap. It was possible to determine the diameter of the apertures at room temperature but we needed to know their dimensions at 4 K, and to solve this we used oxygen-free high-conductivity copper. This was because state-of-the-art measurements had been made in the USA by Swenson, who had determined the thermal contraction of this sort of copper at different temperatures down to 4 K with an accuracy of 1 in 104. Also at this stage John Compton and I used the gas lasers at the NPL to determine the reflectivity of 3M's Nextel black paint at long wavelengths, an essential procedure for determining the emissivity of the black bodies. We published a paper on this work.3
During the construction phase I visited the workshops about two or three times a day—the staff valued the interest and I could immediately make decisions if they were having problems. One also had to have control over what they did. I was one of the few who did all my own drawings, because the drawing office was always under pressure as a result of staff shortages.
The west workshop, which was a massive building, gave us very good service in the 1960s and 1970s. In the 1960s it was one of five big workshops that were devoted to the NPL. Another was devoted to electrotechnics, and another to woodworking, which was vital to the Aerodynamic Division. In the 1960s and 1970s the ability to make things with lathes was very important, although nowadays the skill of a craftsman has been taken over by the ability to run a numerically controlled machine. One needed very close collaboration with the mechanics, who did many things on-site that today would be done by outside companies. In the early days we went with a sketch to a mechanic, but later we had to go through the design office. Highly skilled people were well regarded, and they had to be shown respect. The apertures they made by ‘lapping’ have never been bettered; this was developed in the nineteenth century and involved rubbing two surfaces together with a fine paste to produce a 5 μm surface, but it is hardly done any more. The key person was Ernie Pimm, who was brilliant at lapping. If he saw you approaching he would pick up your aperture off the bench and start working on it immediately. The entire project would have stopped without these craftsmen, and in that sense they had you ‘over a barrel’. The only way in which craftsmen could be promoted was if they were moved into the management side of the workshop, at which point they were known as Technical Officers; when this happened, their abilities as craftsmen were lost to the NPL. One exception was Ernie Pimm, who was allowed to continue lapping even though he was promoted to Technical Officer (and he later became a Principal Technical Officer).
I had much experience in knowing what would ‘work’, and this was very useful in the assembly and commissioning of the radiometer. I knew a lot about the properties of liquid helium and I used to come in on weekends to put this into the radiometer to keep it cold. There were many discussions about how platinum resistance thermometers worked in a vacuum, and many things went wrong—for instance, the electrical connecting wires had to be thermally anchored to prevent dissipation, and we had to invent new ways of doing this. The thermometers were calibrated about once every year, against the triple point of water, both inside and outside the radiometer.
There was an apprenticeship that is not learnt in textbooks; once learnt, it can be passed on. For example, GE varnish was required to wrap around the cryostat copper wires—which had a diameter of 0.1 mm—and you needed vacuum grease to improve the thermal contact between the thermometers and their mounts.
Ultimately we had to do away with our original idea of a copper base welded to stainless steel. We sent it to a specialist firm, but the factory burned down and we were unable to retrieve it until the insurance assessors came. It had been annealed by the flames. There were massive leaks between the stainless steel and the copper, and the soldering material corroded very easily. We could not get serious results for six years because of this fundamental problem, but eventually I changed the design. This certainly would not be funded now, because one needs to reach milestones of two or three years. However, in those days the NPL was managed by directors, and funds were allocated differently; in our case the power to do this was in the hands of the Superintendent, Michael Peover, who was always supportive of the project.
After Terry Quinn left, the head of the Temperature Section was Peter Coates, and he was succeeded by Tony Colclough. At about that time (1974) I was promoted to SEO; this was as far as I could go because there was little chance of being promoted to the position of Chief Experimental Officer. In the mid-1980s the EO and SO grades were amalgamated by the Civil Service, and at that point I became an SSO. There were still divisions between those with degrees and those without, because the former could reach the level of Principal Scientific Officer (PSO). Nevertheless, the Civil Service was becoming more interested in what you did rather than the qualifications you had.
I continued to manage the cryogenic radiometer project with continuous support from Terry Quinn (who by that time had moved to the Bureau International des Poids et Mesures), determining the Stefan–Boltzmann constant and temperatures in the range −30 to 110 °C. This work was of great interest to the radiometry section, headed by Peter Key, and it was suggested that we could use the radiometer for the accurate measurement of a visible laser beam that could then be used to calibrate photodiodes. Two Americans at NIST, Jon Geist and Ed Zalewski, were also involved, and we designed a modification to the lower end of the radiometer. The black body was removed and a new lower vacuum chamber was constructed with a Brewster-angled window through which we could pass the laser beam. I made the measurement with Nigel Fox (from Radiometry), and we demonstrated that we could determine the power of the laser to 1 in 104, although the photodiode measurements were inconclusive.
