Co ntents · 2003-10-02 · Co ntents Articles The Reduction of Uncertainties for Absolute B. E....

Volume 94, Number 6, November-December 1989

Journal of Research of the National Institute of Standards and Technology

Co ntents

ArticlesThe Reduction of Uncertainties for Absolute B. E. Welch, R. E. Edsinger, 343

Piston Gage Pressure Measurements in the V. E. Bean, and C. D. EhrlichAtmospheric Pressure Range

Absolute Isotopic Abundance Ratios and Atomic J. W. Gramlich, L. A. Machlan, 347Weight of a Reference Sample of Nickel I. L. Barnes, and P. J. Paulsen

The Absolute Isotopic Composition and Atomic J. W. Gramlich, E. S. Beary, 357Weight of Terrestrial Nickel L. A. Machlan, and I. L. Barnes

Special Report on Standards for RadioactivityReport on the 1989 Meeting of the Radionuclide Dale D. Hoppes 363

Measurements Section of the ConsultativeCommittee on Standards for the Measurementof Ionizing Radiations

On Measuring the Root-Mean-Square Value E. Clayton Teague 367of a Finite Record Length Periodic Waveform

A Search for Optical Molasses in a Vapor Cell: A. L. Migdall 373General Analysis and Experimental Attempt

SUBJECT INDEX TO VOLUME 94 397

AUTHOR INDEX TO VOLUME 94 401



News Briefs

GENERAL DEVELOPMENTS 379

Gravimeters Used SuccessfullyDuctile-to-Brittle Fracture Behavior of Steel StudiedCertification of Aluminum in Serum Albumin for FDA

CAC QA Program Improves the Quality of Measurements Made in Cancer

Chemoprevention StudiesAtomic Liquid and Solid Behavior Observed in PlasmasNIST Service for Setting Computer ClocksProposed Revision of Standard for POSIX: Portable Operating System Interface for

Computer EnvironmentsPresence of Toxic S2F10 Confirmed for First Time in SF6 Exposed to DischargeNCWM Training Module PublishedHigh-Speed Metallizing of FibersElectrochemical Deposition of Al-Mn Quasicrystals and IntermetallicsStructural Ceramics Database DevelopedMagnetic Order Observed in Artificial SuperlatticesCollaborative Research Effort on Materials Research Renewed

NY Firm Joins With NIST to Investigate CID Sensors

Cold Neutron Workshop on Analytical Measurements With Cold Neutron Beams

High Efficiencies of Soft X-Ray Mirrors Measured at Surf II

Laser Ablation Source for Structural StudiesFrequency Tables for Carbon Monoxide Lasers

New Record Beam Current, 300 mA, Accelerated to Full Energy in SURF-II,the NIST Electron Storage Ring

Agreements of Open Systems Interconnection (OSI) Protocols Published

NCSL Releases Revised Structured Query Language (SQL) Test Suite

Optional Module for Information Resource Dictionary System (IRDS) Standard

to be DevelopedNSF Grant for NIST Senior Research FellowshipFirst Demonstration of Soliton-Like Compression of Pulses from Optical Fiber Laser

EIA Committee Asks NIST to Develop Standards for Fiber GeometryGMAP System Architecture Installed at the National PDES Testbed

NIST Quality-in-Automation Project Draws Machine Tool Company CollaboratorsCBT Develops New Design Procedure for Heat Losses from Underground Piping Systems

CBT Monitors the Source Strengths of Volatile Organic Compounds in a MajorNew Federal Office Building

X-Ray Method Removes Aids Virus from EvidenceBaldrige Quality Award Program Seeks ExaminersInternational Directory of Standards Groups UpdatedEquation Developed for Predicting Steel Behavior in FireHigh-Quality Amorphous Silicon Films ProducedNon-Toxic Acids are Basis for New Class of Cements



Quick Way to Measure Heat ConductionIs There a Robot in Your Future?Properties of Methane DevelopedFiber Laser Research ProgressingHAZARD I Released to Help Reduce Fire Deaths, CostNIST Launches Consortium to Develop Futuristic LabIndustry to Study Wear of Ceramics at NISTEvaluation of High-Frequency Power MetersPortland Cement Clinker Materials AvailableOh, Say Can You See?New NIST Standard Reference DatabaseCeramic Powders Synthesis-Magnetic MediaExpert System for Material Selection Advice for the Petroleum IndustrySandia to Participate in Cold Neutron Facility Instrument DevelopmentIndustry Supports Research on the Development of Optical Waveguide SensorsFederal Information Processing Standard (FIPS) Proposed for the User Interface

Component of the Applications Portability Profile (APP)Open Systems Interconnection/Integrated Services Digital Network (OSI/ISDN) Trial HeldNIST Break-Junction Measurement Provides First Determination of SIS Tunneling

Gap of a Thallium-Based High-Temperature SuperconductorNIST to Participate in Consortium Organized to Study S2F1oNIST Miniature Broadband Electric-Field Probe Design CommercializedNational PDES Testbed in CME Applies Hypertext to Support PDES/STEP Effort

CALIBRATION SERVICES 393New Automated Gage Block Calibration System at NIST

Calendar 395



[J. Res. Natl. Inst. Stand. Technol. 94, 343 (1989)]

The Reduction of Uncertainties for AbsolutePiston Gage Pressure Measurements in the

Atmospheric Pressure Range

Volume 94 Number 6 November-December 1989

B. E. Welch, R. E. Edsinger, NIST pressure calibration services with Key words: manometer; piston gage;V. E. Bean, and C. D. Ehrlich nitrogen are now based on two transfer pressure; pressure measurement; pres-

standard piston gages for which the ef- sure metrology; pressure standards.National Institute of Standards fective areas have been determined byand Technology, calibration with the manometer devel-and Technology, ~~~oped at NIST for gas thermometry.Gaithersburg, MD 20899 Root-sum-squared three sigma uncer-

tainties for the areas for the two gagesare 3.05 ppm and 4.18 ppm. Accepted: September 5, 1989

1. Introduction

There are presently only two technologies thatcan be developed into practical primary pressurestandards in the atmospheric pressure range. Theyare the piston gage (pressure balance, dead weighttester) and the manometer.

The essential features of the piston gage includea vertical hollow cylinder which is closed at thetop by a close-fitting piston and closed at the bot-tom with appropriate plumbing to admit the pres-surizing fluid. The piston is loaded with knownweights to counter-balance the effect of the pres-sure so that it floats at a specified reference leveland is rotated to relieve friction. The pressure iscalculated as the ratio of the force due to theweights to the effective cross-sectional area of thepiston. The piston gage measures the difference be-tween the pressure at the bottom and the top of thepiston. When the top of the piston is at ambientatmospheric pressure, we speak of using the devicein the gage mode. The absolute mode requires thatthe top of the piston be in a vacuum. The accuracy

of a piston gage is limited by our ability to deter-mine the effective area. For a primary standard pis-ton gage, the area must be determined fromdimensional metrology. The highest accuracy re-ported in the literature is on the order of 15 partsper million (ppm) [1] while repeatability of 1 ppm iscommon. The limitation is in the accuracy of thedimensional measurements and in the manufactureof truly straight and round cylinder bores and pis-tons.

The essential feature of a manometer is a verticalcolumn of fluid, suitably contained and supportedat the bottom by an applied pressure. The magni-tude of the pressure is the product of the columnheight, the density of the fluid, and the accelerationdue to gravity plus whatever pressure is applied tothe top surface of the fluid column. Thus, themanometer is also a differential pressure measure-ment device and is used in the absolute mode whenthe space above the top surface of the field columnis evacuated or in the gage mode when that space is

343



at ambient atmospheric pressure. Limitations in ac-curacy are due to the uncertainties in the density ofthe fluid and in the column height measurements.In general, at atmospheric pressure, a state-of-the-art manometer will have a lower uncertainty than astate-of-the-art primary standard piston gage.

The piston gage has two advantages over themanometer: portability and ease of use. These twoproperties coupled with the stability of the pistongage make it an excellent transfer standard. In or-der to meet the demand for reduced uncertainties,NIST has used the manometer that was developedat NIST for gas thermometry [2] as a primary stan-dard to calibrate the two piston gages that serve asthe reference standards for NIST piston gage cali-bration service. Calibration services are now of-fered through these transfer standard piston gages.

Herein we report the results of the calibration ofthe piston gage calibration reference standards us-ing the manometer in the absolute mode.

2. Apparatus

A complete description of the gas thermometermanometer (GTM) and a thorough evaluation ofits uncertainties at the 99 percent confidence levelare found in the literature [2,3,4]. Recently,Edsinger and Schooley, as a part of their new gasthermometry measurements up to 933 K, have re-considered the uncertainty of the manometer andhave determined that the value given in [2] of 2ppm at the 99 percent confidence level is satisfac-tory [5].

The two piston gages are commercial units, iden-tical in make and model. They were modified intwo ways: in an effort to eliminate every part hav-ing even a remote possibility of being hard to cleanand thus becoming a source of dirt, we replacedthe entire mechanism for the top and bottom stopswith parts made of Kel-F. Also the rotative mecha-nism was modified such that the cylinder and thepiston with its weight stack could be rotated to-gether up to the desired angular velocity, on theorder of 60 revolutions per minute (rpm), and thenthe piston was floated and the cylinder wasstopped allowing the piston to continue. Pressuremeasurements were then made with the cylinderstationary and the piston coasting. The Kel-F bot-tom stop was coated with a robust conductive alu-minum film which was grounded to the cylinder toprevent electrostatic charges from accumulatingon the surface when the rotating weight hangerrubs on the bottom stop. Without the conductive

film, the resulting electrostatic forces can produceerrors of several ppm at atmospheric pressure.

The temperature of the piston and cylinder as-sembly was monitored with an array of ten ther-mistors, connected in parallel, mounted on thestationary portion of the cylinder mount, and cali-brated in situ.

Figure 1 is a schematic diagram of the experi-mental arrangement. Item A is a calibrated differ-ential pressure transducer having a full range of130 Pa and a sensitivity of 7X 10-2 Pa. It was cali-brated by using two piston gages, one supplyingpressure to each pressure port of the transducer.The use of the transducer provides two advan-tages: it allows both instruments to work against alimited-volume pressure system which makes theapparatus easier to operate. The transducer wasalso used to read small differential pressures be-tween the manometer and the piston gage whichmakes the establishment of a perfect equilibriumunnecessary. The zero was checked before eachmeasurement by applying the identical pressure toboth sides of the transducer via the bypass valve.Typically, the differential pressure for a given mea-surement was a fraction of a pascal.

Item B of figure 1 is another calibrated differen-tial pressure transducer of the same make, model,range, and sensitivity as item A. It was used tomeasure the pressure in the bell jar surrounding theweights, which is evacuated for absolute mode op-eration.

The pistons and cylinders were examined forresidual magnetism using a Hall-effect detector andwere demagnetized as necessary.

Figure 1. Schematic diagram of the pressure system connectingthe manometer to the piston gage. The items marked A and Bare calibrated differential pressure transducers. The circled X'sare valves.

344

Volume 94, Number 6,

Journal of Research of the NationalNovember-December 1989

Institute of Standards and Technology

3. Results

Both gages were repeatedly calibrated with themanometer at 27 and 95 kPa in the absolute modeusing nitrogen; a total of 107 measurements forPG 28 and 149 measurements for PG 29. The effec-tive areas for the gages were calculated from theexpression

0FE 15

E 10

.0 5Ez

A= (P-Pb-pgh +Mg)[1+(Xp+ -c)(T-TR)] (1)

El 95 kPa

o 27 kPa

PG 28Nitrogenu=0.63

- 1.6 -. 8 0 .8 1.6Deviation from the mean, ppm

Figure 2. Histogram showing the deviation of the measuredarea from the mean for PG 28.

M is the total mass supported by the pressureincluding the weights and piston

g is the local acceleration due to gravityP is the pressure at the lower mercury sur-

face in the manometerPb is the pressure in the bell-jar surrounding

the weightsp is the density of the nitrogenh is the height between the reference level of

the piston gage and the level of the lowermercury surface in the manometer

e is the differential pressure measured bytransducer A

cc is the linear thermal expansion coefficientfor the piston

:c is the linear thermal expansion coefficientfor the cylinder

T is the temperature of the operating pistongage

TR is the reference temperature and is definedto be 23 0C.

The effective area of each gage was found to beconstant over this pressure range. Figures 2 and 3are histograms showing the deviation from the av-erage areas expressed in ppm. The measurements at95 kPa are marked with a dot to distinguish themfrom the 27 kPa measurements. There is no appar-ent significant pressure dependence on the area ofeither gage. The areas and the tripled standard de-viations are given in table 1.

Table 1. Areas and tripled standard deviations of PG 28 andPG 29

PG 28 PG 29

Area, 10-4 m2 3.3582249 3.3572390

Triple standard deviation 1.89 ppm 3.42 ppm

E 95 kPaEl 27 kPa

20F

Ez

15

10

5

PG 29Nitrogen_ =1.14

-- . :- .. .. .n

-3.2 -2.4 -1.6 -. 8 0 .8 1.6 2.4Deviation from the mean, ppm

3.2

Figure 3. Histogram showing the deviation of the measuredarea from the mean for PG 29.

The uncertainty in the effective area of each pis-ton gage arising from the uncertainties in the vari-ous parameters of eq (1) can be expressed as [6]

d =A r[ (Il4 4 v xz 1/2A [ZA 'aXSJJ (2)

where the Xi are the parameters given in eq (1).These uncertainties are given in table 2. The tripledstandard deviations about the means given for theareas of the two gages in table 1 are combined withthe uncertainty given by eq (2) by the root-sum-square method; they are also listed in table 2.

The sources for the values listed under the head-ing "dxi" in table 2 are as follows:

The value of dM is the result of calibration of thepiston gage weights using NIST standards and isdiscused elsewhere [7]. The value of dg is the resultof on-site measurements. Reference 2 discusses thevalue of dP. The values of dPb are based on calibra-tion measurements. The value of dp is based on anuncertainty in temperature measurement of 0.5 K

345

where

........



Table 2. Uncertainties in area for gages PG 28 and PG 29 calculated at a pressure of 100 kPa based on threestandard deviations

Parameter

Parameter Uncertainty Differential 1 aA dXA aX,

X} Units Value dXj A aXI Value ppm

M kg 3.4 l.1X10- 6 I/M 3X10-' 0.33g m/s2 9.8 1.5X10- 6 lhg IX10-1 0.15P Pa I X lo, 2.0X 10'- 1/P I X 10-

5 2.0Pb Pa 1.3 6x 10-2 I/P I X 10-5 0.6p kg/m3 1.2 1.8X 10-3 gh/P 5 x 10- 4 0.9h m 5 2.0X 10-3 pg/P 1.2X 10-4 0.24F. Pa 0.5 6X 10- 2 1/P 1 X 10- 5 0.6('cp+Kcc) /'C 8.22X10-6 1.5X 10-8 T-TR 1 0.015T-TR C TR=23 3 X 10- 2

c p+ cc,: 8.22X 10-6 0.24

-A arising from the parameters of eq (1) 2.40

PG 28 PG 29

Uncertainty in eq (1) 2.40 2.40Tripled standard deviation of area about the mean 1.89 3.42Combined total uncertainty 3.05 4.18

along the tube connecting the manometer to thepiston gage [2]. On-site height measurements pro-duced the value of dh. Separate thermal expansioncoefficient measurements on the materials fromwhich the piston and cylinders were made result inthe value of d( P cc). The value for the uncer-tainty in the temperature measurements is based oncalibration data of the thermistors.

The uncertainty in P includes the uncertainty inthe head correction, pgh, with the assumption thatthe gas is helium [2]. Since, in this case, nitrogenwas used, we have explicitly included the uncer-tainty for the nitrogen head correction without re-moving the helium head correction uncertainty inthe value of dP, a conservative approach.

