the mean sphere diameter measured with conventional array sizing, uncertainties that are avoided when particle chains are used. If the microspheres are illuminated from below and the microscope is focused just above them, then the tiny focal spots can be photographed and the distance between the spots determined very accurately. Using CDF, the mean sphere diameter for SRM 1960 was determined to be 9.89±0.04 mm: this was the certified mean diameter for the SRM.

Supporting measurements were made using RLS and MEM. In the RLS technique, developed by Tom Lettieri, a tunable dye laser was used to generate resonance light-scattering intensity patterns from the microspheres as they were suspended in water. As the laser was tuned through its wavelength range, relatively sharp peaks in the light-scattering intensity appeared at certain wavelengths. Although this technique had been used before by others to investigate sharp resonances from single microspheres, this was the first time that such resonances were detected in a suspension of many microspheres: this could be done because the spheres had a very narrow size distribution (<1% standard deviation). By comparing the experimental resonance wavelengths with those calculated by Egon Marx of the

PED using Mie scattering theory, the NBS researchers were able to use RLS to arrive at a mean diameter of 9.90±0.03 μm for the SRM 1960 microspheres. This result was in excellent agreement with those from both CDF and MEM.

The other supporting metrology technique was MEM. Gary Hembree of the PED used MEM to measure the microspheres in a scanning electron microscope (SEM). With MEM, the spheres are mounted onto the microscope stage as in a conventional SEM. However, in a conventional SEM the electron beam is rasterscanned past the stationary spheres, whereas in the MEM technique developed at NBS, the electron beam is held stationary while an individual particle is moved through the beam via a piezoelectric scanning stage. The scattered electron intensity is measured versus stage position and, from this, a particle profile is generated. These profiles could then be used to determine the diameter of the individual particles. Using MEM, a mean diameter of 9.89±0.06 μm was obtained for the SRM 1960 spheres, in excellent agreement with the other two metrology techniques. Indeed, the remarkable agreement among the three metrology techniques, and the relatively small uncertainty in the diameter, likely made SRM 1960 the best characterized particle-size standard in the world at the time. An important point in these measurements was that the three metrology techniques were independent of each other, in that none of the measurements relied in any way on measurements from the other techniques. This is always good practice when certifying an SRM.

In 1985, SRM 1960 was first offered for sale to the public through the NBS Office of Standard Reference Materials (OSRM), making the SRM the first commercial product to be manufactured in space (SRM-1961, the nominal 30 μm spheres, was the second space-made product). The sale of the space beads was reported in hundreds of newspapers, magazines, and television/radio news stories around the world, from the New York Times to the CBS Nightly News to ABC's Good Morning America. The work was also featured on a National Public Radio program and on an Australian science show, Beyond 2000. Later on, SRM 1960 won an IR-100 award from Research & Development magazine for being among the top 100 products of 1985.

Over the years, SRM 1960 has proven to be a valuable tool for the calibration of particle-sizing instruments in the United States and around the world. Samples have been purchased by dozens of U.S. and foreign companies for use as primary particle-sizing standards. These companies include not only the makers of particlesizing standards and instrumentation, but also "every day" users who need to maintain accuracy and traceability of their measurements. Among the primary users are

major pharmaceutical companies, Fortune 500 petrochemical and chemical companies, small-to-midsized biomedical instrumentation companies, particle standard supply houses, and numerous firms, of all sizes, involved in the measurement of particle size. Several U.S. Government and non-profit labs, including NASA Ames, Battelle Northwest, EPA, FDA, Los Alamos, Sandia, and the USGS have also purchased SRM 1960. In the international arena, the material has been used by research laboratories in Australia, Austria, Brazil, Canada, England, France, Germany, India, Italy, Japan, Korea, Mexico, Norway, Spain, Switzerland, and Thailand. Sales of SRM 1960 average about 25 vials per year through the NIST OSRM. The OSRM also offers for sale another version of the space beads, SRM 1965, which are standard microscope slides with small patches of particles in both regular arrays and microsphere chains. These have been used throughout the world for educational and training purposes, as well as to satisfy the needs of those who wish to own something

"made in space." In addition to OSRM, the microspheres are also sold through the European Community's Bureau Communautaire de Reference (BCR).

The demand for SRM 1960 is spurred, in part, by its incorporation into several document standards in the United States. For example, the U.S. Pharmacopeia, in its test entitled "Particulate Matter in Injections," specifies the use of SRM 1960 for calibrating the liquidborne particle counters used in the test. As a waterquality standard, SRM 1960 is listed by the National Oceanic and Atmospheric Administration (NOAA) in its compilation of Standard and Reference Materials for Marine Science as a physical standard for the assessment of water and sediment quality. The particles have also found application in the monitoring and assessment of air quality, especially with regard to the Environmental Protection Agency (EPA) PM 10 standard that specifies a cutoff of 10 μm for the aerodynamic diameter of particulate emissions from motor vehicles, smokestacks, and

other industrial emission sources. In the medical field, SRM 1960 has been valuable as a calibration standard for blood-cell counting and sorting, as the mean diameter of the SRM is very near that of human red blood cells. Thousands of hospitals and medical testing laboratories throughout the U.S. use blood-cell counters to check for sickle cell anemia, Tay-Sachs disease, and other blood abnormalities, and SRM 1960 helps to ensure the quality and accuracy of such tests. SRM 1960 will, undoubtedly, find more applications in the medical field as biotechnology and medical diagnostics become more pervasive in our daily lives.

