Diamond anvil cell, 50th birthday - Compres

High Pressure ResearchVol. 29, No. 2, June 2009, 163–186Diamond anvil cell, 50th birthdayWilliam A. Bassett*Department of Earth and Atmospheric Sciences, Cornell University, Ithaca, NY, USA(Received 4 November 2008; final version received 4 November 2008 )The year 2008 marked the fiftieth birthday of the diamond anvil cell. Its birth took place when Alvin VanValkenburg, while working with his colleagues, Charles E. Weir, Ellis R. Lippincott, and Elmer N. Bunting,first realized that he could look right through one of the diamond anvils and see a sample while it wasat high pressure. In the following years, these scientists and many others adapted the diamond anvil cellto a wide variety of analytical techniques that have provided an impressive amount of information aboutmaterials at high pressures and high temperatures. But, virtually all of those techniques start with lookinginto the diamond anvil cell.Keywords: diamond anvil cell; high pressure; high temperature; history1. Meeting the inventorsThe year 2008 marked the fiftieth birthday of the diamond anvil cell. My colleague, Taro Takahashi,and I were very fortunate to be among the earliest to learn of it and to embrace its extraordinarycapabilities. In the first part of this article, I hope to convey the thrill that we experienced whenwe had the opportunity to see it for the first time and to get to know its remarkable inventors.And then I tell of the many applications of diamond anvil cells that have evolved over the yearsfollowing its invention. In doing so, I try to give a balanced history, but have been able to mentiononly a fraction of the many ingenious and innovative users. I hope I will be forgiven if I seem toemphasize those with which I was involved; this is only because of my greater familiarity withthem.I joined the geology faculty at the University of Rochester in the fall of 1961. When the chairmanasked who I thought we should hire next, I responded immediately with the suggestion of a fellowgraduate student, Taro Takahashi. I liked the kind of research he was doing, subjecting minerals tohigh pressures and temperatures to learn about their properties at conditions in Earth’s interior. Weoffered him the position, and he accepted. He and I set out to lay plans for our joint research evenbefore he made the move to Rochester. He described running experiments in his large tetrahedralpress at Alfred University. They sounded very interesting, but difficult, and together we wonderedwhat new directions we might take. When I was doing graduate work at Columbia University,*Email: wab7@cornell.eduISSN 0895-7959 print/ISSN 1477-2299 online© 2009 Taylor & FrancisDOI: 10.1080/08957950802597239http://www.informaworld.com

164 W.A. Bassettanother student and I designed and built a hot stage for the X-ray diffractometer, so we couldcollect X-ray data from a sample as it was being heated. Now with Taro joining me, I saw anopportunity to make diffraction studies of samples not only at high temperatures but at highpressures as well.The large-volume press Taro was using at Alfred was invented by Tracy Hall, who hadsynthesized diamonds at General Electric. Obviously, quantity matters to someone synthesizingdiamonds. But quantity is not so important to someone wanting to collect X-ray data on asample. While a student, I had learned to use a micro X-ray camera to examine microscopicsamples at ambient conditions. Now I wondered if it could be done on microscopic samples athigh pressures and temperatures. It was just as Taro and I were turning these matters over in ourminds that Taro saw the title of a talk to be given at a meeting in NewYork City – something abouta diamond anvil cell that could be used to subject small samples to very high pressures. Therewas no hesitation; we climbed into my car and drove there so we could hear about this device thatseemed to have exactly the capabilities we were looking for. We were not disappointed. In fact, wewere wowed by what we saw and heard. We asked the speaker, Alvin Van Valkenburg (Figure 1),if we could visit him in his lab at the National Bureau of Standards (NBS) in Washington, DC.Van, as he was known to his friends, was more than agreeable; he seemed to welcome the opportunityto show off the device. Over the next several years, we benefited immeasurably from theenthusiasm and generosity of this extraordinary man and the co-inventors of the diamond anvilcell, Charles E. Weir, Ellis R. Lippincott, and Elmer N. Bunting. It was clear that they had alreadyrecognized the great potential of the diamond anvil cell and were using it to study samples underpressure by infrared spectroscopy and X-ray diffraction (XRD). Taro and I were so impressedthat we asked the NBS group if they would object if we made a diamond anvil cell similar to theone they showed us (Figure 2). In spite of the existence of a patent, they gave their approval.It was Charlie Weir (Figure 3) who in 1958 made the first diamond anvil cell (Figure 4) withthe tools at hand in the NBS laboratory. It is on display in the museum at the National Institutefor Standards and Technology (NIST). But, it was on a day in 1958 that something happenedthat was a true ‘eureka’ moment. Apparently, Van wanted to squeeze a sample between the twodiamond faces in the pressure cell so an infrared beam could be directed through it. However,he was having difficulty and was not sure that the faces of the diamonds were perfectly parallel.So, he placed the cell on the stage of his microscope and looked at the sample through the upperFigure 1. Alvin Van Valkenburg, 1962 (Biographical File, NIST Archives, Information Services Division, NationalInstitute of Standards and Technology, Gaithersburg, MD, USA).

High Pressure Research 165Figure 2. The lever-arm diamond anvil cell that was in use at the National Bureau of Standards when Taro Takahashi andI visited in 1962. Tightening the screw caused the lever to drive the diamonds together with a sample squeezed betweenthem (Courtesy of G.J. Piermarini [9]).Figure 3. Charles Weir, ca. 1960 (Biographical File, NIST Archives, Information Services Division, National Instituteof Standards and Technology, Gaithersburg, MD, USA).diamond. In the next moment, he realized that something very special had just occurred; not onlycould he see that the diamond faces were out of parallel but, more importantly, he could seewhere the highest pressure was because the sample there looked different (Figure 5). Althoughthere had been earlier high-pressure experiments using diamonds, this was such a major turning

166 W.A. BassettFigure 4. The original diamond anvil cell made by Charles Weir in 1958 and now on display at NIST. The ability tolook at the sample through the diamonds was the breakthrough that made this instrument so valuable (Courtesy of G.J.Piermarini [9]).Figure 5. One of Van’s original pictures of CuBr under pressure in the diamond anvil cell. Although the sample he waslooking at that first time in 1958 was not CuBr, it probably looked somewhat like this. The fact that the dark region was notcentered indicated that the anvil faces were not parallel. Nonetheless, it was perfectly clear that the change in appearancewas due to pressure (Courtesy of Robert Hazen).point that it soon came to be considered the moment of birth of the diamond anvil cell. Fromthat point forward, people could see samples under pressure as no one had ever before beenable to see them. As soon as Van had recovered from the surprise of his discovery, he calledthe others in the lab to come and see what he had seen. One by one, they looked in awe at thehigh-pressure sample under his microscope. Other samples caused even greater amazement; somedisplayed incredible color changes with pressure. The ability to see the sample not only made itpossible to check alignment and determine what part of the sample to send a beam through, itprovided one of the most valuable methods for making direct observations on the properties of a

High Pressure Research 167Figure 6. Alvin Van Valkenburg seated at his microscope preparing to examine a sample at high pressures in his diamondanvil cell, ca. 1964 (Courtesy of Robert Hazen and Eric Van Valkenburg).sample, a capability that had never before been possible. Phase boundaries could be observed, colorchanges were immediately obvious, rates of phase transitions could be measured, recrystallizationcould be seen, even birefringence could be observed between crossed polars – and all of this inaddition to the XRD or spectrographic measurements a person might have set out to make.No wonder Van and his colleagues were so enthusiastic. During the ensuing years, Van lookedat many high-pressure samples under his microscope (Figure 6), and when I would see him atmeetings, he would have his photo album under his arm just waiting to show his friends the latestmarvels he had seen.In the early days of diamond anvil cell work, a pile of sample was placed on the face of thelower diamond anvil. Then, the upper anvil was lowered onto it and the sample was squeezedbetween the anvil faces. As force was applied, the sample extruded from between the anvils untilit was so thin that it ceased to flow. In other words, the sample near the edges served as a kind ofgasket. This usually led to the forming of a very smooth pressure gradient from highest pressurein the center to low pressure at the edges, similar to the sample shown in Figure 5. The samplein the center was not only at the highest pressure, but there was also more of it, as the diamondfaces actually became cup-shaped, trapping more of the sample toward the center. The pressuregradient made it possible to observe the properties of the sample over an extensive range ofpressures and to select exactly what portion of the sample an infrared or X-ray beam should bedirected through. It was much later that diamond cell users discovered the disadvantages of thisapproach.While Van was interested mostly in visual observation, Charlie Weir was making XRD patternsof samples under pressure in the diamond anvil cell and Ellis Lippincott was directing research