The determination of effective areas of pistongages PG 28 and PG 29 as reported in this paperreflect state-of-the-art measurements and theyprovide an improved basis for NIST calibrationservices for the practical use of gas operated pistongages. It is recognized that reasonably frequentverification of the effective areas by the use of theGTM facility is important and necessary to estab-lish a new body of control history to support im-proved levels of calibration service accuracy. Theintrinsic stability of well constructed gas operatedpiston gages combined with the accuracy of theGTM provides a clear basis for significant im-provement in the accuracy levels of practical pres-sure measurements.

About the authors: B. E. Welch and R. E. Edsingerare physicists who were in the Temperature and Pres-sure Division of the NIST Center for Chemical Tech-nology. Both have retired. V E. Bean and C. D.Ehrlich are physicists in that same division.

4. References

[1] Maghenzani, R., Molinar, G. F., Marzola, L., andKulshrestha, R. K., J. Phys. E: Sci. Instrum. 20, 1173(1987).

[2] Guildner, L. A., Stimson, H. F., Edsinger, R. E., andAnderson, R. L., Metrologia 6, 1 (1970).

[3] Guildner, L. A., and Edsinger, R. E., J. Res. Natl. Bur.Stand. (U.S.) 80A, 703 (1976).

[4] Schooley, J. F., Proceedings of the 1988 Workshop andSymposium of the National Conference of Standards Labo-ratories, p. 50-1, August 14-18, 1988.

[5] Edsinger, R. E., and Schooley, J. F., Metrologia 26, 95(1989).

[6] Heydemann, P. L. M., and Welch, B. E., Piston Gages, inExperimental Thermodynamics, Volume II, edited byLeNeindre, B., and Vodar, B., Butterworths, London(1975) p. 183.

[7] Davis, R. S., and Welch, B. E., J. Res. Natl. Bur. Stand.(U.S.) 93, 565 (1988).

346




Absolute Isotopic Abundance Ratios andAtomic Weight of a Reference Sample

of Nickel


J. W. Gramlich, L. A. Machlan, Absolute values have been obtained for 61Ni=1.139894±0.000433, 6 2NiI. L. Barnes, and P. J. Paulsen the isotopic abundance ratios of a refer- =3.634528+0.001142, and 64Ni

ence sample of nickel (Standard Refer- =0.925546±0.000599. The atomicNational Institute of Standards ence Material 986), using thermal weight calculated from this isotopicand Technology, ionization mass spectrometry. Samples composition is 58.693353±0.000147.Gaithersburg, MD 20899 of known isotopic composition, pre- The indicated uncertainties are overallpared from nearly isotopically pure sep- limits of error based on two standard

arated nickel isotopes, were used to deviations of the mean and allowancescalibrate the mass spectrometers. The for the effects of known sources of pos-resulting absolute isotopic ratios are: sible systematic error.5sNi/6wNi=2.596061 ±0.000728,6 lNi/6Ni=0.043469±0.000015 Key words: absolute ratio; assay; atomic6 2Ni/6Ni=0.138600+0.000045, and weight; dimethylglyoxime; isotopic6 4 Ni/6 0 Ni=0.035295±0.000024, which abundance; mass spectrometry; nickel.yield atom percents of 5 8 Ni=68.076886±0.005919, 6°Ni=26.223146±0.005144, Accepted: August 28, 1989

1. Introduction

The Inorganic Analytical Research Division ofthe National Institute of Standards and Technol-ogy, has been conducting a long-term program ofabsolute isotopic abundance ratio and atomicweight determinations using high precision isotoperatio mass spectrometry. Previous atomic weightdeterminations include silver [1,2], chlorine [3],copper [4], bromine [5], chromium [6], magnesium[7], lead [8], boron [9], rubidium [10], rhenium [11],silicon [12], potassium [13], thallium [14], strontium[15], and gallium r16].

To obtain absolute isotopic ratios from the ob-served or relative measurements made on a massspectrometer, it is necessary to calibrate the instru-ment using samples of accurately known isotopicratios of the element under study. These syntheticisotopic standards, assayed and gravimetrically

prepared from chemically pure and nearly isotopi-cally pure isotopes, provide a bias correction (cal-culated isotopic ratio/observed isotopic ratio)which, when applied to the observed isotopic ra-tios of the reference sample being calibrated, al-lows absolute ratios to be calculated for the sample.The atomic weight is then obtained by multiplyingthe fractional abundance of each isotope by its nu-clidic mass [17] and summing the resultant prod-ucts. A more detailed description of the methodused for the determination of isotopic abundanceratios and atomic weights at NIST is given else-where [2].

In 1961, the IUPAC Commission on AtomicWeights recommended a value of 58.71 for theatomic weight of nickel. That value was based onthe isotopic abundance measurements of White and

347



Cameron [18]. The Commission noted in its reportthat all chemical determinations that had been re-ported and believed to be significant gave a meanvalue for the atomic weight of 58.69. The bestchemical determinations appeared in a series ofpublications by Baxter and associates in the 1920s[19,20,21]. These measurements yielded an averageatomic weight of 58.694, with credibility increasedby proof of accuracy of parallel work on theatomic weight of cobalt [22] which is now knownaccurately because of its mononuclidic state.

In 1973 the Commission reexamined both thechemical and mass spectrometric measurementsand so recommended a lower value of 58.70 for theatomic weight of nickel. Later that year, Barnes etal. [23] published a superior but not absolute massspectrometric measurement which produced anatomic weight value of 58.688, in good agreementwith the chemical determinations. After reexami-nation in 1979, the Commission recommended thepresent standard atomic weight value of58.69±0.01 [24]. Nickel is currently listed by theIUPAC Commission on Atomic Weights and Iso-topic Abundances as one of the elements with anatomic weight with a large uncertainty [25].

Since no significant variations in the isotopiccomposition of nickel in nature have been ob-served, either in this study or in previous work, ahigh accuracy measurement of the atomic weightof a reference sample of nickel (SRM 986) will al-low IUPAC to recommend a value with a muchsmaller uncertainty.

2. Experimental Procedure2.1 Mass Spectrometry

Isotope ratio measurements were made on twodifferent thermal ionization mass spectrometerswith separate operators. One instrument was anNIST designed mass spectrometer equipped with a30-cm radius of curvature, 90° magnetic sector(designated inst. #1, operator #1). The second in-strument was a Finnigan-MAT 261 mass spectrom-eter' (designated inst. #2, operator #2). The NISTinstrument employed a shielded Faraday cage col-lector with a double slit collimator. The remainderof the measurement circuitry consisted of a para-

' Certain commercial equipment, instruments, or materials areidentified in this paper to specify adequately the experimentalprocedure. Such identification does not imply recommendationor endorsement by the National Institute of Standards andTechnology, nor does it imply that the materials or equipmentidentified are necessarily the best available for the purpose.

metric electrometer [26], a precision voltmeter, anda computer. Timing, magnetic field switching, anddata acquisition were controlled by the computer.The Finnagin-MAT 261 is a 23-cm radius, 90° mag-netic sector instrument which uses a non-normalentry and exit ion path. This arrangement gives thedispersion of an instrument of 2.78 times (64 cm)the radius. It is equipped with seven deep Faradaycup collectors, six of which are externally ad-justable. Each cup has an individual amplifier con-tained within an evacuated, thermally controlledchamber. The chamber temperature is maintainedto ±0.02 'C.

Nickel was thermally ionized from a platinumfilament fabricated from 0.001 X 0.030 in(0.0025X0.076 cm) high purity platinum ribbon.Prior to filament fabrication, the platinum ribbonwas heated for several hours in dilute hydrochloricacid to reduce any iron impurities that might causeisobaric interferences. After fabrication, the fila-ments were degassed for 1 h by passing a current of3 A through them in a vacuum and under a poten-tial field of 45 V. Filaments cleaned in this mannergenerally exhibited no detectable emissions for thenickel isotopes using blank filament loadings. Themaximum background observed was 1.2X 10-6 Aat mass 58, which appeared to be a natural nickelbackground. This background would be insignifi-cant for measurement of the natural standard andthe point calibration mixes. Filaments used formeasurements of the nickel separated isotopes wereexamined in the mass spectrometer prior to sampleloading. Only those filaments which showed no de-tectable background were used for analysis of theseparated isotopes.

All sample loadings were conducted in a Class100 clean air hood. Pipets made of Pyrex tubingwere used to transfer the samples from their con-tainers to the filaments. The tubing was cleaned byheating in 8 mol/L HNO3 for 24 h, rinsing withultra-high purity water, heating in 6 mol/L HC1 for24 h, followed by several rinsings in ultra-high pu-rity water.

Samples were loaded onto the filaments underthe following conditions: approximately 5 pg ofsolution (5 pL of 1 mg/mL Ni as NiNO3 in 1+49HNO3 )2 was added to the filament and dried for 5min at 1 A. Five pL of a solution containing 17 mg

2A reagent dilution of (1+49) indicates 1 volume of concen-trated reagent diluted with 49 volumes of pure water. If nodilution is specified, use of concentrated reagents is implied. Theacids and water used for these dilutions were produced at NISTby sub-boiling distillation [26].

348



Aerosil 300 (Degussa, Frankfurt, FGR) powder/gof solution, 0.34 mg/g AlCl 3 /g solution, and 0.1 gof high purity H3 PO4/g of solution were added tothe filament and dried at a current, through thefilament, of 1, 1.3, and 1.5 A, each for a time periodof 5 min. The filament was then slowly heated tofume off the excess H3PO4 and then heated for afew seconds at approximately 700 'C. After a 5 mincooling period, the filament was loaded into themass spectrometer.

The analysis procedure for inst. #1 was as fol-lows: the initial filament current was set to pro-duce a filament temperature of 1000'C. At 4 minintervals the filament temperature was increased by50 'C until at 20 min a final temperature of 1250 'Cwas achieved. After measurement of baselines onboth sides of each mass of interest, data were col-lected between 30 and 70 min into the analysis. Forthe reference material, 5 min sets of ratio measure-ments were made in the following order: 58/60,61/60, 62/60, 64/60, 64/60, 62/60, 61/60, 58/60.

The analysis procedure for inst. #2 was slightlydifferent. The heating pattern was the same with amaximum temperature of 1250 'C (as measuredwith a two color optical pyrometer) used. The sixmass unit range covered by the nickel isotopes isjust beyond the range of this instrument so that itwas necessary to jump the magnetic field once tocollect all of the ratios. Thus, the 58/60, 61/60, and62/60 ratios were collected simultaneously duringa 32 s integration, a jump of one mass unit wasmade and the 64/60 ratio was collected. Five setsof twenty such ratios were collected for each sam-ple. The amplifiers were calibrated at the beginningof each sample run; baselines were collected at thebeginning of each block of twenty ratios.

2.2 Purification of Separated Nickel Isotopes

Electromagnetically separated 5 8 Ni, 6 0Ni, and62Ni isotopes were obtained from the Nuclear Divi-sion, Oak Ridge National Laboratory. The 58Niwas designated sample 121426, the 60Ni was desig-nated sample 121501, and the 6 2Ni was designatedsample 158201. The certificate which accompaniedeach sample showed isotopic enrichment to ap-proximately 99.8% for the 58Ni material, 99.8% forthe 60Ni material, and 99.0% for the 62Ni material.The certificates included a semi-quantitative spec-trographic analysis which showed that the princi-pal impurities were Cu and Zn at the 0.1% leveland several other elements at the 0.05% level.

To reduce these impurities to a level low enoughso that they would not cause a significant error in

either the assay procedure or the mass spectromet-ric ratio measurements, the separated isotopes werepurified by a combination of cation exchange chro-matography, ammonium hydroxide precipitation,electrodeposition and anion exchange chromatog-raphy.

Each separated isotope was treated as follows:the nickel (approximately 2 g of 58Ni and 60Ni, and1.8 g of 62Ni) was dissolved in 20 mL of HNO3(1+1), evaporated to dryness and then 25 g ofHC104 were added. The solution was evaporatedto HC10 4 fumes, and after cooling, 5 g of 10 mol/LHC1 was added. This step was used to help elimi-nate any Cr that was present. The sample solutionwas evaporated to dryness and the residue was dis-solved in 50 g of H20. This solution was passedthrough two cation exchange columns (a singlecolumn of the necessary length was not available)in series (each column, 19 X 1.6 cm, was filled withAG 50X8, 100-200 mesh resin and cleaned with120 g of 4 mol/L HC1 followed with H20 until theeluate was neutral). After adding the nickel, someimpurities were eluted from the columns with 200g of 0.4 mol/L HCl. The nickel was eluted with120 g of 3 mol/L HC1 and this solution was evapo-rated to dryness. The residue was dissolved in 50 gof H2 0, an excess of NH 4OH was added (approxi-mately 20 mL) and the solution was filtered to re-move insoluble hydroxides. The solution wasevaporated to dryness, 25 g of H20 were added andthen 25 g of H2SO4 were slowly added. The solu-tion was evaporated to fumes of H2 SO4 and, aftercooling, 5 g of HNO3 were added. The beaker wascovered and heated, then the cover was removedand the solution evaporated to fumes of H2SO4.The sides of the beaker were rinsed down withH20 and the solution again evaporated to fumes ofH2SO4. The solution was diluted to 150 mL withH20, two clean platinum gauze electrodes wereplaced in the solution and a 2 V dc potential wasapplied between the electrodes for 16 h. The elec-trodes were removed from the solution and re-cleaned in HNO3 (1 + 1). The electrodes wereplaced back into the separated isotope solution,NH40H was added to approximately 20 mL excessand a 2.0 V dc potential was applied until thenickel color had disappeared from the solution (thisrequired up to three days, with an additional 5 mLof NH40H being added each day). The electrodewith the nickel deposit was removed from the solu-tion, rinsed with H2 0 and the nickel was dissolvedby heating in 40 g of 10 mol/L HNO 3 . The nickelsolution was evaporated to dryness, 20 g of 5 mol/L HC1 were added and the solution was again

349



evaporated to dryness. The 5 mol/L HCI additionand evaporation to dryness was repeated. Theresidue was dissolved in 10 g of 5 mol/L HCl andall but 4 g of this was evaporated. Thirty g of 9.5mol/L HCl were added and the resulting solutionwas passed through two anion exchange columnsin series (each column consisted of a 5-mL plasticsyringe filled to 5 mL with AGI X 8, 100-200 meshresin and cleaned with 40 g of 9 mol/L HCl, 50 gof H20, and 10 g of 9 mol/L HCl). Ten g of 9mol/L HCI were used to rinse the beaker and com-plete the elution of the nickel from the column.The sample was converted to the nitrate form byadding sequentially 20, 15, and 10 g portions ofHNO3 to the beaker and evaporating the sample todryness between each addition of acid.

2.3 Preparation and Analysis of Separated IsotopeSolutions

The purified 5"Ni and 'Ni were transferred to500-mL Teflon bottles and diluted to approxi-mately 400 g with HNO 3 (1+49). The purified 62Niwas transferred to a 2-L Teflon bottle and dilutedto approximately 1700 g with HNO 3 (1+49).

A preliminary assay of the nickel concentrationof each separated isotope solution was accom-plished by isotope dilution mass spectrometry.Two weighed portions of the 'Ni separated iso-tope solution were spiked with known amounts ofnatural nickel. Two weighed portions of each ofthe other two separated isotopes (58Ni and 62Ni)were spiked with weighed portions of the fi°Ni sep-arated isotope. After mixing, evaporation and dilu-tion to 1 mg Ni/g of solution with HNO 3 (1+49),the 5 8Ni/60 Ni or 6 2Ni/6'Ni ratios were determinedby thermal ionization mass spectrometry. The con-centration of nickel was then calculated for eachsolution and used to determine the amount of eachseparated isotope solution required for the calibra-tion mixes. Samples of the three separated isotopesolutions were analyzed for impurity elements byinductively coupled plasma mass spectrometry(ICP-MS).