After this, I had a meeting with section heads and suggested that I could design a cryogenic radiometer dedicated to measuring laser power. This was agreed, and a radiometer was constructed in accordance with this design by the Oxford Instrument Company (OIC). It became the primary standard at the NPL and was sold commercially by the OIC to other national standards laboratories in Germany and the USA. The original drawing is now proudly framed in my study at home.4
When we had finished the work on temperature measurements and the Stefan–Boltzmann constant, Terry Quinn and I reported the results in Philosophical Transactions of the Royal Society series A in 1985 (figures 1 and 2),5 and I was transferred to the Radiometry Section. There I worked with Nigel Fox and continued to work on photodiodes and laser radiometry with the cryogenic radiometer. I also constructed a mechanically cooled radiometer that did not require liquid helium.6 Other people's expertise was needed for solving the problem of the black paint. We tried to obtain special paint from the Japanese company Amritsu but this did not work out. We went to several university departments, and ended up hiring someone from Newcastle University to reverse engineer what the Japanese had done. After some years we had some success with paint produced with a nickel-phosphorus plating process, but making optical black paint and measuring its reflectance was always extremely difficult.
It was at this point in the early 1990s that we became interested in measuring the solar constant, and we made proposals about putting a mechanically cooled radiometer on a space platform.7 In the past 7 years there have been various moves to put the radiometer into space. Richard Willson found that the Sun's output (Total Solar Irradiance) was falling through ACRIM 1 (Active Cavity Radiometer Irradiance Monitor 1), and this was corroborated by ERBS (Earth Radiation Budget Satellite). All this has serious repercussions for analysing and calibrating the absolute magnitude of the total solar radiance and its variations, which we thought could be a major source of climate change. This work continues today, led by Nigel Fox.
In the late 1980s and early 1990s performance bonuses came in; I, Mike Downes and Dickie Bird, among others, were then put forward to become PSOs. Mike and Dickie were promoted but I did not pass the interview; however, I was put forward a second time and this time succeeded. The interview process lasted about 30 minutes, and I spoke about my work on the cryogenic radiometer.
I had no university qualifications but knew enough about the scientific theory connected with the work I did. Cecil Barber had apparently done all his studies at night school at Birkbeck College in the 1940s, but I preferred to read academic papers and books rather than study formally for a qualification. I gradually took on more responsibility for different tasks, especially after Terry left. I hated management meetings but by the late 1970s I had a job that had much more responsibility than one would have expected given my lack of a university education. By that time a scientist would not have known more than me about the issues connected with making a cryogenic radiometer work, and I began to give papers on the machine at conferences. I lacked the theoretical background that gives you a broad understanding of the principles of quantum physics, but my knowledge of radiometry was excellent. This was a practical knowledge that I had acquired in a different manner from a university training; among other things it gave me the ability to see that occasionally results given in other papers were overstated.
Informal socializing over a tea break was important, as were meetings between groups in radiometry and thermometry. People were encouraged to talk about their work, and I sometimes found that people had solved practical problems I was facing, or had good ideas. Another important part of my job was the social club, because it was only when playing sports that one could meet people from other parts of the NPL—or talk to senior people informally over a drink.
In the 1960s I received instruction from the scientists I was assisting. I helped to run the experiments, using my workshop experience with a lathe, etc.—there was always a ‘staff workshop’ in each building—to make modifications to the apparatus. It also involved recording the data, which were then analysed by the SOs. In the 1970s I was allowed to contribute to discussions on new projects. I spent more time on designing but I was still under instruction from the SOs—I now shared an office and had a desk! By the 1980s I had taken charge of the radiometry project and was responsible for the calibration of the instrument to make measurements. I identified problems that arose and was now able to analyse the data being produced by the equipment. I wrote academic papers on the results we produced, and was allowed to attend conferences in other countries. The NPL projects were now more closely ‘managed’ than previously—we had to justify the money spent and I was required to attend management meetings. By the 1990s I was a project manager, a job that involved writing the programme submission along with various milestones. I was in charge of several staff but tried to spend as much time as I could in the laboratory; I also passed on practical expertise to other scientists who were moving into the temperature and radiometry sections.
I received two awards for my work on the laser-beam cryogenic radiometer, and also a monetary award for it. The NPL acquired financial rewards from royalties. I have been an author or co-author of more than 50 scientific papers, and given numerous presentations at national and international conferences. In the early days at the NPL only SOs were accredited for the work, although technicians were usually mentioned in the acknowledgements. From the beginning, Linc Smith always gave me credit for our work, and I was fortunate that this continued throughout my career.
Most of the work I did was what I would call routine, such as calibrating and checking various parts of the radiometer. Almost every time the results were discrepant it was because the instrument needed fine tuning; if there had been serious differences between what the numbers told us and what we expected, then we should probably have had to change a fundamental law of physics. What mattered was understanding what you were supposed to be doing, making the equipment work, and getting results—especially within a time frame. I am not sure why we were successful, whereas others were not. We were dedicated and had some degree of luck, and we received excellent support from workshop and management alike. Finally, I always enjoyed my career at the NPL and was very, very fortunate to work with many fair-minded scientists who taught me a great deal. Without academics, technicians would not exist. In most cases it is the academics who come up with the ideas, but one needs the technical classes to put them into practice.
I would like to thank Rob Iliffe for his help in the preparation of this article.
↵* Both authors can be contacted at firstname.lastname@example.org
- Received November 12, 2007.
- Accepted November 12, 2007.
- © 2008 The Royal Society