168 W.A. BassettFigure 7. Miniature, hydraulically loaded diamond anvil cell for X-ray powder diffraction used by Weir and Piermariniin 1960. Hydraulic fluid introduced through fitting (M) into the volume between O-rings (J) forces piston (H) to moveand drive the ∼0.15 ct diamond anvils (F) together. Three steel screws (I) are used for adjusting the tilt of the anvil on theright. X-rays enter through two 0.355 mm collimating pinholes (K) and pass through the sample between the two diamondanvils. Diffracted X-rays are recorded on film (D) behind protective paper (P) (Courtesy of G.J. Piermarini [9]).using infrared spectroscopy. By the time we visited in 1962, Charlie Weir and Gasper Piermarini,who had just joined the group, had developed a new design for use with X-rays. The new cell wassmaller, and the diamonds were driven together by hydraulic fluid (Figure 7).2. Events leading up to the inventionThe instrument that Charlie Weir made and Van looked through in 1958 was not the first versionof a high-pressure device employing diamonds; the idea of using diamonds had evolved throughseveral stages before arriving at the simple version that came to be known as the diamond anvil cell.As early as 1949, Andrew W. Lawson and Ting Yuan Tang of the University of Chicago realizedthat diamond would make an ideal material for containing a high-pressure sample because it is notonly strong but is also transparent to X-rays. They took a diamond, cut it in half, and then clampedthe two halves together. Then, they drilled a hole along the contact between the two pieces ofthe diamond. The hole was just big enough for rods about the size of pins to be pushed in fromboth ends to squeeze a sample between them. The device, which they called the ‘split-diamondbomb’, was able to subject samples to pressures as high as 20 kbar (2 GPa), and very successfullyprovided a clear window for X-rays.John C. Jamieson, a young graduate student at the University of Chicago, joined the effort.He reasoned that a hole drilled through a single diamond crystal would be better yet. He chosea three-carat diamond and asked a Chicago firm to drill a hole right through its middle. We donot know if John knew just how slow a process this was going to be. He had a long wait, but

High Pressure Research 169after 8 months, he had his new high-pressure cell. This cell was indeed better; he was able tosubject samples to 30 kbar (3 GPa) and obtain X-ray data on those samples. John remained atthe University of Chicago as a faculty member in Geology, where he designed and built severalingenious high-pressure devices and contributed many important papers to our science, inspiringthose of us who followed in his footsteps [1].When Van Valkenburg first arrived at NBS and decided to work with Charlie Weir doing thingswith diamonds, they found themselves in a truly enviable situation. Customs officials occasionallyconfiscated diamonds from people attempting to smuggle them into the country. Disposing of suchvaluable confiscated materials could be problematic given rules and regulations. A solution wassimply to make such materials available to people at other government agencies if they couldmake a convincing case for their use. Thus it was that Van and Charlie and others who had joinedtheir group had an almost endless source of what they recognized as one of the most desirablematerials for their needs. They were aware of the devices developed at the University of Chicagoand decided that since they had this extraordinary opportunity, they would take advantage of itand make a diamond cell from an even larger diamond, 7.5 carats, probably worth about a quarterof a million dollars. The hole, 1/4 inch long and 1/64 inch in diameter, took 4 months to drill,a bit faster than Jamieson’s, but an agonizing wait just the same. The General Electric ResearchLab also made a similar cell. Now there was something of a competition under way. This endedrather abruptly for the NBS group, when they pushed their cell a little too far and the 7.5-caratdiamond split wide open.Unfortunately, tensile strength is not one of the desirable properties of diamond; fortunately,compressive strength is. The NBS team decided to look for a new approach to subjecting samplesto very high pressures. Van had followed the work of Percy Bridgman at Harvard years earlierwhen he was a student there. One of Bridgman’s devices consisted of opposed anvils. He hadmade extensive use of the simple principle of squeezing a sample between flat faces on steel ortungsten carbide anvils. To the NBS team it seemed that Bridgman’s anvils might make betteruse of diamond’s properties than a hole through the diamond. Contraband diamonds continued tobe delivered to the NBS team, and it was very fortunate that the properties that make diamondsattractive to the jewellery industry (and, of course, the smugglers) were also the properties mostdesirable for the diamond anvil cell. Even the shape of the brilliant-cut gem made a good anvilwith a relatively minor modification. Van has been known to comment that he owed his career tothe smugglers.Jamieson and Lawson at the University of Chicago had the same idea at about the same time.In 1958, they placed a sample between the flat faces of two diamonds located in a vise thatcould push the diamonds together with the turn of a screw. With this device, they were able tosubject a sample of bismuth to 35 kbar (3.5 GPa) and identify a new high-pressure phase usingXRD. They did this by passing the X-ray beam through the side of the anvils rather than throughthe anvils themselves. If they had chosen to send the X-ray beam through the diamonds andhad looked to see if their anvil faces were parallel as Van did, they would almost certainly havereceived the credit for inventing the diamond anvil cell. Ironically, passing the X-ray beam throughthe side of the cell would several years later prove to yield valuable information on strengthand elastic properties of samples, information not available by passing the beam through theanvils.Once the NBS team realized that they could take advantage of all the desirable properties ofdiamond for achieving high pressures, see their samples, and analyze their samples by a varietyof techniques, the door was thrown wide open for a whole new era in high-pressure research. Ilike to think of the evolution of diamond anvil cell research as a tree with visual observation as thetrunk (Figure 8). Virtually all of the analytical techniques shown in Figure 8 begin with lookingthrough the diamond anvil cell to align the anvils and to position the sample; and many of thosemethods benefit from continued visual observation during analytical runs.

170 W.A. BassettFigure 8. Tree in which each branch represents a technique that has been used to analyze samples under pressure indiamond anvil cells. Visual observation, represented as the trunk, is basic to all of them.3. Properties of diamondCompressive strength, hardness, and transparency to visible light were not the only properties ofdiamond that made it such a good choice for anvils. Diamond is also extraordinarily transparent toparts of the electromagnetic spectrum other than visible. As mentioned above, X-rays representeda portion of the spectrum that offered and still offer one of the most valuable opportunitiesfor studying the properties of samples under pressure; and diamond is very transparent to X-rayshaving energies suitable for determining crystal structure and lattice dimensions. These are two ofthe most important properties of a sample, especially for someone wanting to make measurementsthat can be used for determining basic thermodynamic quantities, such as the effect of pressureand temperature on molar volume. Diamond makes an ideal anvil for X-ray analyses; it has a lowatomic number and therefore a very low absorption. But, there is another aspect that is usuallyless appreciated; gem diamonds are single crystals with a high degree of perfection. As a result, arelatively low-energy monochromatic X-ray beam can pass through a properly oriented diamondwith minimal attenuation due to Rayleigh scattering, if the diamond is oriented so that the Braggcondition is not satisfied for any of the lattice planes or is satisfied for only a few of the latticeplanes. That makes it possible to have far less attenuation than would occur in a polycrystalline