Samples, equivalent to approximately 1.5 mg ofNi, were spiked with 1.5X(10-' g each of 206Pb,20 3TI, 20111g, 1

9 5 Pt, 18 3 W, '45 Nd, '42 Ce, 37Ba, '2 5Te,

123 Sb, 113 In, ttlCd, "°0Pd, 97Mo, 9'Zr, 87 Rb, 86 Sr , 73 Ge,7"Ga, 67Zn, 65Cu, 5 7 Fe, 53Cr, 5 0V, 47 Ti, 26Mig, and 6Li.

The solutions were diluted to 1 mg Ni/mL withHNO3 (1+49). Table 1 shows the results of theseisotope dilution analyses as well as concentrationsestimated from relative sensitivity factors (rsf). Thersf values are derived from a mass vs sensitivity

response curve obtained from an external standardcontaining 20 elements spaced across the massscale from Li to Hg. Table 2 shows the results ofthe ICP-MS analyses of SRM 986 and the same

Table 1. Analysis of impurities in the nickel separated isotopes

Element Spike isotope 58Ni 60Ni 62Ni

Concentrations in ttg/g, determined by isotope dilution

PbTlHgPtWNdCeBaTeSbInCdPdMoZrSrRbGaGeZnCuFeCrVTiMgLi

206Pb2 0 3

T120 1Hg'95Pt183w

145

Nd

142

Ce

'37Ba'25Te123Sb113 In

"'Cd1"0Pd97mo

9'Zr36 5r86 Sr87Rb7'Ga73 Ge67 Zn65 cu57Fe53 Cr50v

47Ti2 6

Mg6Li

1.30.166.50.90.10.10.10.60.90.10.91.61.520.10.30.320.32

0.7333

(14

2.20.129.4210.20.20.510.5220.520.20.41

2(5

<12<35

0.4134

<15

3.10.32

151

0.10.30.220.10.1320.70.50.20.20.4I1a

<160.613I

(20

Concentrations in jxg/g, determined by rsf and/or external std

U 0.1 0.1 0.1Th 0.1 0.1 0.1Bi 0.1 0.1 0.1Au 0.2 0.2 0.2Ir-I 0.3 0.3 0.3Sn 0.2 0.9 0.8Ag 0.3 0.1 0.1Rh 0.3 0.1 0.1Ru 0.3 0.3 0.3Y 0.3 0.1 0.1Se 1 4 1As 1 1 1Cob <240 <2000 13Mn 1 2 1Sc 1 1 0.1Al (3 (3 <3Na (10 <10 <10B 1 1 1Be 1 1 1

a Matrix interference.b "8Ni and '"Ni tails at 59Co.

350



Table 2. Analysis of impurities in the pure nickel (SRM 986)and the doped and purified nickel standard

Element Pure nickel Doped and purified Ni

Isotope dilution External standard in Ni

Pb (0.4 2Tl <0.2 <0.2Nd <0.8 (0.7Ba (0.4 9Te (0.6 (0.5Sn 2 0.6Cd (0.3 (0.2Pd 9 (0.1Sr (0.1 (0.1Se (2 <2Zn (0.7 (0.7Cu (3 <1Cr 0.7 76Mg (0.1 4

External standard in Ni

Au (2 (2Pt (0.8 17Co (4 <4Mn 0.4 9Ca (4 (4Al 4 11

nickel material doped with 0.1 % of natural Pb, Au,Pt, Ba, Cd, Pb, Zn, Cu, Co, Fe, Mn, Cr, Ca, Al,and Mg, and then cleaned up with the same separa-tion procedure developed for the separated nickelisotopes. The pure nickel was analyzed by a combi-nation of stable isotope dilution and comparison toan external standard. The doped and cleaned sam-ple was analyzed only by comparison to the exter-nal standard. The external standard was made byadding 10 ppm (10 pg/g Ni) of each of the dopingelements to the pure nickel solution, thus produc-ing a matrix matched standard.

2.4 Assay of Separated Isotope Solutions

Four weighed portions containing approxi-mately 1.7 mmol of nickel were withdrawn fromeach separated isotope solution in the followingmanner: a polyethylene stopper with a 20-cmTeflon needle inserted through it was used to re-place the cap on the bottle. A 20-mL allpolypropylene-polyethylene syringe was attachedto the hub of the needle and the desired amount ofsolution was withdrawn. The syringe was then dis-connected from the hub and the tip was cappedwith a plastic cap. Any static charge that might be

present on the plastic syringe was dissipated bywiping it with a damp lintless towel and placing iton the balance pan that was surrounded by severalpolonium anti-static sources. The syringe and con-tents were weighed on a semimicro balance to±0.02 mg. The solution was then delivered fromthe syringe into a 600-mL pyrex beaker and thesyringe was again capped, wiped and weighed. Theweight of the sample was determined from theweight of the syringe before and after the deliveryof the sample. Two assay samples were withdrawnfrom each solution before after the calibration sam-ples and two after to ensure that no change in con-centration occurred during the time interval (about6 h) required for the aliquanting. Each weighedsample was assayed as follows: the sample wasevaporated to dryness and converted to the chlo-ride by adding 10 mL of 4 mol/L HCO and evapo-rating slowly to dryness. The addition of 10 mL of4 mol/L HC1 and the evaporation were repeatedtwo more times. Two g of 4 mol/L HCO, 4 g ofammonium citrate solution (prepared by dissolving25 g of (NH4 )2HC2H507 in 200 mL of water, filter-ing and diluting with water to 250 mL) and 250 mLof water were added to the sample. A weighedportion of dimethylglyoxime reagent solution (pre-pared by dissolving 10 g of dimethylglyoxime inn-propanol, filtering and diluting with n-propanolto 1 L) equal to the amount required to form nickeldimethylglyoxime and 10 g excess was added toeach assay sample. Ten g of dimethylglyoximereagent solution and 30 g of n-propanol wereadded to each blank. The sample was heated in awater bath maintained at 65+2 'C. Five drops of 4mol/L NH40H were added, the sample was stirredwith a glass stirring rod and after 5 min the addi-tion of ammonia and stirring were repeated. Theaddition of 4M NH40H was repeated, graduallyincreasing the number of drops to ten, until the Niwas completely precipitated as determined by thesolution remaining colorless with the addition ofNH40H. Three more additions of ten drops of am-monia were made to ensure the precipitation of thenickel. The sample was removed from the waterbath 30 min after the last ammonia addition. A totaltime of approximately 3 h was required for the pre-cipitation. The nickel dimethylglyoxime crystalsthat are formed by this procedure are relativelylarge when compared to the usual method of pre-cipitation.

The sample was allowed to stand approximately48 h and then was filtered through a tared 15 mLglass filtering crucible of medium porosity. Asmuch of the nickel dimethylglyoxime crystals as

351



possible was transferred to the crucible using a wa-ter wash. The material in the crucible was washedthree times with water (a total of 55-60 g of waterwas used for transfer and washing). The crucibleand contents were dried at 150 'C for 16 h. Thefiltrate was transferred back to the original beakerand reserved for the determination of dissolved anduntransferred nickel.

The filtering crucible and contents were cooledin a desiccator, transferred to the case of a mi-crobalance and allowed to stand for at least 2 h.The crucible and contents were weighed to±0.002 mg. A combination blank and buoyancycorrection was made by averaging three cruciblesthat had been used to filter blank samples whichhad been carried through the procedure. Thecrucible and contents were dried an additional 3 hat 150 0C and the cooling and weighing were re-peated. The additional drying, cooling and weigh-ing were repeated until constant weight wasreached. The air weight of the nickel dimethylgly-oxime was then determined and converted to vac-uum weight using 1.606 for the density of58NiCH1 4O4 N4 , 1.617 for the density of6 0NiC8H,4O4N4 , and 1.628 for the density of62 NiC8H, 4O4N4 . These densities were calculated byassuming that they were proportional to the den-sity of natural NiCgH1 404 N4 , 1.611, in the same re-lationship as their molecular weight. The vacuumweight of the nickel dimethylglyoxime was con-verted to millimoles of nickel using a calculatedatomic weight for nickel and the 1987 atomicweight values for carbon, hydrogen, oxygen, andnitrogen. The formula weights used were 288.1616for the 58NiCH 140 4N4, 290.1511 for 6°NiC8 H1 404 N4,and 292.1279 for 6 2NiC8HI40 4N4.

The filtrate from the precipitation of the6 2NiC8HI404N4 was spiked with approximately 0.5jtmol of 'Ni and the filtrate from the 5"Ni and 'Niprecipitation were spiked for determining solubleand untransferred nickel by isotope dilution massspectrometry with approximately 0.5 ,umol of 62Ni.After adding the spike, the pH of the filtrate solu-tion was adjusted to 1.65±0.05 with 4 mol/L HCL.The solution was heated for 2 h to ensure equilibra-tion of the spike and sample nickel. After cooling,this solution was passed through a cation exchangecolumn (7.0X0.45 cm filled to 4.5 cm with AG50X8, 100-200 mesh cation exchange resin andcleaned with 25 g of 4 mol/L HCl and H2 0 untilthe eluate was neutral), washed with a few mL ofH2 0 and then 150 mL of 0.25 mol/L HCL. Thenickel was eluted with 18 g of 4 mol/L HC1 andthe eluate was evaporated to dryness on a hot

plate. A few drops of HNO 3 and HC10 4 wereadded to the residue and it was heated to help de-compose the organic material. The sample wasevaporated to dryness, cooled and the residue dis-solved in 10 g of H20. This solution was passedthrough the same cation exchange column aftercleaning the column as before and washed with afew mL of H 2 0 and 20 g of 0.25 mol/L HCL. Thenickel was eluted with 18 g of 4 mol/L HCl andthe eluate was evaporated to dryness on a hotplate. A few drops of HNO3 and HCl04 wereadded to the residue and it was heated to dryness.The residue was dissolved in a few drops of HNO 3(1+49) and the nickel isotopic ratios were deter-mined by thermal ionization mass spectrometry.The nickel found as soluble Ni was added to thenickel determined by gravimetry to yield the totalnickel in the sample. Table 3 shows the results ofthese analyses.

This method of determining the concentration ofnickel solutions was previously tested on solutionscontaining known amounts of nickel. Solutionswere prepared from high purity nickel metal. Thenickel concentration in five sets of four samples,each containing 1.58 to 1.79 mmol of nickel wasdetermined as described above. Comparison of thecalculated and measured concentrations detected apositive bias of about 0.03 percent, but this wouldhave a negligible effect on ratios.

2.5 Isotopic Analysis of the Separated IsotopeSolutions

Each of the three separated isotope solutions wasanalyzed eight times by operator #1 on instrument#1. The ion source was cleaned between analysesof the solutions as a precaution against the possibil-ity of cross-contamination from the source parts.Preliminary measurements showed that the differ-ent separated isotope solutions could be analyzedback-to-back on the same source with no de-tectable cross-contamination.

As mentioned in section 2.1, preliminary mea-surements to evaluate possible sources of system-atic errors indicated that a small natural nickelbackground could be present on some filaments.Acid leaching of the filaments and optimized fila-ment outgassing procedures minimized the magni-tude and frequency of this problem. Thisoptimization was achieved by monitoring the 5"Nion increasingly smaller sample sizes of the 60Ni sep-arated isotope, and measuring the nickel signalfrom filaments cleaned and degassed under variousconditions using an ion counting detection system.

352



Table 3. Concentration of nickel in separated isotope solutions

Weight Ni from Ni from Total WeightSample NiDMGa NiDMG filtrate Ni sample Concentration

Solution no. (g) (mmol) (mmol) (mmol) (g) (mmol Ni/g)

"Ni-58" 1 0.487793 1.692775 0.000444 1.693219 19.64900 0.08617332 0.492528 1.709207 0.000480 1.709687 19.83881 0.08617893 0.486624 1.688720 0.000494 1.689214 19.60097 0.08618014 0.488016 1.693550 0.000544 1.694094 19.65837 0.0861767

Total 0.0861773SD 0.0000030

"Ni-60" 1 0.506874 1.746931 0.000449 1.747380 20.31729 0.08600462 0.497094 1.713225 0.000559 1.713784 19.92565 0.08600893 0.501435 1.728185 0.000451 1.728636 20.09839 0.08600874 0.599604 1.725323 0.000563 1.725323 20.06610 0.0860100

Total 0.0860081SD 0.0000024

"Ni-62" 1 0.509754 1.744969 0.000397 1.745366 101.53728 0.017112672 0.512447 1.754186 0.000465 1.754651 102.53728 0.017112323 0.510848 1.748713 0.000458 1.749171 102.21630 0.017112454 0.511917 1.752374 0.000497 1.752871 102.43149 0.01711264

Total 0.01711253SD 0.00000017

a NiDMG=nickel dimethylglyoxime.

All filaments used for measurement of the sepa-rated isotopes were examined in the mass spec-trometer prior to sample loading. In addition tomonitoring all nickel masses for contamination,5 6Fe and 6̀ Zn were examined to insure the absenceof isobaric interferences. Only those filamentswhich showed no detectable (< 1 X 10-6 A) signalwere used for the measurement of the separatedisotopes.

The corrected isotopic compositions of the sepa-rated isotopes are given in Table 4.

mol/L HNO 3 and evaporated to dryness on a hotplate. The residue was dissolved and diluted withHNO 3 (1+49) to 5 mg Ni per gram of solution.After thorough mixing, a portion of this solutionwas diluted with HNO3 (1+49) to 1 mg Ni pergram of solution and transferred to a small Teflonbottle. The isotopic compositions of the calibrationmixes are given in table 5.

Table 4 Isotopic composition of separated nickel isotopes usedin cal o'ration samples

2.6 Preparation of Calibration Samples

Five calibration samples were prepared by mix-ing weighed portions of "Ni-58", "Ni-60", and "Ni-62" solutions. The portions were withdrawn fromthe bottles and weighed in the manner previouslydescribed for the assay of the solutions. The por-tions weighed from 4.9 to 110 g and each wasweighed to +0.05 mg. It is therefore estimated thatthe weighing error for each mix should not exceedtwo parts in 105. To minimize any significant possi-bility of change in concentration of the isotope so-lutions with time, the portions for the calibrationmixes were withdrawn from the bottles betweenthe samples taken for assay, over a period of about6 h.