In addition to the above applications of SRM 1960, other areas of scientific research where the material has found use include: electron microscopy; chemical chromatography; powder metallurgy; ceramics; food processing; photographic films; and basic particle research, among many others. The rigid demands of ISO9000 will likely increase the importance of particlesizing standards such as SRM 1960, as companies become more concerned with quality control, conformance assessment, and reliability issues.

Of the four authors of the space beads paper, two, Tom Lettieri and Egon Marx, are still at NIST. Tom started at NBS in 1978 after graduate school at the University of Rochester, joining the Pressure Group to do high-pressure optical studies of liquids. Less than two years later, he joined the PED to conduct optical metrology studies of small particles, rough surfaces, and noncontact methods for dimensional measurement. Tom is now a Program Manager with the NIST Advanced Technology Program, where he helps select and manage industrial technology projects in the area of optics/photonics. Egon Marx did his graduate studies in nuclear physics under the direction of Murray Gell-Mann at the California Institute of Technology. After a few years of teaching at Drexel University,

Egon went to work at Harry Diamond Laboratories, conducting theoretical investigations in electromagnetic (EM) interference. His interests at NIST have included EM scattering, surface roughness, linewidth metrology, and quantum electrodynamics. Two authors, Gary Hembree and Ike Hartman, left NIST not long after the space beads project was finished. Gary came to NBS after completing graduate studies in electron microscopy at Arizona State University (ASU). After several years at NBS, he returned to ASU to work with the world-renowned electron microscopy group there. Ike Hartman came to NBS after spending a number of years at General Electric doing optics and imaging research. At NBS, he worked in various areas of optical metrology, including microscopy, particle sizing, electro-zone particle counting, and linewidth metrology. He retired from NIST after a long and satisfying career in optics, both in his native Netherlands and in the United States.

Prepared by Tom Lettieri.

Bibliography

[1] T. R. Lettieri, A. W. Hartman, G. G. Hembree, and E. Marx, Certification of SRM 1960: Nominal 10 μm Diameter Polystyrene Spheres ("Space Beads"), J. Res. Natl. Inst. Stand. Technol. 96, 669-691 (1991).

[2] G. Mulholland, G. Hembree, and A. Hartman, Sizing of Polystyrene Spheres Produced in Microgravity, NBSIR 84-2914, National Bureau Standands, Gaithersburg, MD (1985).

[3] T. R. Lettieri, Optical Calibration of Accurate Particle Sizing Standards at the U.S. National Bureau of Standards, in Optical Particle Sizing: Theory and Practice, Gérard Gouesbet and Gérard Grehan (eds.), Plenum Press, New York (1988). [4] Nancy M. Trahey (ed.), NIST Standard Reference Materials Catalog 1998-1999, NIST Special Publication 260, National Institute of Standards and Technology, Gaithersburg, MD (1998). [5] D. Kornfeld, Monodisperse Latex Reactor (MLR), NASA TM86847, NASA, Washington, DC (1985).

An unavoidable consequence of quantum mechanics is that, for sufficiently short length scales, all objects appear to be "wavy." We do not notice this effect in our everyday lives because, for objects larger than an electron, the length scale over which the waviness occurs is fantastically short, far too small to be observed by the unaided eye. Nature makes an exception to this rule, however, in the case of extreme cold. As objects are cooled very close to absolute zero, their characteristic quantum-mechanical wavelengths become increasingly long. This tendency towards ever-expanding wavelength culminates in a dramatic phenomenon known as "BoseEinstein Condensation" (BEC).

BEC was originally conceived in 1925 by Albert Einstein, who calculated that if a gas of atoms could be cooled below a transition temperature, it should suddenly condense into a remarkable state in which all the atoms have exactly the same location and energy-in modern language, the wave-function of each atom in a Bose-Einstein condensate should extend across the entire sample of gas. For a dilute gas, the requisite transition temperature is so low as to be unachievable by the technology of Einstein's day. By the 1980s and early 1990s, however, cooling techniques had advanced to the point where a number of experimental groups around the world felt emboldened to attempt to realize Einstein's original vision. Many of the necessary advances came from NBS/NIST atomic physics laboratories. The first successful creation of dilute-gas BEC, announced in the NIST publication Observation of Bose-Einstein Condensation in a Dilute Atomic Vapor [1], was both a natural continuation of a 75-year tradition of NBS pre-eminence in spectroscopy (which is detailed in several other entries in this book [2-5]) and a striking confirmation that present-day NIST research is at the cutting edge of modern technology.