High Pressure Research 171diamond anvil. This feature is especially valuable for studies that require the use of low-energy(long wavelength) X-rays.Infrared spectroscopy can be performed on a sample under pressure in a diamond anvil cell.In fact, it was Ellis Lippincott’s desire to make infrared spectra of high-pressure samples thatmotivated much of the early diamond anvil cell work at the NBS. However, diamonds differsignificantly in their transmission in the infrared portion of the spectrum. In general, the rarertype II diamonds are much more transparent. Various types of visible-light spectroscopy takeadvantage of the clear, colorless transmission by diamonds. Ultraviolet light has been used toexcite fluorescence in samples under pressure. Gamma rays and even protons can pass throughthe diamonds. Because of diamond’s extraordinary elastic properties, sound waves can easily passthrough the anvils to the high-pressure sample.All of these properties of diamond would eventuallyprove to open the way for the diamond anvil cell to provide new and valuable observations.4. Branching into other applicationsAs for Taro and me, after visiting the NBS team and learning of their marvelous diamond anvil cell,we returned to Rochester determined to make our own diamond anvil cells. Taro knew a machinistwho had a shop in a drafty old abandoned factory in Hornell, NY. We wasted no time makingdrawings of the cells we wanted to have him make for us. Within a few months, we had the firstone and could not wait to try it out. It worked like a charm. Silver iodide, one of the favoritesamples to look at, did its thing. It changed from lemon yellow to dark brown with increasingpressure, then back to a lighter yellow-brown at still higher pressures with sharp circular linesmarking changes in refractive index where the sample passed through phase transitions. But wewere interested in making measurements that would give us information about properties at highpressures, which could be used to interpret what goes on inside the Earth. Our first graduatestudent was Ho-kwang (Dave) Mao. He immediately mastered the instrument and grasped itsimportance in learning about properties of minerals for understanding the Earth’s interior. Closeto a week was required to obtain each XRD pattern on a film, so we simplified the design of thediamond anvil cell and had our machinist make more of them. We bought three X-ray machines,each with four ports. There were one or two new films each day. That provided just about the rightlevel of research activity for the students while they took courses. We confirmed John Jamieson’sidentification of the high-pressure phase of iron as hexagonal close packed, and then measuredthe compressibilities of the iron phases. We studied olivine and its high-pressure phase (only onewas known at the time) as well as other minerals pertinent to the Earth’s interior. We could counton just about everything we did being new. Dave satisfied his PhD requirements in record timeand earned his degree but went right on collecting data. A position at the Geophysical Laboratoryin Washington, DC seemed like just the right place for him to find the freedom to pursue morehigh-pressure research. Since his arrival there in 1968, diamond anvil cell research has flourishedand many young trainees have gone on to establish their own high-pressure labs. Dave remainsthere today as one of the most active and successful diamond anvil cell users in the world. Hisaccomplishments during his stay at the Geophysical Laboratory are numerous.5. Gaskets, a big step forwardA few years after Taro and I made our own diamond anvil cells, I paid Van another visit. Thistime he took me to his home. He had built a sun room above the garage, which served as aninformal lab where he searched for new wonders that could be observed in his cells. He sat me

172 W.A. Bassettdown in a comfortable chair while he worked at a microscope on a large handsome walnut desk.He dropped a tiny cleavage fragment of calcite into the hole of a metal gasket perched on theanvil face of the lower diamond. He then dropped some glycerin into the hole, so the calcitecrystal was free to swim around. When he positioned the upper diamond so that it trapped theglycerin and calcite rhomb in the hole, he was able to apply pressure simply by turning a screw.When he was satisfied he had turned the screw far enough, he called me over to peer into themicroscope. There the calcite crystal sat with a sharp line running across its middle (Figure 9).‘That’, he told me, ‘is the boundary between calcite and its high-pressure phase, calcite-II. Nowturn the screw’. As I gingerly did his bidding, I could see the boundary drift across the crystal.Needless to say, I was amazed. I had no idea that a crystal could go through a phase transitionand remain intact. He showed me several other wonderful things that day, such as turning waterto ice at room temperature (Figure 10), but the calcite is the one that stuck in my mind; the crystalstayed intact; it was so incredibly beautiful.It was not just the sight of the calcite rhomb staying intact through the phase transition thatso impressed me that day. It was also the ease with which Van had encapsulated a fluid, makingFigure 9. One of Van’s original pictures of a calcite rhomb under hydrostatic pressure in his gasketed diamond anvilcell. The very sharp, straight line running from southwest to northeast is the phase boundary between calcite-I and thehigh-pressure phase, calcite-II. It was this boundary that drifted across the crystal as pressure was increased (Courtesy ofRobert Hazen).Figure 10. One of Van’s original pictures of a perfect octahedron of ice VI produced by applying pressure to water in agasket in the diamond anvil cell at room temperature (Courtesy of Robert Hazen).

High Pressure Research 173it possible to observe the calcite rhomb under perfect hydrostatic pressure. Van’s use of gasketsespecially for encapsulating fluids was almost as important a contribution as the invention ofthe diamond anvil cell itself. During the following years, there were to be many other lines ofresearch taking advantage of it. Although providing a hydrostatic pressure medium is certainlyone of the most valuable, other applications include viscosity measurements, chemical reactions,structure determinations of molecular species in solution, and eventually a way to reach the highestpressures achieved in the diamond anvil cell.6. Single-crystal studies at high pressureThe first to benefit from the fluid encapsulation technique were Van’s close colleagues, CharlieWeir, Gasper Piermarini, and Stan Block. Using a precession camera, they pursued single-crystalXRD studies of numerous high-pressure solids produced by squeezing liquids to grow singlecrystals at room temperature in the diamond anvil cell. Many of these were common liquids suchas water, ethanol, benzene, carbon tetrachloride, and toluene. They called the long ethanol crystalstheir ‘gin-sickles’ (Figure 11) and wondered if they were subject to the same laws that governedethanol in its liquid form.In addition to studying the structure of the crystalline form of these liquids, the NBS group alsoused the method to measure, for the first time, anisotropic compressibilities of several energeticmaterials (inorganic azides) by encapsulating single crystals of these compounds in an inerthydrostatic liquid medium. The single crystal X-ray data were obtained at the known freezingpressures of the various liquid media used. Such data had never been collected before to suchextreme pressures. This was a new and very important application of the diamond anvil cell. Once,when I visited them, they told me of their azide studies. At the time, I did not really appreciatethe importance of what they were doing. Prior to that visit, I had assumed that the only value ofthe very small sample size in diamond anvil cells was simply to allow users to go to very highpressures with small (and not so expensive) diamonds. But because the samples in the diamondanvil cell are so small, explosive materials can be safely worked on without the risk of blowing upFigure 11. One of Van’s original pictures of crystals of ethanol produced by pressure in the diamond anvil cell at roomtemperature. He called these crystals his gin-sickles. The colors result from the birefringence of the crystals as viewedbetween crossed polars in a polarizing microscope (Courtesy of Robert Hazen).