Each calibration mix was thoroughly mixed, thesides of the beaker were washed with H20 and 0.3

Separatedisotope

"Ni-58"

"Ni-60"

"Ni-62"

Isotopic compositionAtom percent 2-sigma uncertainty

58 99.87368460 0.11672161 0.00212662 0.00536464 0.002105

58 0.22687660 99.70270961 0.02717562 0.03380664 0.009434

58 0.35595460 0.47822261 0.09643362 99.00738064 0.062011

0.0004780.0003080.0001560.0002380.000164

0.0007000.0011760.0005280.0002560.000456

0.0004540.0006980.0004380.0012720.000152

353


Journal of Research of the National Institute of Standards and Techr

Table 5. Isotopic composition of calibration samples

Ni 58Ni WONi 62NiSolution Isotope Weight from from from from Total Total Total Ratio Ratio

no. solution solution solution solution solution solution ssNi W0Ni 62Ni 58 Ni/60Ni 62Ni/60Ni(g) (mmol) (mmol) (mmol) (mmol) (mmol) (mmol) (mmol)

1 "Ni-58" 19.16757 1.651809 1.649723 0.001928 0.000089"Ni-60" 7.22959 0.621803 0.001411 0.619955 0.000210"Ni-62" 5.00563 0.085659 O.O00305 0.000410 0.084809

1.651439 0.622292 0.085107 2.653798 0.1367642 "Ni-58" 18.41320 1.586800 1.584795 0.001852 0.000085

"Ni-60" 7.02949 0.604593 0.001372 0.602796 0.000204"Ni-62" 5.07309 0.086813 0.000309 0.000415 0.085951

1.586476 0.605063 0.086241 2.622002 0.1425323 "Ni-58" 18.46549 1.591306 1.589296 0.001857 0.000085

"Ni-60" 7.12359 0.612686 0.001390 0.610865 0.000207"Ni-62" 5.00786 0.085697 0.000305 0.000410 0.084846

1.590991 0.613132 0.085139 2.594858 0.1388594 "Ni-58" 18.42601 1.587904 1.585898 0.001853 0.000085

"Ni-60" 7.19758 0.619050 0.001404 0.617210 0.000209"Ni-62" 4.93759 0.084494 0.000301 0.000404 0.083656

1.587603 0.619467 0.083950 2.562852 0.1355205 "Ni-58" 18.11755 1.561322 1.559349 0.001822 0.000084

"Ni-60" 7.14866 0.614843 0.001395 0.613015 0.000208"Ni-62" 5.08622 0.087038 0.000310 0.000416 0.086174

1.561054 0.615253 0.086466 2.537254 0.140536

2.7 Isotopic Analyses of the Calibration Mixes and 3. Results and Discussionthe Reference Sample

Two sets of analyses of the calibration mixes and The results of the measurements of the calibra-tion mixes are shown in Table 6. Table 7 summa-reference sample were performed on istrument w u rizes the observed and corrected nickel isotopic

and instrument #2. Instrument # 1 was used to raisfrteeeeneamlfoopaosIadmeasure 28 samples of the reference material and 4 resp v a e the absole isto pIcsamples each of mixes 1-5. Instrument #2 was used

to eaur21saplsoteefrec m l aabundance ratios for nickel and their uncertainties.to6measure21 samples of mixes The ramplef n e matalynd Table 8 gives summary calculations for the refer-6 samples of mixes1 T e aml werefana e ence sample. The atomic weight is calculated fromin a random pattern of mixes and the reference teaslt stpcaudneb umn hsample. Deviations from a totally random pattern . .were the initial successive analyses of the reference product of the nuclidic masses [17] and the corre-material to insure instrument and statistical control sponding atom fractions.

andthesame mix solution not As mentioned in the introduction, the presentlybeandatheyrequiremen t thatessrecommended atomic weight of nickel is

58.69+±0.01. The IUPAC Commission on Atomic

Table 6. Determination of mass spectrometer bias

Calibration Isotopic ratios Correction factorssample Calculated Operator I Operator 2 Operator 1 Operator 2

no. SSNi/6Ni 6 2 Ni/60Ni 5 SNi/WNi 62 Ni/60Ni iSNi/6WNi 6 2Ni/60 Ni 5 8 Nit6ONi 62Ni/60Ni 5sNi/wNi 62Ni/WNi

1 2.653798 0.136764 2.695878 0.134686 2.694002 0.134730 0.984391 1.015430 0.985077 1.0151022 2.622002 0.142532 2.663804 0.140356 2.659318 0.140512 0.984307 1.015506 0.985968 1.0143783 2.594858 0.138859 2.635760 0.136696 2.633919 0.136854 0.984483 1.015826 0.985170 1.0146514 2.562852 0.135520 2.604370 0.133421 2.600775 0.133575 0.984056 1.015734 0.985419 1.0145635 2.537254 0.140536 2.577182 0.138349 2.575263 0.138502 0.984508 1.015817 0.985241 1.014687

Average correction factors 0.984349 1.015663 0.985375 1.014676

Correction factor standard deviations 0.000182 0.000183 0.000354 0.000267

354



Table 7. Corrected isotopic ratios of the SRM

Ratio Operator 1 Operator 2 Average ratio

58Ni/uONi 2.5961255 2.5959968 2.59606126 1Ni/6 0Ni 0.0434680 0.0434703 0.04346916 2Ni/6 0Ni 0.1385976 0.1386025 0.138600164Ni/J Ni 0.0353113 0.0352784 0.0352949

Table 8. Atomic weight, atom percent, and isotopic ratios of nickel

Due to Due to Due toDue to uncertainty uncertainty uncertainty

Total uncertainty in mass in mix in nuclidicuncertainty in assay spectrometry preparation mass

Value (2-sigma) (2-sigma) (2-sigma) (2-sigma) (2-sigma)

Atomic weight 58.693353 0.000147 0.000091 0.000104 0.000049 0.000002

Atom percent58Ni 68.076886 0.005919 0.004045 0.003762 0.0021256ONi 26.223146 0.005144 0.003776 0.002724 0.0021866'Ni 1.139894 0.000433 0.000138 0.000402 0.00008362Ni 3.634528 0.001142 0.000615 0.000817 0.00050864 Ni 0.925546 0.000599 0.000317 0.000420 0.000287

Isotopic ratios58 Ni/60Ni 2.596061 0.000728 0.000525 0.000410 0.00029361Ni/60Ni 0.043469 0.000015 0.000004 0.000014 0.00000462 Ni/60Ni 0.138600 0.000045 0.000028 0.000025 0.00002564Ni/60Ni 0.035295 0.000024 0.000014 0.000015 0.000013

Weights and Isotopic Abundances lists this as oneof the least well known atomic weights and alsoone of the few remaining elements where theatomic weight is based, at least in part, on chemicaldeterminations made in the early 1920s. Based onthis work, a value of 58.6934±0.0002 will be rec-ommended which is two orders of magnitude moreprecise and, more importantly, is now known on anabsolute scale.

The reference material is issued by the NIST Of-fice of Standard Reference Materials as SRM 986,Nickel Metal Isotopic Standard, and is certified forisotopic composition.

4. Acknowledgments

We are greatly indebted to Keith R. Eberhardtand Susannah B. Shiller of the Statistical Engineer-ing Division, who spent more time than they hadavailable on the statistical analysis of the data andto William A. Bowman III, Ronald W. Shideler,and the late George M. Lambert for instrumental

maintenance support. We are especially grateful toProfessor John De Laeter of Curtin Universitywho shared with us his experiences and those of hiscolleagues with the Aerosil procedure and pro-vided the material used in this research.

About the authors: J. W Gramlich was associatedwith the High Accuracy Measurement Program in theInorganic Analytical Research Division of NBS formany years as a research chemist. He is currently withthe U.S. Department of Energy, New Brunswick Lab-oratory, Argonne IL 60439. L. A. Machlan was asso-ciated as a research chemist with the Atomic WeightsProgram in the Inorganic Analytical Research Divi-sion of NBS for many years. He is currently retiredbut continues to serve NIST as a contractor. L L.Barnes is a research chemist with the Inorganic Ana-lytical Research Division in the NIST Center for An-alytical Chemistry. P. J. Paulsen is a research chemistin the Inorganic Analytical Research Division in theNIST Center for Analytical Chemistry.

355



5. References

[I] Shields, W. R., Garner, E. L., and Dibeler, V. H., J. Res.Natl. Bur. Stand. (U.S.) 66A, 1 (1962).

[2] Powell, L. J., Murphy, T. J., and Gramlich, J. W., J. Res.Natl. Bur. Stand. (U.S.) 87, 9 (1982).

[3] Shields, W. R., Murphy, T. J., Garner, E. L., and Dibeler,V. H., J. Amer. Chem. Soc. 84, 1519 (1962).

[4] Shields, W. R., Murphy, T. J., and Garner, E. L., J. Res.Natl. Bur. Stand. (U.S.) 68A, 589 (1964).

[5] Catanzaro, E. J., Murphy, T. J., Garner, E. L., andShields, W. R., J. Res. Natl. Bur. Stand. (U.S.) 68A, 593(1964).

[6] Shields, W. R., Murphy, T. J., Catanzaro, E. J., andGarner, E. L., J. Res. Natl. Bur. Stand. (U.S.) 70A, 193(1966).

[7] Catanzaro, E. J., Murphy, T. J., Garner, E. L., andShields, W. R., J. Res. Natl. Bur. Stand. (U.S.) 70A, 453(1966).

[8] Catanzaro, E. J., Murphy, T. J., Shields, W. R., andGarner, E. L., J. Res. Natl. Bur. Stand. (U.S.) 72A, 261(1968).

[9] Catanzaro, E. J., Champion, C. E., Garner, E. L.,Marinenko, G., Sappenfield, K. M., and Shields, W. R.,Natl. Bur. Stand. (U.S.) Spec. Publ. 260-17; 1970 Febru-ary, 60 pp.

[10] Catanzaro, E. J., Murphy, T. L, Garner, E. L., andShields, W. R., J. Res. Natl. Bur. Stand. (U.S.) 73A, 511(1969).

[11] Gramlich, J. W., Murphy, T. J., Garner, E. L., andShields, W. R., J. Res. Natl. Bur. Stand (U.S.) 77A, 691(1973).

[121 Barnes, I. L., Moore, L. J., Machlan, L. A., Murphy,T. J., and Shields, W. R., J. Res. Nat]. Bur. Stand. (U.S.)79A, 727 (1975).

[13] Garner, E. L., Murphy, T. J., Gramlich, J. W., Paulsen,P. J., and Barnes, I. L., J. Res. Natl. Bur. Stand. (U.S.)79A, 713 (1975).

[14] Dunstan, L. P., Gramlich, J. W., Barnes, I. L., and Purdy,W. C., J. Res. Natl. Bur. Stand. (U.S.) 85, 1 (1980).

[15] Moore, L. J., Murphy, T. J., and Barnes, I. L., J. Res.Natl. Bur. Stand. (U.S.) 87, 1 (1982).

[16] Machlan, L. A., Gramlich, J. W., Powell, L. J., andLambert, G. M., J. Res. Natl. Bur. Stand. (U.S.) 91, 323(1986).

[17] Wapstra, A. H., and Audi, G., Nucl. Phys. A432, 1 (1985).[18] White, J. R., and Cameron, A. E., Phys. Rev. 74, 991

(1948).[19] Baxter, G. P., and Parsons, L. W., J. Amer. Chem. Soc.

43, 507 (1921).[20] Baxter, G. P., and Hilton, F. A., Jr., J. Amer. Chem. Soc.

45, 694 (1923).[21] Baxter, G. P., and Ishimaru, S., J. Amer. Chem. Soc. 51,

1729 (1929).[22] Baxter, G. P., and Dorcas, M. J., J. Amer. Chem. Soc. 46,

357 (1924).[23] Barnes, 1. L., Garner, E. L., Gramlich, J. W., Machlan,

L. A., Moody, J. R., Moore, L. J., Murphy, T. J., andShields, W. R., Geochim. Cosmochim. Acta Suppl. 4, 2,1197 (1973).

[24] Atomic weights of the elements 1987. Pure Appl. Chem.60, 841 (1988).

[25] Peiser, H. S., Anal. Chem. 57, 511A (1985).[26] Kuehner, E. C., Alvarez, R. A., Paulsen, P. J., and Mur-

phy, T. J., Anal. Chem. 44, 2050 (1972).[27] Shideler, R. W., and Gramlich, J. W., in preparation.

356




The Absolute Isotopic Composition andAtomic Weight of Terrestrial Nickel


J. W. Gramlich, E. S. Beary, Twenty-nine samples of high-purity tainty for SRM 986 is applicable to ter-L. A. Machlan, and I. L. Barnes nickel metals, reagent salts and minerals, restrial nickel samples. The atomic

collected from worldwide sources, have weight calculated for SRM 986 isNational Institute of Standards been examined by high-precision isotope 58.69335±0.00015 121. The currentlyand Technology, ratio mass spectrometry for their nickel recommended IUPAC value for terres-

Gaithersburg, MD 20899 isotopic composition. These materials trial nickel is 58.69±0.01.Gaithersburg, MD 20899 were compared directly with SRM 986,

certified isotopic standard for nickel, us- Key words: absolute ratio; atomicing identical measurement techniques weight; isotopic abundance; mass spec-and the same instrumentation. This sur- trometry; nickel; SRM 986; Standardvey shows no statistically significant ' '. X .

variaionsamongthe ample invsti- Reference Materials; terrestrial samples.variations among the samples investi-gated, indicating that the certifiedatomic weight and associated uncer- Accepted: September 1, 1989

1. Introduction

For several decades, the Inorganic AnalyticalResearch Division of the National Institute of Stan-dards and Technology has conducted a continuingprogram of absolute isotopic abundance andatomic weight determinations using high-precisionisotope ratio mass spectrometry. Although thisprogram has yielded extremely accurate atomicweights for reference samples, the uncertainty as-sociated with a generally useful atomic weight islimited by the isotopic variations among materialsreadily available to the scientific community or bya lack of knowledge regarding such variations.This work compares the isotopic composition andatomic weight of nickel from a global population ofterrestrial nickel sources, and samples that havebeen purified industrially into nickel metal andreagent salts, to a Standard Reference Material(SRM 986) of accurately known isotopic composi-tion and atomic weight [1]. This reference sample

is pure nickel powder, lot F-3625 obtained fromAtomergic Chemicals Corporation' and is availablefrom the Office of Standard Reference Materials atNIST.

2. Experimental Procedure2.1 Chemical Purification of the Samples

Nickel ores, such as niccolite, linnaeite, gersdorf-fite, and carrollite, were obtained from the U.S.National Museum. These samples were representa-tive of ores from mines in North America, South

' Certain commercial equipment, instruments, or materials areidentified in this paper to specify adequately the experimentalprocedure. Such identification does not imply recommendationor endorsement by the National Institute of Standards andTechnology, nor does it imply that the materials or equipmentidentified are necessarily the best available for the purpose.

357

Volume 94, Number 6,

Journal of Research of the NationalNovember-December 1989

Institute of Standards and Technology

America, Europe, Asia, and Africa (table 1). Alsolisted in table 1 are commercially purified nickelmetal and nickel compounds from European andAmerican sources which were procured for thisstudy. The reagent nickel compounds required lit-tle pretreatment, while the ores required extensivechemical separations to obtain nickel essentiallyfree of contamination, as required for this high-pre-cision mass spectrometric technique.

Samples of nickel ore weighing 200-600 mgwere treated with about 10 g of aqua regia (3:1HCI, HNO3), covered and heated until the reactionsubsided. Although some undissolved material re-mained, the samples were evaporated to dryness,and then taken up in 10 g of HCI ( 1+2). The sam-ples were then filtered into 400-mL beakers and theprecipitate was discarded. Ten to fifteen g of am-monium citrate solution (400 g/L) were added toeach sample filtrate to complex the Fe, and the re-sulting mixture was diluted to 250 mL. Thirty-fiveg of 1% dimethylglyoxime solution in n-propanolwere added to each beaker and heated slowly toabout 70 'C while adding NH 4 0H until the pH wasbetween 8 and 9. The samples were heated for 30min while maintaining the temperature at 70 'C,then cooled and filtered. The precipitate was redis-solved in 4 g HNO3 (1+3), covered and heateduntil all the nickel dimethylglyoxime was dis-solved. Twenty-five g of water were added to thesamples as a preparation for cation exchange chro-matography. Five mL of AG 50 X 8, 100-200 meshcation exchange resin were placed in plasticcolumns about 1 cm in diameter, and the resin wascleaned using 25 g of 3 N HNO 3 and then H2 0 untilthe eluate was neutral. After the sample was loadedonto the column the impurities were eluted using25 g of 0.3 N HNO3 . The Ni was eluted with 15 gof 3 N HNO3 , and evaporated to a small volume(about 1 mL). Four g of 9 N HCl were added as apreparation for anion exchange chromatography.Five mL of AG 1 X 8, 100-200 mesh anion ex-change resin was added to plastic columns 1 cm indiameter, and the resin was cleaned using 9 NHCL, then neutralized with water. The Ni waseluted using 12-15 g of 9 N HCI, while impuritiesremained on the column. The purified Ni solutionwas evaporated to dryness and redissolved in 4 g of5 N HCL. The dissolved sample was then diluted toa 35-mL volume and neutralized with NH4 0H, fol-lowed by 2 mL excess NH 4 0H. This solution wasthen electrolyzed at 2-2.5 V using platinum gauzeelectrodes until the Ni color disappeared from thesolution. The electrode with the Ni deposit wasremoved from the solution and rinsed with H2 0,

allowed to air dry, and the electrode was weighed.The Ni was removed using concentrated HNO 3 ,and the electrode was air dried and reweighed.Typically, 50-150 mg of Ni metal were recoveredfrom each ore. Sample solutions using the purifiedNi were prepared in HNO 3 (1+49), at a concentra-tion of 1 mg Ni/g.