The scientific motivation to create and study BEC in a gas stemmed from the long-held belief that the mechanism underlying BEC is the same mechanism responsible for the mysterious effects of superconductivity and superfluidity. Indeed, in the broadest sense the electrical currents that flow (without resistance) in a superconducting metal and the liquid currents that persist (without viscosity) in superfluid helium are basically Bose condensates. But liquids and solids are much more complicated than the relatively simple gas-phase system that Einstein first envisioned, and it is

not easy to connect the elegant mechanism that Einstein proposed with the complex behavior of solids and liquids. If one could create a Bose condensate in a gas, it was reasoned, one would have a well-characterized model system, a system that might illuminate the counter-intuitive behavior of its liquid and solid prede

cessors.

The technical motivation for creating a BEC was equally compelling. Much of the standards and metrology work that NIST is charged with performing relies on precise spectroscopy of various internal resonances in atoms. When it comes to spectroscopy, the general rule of thumb is “colder equals more accurate." Colder atoms move more slowly, which means they can be probed longer, with correspondingly narrower resonance lines. In addition, systematic errors are often more easily controlled at lower temperatures. For a gas of atoms, the natural and obvious limit of improved cooling is exactly the Bose-condensed state. Thus from both technological and scientific viewpoints, there were compelling reasons to push the techniques of refrigeration to the ultimate limits with the goal of creating BEC.

The first condensates were formed at NIST at temperatures well under a microkelvin. To reach these unprecedented temperatures required a two-stage cooling technique. The first stage of refrigeration is provided by laser cooling. Of the three or four most prominent players in the development of laser cooling, two (David Wineland and Bill Phillips) are long-standing Bureau scientists; two of their most influential papers are described in this volume [6,7]. As powerful as laser cooling is, it is not sufficient on its own to reach BEC temperatures. The second stage of cooling is known as evaporative cooling. The laser-cooled atoms are collected in a magnetic trap (another NIST development [8]) which provides near-perfect thermal isolation from the surrounding environment. Via a technique known as rf evaporation [9], the trapped atoms with the most energy are ejected from the magnetic trap. The remaining atoms have, on average, less energy per atom, and are therefore colder. After evaporation has cooled the atoms to a temperature perhaps another factor of a hundred colder than the laser-cooled sample, the condensate begins to form.

The presence of condensates was originally detected by velocity-distribution information observed in timeof-flight images. The magnetic fields used to confine

the atoms were very suddenly turned off. The residual thermal and quantum energy of the atoms caused them to fly apart. After a brief delay, the atoms were illuminated with a strobed flash of laser light, and their image was captured on an electronic screen. The atoms with large thermal velocities in the trapped cloud ended up far from the center of the image; atoms with relatively low velocities did not travel as far during the delay time and contributed to the central portion of the recorded density. Fig. 1 shows a series of three such images; from left to right they correspond to images taken of three clouds at progressively lower temperatures [10]. In the left-most image, the atoms are not yet condensed; the distribution of velocities is well approximated by a conventional Maxwell-Boltzmann thermal distribution. In the center image, the condensate has begun to form; the central spire corresponds to the near-stationary atoms of the condensate. The final, right-most cloud is a near-pure condensate. The central feature amounts to a photographic image of a single, macroscopicallyoccupied quantum wavefunction.

The original observation of BEC in a gas of atoms occurred in June of 1995. A few months earlier, several groups (most notably the NIST/CU collaboration in Boulder, and groups at Rice University and at MIT) were very close to achieving Bose-Einstein condensation. All three groups presented their progress in invited talks at the May 1995 meeting of the American Physical Society. The audience was left with the impression that the long-standing goal of BEC might be realized quite

soon. There was a pronounced sense of keen, but goodspirited, competition that added to the general anticipation felt in the physics community.

Ultimately, the NIST group prevailed and its paper [1] appeared, as the cover article, in Science magazine on July 14th, 1995. In the same issue, Science also ran a "perspective" piece by Keith Burnett, of Oxford University, in which he referred to the achievement of BoseEinstein condensation as a sort of "Holy Grail" of physics. The announcement of Bose-Einstein condensation attracted an unusual amount of attention from the lay public. There were front-page articles in the Washington Post, the New York Times and the Los Angeles Times, and even professional entertainers made remarks about scientists creating new states of matter. The scientific press was also duly impressed: the work was written up in all the major science magazines; the paper won the AAAS Newcomb-Cleveland award; and in December 1995 Science deemed BEC the "Breakthrough of the Year."

In the years immediately following NIST's breakthrough result, there was an enormous surge of interest in the field of BEC. Within a few months, the group at MIT had successfully created a sample of BEC over a hundred times larger than the initial NIST result [11]. Theoretical calculations performed at NIST predicted that the condensate clouds should support standingwave acoustic modes [12], with resonance frequencies determined by solutions to a macroscopic quantum wave equation. Within a year, these predictions were