174 W.A. Bassettthe cell or the researchers or the lab, a very real concern for those using large-volume high-pressuredevices. This continues today to be one of the important benefits of diamond anvil cell studies.But, it was the potential of being able to drop a single crystal into a liquid and make it passthrough a phase transition while remaining intact that most impressed me. In 1969, I invited LeoMerrill of Brigham Young University to join me at the University of Rochester for graduate workin high-pressure studies. He too was fascinated by the way a calcite crystal could survive a phasetransition and remain a perfect single crystal. His thesis research was first to design a simplediamond anvil cell small enough to be mounted on a single-crystal X-ray goniometer and thento collect single-crystal XRD data that could be used to solve the structure of a high-pressurephase. The structure of calcite-II was not known and so it was the obvious choice for a sampleto study. There was a new diffractometer suitable for single-crystal analysis on the University ofRochester campus that would serve the need well. Computerized diffractometers did not yet exist,and so all the settings and readings had to be made by hand. Although Leo’s determination ofthe structure of calcite-II was an important contribution, the simple little diamond cell consistingof two triangular platens with the diamond anvils mounted on beryllium windows seated in theplatens that were pulled together by three screws and guided by rods, has proved to be one ofthe most popular diamond-cell designs over the years. The beryllium supports allowed X-rays topass through the sample over a wide range of angles, a capability very important to single-crystalXRD. The principle of this cell later morphed into a four-screw system and even grew in size tobecome the basis for the hydrothermal diamond anvil cell (HDAC) many years later.Leo’s little diamond anvil cell was put to very good use by Bob Hazen and Larry Finger at theGeophysical Lab. I had met Bob Hazen while he was still a graduate student in the Geology Departmentat Harvard. He was fascinated by Leo Merrill’s little diamond anvil cell and asked me to makea couple for him. When he moved to the Geophysical Laboratory after receiving his PhD degree,he joined forces with Larry Finger. Together, they made many important XRD studies of singlecrystals at high pressures. The most remarkable undertaking was the determination of the crystalstructure of solid hydrogen at 50 kbar (5 GPa). This was truly a tour de force, as hydrogen is theweakest scatterer of X-rays from among all elements and compounds. In 1993, Bob Hazen wrotea book entitled ‘The New Alchemists’ [2]. Although the first part of the book tells the fascinatingstory of diamond synthesis, the final third offers an excellent history of the diamond anvil cell.7. Measuring pressure7.1. X-ray diffractionDetermining pressure in the diamond anvil cell or in any other high-pressure apparatus, for thatmatter, is of critical importance. Those of us using the diamond anvil cell for XRD studies wereable to measure the lattice parameters of some calibrated crystalline material mixed with oursamples. That was usually NaCl, although other compressible materials were often used as well.This, of course, required an accurate compressibility scale for the material serving as a calibrant.Fortunately, in 1968, Dan Decker of Brigham Young University, working from first principles,provided the most reliable NaCl compressibility scale at that time. That worked well for those ofus who were using XRD but left others without an easily applied method for determining pressure.Making an XRD pattern just to determine pressure was inconvenient to say the least.7.2. Birth of the ruby methodIt was in 1971 that NBS Division Chief Jack Wachtman urged the high-pressure group to finda more spectroscopic way to measure pressure. The high-pressure group at that time was a very

High Pressure Research 175strong team, consisting of NBS scientists Stanley Block, Gasper Piermarini, and Richard Forman,plus J. Dean Barnett, who had decided to take his sabbatical leave from BrighamYoung Universityto work at the NBS. They put their heads together to pursue the suggestion Wachtman had made.The result was the ruby method for determining pressure. A small chip of ruby placed in a samplein the diamond cell can be made to fluoresce. They found that the wavelength of the fluorescenceemission changed with pressure and could easily be calibrated against something like Dan Decker’sNaCl scale. The ruby method had another very important benefit recognized by the high-pressuregroup at NBS. There are two closely spaced peaks in the ruby emission spectrum. The depth of thetrough between the two peaks is very sensitive to peak broadening. Peak broadening, in turn, isvery sensitive to deviatoric stress or non-hydrostaticity. In other words, a quick check of the rubyspectrum could indicate immediately whether the pressure medium was continuing to act as afluid. It could even give an indication of how soft a solid the medium had become if it did solidify.7.3. Ruby method and extremely high pressuresIn the 1970s, there was a competition to see how high a pressure could be achieved in thediamond anvil cell. There were claims, there were counterclaims, and there were skeptics. DaveMao and Peter Bell of the Geophysical Laboratory realized that in order to convince the restof the high-pressure community of the high pressures they had reached, they would need veryconvincing evidence. They turned to the shock-wave work of John Shaner and Daniel Steinberg,who measured the compressibilities of Cu, Mo, Pd, andAg.Although shock-wave experiments cangenerate extremely high pressures, they are challenging because samples are at those pressures foronly a millionth of a second or so. However, the remarkable thing about shock-wave measurementsis that during that very short time, it is possible to collect data which can be used to make anindependent calculation of density versus pressure. As a result, Dave and Peter had a way todetermine static pressures in the range of megabars (hundreds of GPa). They could measurethe lattice parameters of the metals by XRD and then, using the compressibilities measured byshock-wave experiments, calculate the pressures. It was Dave who reasoned that a metal gasketcontaining a sample between diamond anvils might serve an even more important function thanjust containing the sample. As the metal of the gasket extrudes from between the anvils, it formsa kind of belt, imparting to the device the characteristics of the belt apparatus that Tracy Hall hadused so successfully to synthesize the first diamonds at General Electric. Dave found that addinga bevel to the diamond anvils could enhance this effect, leading to a substantial increase in thepressures that could be achieved with the diamond anvil cell. Dave and Peter and their colleaguesreported 1.2 Mbar in 1976, 1.5 Mbar in 1979, 2.5 Mbar in 1985, and 5.5 Mbar in 1987. They wereable to extend the ruby scale to these megabar pressures by mixing ruby chips with the metalsthat had been measured by Shaner and Steinberg.The story of the early development of the diamond anvil cell and especially the ruby method istold on the NIST website (http://nvl.nist.gov/pub/nistpubs/sp958-lide/100-103.pdf). The rubymethod has been so successful that it continues to be the favored way to measure pressure in thediamond anvil cell even for XRD studies. Although it remains a secondary method, i.e., one thatmust be calibrated, it has been greatly improved and extended over the years as increased intensityof lasers has allowed ever smaller ruby chips to be used and better primary standards have yieldedmore reliable sources of calibration to much higher pressures. Karl Syassen gives a very detailedanalysis and review of the ruby method [3].The ruby method does not work well for measuring pressure at high temperature because thefluorescence signal fades with temperature. For those using the diamond anvil cell for XRD,there are pressure markers for which the effects of both pressure and temperature on latticedimensions are well known. Gold is often favored for this purpose because it is chemically inert

176 W.A. Bassettand because it is an efficient scatterer of X-rays. The search for even more fundamental and moreaccurate methods for measuring pressure and especially pressure at high temperature remains avery active area of investigation today.8. High temperature in the diamond anvil cell8.1. External resistance heatingOne of the earliest additions to the diamond anvil cell was the ability to raise the temperature ofa sample while under pressure. Even in the very early days, Van occasionally heated his diamondanvil cell to observe the effect of temperature on crystals as they grew from a liquid. We, aswell as the NBS group and Roger Burns’ group at MIT, found that we could heat our sampleswith a cylindrical heater placed around the anvils. Later, we found that wrapping resistancewires around the diamond supports worked just as well, was less interfering, and could routinelyprovide temperatures up to 1000 ◦ C. In this temperature range, a thermocouple attached to eachdiamond is the most satisfactory way to measure temperature. The fact that diamond is the bestthermal conductor known makes it possible to obtain accurate sample temperatures by attachingthermocouples to the outside surfaces of the diamond anvils. This approach is still in use andis the basis for the HDAC that I describe below.8.2. Laser heatingStill higher temperatures can be reached with laser heating. One day in 1966, while we werehaving lunch, Taro Takahashi said to me, ‘why not use a focused laser beam to heat our sampleswhile they are under pressure in the diamond anvil cell?’ Lasers were still very new at that timeand our first attempt at laser heating employed a pulsed ruby laser. With it we were able toconvert graphite to diamond and we were able to drive other phase transitions in samples wheretemperature was needed to overcome kinetic barriers. However, control was not this laser’s strongpoint. These experiments worked best when dark-colored particles of the sample were surroundedwith a transparent, insulating medium such as NaCl. We eventually switched to aYAG laser, whichemitted a continuous infrared beam with λ = 1.06 μm. With it, my student, Li-chung Ming, wasable to show that the mineral fayalite breaks down to mixed oxides at pressures higher than thestability field of the spinel phase. This laser could also be run in Q-switched mode, which producedincredibly intense pulses. I once measured a temperature on the order of 7000 ◦ C, an impressivetemperature but not very useful when one considers that the duration was only about 16 ns and thetemperature was more an estimation than a measurement. But the Q-switched pulses were intenseenough to melt the surfaces of our diamond anvils when pressure was high enough to preventgraphitization. Both continuous and pulsed operations had their good and their bad points.Temperature measurement was a definite challenge. We started by using a handheld opticalpyrometer to look through the microscope at the intensity of the light from the incandescentsample. However, it quickly became clear that the color of the incandescent light would be amuch better indication of temperature. And so, we added a spectrometer that could plot the entirevisible spectrum which we could then match to published blackbody spectra. At the same time,Raymond Jeanloz and his students at UC Berkeley were developing even better ways to utilizeblackbody radiation for determining temperature and were applying the method to a variety ofquestions about the Earth’s interior. Laser heating, while capable of very high temperatures, hasthe disadvantage of creating non-uniform temperatures and very steep thermal gradients. It wasthe Geophysical Lab group that devised the double-sided laser heating approach to give a more