The four nickel metal samples and four nickelcompounds were simply dissolved, converted tothe nitrate, and sample solutions were prepared as 1mg Ni/g in HNO 3.

2.2 Mass Spectrometry

Isotope ratio measurements were performed onthe same NIST designed mass spectrometer thatwas used to determine the absolute isotopic abun-dance and atomic weight of a reference sample ofnickel (SRM 986) [1]. The measurements on terres-trial nickel materials, reported in this paper, wereconducted in conjunction with, but following themass spectrometric measurement associated withthe certification of SRM 986. Analyses of SRM 986that were used to assure measurement control andto correct for isotopic fractionation in this experi-ment, were also used in the calculations used tocertify SRM 986. The bias-corrected isotopic ratiosreported in this paper are thus based on the exten-sive measurements made on SRM 986 both to cer-tify the SRM and in this work.

Samples containing approximately 5 jug of nickel(as NiNO 3 in 2% HNO3 ) were loaded onto a plat-inum filament with a mixture of silica gel- AlCl3-H3 PO4 as an ionization enhancement agent.Samples were analyzed by thermal ionization massspectrometry using procedures and parametersidentical to those used for the certification of SRM986. The details of filament preparation, sampleloading and the analysis procedures have been pub-lished [1].

3. Results and Discussion

The corrected abundance ratios for the 29 sam-ples are given in table 2. The same correction fac-tors as used for SRM 986 were used. In the designof the survey both mineral type and geographicorigin were considered to obtain the widest possi-ble distribution.

The isotopic abundances reported in this workindicate no significant trend or variation in the iso-topic composition of nickel in terrestrial sources.

358



Table 1. Nickel reagent-mineral survey (description of samples)

NIST no. Description of sample Source

NIST SRM 986 (reference sample)

Linnaeite (with tetrahedraite)U.S. Natl. Museum no. B2981

MilleriteU.S. Natl. Museum no. 66501

PentlanditeU.S. Natl. Museum no. 96387

GersdorffiteU.S. Natl. Museum no. 113044

GersdorffiteU.S. Natl. Museum no. B3310

CarrolliteU.S. Natl. Museum no. 122349

RammelsbergiteU.S. Natl. Museum no. C693

GersdorffiteU.S. Natl. Museum no. 120381

Millerite (with gaspeite and polydimiteU.S. Natl. Museum no. 121729

MilleriteU.S. Natl. Museum no. B1596

MilleriteU.S. Natl. Museum no. 113065

Niccolite (and vaesite)U.S. Nati. Museum no. 86655

NiccoliteU.S. Natl. Museum no. 105150

NiccoliteU.S. Natl. Museum no. 121284

Niccolite and ChalcanthiteU.S. Natl. Museum no. 103642

SiegeniteU.S. Natl. Museum no. B3169

Niccolite (with rammelsbergite)U.S. Natl. Museum no. 114458

RammelsbergiteU.S. Natl. Museum no. 117334

RammelsbergiteU.S. Natl. Museum no. 94556

HeazlewooditeU.S. Nat. Museum no. 115427

Pentlandite in PyrrhotiteU.S. Natl. Museum no. Ri 1297

LinnaeiteU.S. Natl. Museum no. 96797

NIST SRM 772 (magnetic moment std.)Ni metal, 99.999% pure

Ni metal, wire

Ni metal, rodlot no. 02801

NiO, 99.999% purelot no. N121G0541

NiCI2 -6H2 0lot no. 634B740817

NiC12-6H2 0lot no. 631216/1

NiSO4-6H2 0lot no. 1453

Schwabengrobe(near Mussen), Germany

Gap Mine, LancasterCounty, PA

Worthington Mine,Sudbury, Canada

Temagamy Island,Ontario, Canada

Merkur Mine, Ems,Germany

Kambove, Congo

Eisleben, Saxony,Germany

Ait Ahmane, Bou Azzer,Morocco

Pqfuri Native Trust,Transvaal, S. Africa

Victoria Mine,Littfeld, Germany

Temagami, Ontario,Canada

Green-Meehan Mine,Cobalt, Ontario, Canada

Alistos, Sinaloa,Mexico

Iran

Schneeberg, Saxony,Germany

Schwaben Mine,Siegerland, Germany

La Sorpresa Mine,Tapacari, Bolivia

Mohawk #2 Mine,Keweenaw County, Michigan

Hudson Bay Mine,Cobalt, Ontario, Canada

Lord Brassey Mine,Heazlewood, Tasmania

Stare Ransko Near-Chotebor,Bohemia, Czechoslovakia

Mineral Hill Mine,Sykesville, Maryland

Leico Industries Inc.,New York, NY 10019

Jarrell Ash

Spex

Atomergic Chemicals, Inc.

Merck

Union Chimique, Belge,SA

Union Chimique, Belge,SA

359

0

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29



Table 2. Isotope ratios for the mineral samples corrected to SRM 986

No. Identification 5 8Ni/6ONi 61Ni/6ONi 6 2Ni/6Ni 64Ni/6uNi

1 SI-B2981 2.596679 0.043485 0.138603 0.0353552 SI-66501 2.596233 0.043458 0.138551 0.0353273 SI-96387 2.596016 0.043460 0.138632 0.0353274 SI-113044 2.596500 0.043472 0.138551 0.0353595 SI-B3310 2.595810 0.043469 0.138649 0.0353526 SI-122349 2.596492 0.043473 0.138555 0.0352827 SI-C693 2.596163 0.043451 0.138559 0.0352958 SI-120381 2.596368 0.043470 0.138571 0.0353089 SI-121729 2.595924 0.043463 0.138598 0.035286

10 SI-B1596 2.596326 0.043469 0.138597 0.03532611 SI-113065 2.596410 0.043493 0.138616 0.03533812 SI-86655 2.596075 0.043468 0.138628 0.03529313 SI-105150 2.596513 0.043467 0.138584 0.03533114 SI-121284 2.596502 0.043477 0.138634 0.03530115 SI-103642 2.596470 0.043469 0.138607 0.03533716 SI-B33169 2.596537 0.043458 0.138604 0.03533917 SI-114458 2.596135 0.043460 0.138559 0.03529018 SI-117334 2.596225 0.043461 0.138599 0.03532619 SI-94556 2.596113 0.043458 0.138603 0.03529320 SI-115427 2.596205 0.043486 0.138602 0.03536821 SI-R11297 2.596508 0.043451 0.138625 0.03535821 SI-R I1297 2.595943 0.043479 0.138570 0.03530522 SI-96797 2.596204 0.043478 0.138595 0.03532022 SI-96797 2.595841 0.043478 0.138556 0.03535323 SRM 772 2.595992 0.043487 0.138606 0.03533824 Jarrell Ash 2.595953 0.043487 0.138617 0.03532925 Spex Ni rod 2.595996 0.043478 0.138634 0.03532226 Atomergic 2.596071 0.043467 0.138615 0.03532427 MCl2-6H 2 0 2.596036 0.043479 0.138614 0.03532028 NiCI 2-6H2 0 2.596096 0.043476 0.138589 0.03531929 NiSO4 -6H20 2.596310 0.043480 0.138587 0.035319

To evaluate formally the data for constancy ofisotopic composition, we used an analysis of vari-ance (ANOVA) procedure to conduct a statisticaltest of the hypothesis that the materials have identi-cal isotopic ratios. The ANOVA procedure wascarried out in two ways: both including and ex-cluding the SRM 986 data obtained from the twodifferent instruments used in the absolute abun-dance ratio work. This was done to be sure thatany differences between the two instrumentswould not affect the conclusions.

For each of the four isotope ratios, the ANOVAtests showed that the variability among the mea-sured ratios for the minerals survey samples isnever significantly greater than variability of themeasurements of the SRM 986 material. Thus, theconclusion of this statistical test is that there is noevidence of real variation in the isotopic ratios forthe nickel samples surveyed. The ANOVA resultsare summarized in Table 3.

Table 3. Results of ANOVA tests of the hypothesis of equalityof isotopic ratios for SRM 986 and the other nickel samples

Isotope ANOVA Significanceratio F-ratio probability'

All data (includinginstrument #2):

58 Ni/6ONi 0.184 1.006'Ni/6ONi 0.067 1.0062Ni/60Ni 0.433 0.996 4 Ni/6 0Ni 1.157 0.32

Instrument #1 data only:

5 8Ni/6ONi 0.113 1.006 1Ni/ 60Ni 0.032 1.006 2Ni/6 0Ni 0.288 1.0064Ni/6ONi 0.649 0.87

a The significance probability is the probability of obtaining anF-ratio as large or larger than the observed value assuming thatthe null hypothesis (equality of isotope ratios) is true. The re-sults of the ANOVA test are generally considered not to showsignificant evidence of a difference unless the significance prob-ability is less than 0.05.

360



Graphical summaries of these results, in a some-what different form, are shown in figures 1-4.These figures, called "bihistograms," provide a vi-sual comparison of the measured isotope ratio datafor the SRM 986 material (as measured by instru-ment #1, see reference [1]) with the correspondingratios for the other nickel samples surveyed. Ineach figure the isotope ratios for SRM 986 are rep-resented in the "upside-down" histogram, with thehistogram of the mineral survey ratios shown on

02nii

LU

Lii

0.3

0.2

0.1

-0.1

-0.2 1 I I I I

2.594 2.59525 2.5965 2.59775 2.599MINERAL NI-58160 RATIO (TOP) - SRM NI-58160 RATIO (BOTTOM)

Figure 1. A bihistogram showing the values obtained for thessNi/WNi ratios (corrected to the absolute values of the ratiosobtained for SRM 986) vs those values. Shown on the top is ahistogram of the mineral SSNi/WNi ratios and on the bottom(inverted) is a histogram of the same ratios obtained for SRM986.

0zLI

0

U.

i4c-S'

0.4

0.3

0.2

0.1

-0.1

-0.2jl- I

0.0433 0.0434 0.0435 0.0436 0.0437MINERAL NI-61/60 RATIO (TOP) - SRM NI-61/60 RATIO (BOTTOM)

Figure 2. A bihistogram showing the values obtained for the6t Ni/6 0 Ni ratios (corrected to the absolute values of the ratiosobtained for SRM 986) vs those values. Shown on the top is ahistogram of the mineral UlNi/WNi ratios and on the bottom(inverted) is a histogram of the same ratios obtained for SRM986.

C.)z

MU

31:

Cc

0

0.2

0z

ILl0aU-

Lii

0.1

0

-0.1

-0.2 1 4 I E

0.1384 0.1385 0.1386 0.1387 0.1388MINERAL NI-62/60 RATIO (TOP) - SRM NI-62/60 RATIO (BOTTOM)

Figure 3. A bihistogram showing the values obtained for the6&Ni/WNi ratios (corrected to the absolute values of the ratiosobtained for SRM 986) vs those values. Shown on the top is ahistogram of the mineral 62 NiA"Ni ratios and on the bottom(inverted) is a histogram of the same ratios obtained for SRM986.

0.0351 0.0352 0.0353 0.0354 0.0355MINERAL NI-64160 RATIO (TOP) - SRM NI-64/60 RATIO (BOTTOM)

Figure 4. A bihistogram showing the values obtained for the64 Ni/60 Ni ratios (corrected to the absolute values of the ratiosobtained for SRM 986) vs those values. Shown on the top is ahistogram of the mineral MNi/uNi ratios and on the bottom(inverted) is a histogram of the same ratios obtained for SRM986.

top. The SRM 986 data always show at least asmuch variability as do the mineral survey data.This fact indicates that the observed variation inthe survey nickel samples is readily accounted forby measurement variability, and therefore it sup-ports the conclusion that there is no real variationin the isotopic composition of the nickel samples. Itappears in these figures that the data for the miner-als actually show less variation than that of theSRM. We beleive that this is only an indication that

361



the procedure for the analyses was even moretightly controlled by the time that these sampleswere analyzed.

Evidence in the literature also indicates no iden-tifiable variations in the atomic weight of nickelamong terrestrial sources [2]. However, many ofthe previous measurements are far less precise thanthis study, and thus many have not been able toidentity small isotopic variations. Further supportof isotopic homogeneity is supplied by a limitedstudy of the isotopic composition of lunar samples,directly compared to SRM 986, which also showedno significant variations relative to the SRM [3].

4. Acknowledgments

The authors are indebted to the staff of the U.S.National Museum for assistance in selecting theminerals used in this study. S. Goldich also pro-vided valuable help in the selection of these mater-als. We gratefully acknowledge the help of KeithEberhardt of the Statistical Engineering Divisionwith the statistical analysis of the data. We thankW. J. Barnes for his assistance in typing themanuscript.

About the authors: J. W Gramlich was associatedwith the High Accuracy Measurement Program in theInorganic Analytical Research Division of NBS formany years as a research chemist. He is currently withthe U.S. Department of Energy, New Brunswick Lab-oratory, Argonne IL 60439. E. S. Beary is a researchchemist with the NIST Inorganic Analytical ResearchDivision in the Centerfor Analytical Chemistry. L. A.Machlan was associated as a research chemist withthe Atomic Weights Program in the Inorganic Analyt-ical Research Division of NBS for many years. He iscurrently retired but continues to serve NIST as acontractor. L L. Barnes is a research chemist with theInorganic Analytical Research Division in the NISTCenter for Analytical Chemistry.

5. References

[1] Gramlich, J. W., Machlan, L. A., Barnes, I. L., andPaulsen, P. J., J. Res. Natl. Inst. Sci. Technol. 94, 347(1989).

[2] Element by element review of their atomic weights, PureAppl. Chem. 58, 695 (1984).

[3] Barnes, 1. L., Garner, E. L., Gramlich, J. W., Machlan,L. A., Moody, J. R., Moore, L. J., Murphy, T. J, andShields, W. R., Geochim. Cosmochim. Acta Suppl. 4, 2,1197 (1973).

362




Special Report on Standards for Radioactivity

Report on the 1989 Meeting of theRadionuclide Measurements Section of the

Consultative Committee on Standards for theMeasurement of Ionizing Radiations


Dale D. Hoppes This report describes the activities dis- tional Bureau of Weights and Measurescussed at the 10th meeting of Section II (Bureau International des Poids et

National Institute of Standards of the Consultative Committee on Stan- Mesures, BIPM), and progress at mem-

and Technology, dards for the Measurement of Ionizing ber laboratories.