High Pressure Research 177uniform temperature distribution from anvil to anvil. They also found that hitting the samplewith a laser beam produced by a laser mode or combination of modes with a larger diameterwould greatly improve the uniformity of the temperature across a heated spot. And, of course,having a more powerful laser helped a lot. There have also been important advances in minimizingtemperature fluctuation during continuous laser heating. These improved methods of heating bylaser are now in daily use in many labs.8.3. Internal resistance heatingAnother important way to heat samples under pressure in a diamond anvil cell is by internalresistance heating. If a person is lucky enough to want to study a sample that is electricallyconducting, the sample itself can be made to serve as the heater. That was the case in 1974 whenour student, Lin-gun (John) Liu, wanted to make measurements on pure iron at simultaneoushigh pressure and temperature. We selected the finest iron wire we could find and made it eventhinner. We ran the wire between two large electric leads and surrounded it with Al 2 O 3 or MgObetween the diamond anvils. When we used 60-Hertz AC to heat the sample, we learned justhow severe the heat loss was. The sample flickered in a most disconcerting way. Not wantingthe sample to pass through its phase transitions 120 times a second, we quickly decided on a carbattery as our power source. That worked well, and we were able to collect valuable data on thephase boundaries and melting curve of iron. John went on to a position at the Australia NationalUniversity in Canberra, where he worked with Ted Ringwood, and was the first to discover thatthe most abundant iron-magnesium silicate phase at pressures and temperatures of the Earth’slower mantle has the perovskite structure. In fact, his discovery led to the remarkable conclusionthat more than half of the Earth consists of this silicate perovskite phase.Heating of a sample by an internal resistance heater has been a specialty of Chang-Sheng Zhafor many years. He had in mind those not so lucky as to have samples that are electrical conductors.His idea was to use a thin metal strip with a hole in it to receive a solid sample. He started workon the idea in 1986 while working with Reinie Boehler at UCLA. It worked and he was able tosubject samples to impressive temperatures and pressures. He returned to his fascination with thisapproach years later while a beamline scientist at the Cornell High-Energy Synchrotron Source(CHESS). I helped him by laser machining a fine rhenium strip with a hole only microns acrossto receive the sample. By then, CHESS was able to produce an extremely small, intense X-raybeam. This allowed him to pass a beam through the very small sample and obtain X-ray data onit. We used it to refine the equation of state (EOS) of platinum, among other things. This is adifficult experimental procedure to apply, but has great potential. I have reviewed this plus othermethods of producing high temperatures in the diamond anvil cell in an encyclopedia article onthe subject [4].9. The pressure environment9.1. The need for hydrostatic pressureThere are many parameters to consider when doing high-pressure research. In addition to the obviousones such as pressure and temperature and molar volume, there are such important parametersas deviatoric stress or its converse hydrostaticity. Deviatoric stress defined, as anisotropic forceswithin a sample, and deviatoric strain, defined as anisotropic displacements within a sample, arefeatures that are of great concern to high-pressure experimentalists. Many of the early XRD measurementsmade on ungasketed samples had systematic errors due to the presence of a deviatoric

178 W.A. Bassettstress. It was soon realized that zero or minimal deviatoric stress was desirable because pressure,i.e., isotropic stress, is a thermodynamic parameter. Therefore, it is no surprise that eliminatingor minimizing deviatoric stress became an important objective. For modest pressures like thatrequired for the calcite phase transition that Van showed me, there are many liquids available. Butas experimentalists desired to go to higher pressures and still have their samples under hydrostaticpressure, much attention was paid to the pressure media used. One of the most successful liquidswas, and still is, a mixture of methanol and ethanol in a ratio of 4:1. This allowed users to takehydrostatic samples up to 90 kbar (9 GPa). However, the upper limit seemed to vary, dependingon the time of year and how long the alcohol mixture sat on the shelf. It was discovered that waterwas sometimes contaminating the mixture and had the effect of increasing the pressure range. Ithas now been determined that a 16:3:1 mixture of methanol:ethanol:water remains truly hydrostaticto 105 kbar (10.5 GPa). But, there was a need to extend the pressure range to much higherpressures. Dave Mao and Peter Bell at the Geophysical Lab realized that the noble gases were theanswer, not because they remained gas or liquid, but because they became very weak solids at highpressures. Argon and neon have the added advantage that their solids yield diffraction patternsthat can be used for pressure measurement. But, helium was the most nearly hydrostatic. Gasloading of samples into diamond anvil cells became an important way to obtain nearly hydrostaticmeasurements to very high pressures. Cold liquid argon could be loaded into a diamond anvil cellby immersing the cell in a Dewar of liquid argon and then closing it there. However, any noblegas can be loaded by placing the cell in a pressurized tank of liquid noble gas and then closingthe cell there. This method, developed by Dave Mao and Peter Bell, has been very successful,but its large volume posed a far greater danger to its users than the extremely high-pressure buttiny samples enclosed in the cell itself. High-pressure gas loading apparatuses have now becomemuch safer, more versatile, and easier to use.9.2. The uses of non-hydrostatic pressureIt was in the early 1970s that Gary Kinsland joined our group as a graduate student. He becameinterested in the possibility of passing an X-ray beam through the side of an ungasketed sample ina diamond anvil cell, much as Jamieson and Lawson had done in 1958. This called for a diamondanvil cell with large openings on the sides. The X-ray beam passed through the entire range ofpressures in the polycrystalline sample, but we used only those scattered by the highest pressureportion of the sample. X-rays scattered by lattice planes perpendicular to the compression axisrevealed the smallest d-spacings, whereas those parallel to the compression axis revealed thelargest d-spacings. In other words, we were able to measure the deviatoric elastic strain as afunction of pressure. From that we could calculate the deviatoric stress and strength as a functionof pressure. Many years later, Singh, Mao, and others at the Geophysical Lab applied this samegeometry in a very novel way to determine elastic moduli in iron at pressures comparable to thosein the Earth’s inner core. I describe the bad and the good aspects of deviatoric stress in an articletitled ‘Deviatoric Stress, a Nuisance or a Gold Mine?’ [5].10. Chemical reactions in the diamond anvil cell10.1. HydrationThe Geophysical Lab occupied a massive old masonry building designed specifically for researchin geophysics. It sat only a few blocks from the NBS, and so it is not surprising that Van spenttime collaborating with people at the Geophysical Lab. One of the early papers to come from the

High Pressure Research 179Figure 12. One of Van’s original pictures of ikaite crystals precipitated from water in which calcite had been dissolvedunder pressure. The intense colors result from the birefringence of the crystals as viewed between crossed polars in apolarizing microscope (Courtesy of Robert Hazen; color online).collaboration concerned a high-pressure hydrous phase of calcite, the mineral ikaite. They foundthat they could dissolve calcite in water under pressure and then precipitate crystals of ikaite(Figure 12). This was the first time that a reaction between a solid and a liquid had been studiedin the diamond anvil cell. Oddly, many years passed before other reactions involving water werestudied using the diamond anvil cell. My student, Anat Shahar, returned to the study of ikaite in2005 and extended the range of pressure and temperature.10.2. High-temperature study of fluids in the diamond anvil cellThe ability to heat aqueous samples at high pressures has made it possible to study hydrothermaland magmatic processes in the diamond anvil cell. Andy Shen, a graduate student at Cornell,and I developed a version of the diamond anvil cell in the early 1990s with the help of I-MingChou of the US Geological Survey. We called it the hydrothermal diamond anvil cell (HDAC).It consists of diamond anvils supported on tungsten carbide seats that are wrapped with heaterwires. A liquid can be placed in a gasket hole between the diamonds and subjected to pressure andtemperature simultaneously. If care is taken, the sample chamber volume can be made to remainessentially constant. This allows us to increase the pressure along an isochore as the temperatureis increased. The isochore can be chosen simply by choosing the size of a bubble included with theliquid. A solid sample can be included as well. If the volume remains constant, the sample can beheated until the bubble disappears. That temperature, known as the homogenization temperature(T h ), defines the bulk fluid density of the sample. Further temperature increase results in pressureincrease because of the constant confining volume; i.e., the sample passes up along an isochorein the P –T space. Temperatures up to 1000 ◦ C can be reached. As in Van’s original discovery,it is the ease with which we can observe the sample that allows us to make observations, whichcan then be used to characterize the sample and map phase relations. I can well remember theexperience of seeing white-hot water. Somehow, it had never occurred to me that that would bepossible. In these experiments, we use the P –V –T EOS of the fluid (usually water or an aqueoussolution) to determine pressure. That is accomplished by determining density from T h and thetemperature at which an observation is made. Knowing the temperature, density, and EOS, wecan then solve for the pressure. Once the visual observations have been made, we can turn tothe other techniques, such as Raman spectroscopy, visible-light spectroscopy, X-ray fluorescence

180 W.A. Bassettspectroscopy (XRF), and X-ray absorption spectroscopy (XAS). More information on the HDACcan be found in a review paper [6].11. Continued research by the foundersIt is interesting that Van and his colleagues devised so many of the diamond cell applicationsthat have gone on to become routine in the arsenal of high-pressure research the world over. Twogood accounts of the early work at the NBS can be found in papers by Piermarini [7,8]. In 1961,Van and his colleagues set up a small business to manufacture diamond anvil cells. They calledit High-Pressure Diamond Optics, HPDO. Oddly, forensics experts have been among the mostactive customers to use the diamond anvil cells made by HPDO because they can use the cells tohold their samples at very uniform thickness, good for making spectroscopic analyses at ambientpressures.WhenVan was asked in 1964 to serve as a Program Director at the National Science Foundation,he accepted, and for 6 years contributed in that capacity. As a program director, he made surethat good ideas for experimental mineralogy and petrology received funding. In 1970, he took onresponsibilities at the US Bureau of Mines as Head of the Office of University Relations. From1974 to 1980, Van spent much time at the Geophysical Lab collaborating with the high-pressurescientists there. In 1980, he moved to Tucson and took HPDO with him. His son, Eric, joinedhim in the business and together they made sure that there would always be good diamond anvilcells available for those wanting to start high-pressure labs of their own. Van’s love of exploringphenomena in the diamond anvil cell continued until his death in 1991. Since his father’s death,Eric has continued to run HPDO, supplying instruments to a wide range of users around the world.The others in the original NBS group followed NBS as it moved to Gaithersburg, MD, in the early1960s, and in 1988 changed its name to NIST. Charlie Weir retired in 1970. Stanley Block, GasperPiermarini, Richard Forman, and others continued to develop a number of novel diamond anvilcell techniques and apply them to important problems.12. The many applications of the diamond anvil cellThe remarkable simplicity and versatility of the diamond anvil cell has led to a proliferation ofanalytical techniques that can be applied to samples at high pressures. Two gem diamonds pushedtogether with a sample between the flat faces on the diamonds; what could be simpler? Becausepressure is defined as force divided by area, very large pressures can be generated between verysmall faces with relatively modest forces. As a result, diamond cell users have ventured to everhigher pressures by reducing sample sizes. Users have found ingenious ways to design diamondanvil cells to permit analyses of these very small samples; and the ever increasing brilliance(greater intensity in smaller beams) of radiation sources has been of great benefit in adaptingdiamond cells to many of the techniques illustrated in Figure 8.12.1. Raman scatteringOne of the most valuable analytical tools has been visible-light Raman spectroscopy. It utilizesfrequency shifts in scattered light that result from interaction with vibrational frequencies in thesample. Because it detects vibrational frequencies rather than atomic spacings, it is sensitive tomany species, such as carbon dioxide, methane, and ammonia, which X-rays are poorly suited todetect. In addition to identifying molecular species, Raman spectroscopy has been used to measure

High Pressure Research 181changes in molecular vibration frequencies as a function of pressure, thus providing informationon the nature of bonding and structural change. One of the earliest uses of Raman spectroscopyfor this purpose took advantage of a fine spectrometer that was handmade by Edward Brody atthe University of Rochester. Ed was in the Department of Optics, and occupied a lab two floorsbelow mine. He had designed and built the five-pass Fabry–Perot spectrometer with the bestspectral discrimination available at that time. In 1977, Tom Lettieri, Ed’s student, and I chose tostudy sodium nitrate because its nitrate groups in the high-pressure phase were known to have alibrational vibration mode that should drop to zero when the nitrate groups became locked in oneorientation at the transition to the low-pressure phase. Our series of Raman spectra taken as weslowly decreased the pressure to the phase transition beautifully showed the mode softening andthe eventual decrease of frequency to zero.12.2. Brillouin scatteringEd’s state-of-the-art spectrometer was even better suited for using Brillouin scattering of visiblelight to measure phonon velocities in single-crystal samples. There seemed to be no reason whythis could not be done in a diamond anvil cell if a light beam were to enter through one diamond andexit through the other. Amazingly, our calculations showed that if the impinging beam entered onediamond at exactly 45 ◦ to all the interfaces, and the scattered light emerged from the other diamondat exactly 45 ◦ to all the interfaces, no correction for refractive indices would be required. A fewextra holes in the diamond cell did the trick, and we were able to use the phonon velocities tocalculate elastic moduli in a single crystal at pressure. Ed’s student, Charles Whitfield, madethe measurements that proved the method could be applied to individual crystals of NaCl inthe diamond anvil cell; soon others adopted the approach. Hiro Shimizu, Ed’s visitor, appliedthe method to the mineral forsterite, as well as fluid hydrogen and deuterium. Don Weidnerat the New York State University at Stony Brook made excellent use of this technique, as didChang-Sheng Zha at the Geophysical Laboratory and Jay Bass at University of Illinois, Urbana-Champaign.12.3. ReflectivityOther optical phenomena have been measured in diamond anvil cells. Reflectivity, in particular,has found an interesting application. Karl Syassen and others showed that phase transitions inmetals and other opaque samples can be observed in reflected light. Chang-Sheng Zha mademeasurements of the reflectivity of the ice–diamond interface in the diamond anvil cell at veryhigh pressures as a way to measure the refractive index of the ice as a function of pressure. It wasvery successful and is a technique that has considerable promise for future studies.12.4. Electrical measurementsA few years after its invention, the diamond anvil cell proved to be a valuable device for makingelectrical measurements. It was challenging to introduce electric leads into a sample, especiallywhen a metal gasket was used. A very ingenious way to accomplish this was developed byYogeshVohra at the University of Alabama and Sam Weir at Lawrence Livermore National Laboratory.With improvements to chemical vapor deposition as a way to synthesize a single-crystal diamond,they found that they were able to imbed electric leads in a diamond anvil by depositing the leadson the diamond surface and then covering the leads by vapor depositing a fresh layer of diamondon top of them. They dubbed these their ‘designer diamonds’.