Gaithersburg, MD 20899 Radiations (Comit6 Consultatif pour lesEtalons de Mesure des Rayonnements Key words: becquerel; comparisons oflonisants, CCEMRI) held in May 1989 activity measurements; "9Cd; Consulta-at Sevres (France). Topics included tive Committee on Standards for thepresent and future international com- Measurement of Ionizing Radiations;parisons of activity measurements, the 1251; International Reference System for

status and possible extension of the ~In- radionuclides; radioactivity standards;ternational reference system for activity 7 5Se.measurements of gamma-ray emittingnuclides, reports from other workinggroups, accomplishments at the Interna- Accepted: August 16, 1989

1. Background

One of the consultative committees of the Inter-national Committee of Weights and Measures(Comit6 International des Poids et Mesures, CIPM)which guides and assists the International Bureauof Weights and Measures (Bureau International desPoids et Mesures, BIPM) in ensuring the unifor-mity of physical measurements is the ConsultativeCommittee on Standards for the Measurement ofIonizing Radiations (Comit6 Consultatif pour lesEtalons de Mesure des Rayonnements Ionisants,CCEMRI) [1]. The variety of measurement typesin the field led to the committee being set up withthree sections. Section II, with the responsibilityfor radioactivity measurements, met on March 29-

31, 1989, with selected representatives from 11countries and the European community laboratoryparticipating.

The challenge in radioactivity standards is asmuch quantity as quality. Although the accuracyrequired and achieved for specifying the activity ofa sample of a radionuclide is usually only of onepercent, a standard must be developed for eachpertinent radionuclide. Currently the National In-stitute of Standards and Technology has receivedjustified requests for over 130 radionuclides, for ex-ample [2]. There is no official international stan-dard for any radionuclide; rather, the standardizinglaboratory in each country develops an appropriate

363



measuring method which would allow a sample ofthe particular radionuclide to be measured with aknown uncertainty.

Almost all countries have adopted the Interna-tional System (SI) unit, the becquerel (Bq), specify-ing the activity of a sample in decays per second(s-'). Goals of the Radionuclides Section ofCCEMRI are to improve measuring methods andbetter define their uncertainties, to compare resultsobtained in different countries, and to bring theseresults into congruence. The 10th meeting madeconsiderable progress towards these goals.

2. Section Activities

The agenda of the meeting emphasized actionscoordinated through the BIPM, the accomplish-ments of working groups, and the significant activi-ties of the small BIPM group.

2.1 Comparisons of Activity Measurement

One mechanism with which national standardiz-ing laboratories can compare results for a radionu-clide is through simultaneous measurements ofquantitative mass aliquants of a common solution.Twenty-eight such comparisons have been com-pleted since 1961. The status of the three followingcurrent exercises was reported.

A 1986 comparison of '`Cd had been reported atthe previous session in 1987, with satisfactoryagreement on the activity [3]. However, 10 of the18 laboratories participating also measured the 88-keV gamma-ray emission rate, which is the desiredquantity when the radionuclide is used to calibratethe counting efficiency of gamma-ray spectrometrysystems at that energy. It was pointed out that thecombined result probably offers the most reliablemeasurement of the gamma-ray probability per de-cay for this radiation. The value obtained wasPy=0.03614±0.00012, where the uncertainty is thecombined standard deviation of the means for ac-tivity and emission rate. It was recognized at thesession that simply quoting this uncertainty mightlead to an unrealistically small value being carriedforth in evaluations, for uncertainties common toall measurements had not been considered. TheBIPM staff was given the challenging task of mak-ing a realistic appraisal of the uncertainty beforepreparing an article describing the comparison forpublication. This exercise emphasizes that nucleardata can be significant in the application of ra-dionuclidic standards, that cooperative measure-

ments can provide an incidental dividend of goodnuclear data, and that standardizing laboratoriesshould lead the way in demonstrating meaningfulanalyses of the uncertainties of such data.

The comparison in 1988 of solutions of "25I [4]also led to session reports of the measurements ofanother important nuclear datum, the half life, by 5of the 19 laboratories participating. The value sug-gested for the comparison, (59.4±0.5) d, perhapscarried such a large uncertainty because of chemi-cal instability or detection problems for the low-en-ergy radiations in past measurements. These factorsalso can cause difficulties in activity measurements,as demonstrated by one laboratory in the compari-son. Three ingenious methods gave very consistentresults in-house with small estimated uncertain-ties-but were conspicuously 1.5% deviant fromthe average of measurement results from other lab-oratories. Sources had been prepared from a dilu-tion of the solution supplied; a subsequent recheckwith the main solution produced a result in agree-ment with the average of other participants. Oneimportant function of the comparisons is to alert alllaboratories to potential problems; another is to testthe variations observed against the combined un-certainties explicitly given in detail.

The third comparison discussed was that of 75Se,

a radionuclide with a 16-ms delayed state in thedaughter 7"As. This radionuclide had been dis-tributed only to five members of a working groupfor a "trial" comparison, a precaution which hasrevealed unexpected problems in some past com-parisons. Only one result was reported at the ses-sion due to a delay in distributing the solutionbecause of a high radionuclidic impurity level inthe first material received from the commercialsupplier. While impurities must always be consid-ered in activity measurements, some effort is madeto minimize the possibility of their beclouding theaccuracy of the basic direct methods being com-pared.

As the above three examples show, comparisonsare made only of radionuclides selected because ofsome special challenge. The same was true of thoseconsidered as possible candidates for future com-parisons, once the "Se exercise is completed. 99Tcemits only beta particles and has a half life suffi-ciently long that the thin solid sources required forsome methods must be of low activity; 2 0Pb mayhave chemical problems; '"Ce-'"Pr has showndiscrepancies in measurements with different ion-ization chambers, perhaps because of the propor-tion of response due to bremsstrahlung; '4 7Pm is abeta-ray-emitting radionuclide with a gamma ray

364



of low probability; and 137Np would be the firsthigh-atomic-number nuclide measured recently.Availability of suitable material of the first threeradionuclides is being checked by section members.

2.2 The International Reference System (SystimeInternational de Ref6rence, SIR) for Gamma-Ray Emitting Radionuclides

Comparisons are time-consuming activitieswhich require a concentrated action by all partici-pants at the same time. Obviously comparisons canonly address a small fraction of the required ra-dionuclides, even supposing that national capabilityfor accurate measurement of a radionuclide can bemaintained, once demonstrated.

Several years ago a more general comparisonmethod was established at BIPM for gamma-ray-emitting radionuclides [5]. Samples of any suitableradionuclide are submitted at a convenient time asa specified volume of solution in uniform ampoules.The response for a stated amount of activity in twore-entrant ionization chambers is compared withthat of a radium reference source, and the ratio perbecquerel is registered on tables for that radionu-clide containing the same information for samplesfrom other laboratories.

At the present time 373 results for 48 radionu-clides from 25 laboratories have been registered.The scheme allows not only a comparison of stan-dards between nations, but also provides a back-uprepository of measurements of a given nation withtime, in case similar ionization chambers used toretain standards in the national laboratory becomeinoperable.

The success of the present SIR, and the demandfor demonstrated international agreement for ra-dionuclides not emitting suitable gamma rays, ledto the formation of a working group at the ninthRadionuclides-Section session to investigate exten-sion. A survey of possible users indicated that de-mand for three further types of desiredradionuclides might be satisfied in a straightfor-ward manner, albeit with considerable testing todetermine reliability and comparison uncertainty.

Gaseous radionuclides which emit suitablegamma rays can be compared in the present cham-bers, once a suitable ampoule is selected. One possi-ble ampoule is being studied.

Measurements in some member laboratories sug-gest that the bremsstrahlung from radionuclidesemitting only beta particles with maximum energygreater than 600 keV can probably be comparedwith the present BIPM chambers, and this can be

tested. For radionuclides emitting photons whichare highly absorbed by the well walls of the presentBIPM chambers, thinner-walled chambers are sug-gested. Members were asked to submit specifica-tions of suitable chambers by the end of 1989.

For radionuclides emitting even lower-energyphotons or only beta-particles, two possibilitieswere suggested. Calorimetry [6] would not requireopening of sealed containers supplied by partici-pants, but required activity levels are much greaterthan most laboratories wanted to supply.

Another possibility, of potentially universal ap-plication, is liquid-scintillation counting. Previousbilateral comparisons using this method involvedthe preparation of extensive sets of samples. How-ever, one laboratory proposed a system using cal-culated response functions for each radionuclide ina system comparing the coincidences between twoand three phototubes viewing the same scintillator[6]. This possibility, which might allow each labo-ratory to prepare sealed samples, is to be furtherdiscussed and possibly tested in a working group.

2.3 Other Working Group Actions

Although they cannot be discussed here in detail,several other actions involving techniques orspecific topics were discussed.

2.3.1 Principles of the Coincidence MethodsSeveral of the direct activity measurement methods[6] make use of the time relation between the initialand subsequent radiations in a decay. Nineteendrafts of papers, reports, or working party noteswere circulated between the 1987 and 1989 ses-sions. Many were concerned with the correctionsfor detector pulses lost in the time a counting sys-tem is insensitive due to a previous pulse [7].

2.3.2 High-Count-Rate Measurements Someof the corrections which generate the major uncer-tainties in activity measurements are better testedwith rates beyond those normally used. Severaltechniques have been developed since the lastworking group test, but questionnaire results sug-gested that other section actions should be carriedout first while the best testing scheme is devised.

2.3.3 226Ra Standards "Radium" is unique inthat international standards, by mass, do exist. Aset was prepared by Honigschmid in 1934. Twolaboratories in countries where large radiumsources find extensive use suggested a repackagingof the original set from glass to metal tubes. How-ever, great difficulty in opening a similar sourceprepared in 1924 showed that fears of breakage dueto pressure buildup and glass deterioration are

365



probably not warranted. Laboratories who nolonger use or desire the H6nigschmid standardsthey have were requested to supply them to thetwo laboratories that use them.

2.3.4 High-Efficiency NaI(T1) Techniques Ifeach decay of a radionuclide results in a cascade ofx rays or gamma rays, a well-type NaI(Tl) detectorcan detect at least one radiation most of the time,and the fraction lost can be calculated accurately ifthe efficiency for different energies is known. Areport on experiences in member laboratories is tobe published.

3. BIPM Activities

In addition to maintaining the SIR, and investi-gating its extension, the BIPM staff participates inthe general comparisons and prepares the detailedreports describing the results. They also make fun-damental contributions to the development of mea-suring techniques and their analysis.

Examples presented at this session were the mea-surement of the half life of a metastable state fol-lowing the decay of "5Se, with a preliminary valueof (16.3±0.3) ins; the development of correctionsfor circuit dead times in series, with possible mix-tures of the two types which further extend when asecond pulse arrives during a dead period, or not[8]; and possible applications of modulo-2 countingfor identifying dead times.

ciency and accuracy of such measurements. Theexamples displayed in this article from the 10th ses-sion of the Radionuclides Section illustrate thescope and range of these activities.

About the author: Dale D. Hoppes is a physicist andleader of the group for radioactivity measurements inthe Ionizing Radiation Division, Centerfor RadiationResearch, NIST National Measurement Laboratory.

6. References

[1] BIPM Com. Cons. Etalons Mes. Ray. Ionisants 11, R 109(1985).

[2] Hoppes, D. D., Env. Intl. 10, 99 (1984).[31 Ratel, G., International Comparison of Activity Measure-

ments of a Solution of `0Cd (March 1986), Rapport BIPM-88/4, Sevres, France, 1988.

[4] Ratel, G., International Comparison of Activity Measure-ments of a Solution of "25I (May 1988), Draft RapportCCEMRI(II)/89-2, Sevres, France, 1989.

[5] Rytz, A., Int. J. Appl. Radiat. Isot. 34, 1047 (1983).[6] Mann, W. B., Rytz, A., and Spernol, A., Appl. Radiat.

Isotopes 38, 717 (1988), chapter 5.[7] Chauvenet, B., Bouchard, J., and Vatin, R., Appl. Radiat.

Isotopes 38, 10 (1988).[8] Muller, J. W., The transmission factor T1(N,N) revisited,

Rapport BIPM-88/12 (1988); and some series expansions ofT., Rapport BIPM 88/13 (1988) (BIPM, Sevres, France).

4. Laboratory Reports

A final portion of this session, and all recentones, has been the presentation of written and oralreports from each attending laboratory coveringtopics of possible general interest not discussedelsewhere in the agenda. These cover not only dis-coveries and new data, but also trends in the de-mands and delivery of standards to users.

5. Discussion

The unusual nature of radioactivity standardiza-tion makes international cooperation especially sig-nificant. The interchange of ideas about basicstandardization during the sessions of the Radionu-clides Section of the CCEMRI at BIPM, and aboutapplied radioactivity measurements at associatedmeetings of the International Committee for Ra-dionuclide Metrology, contributes to both the effi-

366




On Measuring the Root-Mean-Square Valueof a Finite Record Length Periodic Waveform


E. Clayton Teague An analysis of the uncertainty in mea- width limited Gaussian waveform to in-suring the root-mean-square, rms or Rq, troduce definitions. Following this is an

National Institute of Standards value of a periodic waveform which re- analysis of the periodic waveform usingand Technology, suits from the use of a finite record the same approach. It is shown that forGaithersburg, MD 20899 length is presented. Even though the re- a large number of periods, n, in the

sults of the analysis are somewhat as ex- record length, the normalized threepected, i.e., that the uncertainty is standard deviation of the rms value isinversely proportional to the number of given by 3/(87rn).periods in the record, the explicit rela-tionship between the magnitude of the Key words: periodic surface profile; pe-uncertainty and properties of the wave- riodic waveform; random error; root-form does not appear to be available in mean-square value; surface roughness;the literature. The paper first presents uncertainty.an introductory example in terms of thereasonably well known case of band- Accepted: August 22, 1989

1. Introduction

This paper presents a discussion of the uncer-tainty in measuring the root-mean-square, rms,value of a periodic waveform which results fromthe use of a finite record length. The analysis wasmotivated by seeking to understand the source of arandom uncertainty component which was presentin some measurements of the absolute arithmeticaverage, R an deviation from a mean line of profilesof precision roughness specimens. The profiles ofthese specimens had an approximately triangularwaveform with two wavelengths and amplitudes.For the longer wavelength specimens the randomphasing of the waveform with respect to therecording interval proved to be a major source ofuncertainty in the measurements. In this case therecord length included 40 periods of the wave-form. Even though the results of the analysis aresomewhat as expected, i.e., that the uncertainty is

inversely proportional to the number of periods inthe record, the explicit relationship between themagnitude of the uncertainty and properties of thewaveform does not appear to be available in theliterature. The discussion first presents an examplein terms of the reasonably well known case of abandwidth limited Gaussian waveform to intro-duce some definitions and relationships. Followingthis is an analysis of the periodic waveform whichuses a similar approach.

2. Bandwidth Limited Gaussian Waveform

By definition the mean square value, q 2, calcu-lated from a finite length L of the waveform y(x) isgiven by the equation:

367



rLql= 1/L y2(X) dX.(1

For convenience and brevity in the following dis-cussion, the definitions listed below will be used

[1].

Sample space=the set of points representing thepossible outcomes of a measure-ment.

4(k)=a real number, called the randomvariable, which represents the out-come of a measurement indexed byk.

p (4_1im Probability [4) < c(k) < 4 + A&c)]AO->O AO

E{sg[(k)]}sthe expected value of any real sin-gle-valued continuous functiong(O) of the random variable 4(k).It is given by:

E{g[4(k)]}= f g(4)p(4) d4.

As an example, the mean square value of y(x) isgiven by:

E[y2(x)]=f y2(x)p(x) dx.

E[y2 (x)] will be defined as q2 which is that valueapproached by q2 as the record length L ap-proaches infinity. Also q2 is the limiting value forthe mean q2 as the number of samplings gets large.The variance of y(x) is defined by:

var {y}=EI[(y E)2]={ (y-y) 2 p(x) dx.

Similarly the variance in q2 is defined by:

var {q2 } =E [(q2 _ q2 ) 2].