182 W.A. Bassett12.5. Synchrotron radiation – X-ray diffractionThe advent of synchrotron radiation as an experimental tool had a profound effect on diamondanvil cell research. The diamond anvil cell and synchrotron radiation were made for each other. Iwas very fortunate to be moving to Cornell University in 1978 just as the synchrotron there wasbeing modified to accommodate CHESS. In the very early days before CHESS was christened asa general user facility, Art Ruoff of the Cornell Department of Materials Science and his studentsalternated with me and my students learning how best to collect XRD data on samples underpressure in our diamond anvil cells. Runs were difficult, and each little success was grounds forcelebration. The staff at CHESS were wonderful in helping us set up and run experiments. Artand his students pushed to very high pressures and discovered many new extraordinary propertiesof materials. We concentrated on applying pressure and temperature simultaneously. Bit by bit,we added new capabilities until we were able to change both pressure and temperature withoutopening the hutch door.A high-energy synchrotron source produces a very intense, very small, and well collimated beamof X-rays. The beam may consist of white X-rays having a broad spectrum or monochromaticX-rays having essentially a single energy. Both have been valuable for diamond anvil cell studies.Some of the earliest studies of polycrystalline samples utilized the white beam and produceddiffraction patterns by the energy-dispersive method, i.e., analyzing energies (thus wavelengths)of X-rays scattered in a specific direction. The extraordinarily rapid data collection that resultedopened the way to making real-time diffraction observations of samples as pressure and/or temperaturewere changed. That is why we concentrated on being able to change both pressure andtemperature remotely. However, resolution was never as good as what could be obtained withangle dispersion when monochromatic X-rays were employed.Dramatic improvements in area detectors that replaced the old film methods of thepre-synchrotron era began to catch up with the energy-dispersive method in speed while offeringsuperior resolution. Computer processing of data collected by area detectors pioneered byShimomura and his colleagues at Photon Factory in Tsukuba, Japan, in the early 1990s addedto the improved performance of area detectors. One of the most valuable attributes of computerprocessing, developed by Richard Nelmes and his colleagues at the University of Edinburgh,was the ability to integrate the intensities around each diffraction ring, greatly improving thesignal-to-noise ratio and the accuracy of intensity measurements even when polycrystalline sampleswere coarse-grained and orientations of sample particles had less than desirable randomness.Structural analysis of polycrystalline samples by the Rietveldt method became an important newtool. Nelmes and his colleagues greatly improved opportunities for taking advantage of these newtechniques by making changes to the diamond anvil itself. At that time, nearly all diamond anvilcell powder diffraction data were collected through a slot in the seat supporting the down-streamdiamond anvil because diffraction data were collected by energy dispersion or because we stillthought of powder diffraction as linear, as in the Debye–Scherrer camera. The Edinburgh groupsaw the value of being able to collect data from the complete diffraction cones for their experimentsat the Daresbury synchrotron source. They used wide conical-aperture cells, starting withMerrill–Bassett cells, in which the diamond anvils were mounted on beryllium seats. However,pure beryllium supports lacked the strength for very high pressures, and so the Edinburgh groupset out to find a stronger beryllium alloy still sufficiently transparent to X-rays. This paid off andeventually led them to achieve pressures as high as 700 kbar (70 GPa) while still providing forlarge cone angles. Subsequently, an even more successful way of achieving high pressures withlarge dispersion angles was designed by Reinie Boehler, now of the University of Mainz. In thisdesign, known as the Boehler–Almax design, the diamonds are supported by a conical surface ina tungsten carbide seat, thus freeing the entire table face for scattered X-rays to exit. This methodcan provide a dispersion cone up to 90 ◦ and megabar (100 GPa) pressures.

High Pressure Research 183The type of diamond anvil cell known as the panoramic cell also provides for very large coneangles. In this cell, the X-ray beam enters through the edge of the sample or through a berylliumgasket containing the sample. Many interesting high-pressure studies have been conducted withthis geometry, but the user needs to be aware of possible effects of deviatoric stress and pressuregradients, as well as interference from the gasket. The panoramic cell is also used for applicationsother than diffraction.12.6. Synchrotron radiation – X-ray fluorescenceAlthough XRD has been the most extensively utilized method at synchrotron sources, it is farfrom being the only one. A popular analytical technique is XRF. It can be used to chemicallyanalyze individual phases within a diamond anvil cell sample by exciting fluorescence emissionsfrom the elements within individual phases as the X-ray beam is directed through them. This isa valuable technique for studying phenomena such as partitioning of elements between phases atvarious pressures and temperatures.12.7. Synchrotron radiation – X-ray absorptionXAS utilizes the structure of the absorption spectrum near the absorption edge of an element. It isan important source of information on the short-range order of atoms surrounding the atom thatis absorbing the X-rays. It has proved valuable for better characterizing amorphous materials andmolecules in solution. XAS has been an exceptionally valuable tool for the study of structure andspeciation of complexes in solution at elevated pressures and temperatures in the HDAC. It is themain tool used by my colleagues, Alan Anderson, I-Ming Chou, and Bob Mayanovic, and myself.For our early XAS experiments, we drilled holes part way through our diamonds to reduce to300 μm the amount of diamond that the X-rays would have to pass through as the beam enteredthrough one diamond and exited through the other. This allowed us to use the absorption edgesof low-z elements such as iron and other first-row transition elements. The spectra we collectedyielded the amount of absorption as a function of energy as we scanned across the absorption edgeof some element in our sample. This worked well; however, it soon became clear that there was abetter way. Measuring the intensity of the fluorescence emission as a function of impinging energycould provide a more sensitive way to measure absorption, and it could be done at 90 ◦ to the beamand in the plane of polarization, thus eliminating Rayleigh scattering from the sample and makingit possible to reduce the amount of diamond the X-rays had to traverse to only 160 μm. In ourHDAC, the lower diamond anvil has a recess 300 μm in diameter in the center, and thin wallswhere the X-rays enter and exit. We reduced the wall thicknesses until we were afraid we couldnot get away with anything less without risking failure. An aqueous solution, with or withouta solid present, can be loaded into the recess. A 50-μm thick metal gasket with a 300 μm holeplaced over the recess provides a seal. Recently, however, we learned that we can run hydrothermalsamples without a metal gasket. This eliminates both contamination of the sample by the gasketand deviation in sample volume caused by gasket deformation. The very constant volume of therecess makes it possible for us to accurately control and measure bulk fluid density and pressure asa function of temperature by the means described elsewhere in this paper. The very thin (80 μm)diamond walls allow us to analyze elements having absorption edge energies

184 W.A. BassettWe have found that setting up a video camera in the experimental hutch so that we can observeour sample continuously during a run is extremely valuable, as it allows us to measure the temperatureat which a bubble vanishes or a solid dissolves or a crystal forms. And, best of all, theuse of very fluorescent diamonds makes it possible for us to see the X-ray beam, thus making iteasy to aim the beam right where we want it to go. This is one more example of the extraordinaryimportance of Van’s discovery 50 years ago – that visual observation of a sample in the diamondanvil cell can be made so easily.12.8. Synchrotron radiation – inelastic scattering and X-ray RamanNew developments just within the past few years have now made spectroscopy at much lowerenergies possible. Inelastic scattering and X-ray Raman take advantage of the principle that lowenergyevents can be probed by measuring energy differences between impinging and scatteredphotons. In other words, the X-rays that enter and exit a diamond anvil cell can have energiesfor which diamond is very transparent, while the energy of the event, be it absorption or someother low-energy phenomenon, is measured by the difference in energy between the impingingand scattered photons. These techniques require extraordinary resolution employing technology,only recently available at synchrotron sources.The application of Mössbauer spectroscopy to the study of iron-bearing phases under pressurein a diamond anvil cell was pioneered by Roger Burns and his colleagues at MIT in the 1970s.Those early studies utilized the highly monochromatic gamma radiation originating in the nucleusof 57 Fe, and were therefore valuable for the study of iron-bearing minerals. Fortunately, mostof the phases important to understanding the Earth’s interior do contain iron. Other isotopessuitable for Mössbauer spectroscopy were few and rather exotic.As synchrotron experiments havebecome increasingly monochromatic, thanks to innovations in monochromators and dramaticallyimproved resolution of detectors, Mössbauer spectroscopy can now be applied to phases with awider range of compositions.12.9. Synchrotron radiation – infrared and visibleX-rays are not the only kind of synchrotron radiation to be utilized for diamond anvil cell research.Russell Hemley and his colleagues at the Geophysical Lab have made extensive use of infraredradiation from the low-energy ring at the National Synchrotron Light Source at BrookhavenNational Laboratory. Others have put synchrotron-generated visible and ultraviolet radiation togood use as well.13. Ultrasonic studiesIn 1993, Hartmut Spetzler of the University of Colorado, Boulder, approached me to see whatI thought of the idea of using gigahertz ultrasound directed right through one of the diamondanvils to measure sound velocities in a single crystal under pressure in the diamond anvil cell. Theidea was to have the sound enter through one of the diamonds, traverse the crystal while underpressure, reflect from the far side of the crystal, return through the same diamond, and be detectedby the same transducer that sent the signal. Sorting out the various reflections would then makeit possible to determine the travel time through the crystal. This was an extremely difficult thingto do and was not really completed until several years later. Many of the problems were solvedby Steve Jacobsen, Hans Reichmann, and others at the University of Bayreuth in Germany. SteveJacobsen started as a graduate student in Geology at the University of Colorado and continued

High Pressure Research 185to work on the problems well after he received his degree. As a faculty member at NorthwesternUniversity, he is now able to make elegant measurements by this method.14. RadioactivityWalter Hensley, a student of John Huizenga at the University of Rochester, put the diamond anvilcell to excellent use measuring the effect of pressure on the radioactive decay rate of 7 Be. Thereasoning was that because the radioactive decay of 7 Be occurs by electron capture and becauseberyllium has few orbiting electrons, pressure would push the K-electrons closer to the nucleus,thus increasing the chance of capture. Sometimes the simplest line of reasoning proves to be right.Walter measured gamma-ray emission rates from a sample of 7 Be under pressure over a periodof several months. The results provided clear evidence that pressure accelerated the decay rate.The cell he used for that study was the one with the large openings on the sides that we haddeveloped for measuring deviatoric stress, thus showing how a particular design can have diverseapplications.15. Magnetic studiesAnother important probe for studying samples at high pressures has been magnetic fields. StanTozer of the National High Magnetic Field Laboratory in Florida has put diamond anvil cellsto excellent use for this purpose and has contributed many important papers on the effects ofpressure on magnetic properties of a wide range of materials. Someone wanting to examine asample under pressure using a magnetic field, however, faces a big challenge because of theinterference caused by the stainless steel body of a typical diamond anvil cell. Stan has takentwo approaches to solving these problems: better materials and smaller sizes. Beryllium–copperalloy for the diamond anvil cell parts is considerably better than stainless steel and has provedto be a good choice. Even better is a cell made with no metal, just plastic. Stan has put both ofthese approaches to good use. As for size, I think no one would question that Stan holds the worldrecord for smallness. His beryllium–copper cell is just 0.25 inches in diameter by 0.37 inches long.The plastic cells are a little larger, but not much. They are 0.45 inches in diameter by 0.52 incheslong and 0.31 inches in diameter by 0.34 inches long. People were impressed years ago whenVan would pull his high-pressure apparatus out of his pocket; today, Stan can put a dozen or soin his pocket. It was while taking a hike in Colorado that he noticed the turnbuckles on the guywires bracing the power line poles nearby. It occurred to him that that same simple principlecould be used to make a diamond anvil cell with only three moving parts, the buckle and the tworods entering the buckle with one having a reverse thread. The rods could be made shorter andthe diamond anvils could go on the ends of the rods. A large press is needed to push the anvilstogether, and then the turnbuckle is used to hold them together, resulting in no further need forthe press. These modifications have made it possible for him to subject samples to DC fields to35 T at temperatures down to 400 mK. By using a miniature coil that fits within a gasket hole andby using an epoxy-diamond gasket, he has been able to extend his studies to 350 mK in pulsedfields and to 10 mK in DC fields.16. The futureJust when I think I have heard of all the analytical methods that can be applied to samples ina diamond anvil cell, I read a paper describing yet another new method. There is no apparent

186 W.A. Bassettend in sight. The future of diamond anvil cell research appears to be very bright indeed. A largevariety of diamond anvil cells are now commercially available, and a search for diamond anvilcell manufacturers on the Internet will quickly turn up information on sources.I wish I had been able to acknowledge more of the ingenious and innovative contributions madeby users of the diamond anvil cell. I would also like to make the point that as wonderful as thediamond anvil cell is, large-volume presses, shock-wave experiments, and theoretical studiescontinue to play very important roles in high-pressure research. The bringing together of thesediverse approaches to complement each other in solving questions in all the sciences has been abigger success story than the accomplishments of any one device or any one approach. I marvelat how many things that seemed to be insurmountable barriers in our science have given wayone by one as the high-pressure community grows, with new members bringing new insights totackle old problems. It would certainly please Alvin Van Valkenburg if he were still alive today.Van always welcomed newcomers. On more than one occasion I can remember him saying ‘I amso glad to see you young fellows taking such an interest in this field’.AcknowledgementsI would like to thank Stefan Klotz, Editor in Chief of High Pressure Research, for inviting me to write this history of thediamond anvil cell and for providing valuable guidance in doing so. I am indebted to Robert Hazen for providing originalphotographs taken by Alvin Van Valkenburg. I especially want to thank Gasper Piermarini, who was there at NBS in theearly days of diamond anvil cell development; he helped fill in some of the important details of the early history andprovided several of the illustrations. Likewise, Eric Van Valkenburg provided some details and informed me that he nowhas the walnut desk Van used in his informal home lab. I am grateful to Stan Tozer and Richard Nelmes also for filling insome details and providing encouragement. I had many valuable comments and suggestions from friends and colleagueswho read the manuscript. In addition to Robert Hazen, Gasper Piermarini, and Eric Van Valkenburg, they include AlanAnderson, I-Ming Chou, Robert Mayanovic, and Elise Skalwold.References[1] W.A. Bassett, John C. Jamieson, pioneer in experimental mineralogy, J. Geophys. Res. 91 (1986), pp. 4621–4624.[2] R.M. Hazen, The New Alchemists, Times Books, Random House: New York, 1993.[3] K. Syassen, Ruby under pressure, High Pres. Res. 28 (2008), pp. 75–126.[4] W.A. Bassett, Diamond anvil cells, inEncyclopedia of Condensed Matter Physics, G. Bassani, G. Liedl, andP. Wyder, eds., Elsevier: Boston, MA, 2005, p. 1.[5] W.A. Bassett, Deviatoric stress: a nuisance or a gold mine? J. Phys. Condens. Matter 18 (2006), pp. S921–S932.[6] W.A. Bassett, I-M. Chou, A.J. Anderson, and R. Mayanovic, Aqueous chemistry in the diamond anvil cell up to andbeyond the critical point of water, inChemistry at Extreme Conditions, M.R. Manaa, ed., Chapter 7, Elsevier: TheNetherlands, 2005, pp. 223–238.[7] G.J. Piermarini, Alvin Van Valkenburg and the diamond anvil cell, High Press. Res. 11 (1993), pp. 279–284.[8] G.J. Piermarini, High-pressure X-ray crystallography with the diamond anvil cell at NIST/NBS, J. Res. Natl Inst.Stand. Technol. 106 (2001), pp. 889–920.