For the special case when y(x) is a bandwidth lim-ited Gaussian waveform with zero mean, Bendatand Piersol [1] show that

var~q 2} 2L * [ 1 2L/L*l

-4 L [I e

+L [2 L*+1 Ie 2L/L _ 1]

where L * is the shift distance required for the auto-correlation function to drop by 63% of its zeroshift value. If L> >L * the relation reduces to

var{q2 } 2 L*4 - L

The propagation of error formulae discussed by Ku[2] may be used to relate the variance of the root-mean-square value=rms=/Vq2 to this variance ofq

2 . Results given in table 1 of this reference showthat if

var q2 var rms E_4 =E, then .

Thus, the normalized three-standard-deviationslimit, 3SD, in the root-mean-square values calcu-lated from randomly sampled lengths, L, of aGaussian profile with a correlation length L* isgiven by:

rms 3V'\2L* V- 3SD limit of-=2 L =2.1L

Whitehouse and Archard [3] report reasonableconfirmation of this result for their test specimenswhich had approximately Gaussian profiles. Theimportance of this result is stated very clearly byWhitehouse and Archard [3]. "The variance ofmeasured rms or Ra values for the roughness of asurface may be found easily if one knows the stan-dard deviation of a large number of such measure-ments made upon the same surface. Alternativelyone may predict the variance from a knowledge ofthe correlation length of a typical profile of the testsurface." The argument just given for the 3SDlimit partially fulfills the "it can be shown" state-ment by Whitehouse and Archard.

3. Periodic Waveforms

Unfortunately these results do not apply for peri-odic waveforms such as that of many precisionroughness specimens. However, the principles ofcalculation are still applicable. Consider the exam-ple of the sinusoidal waveform:

y(x)=A sin(ox + O). (2)

The waveform is sampled for a length L with thephase angle 0 considered as the random variable of

368



the sample space. The angle 0 will be considered ashaving a uniform distribution over values from 0 to27r, i.e.,

p () =I , 0<0 <2i and p(0)=0, for all other

values of 0

According to definitions in eqs (1) and (2):

=1 A2q2 =L-I A 2sin'(cox + O)dx .

Evaluation of the integral yields

q2 A42sin 2(5+ ) ±sin 20]2 [ 2coL +2oL ]

where 8 = L- 2n 7r. The conditional variance ofq 2, given 8, is (noting that i2 =A 2/2):

Var(q2 18)=E[(q2 q2)2 1

f [A 2 (sin2(8+0) sin20 A]2

-(I- + -w -~ dO.Evaluation of the integral yields:

var(q2 1 )=A (1-cos 28)var~~qj 1 6&)2 L

2

Thus, if the record length happens to include anintegral number of periods, var q2=0. But if wetake the more likely case in which S has a distribu-tion of values like that of 0 due either to variationsin the measured wavelength, a)-', or in the recordlength, L, the result integrated with respect to 8 isobtained as follows.

var(q2)8 =2 f2 var(q2 1 8)dS

1 C2' A 4(l-cos28)2

1T J 1 6o2CL d2

Substituting otL = 8 + 2n r followed by the intro-duction of dummy variables x and y reduces theintegral to:

1 t2

(n+1)irvar(q2)a= I X 2 dx

2 J 2#rn

4 (n + )1T--J(cos y/y2)dy.

47T nir

Solving these integrals yields:

/A A4 1 (var q2)a(var q ) 16 47r2n(n + 1) 4o

1

16'7r2 n(n +1)

where n is again defined by the equation8=coL-2nvr. This result for (var q2)8 does ac-count for the relationship between ctoL and 5. Useof the same relationships between var q2 andvar rms given earlier yields:

(var rms)s 1q2 647n22n(n + 1)

An averaged normalized three-standard-deviationlimit for variations of the rms values is then givenby:

3SD (rms) = 3

or for large n,

3SD(rms)= 3(I ) 8~rn

(3)

(4)

Finally, if the record length L and measured wave-lengths, &)-', are stable and such that 8 has a nar-row distribution about 7r/2 or 37T/2 the3SD(rms/4) could be as much as twice this averagevalue.

4. Comparison with Variations ofRoughness Measurements

For surface profiles obtained from roughnessmeasurements, one would expect that the normal-ized 3SD of a set of measured Ra values would beapproximately equal to that computed for the Rq(rms) values since both quantities are similar mea-sures of the sampled profile's empirical probabilitydensity function. The measurements of Ra valueswhich initiated this analysis should therefore be un-derstandable in terms of the results just derived.

For the measurements, two types of specimenswere used; one set had an Ra value of 3.2 ptm and awavelength such that n =40; the other set had anRa value of 0.5 pm and a wavelength such thatn =250. Record lengths for the measurements wereapproximately 3.8 mm. Predicted uncertainties for

369



the variations of a measured value about the meanRa are therefore:

3SD3.2 ,m =0. 3 % of mean Ra,

and

3SDO5,m=0.05% of mean Ra,

if one assumes the averaged three standard devia-tions limit. The maximum values for the analyticaluncertainties are 0.6 and 0.1% of the mean Ra val-ues for the 3.2 and 0.5 .im specimens, respectively.

In a series of 80 sets of three Ra measurements,the uncertainty (3 standard deviations) of thesetriplicate measurements with respect to their aver-age values was:

3SD 3.2 = 1.0% of mean Ra,

and

3SDO5 =0.7% of mean Ra-

The intent of this comparison between the analyti-cal and experimental values is not that of justifyingthe theoretical analysis. However, in terms of un-derstanding the experimental variations the agree-ment is sufficient to confirm that, for the longerwavelength specimen, a major source of uncer-tainty for the roughness measurements results fromthe finite sampling length.

The imperfect waveforms of the precisionroughness specimens are the most likely sources ofthe residual uncertainty. As illustrated in figure 1the specimen waveforms have distortions whichare produced by either an additive random func-tion together with phase modulation or by the

3.25 SLm| /

H = 1.06 um

la. Waveform of 3.2 pRm R. Specimen

0.82 pRmj

i-i = 25.5 p1mlb. Waveform of 0.5 p.m R. Specimen

Figure 1. Waveforms of roughness specimens which were usedfor statistical studies.

combination of random amplitude and phase modu-lation. The effect of an additive random function isthe only one which can be readily analyzed. Forthis case the rms value of the sum of two uncorre-lated waveforms adds in a rms manner. Assumethat the waveforms can be represented by the sumof a random component, with an rms value of qRand a correlation length of L *, and a periodic com-ponent, with an rms value of qp and angular fre-quency oi. Then the use of propagation of errorformulae derived by Ku [2] yields:

3SD q} )=1.5 [2aL+ 16 2 2 ]

where n and L are as defined earlier and

a- =(1 + -!, b=U 1+2

Estimates of qR and L * from the calculation of au-tocorrelation functions of the waveforms for the0.5ttm-Ra specimen are: L*=0.05 Land qR=0.01qp. Effects of random amplitude and phase modula-tion are not accounted for in these estimates. Withthese values of qR and L * the dominant source ofuncertainty for the waveform sampled for 250 peri-ods is the random additive component which in-creases the uncertainty limits to 0.3% of the meanvalue. Approximately the same values for L * and2qR were obtained from studies of the 3.17pxm-Ra

specimen. Since qR = 2.8 X 10-4 q2, in this instance,the uncertainty produced by the random compo-nent is insignificant for measurements on this speci-men.

5. Conclusion

The uncertainty in measuring the root-mean-square value of a finite record length of a periodicwaveform has been obtained and is given in eqs (3)and (4). The theoretical results were comparedwith data obtained from surface profile measure-ments of precision roughness specimens. However,eqs (3) and (4) are quite general and will apply torms measurements of the amplitude of periodicwaveforms arising from other areas.

During customary announcements of this paperto NIST staff, an analysis of the errors of measure-ment due to variations in sampling interval length,(complimentary to that presented here), wasbrought to the author's attention. The effects of

370



adjusting the sample interval length relative to anintegral number of cycles of the test waveform aredetermined and applied to measurements of electri-cal power with a wattmeter [4].

6. Acknowledgments

The author gratefully acknowledges helpful dis-cussions with Charles P. Reeve of the StatisticalEngineering Division. His suggestions clarifiedmany of the ideas presented in the paper; in partic-ular, the use of the concept of conditional varianceenabled the basic calculation to be expressed inmore conventional statistical terminology.

About the author: Dr. E. Clayton Teague is GroupLeader of Micrometrology in the NIST Precision En-gineering Division of the Center for ManufacturingEngineering, National Engineering Laboratory. Hehas worked in the area of surface texture measure-ment for 18 years.

7. References

[1] Bendat, Julian S., and Piersol, Allan G., Random Data:Analysis and Measurement Procedures, Chapters 3 and 6,Wiley Interscience (1971).

[2] Precision Measurement and Calibration, Edited by Ku,H. H.; pp. 331-341, NBS Special Publication 300, Volume1, U.S. Government Printing Office, Superintendent ofDocuments, 1969.

[3] Whitehouse, D. J., and Archard, J. F., Proc. Roy. Soc.London A316, 97 (1970).

[4] Stenbakken, G. N., IEEE Trans. on Power Apparatus andSystems PAS-103, 2919 (1984).

371




A Search for Optical Molasses in a Vapor Cell.General Analysis and Experimental Attempt


A. L. Migdall We analyze the application of optical are given to illustrate which fundamen-molasses to a thermal vapor cell to tal parameters are most important in the

National Institute of Standards make and collect cold atoms. Such an production of cold atoms in a vapor

and Technology, arrangement would simplify the produc- cell.Gaithersburg, MD 20899 tion of cold atoms by eliminating the

difficulty of first having to produce and Key words: laser cooling; optical mo-slow an atomic beam. We present the lasses; vapor cell.results of our calculations, computermodels, and experimental work. As aguide for future work, general results Accepted: August 25, 1989

1. Introduction

We have proposed using the technique of lasercooling to directly cool and collect gas atoms in athermal vapor cell. This allows a high density re-gion of very cold atoms to be built up [l]. Thiseliminates the difficulty of first having to produceand slow an atomic beam, and thus holds thepromise of greatly simplifying the production ofcold atoms. We present here an analysis of the pro-cesses involved in this proposal, as a guide for fu-ture work, and give the results of our attempts torealize it experimentally.

The proposal is to set up a region of optical mo-lasses directly in an atomic vapor cell, allowing theaccretion of a high density of cold atoms. The mo-lasses consists of three pairs of counterpropagatinglaser beams tuned below resonance in an atomicvapor cell. This arrangement [2] of opposing lightbeams provides a strong damping or viscous forcefor slow atoms. This effect, which was first demon-strated with a cooled atomic beam several yearsago [3], is well described in the literature [4,5] and

so will only be discussed briefly here. Atoms thatenter the molasses region with slow enough veloc-ities have a high probability of being slowed andviscously captured before exiting the other side.The motion of these captured atoms becomes diffu-sive, because any velocity the atoms acquire in aparticular direction (due to random scattering ofphotons) is quickly damped by the viscous force[3]. Because of this random walk type motion,these atoms remain in the molasses region for muchlonger times than they would have under ballisticmotion at their original thermal velocities or eventheir ultimate cooled velocity (hence the term opti-cal molasses). After a time, all the atoms in the cellwith "catchable" velocities (most probably thosewith velocities less than some critical value) be-come mired in the molasses region. Thus the den-sity of cold atoms in the molasses region builds up,as the lowest velocity atoms in the rest of the cellare depleted. The density of cold atoms continuesto increase as the low velocity tail of the thermal

373



distribution is replenished by processes that movethe velocity distribution back toward thermal equi-librium. It is the fluorescence of the accumulatedcold atoms that would be observed.

2. Theoretical Analysis

We calculate how fast the cold atom densityshould increase with time and estimate the pro-cesses limiting residence time in classical two-statemolasses [5]. This is done with the help of com-puter models of the cooling process for the specificcase of a sodium atom in a vapor cell containing amolasses region. In addition, we calculate the visi-bility of the effect by comparing the fluorescencesignal expected from the cold atoms to the back-ground signal that is also present. We also examinethe general dependences of the signal size on theatomic parameters, so that comparisons can bemade of the relative merits of trying to cool otheratoms in a vapor cell or using other transitions.

The initial rate of increase of the cold atom den-sity is governed by the rate at which capturableatoms enter the molasses region. The rate of entry(and subsequent capture) of atoms is just the flux PD,

<D = I nC Vcavg (1)

this case, the catchable density, nc found by inte-grating the Maxwell-Boltzmann distribution fromV=0 to Vc (for Va< <V(kBT/M), where kB is theBoltzmann constant, T is the gas temperature, andM is the atomic mass) is given by:

'C _ 4 no M ) 3/2v

3

(2)

where no is the total density of the thermal atoms inthe cell. The average velocity of these atoms,found by performing a similar integration, is justequal to (3/4) Vc. The fractional increase in the to-tal density with time, F(t) (for t much less than theaverage time for leaving the molasses) equals therate of new atoms getting caught in the regiontimes t divided by the volume of a sphere of radiusR and the original density, no:

1 nc Vcavg 47rR 2n4 3 4rR3

no ~ r3t = 9nc VC t

16no R (3)

To estimate this increase, we use the results ofour computer simulations, including those shownin figure 1. This figure shows the general resultthat the average distance needed to snag an Na

integrated over the surface area of the molassesvolume, where nc is the density of catchable atomsand Vcavg is the average velocity of those atoms.The probability of capture versus velocity for atwo-state atom entering the molasses region wasdetermined by a Monte Carlo simulation. The useof this simple two-state model is justified in thissituation, because during most of the capture pro-cess, the velocities correspond to detunings that arelarge relative to atomic linewidth. (It is importantto note that it is the capture rate that is more im-portant in estimating the final signal rather than theultimate temperature of the molasses, since the resi-dence time in the molasses is limited by other pro-cesses.) A 1-cm region of 3-D molasses is modeled,where each beam has a saturation parameter of one(defined as I/1o, where I is the light intensity and 1ois the on-resonance intensity which power broad-ens the natural transition width by a factor of \/2).The laser frequency is tuned rn/2 below reso-nance, where r], is the natural width of the coolingtransition. For this model, the result shows that thecatchable atoms are simply those atoms with veloc-ities less than some maximum cutoff value, VK. In

20

E

0)E

Cu0~

r_

a

._ICL

0

15

10

5

0250 5 10 15 20

Initial Velocity (M/s)

Figure 1. The distance required to bring a Na atom from itsinitial velocity to the point where its motion is completely diffu-sive. The points are the result of a model of 3-D molasses witheach beam having a saturation parameter of one. The size of themolasses region needed to hold an atom for an extended timemust be somewhat larger than the stopping distance, becausethe atom must stop away from the boundary of the region toremain caught for an appreciable time. The line is a fit to apower law, a Vb. The best fit parameters are a = 4.7 X 10-6 andb =4.87, very nearly the expected 5th power form.

374



atom in uniform molasses is proportional to the 5thpower of the initial velocity. (Note that the size ofthe molasses region required to catch and hold anatom would be somewhat larger than these stop-ping distances, since the atom must not stop righton the boundary if it is to remain caught for anappreciable length of time.) This velocity depen-dence agrees with the calculated dependence ob-tained by expanding to first order, the velocitydependence of the damping force for a red detun-ing of rn/2 and kv>rF, where k is the wave vec-tor in wavenumbers of the laser. By integrating, weget the stopping distance, D as a function of veloc-ity. The result of this calculation has the form

Mk2V55ir4 1- (4)

According to our simulations for Na atoms on typ-ical trajectories traversing a molasses region con-sisting of the intersection of three 1-cm diameterbeams, 20 m/s is roughly the critical velocity, be-low which the probability of stopping before exit-ing the other side is nearly unity. The fraction ofatoms with velocities less than 20 m/s, as given byeq (2), is 5 X lo-5 of the total Na density at 75 'C.

We also investigated how the capture rate is af-fected by having 1-D molasses surrounding thecentral 3-D molasses region [6]. This is importantbecause, before an atom enters the central region itis likely to pass through a region of only one pair ofcounterpropagating beams. In this region, only onecomponent of velocity is damped, but it is just thecomponent of velocity needed to reach the centralregion. Thus some atoms headed for the centralregion may be, in effect, deflected away, therebyreducing the velocity capture range. This processis modeled and found to produce a minimum veloc-ity below which atoms on certain trajectories can-not reach the central volume. The circles in figure2 show an example of a trajectory where onlythose atoms with velocities between 17 and 19 m/sare stopped in the molasses region. The trianglarpoints in figure 2 represent a nearly equivalent caseto that of no surrounding 1-D molasses for reasonsto be described later. 1-D molasses surrounding thecentral 3-D region should also produce a slight in-crease in the maximum catchable velocity, due tothe extra length of the cooling region. Some effectof this is seen on those velocities just slightly toofast to be caught. The circular points show signifi-cant cooling of velocities in the 20 to 23 m/srange, whereas the triangular points show minimalcooling in that range. The number of atoms within

30

25

-aa)aC,,Cnco

.

20

15

10

5

00 5 10 15 20

Initial Velocity (m/s)

25 30

Figure 2. The final speed of Na atoms incident on the molassesregion vs starting velocity. The molasses region has a centralvolume of 3-D damping surrounded by 1-D damping due to asingle pair of beams. The insert shows the particular trajectoryused, which passes through the l-D region before entering the3-D region. The circles are for the configuration with both laserfrequencies in all beams. The triangles indicate that one laserfrequency is in one pair of beams and the second frequency inthe other two pairs. For this curve, only atoms with velocities inthe range between 17 and 19 m/s were caught. Atoms in therange from 20 to 23 m/s were greatly slowed but did not in factstop in the 3-D molasses region. The difference between thedashed line and the data points is the total slowing experiencedby the atom.

this band of capturable velocities is about one-third of the number that is caught in the central3-D molasses without the surrounding 1-D mo-lasses. We describe a means to reduce this effectlater.

What ultimately limits the maximum density thatcan be built up are the processes that remove coldatoms from the molasses. Atoms leave the regionby diffusion, by aquiring a drift velocity due to im-balance of opposing beams [5], and by collisionswith the thermal gas atoms which transfer a rela-tively large amount of kinetic energy. Here we esti-mate the time for each of these processes, assumingclassical two-state optical molasses damping.

The time for an atom to random walk out of themolasses is proportional to R2, where R is a char-acteristic radius of the molasses region. Assumingfor Na a random walk with a time step size equal tothe velocity correlation time of 4X 10- s and avelocity equal to the simple Doppler cooling limitof 50 cm/s, the average time to leave is long, about0.45 s for a 1-cm diameter region. The calculatedtime [5] for an atom to drift out due to a beamintensity imbalance is also long, 0.25 s for a 1%intensity imbalance, I/1o= 1, and a detuning of

375



rP/2. Our final estimate is not hurt by using such asimple model of molasses since it is known to un-derestimate these times which, in our case, are al-ready long enough not be the process limiting thefinal density buildup.

The time for a cold atom to be knocked out ofthe molasses depends on the density of the thermalgas atoms and their collision cross sections. Thedominant cross section for this is the elastic crosssection for collisions between ground-state and ex-cited-state atoms. We estimate this cross section tobe on the order of 10-'3 to 10-2 cm2, which for Naat 75 0C, gives a time to be knocked out of themolasses of 1.0 to 0.1 s, assuming that half theatoms are in the excited state.

The visibility of the cold atom signal depends onthe size of that signal relative to any backgroundsignal. A background signal exists because each ofthe laser beams is in resonance with a fraction ofthe uncooled atoms. This fraction is approximatelyequal to rn/rD where rD is the Doppler width ofthe resonance transition at the cell temperature.This will produce a sizable background fluores-cence signal. The total background fluorescencesignal will be about three to six times the signal dueto a single beam, since the three pairs of beams areorthogonal, and the tuning is such that the sixbeams interact with nearly distinct velocity classes.

3. Signal Estimate

To understand better how cell molasses works, itis useful to estimate the signal size in terms of fun-damental atomic parameters so that the estimatecan easily be extended to an arbitrary atom. In thisway, we can see best how to maximize the coldatom signal. To do this, we eliminate Vc and n,,from eq (3) by writing them in terms of the massand transition width of the atom. Substituting eqs(2) and (4) into eq (3) and using D =2R we get

i ofirn4/14

F(t) =3 k (2T)3/2 V t (5)

To determine the size of the signal expected fromthe increase in density of the cold atoms, we takethe ratio of this density to the density of the hotatoms that are also fluorescing. The total fractionof hot atoms that are on resonance with any of thelaser beams is estimated to be 5rn/rD, for a lasertuned r/2 below resonance. Since rD is propor-tional to the thermal velocity of the atoms, the ra-

tio S of the cold atom signal to the backgroundfluorescence can be written as:

S(t)=- 5[2( T lF(t) (6)

Substituting eq (5) into eq (6) and extracting justthe proportionalities with respect to the atomicparameters and cell conditions, we find

S(t) a Ma.2 P2.2 R -0.2 T-'. (7)

From this form we can see that S(t) depends onlyweakly on such parameters as M, R, and T but isstrongly dependent on rn. Thus, to enhance thedensity gain and signal visibility in cell molasses itis important to select an atom with a large transi-tion width. Alternately, it may be possible toachieve this artificially, possibly by power broad-ening the transition, if this can be done withoutgreatly degrading the cooling process.

The above analysis has assumed a two-stateatom. In actuality since Na has two ground states(F= 1 and F=2), two frequencies are necessary toprevent optically pumping the atom to the otherground state and losing it to the cooling process.The two-state ideal can be approximated by havingboth frequencies present in the molasses region.This can be accomplished by two different ar-rangements. The first uses both frequencies in allbeams. This is the usual arrangement for opticalmolasses in beam experiments. The second arrange-ment uses one frequency in two pairs of beams withthe other frequency in the third pair of beams. Thissecond optical arrangement, by separating the fre-quencies, nearly eliminates the effect of the sur-rounding 1-D molasses region while possiblyenhancing the signal-to-background fluorescenceratio [7]. One particular arrangement might be tohave only light resonant with the F=2 to F'=2transition in two pairs of counter propagatingbeams, while the third pair contains only light reso-nant with the F= 1 to F'=2 transition. With thissetup, the cooling in the 1-D molasses regionswould be effectively turned off after only a fewtransitions as the atom is optically pumped to theother ground state. Only in the intersection of thebeams would both frequencies be present to pre-vent optical pumping out of the cooling process.The triangular points of figure 2 show the result ofoptical pumping, the lower velocities becomecatchable again as the effect of the l-D molassesregion is nearly eliminated. For this arrangement,

376



the cooling in the intersection region will proceedat a reduced rate (since cooling on the F= 2 toF'=2 transition is not as effective for not clearlyunderstood reasons), but 3-D cooling will still oc-cur.

The big advantage with this technique is that thevisibility should be greatly enhanced. While boththe cold atom signal would be smaller (due to lesseffective damping using these transitions) and thebackground signal would be smaller (due to opticalpumping), it is expected that the background signalreduction would be greater than the cold atom sig-nal reduction. The rms thermal velocity of theatoms in the cell is about 6 X I04 cm/s. This gives atransit time across the 1 cm beams of 1.6X lo-5 s.Using either the F= I to F'=2 or F=2 to F'=2frequency alone will optically pump the atoms af-ter only a few transitions or about 100 ns. Since ahot atom is in the beam for 160 times as long as thisoptical pumping time, the background fluorescenceshould be reduced by nearly this factor.

4. Experiment

To look for evidence of molasses experimentally,we set up a Na vapor cell with a base pressure atroom temperature of 2X 10`o Torr. The cell wasoperated at 75 'C. An effort was made to make thetemperature distribution as uniform as possible.This uniformity was important to prevent a non-Maxwell-Boltzmann velocity distribution resultingfrom local hot spots in the cell. The Na vapor pres-sure at this temperature was measured to be in therange of 1 to 3 X IO-' Torr. The cell windows wereantireflection coated to reflect less than 0.25% persurface. High reflectivity (R >99.85%) mirrorswere used to retroreflect the beams. Each of thethree mutually orthogonal pairs of counterpropa-gating beams were linearly polarized with their po-larizations mutually orthogonal. The Na D2 linewas used for the cooling transition. The power ineach beam was varied from about 1 to 10 mW withthe beams apertured from 0.5 to 1 cm in diameter.The collimation of the molasses beams were set us-ing a shearing interferometer. With this interferom-eter, we were able to set the radius of curvature ofthe wavefront to be greater than 100 m. A photo-multiplier tube was used to collect the light fromthe region of intersection of the beams.

In setting up this experiment, there is the practi-cal problem of finding the initial evidence of a sig-nal in the presence of the background fluorescence.It is our experience that in a beam experiment, mo-

lasses lifetimes of about 100 ms can be set up blind(i.e., with no need to see the signal). For molassesin a cell, the maximum lifetime may be somewhatsmaller than this due to collisions with backgroundatoms, so we need a signal that is visible at thislifetime to observe the effect. Our calculations ofdensity buildup are valid when the rate of increaseis greatest, that is, for times much less than any ofthe limiting lifetimes. Therefore, we chose to lookfor density buildups on a time scale of less thanabout 50 ms. We attempted to observe the signal(with both frequencies in all beams) using a lock-inamplifier and modulating the F=1 to F=2 laserfrequency on the order of 10 MHz at a 20 Hz ratewhile slowly scanning the F = 2 to F = 3 laser. Thiswould allow the cold atom signal to build up for 25ms. According to our calculations, using Vc= 20m/s, a 1-cm region, and allowing for the reductondue to 1-D molasses surrounding the central 3-Dregion, the initial rate of density increase for theabove parameters is about 3.5% per second. Sincethe background signal results from the fluores-cence of about 3.6% of the total number of atoms(for our Na cell at 75'C, the natural width is-0.7% of the Doppler width), the rate of increase

in cold atom fluorescence signal is predicted to beabout 100% of the background signal per second.For the 25 ms build up time, the expected fluores-cence increase due to laser cooled atoms is approx-imately 2.5% over the background fluorescence.The signature of this signal should be a peak with afrequency width on the order of the natural widthof the transition and located just to the red of theF=2 to F = 3 transition. The signal should, ofcourse, disappear when one of the six beams isblocked. We estimate that our sensitivity is roughlyat the 1% level, which is comparable to our ex-pected signal. We saw no evidence of laser cooledatoms in the cell using the above setup, or the alter-nate arrangement of separate frequencies in sepa-rate beams, so we are only able to conclude thatour signal estimates give a limit to the size of thecold atom density buildup.

5. Discussion

One important assumption in our calculations isthat the velocity distribution is quickly rethermal-ized as the low velocity atoms are caught in themolasses. In our cell, the temperature and densityare such that, for thermalization purposes, there areessentially no collisions between atoms in the gasphase. In addition, the sticking coefficient for Na

377



atoms hitting the walls should be nearly unity. Thismeans that any atom in the gas is a fresh atom emit-ted from the wall with no memory of any previousvelocity. As a result, our calculations depend onthe velocity distribution of the atoms emitted fromthe walls being Maxwell-Boltzmann. This shouldbe a valid assumption since, at the low densities ofour cell, the low-velocity end of the distributionshould not be suppressed. This is in contrast tobeam experiments, where the low velocity atomscan be depleted by collisions with a large flux offaster atoms [8]. We mention here, for complete-ness, that there may be other mechanisms thatcould drastically reduce the rate of emission ofslow atoms from the walls, such as a surface barrierpotential. If such a barrier did exist, the slowestatoms emitted by the wall would not make it overthe hill to enter the molasses volume. In this case,rethermalization would occur through inelastic in-teractions between the atom and the barrier. Thiscould be a very slow process which would reducethe cold atom density that could be achieved. Also,if for some reason the sticking coefficient wasanomalously low, the rate of rethermalizationwould be slow, reducing the rate of cold atombuildup.

ject. We also thank all the members of the NISTlaser cooling and trapping project for their usefulsuggestions on this work. The support of the Officeof Naval Research is gratefully acknowledged.

About the author: Alan L. Migdall is a physicist inthe Radiometric Division of the Center for RadiationResearch at the National Institute of Standards andTechnology, Gaithersburg, MD.

8. References

[1] Migdall, A. L., Metcalf, H., and Phillips, W. D., J. Opt.Soc. Amer. A 2 (13) P51 (1985).

[2] Hansch, T. W., and Schawlow, A. L., Opts. Comm. 13 68(1975).

[3] Chu, S., Hollberg, L., Bjorkholm, J. E., Cable, A., andAshkin, A., Phys. Rev. Lett. 55 48 (1985).

[4] Gould, P. L., Lett, P. D., and Phillips, W. D., in LaserSpectrocsopy VIII, Persson, W. and Svanberg, S., eds.,Springer, Berlin (1987) p. 58.

[5] Lett, P., et al., submitted to J. Opt. Soc. Amer. B.[6] Migdall, A. L., Phillips, W. D., and Metcalf, H. J., Bull.

Amer. Phys. Soc. 31 973 (1987).[7] Migdall, A. L., Metcalf, H. J., Phillips, W. D., and Gould,

P. L., Bull. Amer. Phys. Soc. 32 281 (1987).[8] Ramsey, N. F., Molecular Beams, Oxford Science Publica-

tions, New York (1985) p. 24 and references therein.

6. Conclusions

We have estimated the processes relevant tolaser cooling atoms in a thermal vapor cell. Wehave shown which parameters are important in at-tempting to observe such cooling. This should be auseful guide when considering the relative meritsof trying to cool other atoms or trying to cool onother transitions. Our experimental results give anupper limit on the size of the cold atom densitybuildup that can be expected for optical molasses ina thermal vapor cell. In particular, we have seenthat it is most important to have a large width forthe cooling transition. It is left open whether anappropriate system can be found with a broadertransition or one that can be broadened artificiallyby some means, such as power broadening, whichwill improve the cold atom density increase andallow cell molasses to achieve its promise as a con-venient source of cold atoms.

7. Acknowledgments

We would like to thank P. Gould and P. Lettwho helped with the experimental part of this pro-

378



News Briefs

General Developments

GRAVIMETERS USED SUCCESSFULLYHighly sensitive and portable instruments to mea-sure the acceleration of gravity, developed by theNIST-University of Colorado Joint Institute forLaboratory Astrophysics (JILA), have performedsuccessfully at various sites around the country infield tests by the National Geodetic Survey. Theresults of this initial year of field testing pave theway for a network of measurements made withthese instruments, both in the United States andabroad. They will be used to monitor local varia-tions in gravity, study vertical ground motions forearthquake detection, and assist in the determina-tion of sea level changes. A paper describing thefield test is available from Jo Emery, NIST, Divi-sion 104, Boulder, CO 80303.

DUCTILE-TO-BRITTLE FRACTUREBEHAVIOR OF STEEL STUDIEDAlthough the fracture behavior of steels is under-stood at high and at very low temperatures, thebehavior at intermediate temperatures has beenhard to characterize. In this ductile-to-brittle tran-sition region, measurements of the fracture tough-ness as a function of temperature show a largescatter, and measurements of different-sized speci-mens at different loading rates give different re-sults. A NIST guest scientist from West Germanyhas clarified the cause of this variability on aquenched and tempered pressure vessel steel (DIN20 MnMoNi 55, similar to ASTM A533B). Fourdifferent types of fracture initiation sites have beenidentified, including cleavage facets, inclusions,clusters of inclusions, and local zones of ductiletearing. The fracture toughness depends on the dis-tance from the crack tip to the initiation site. The

study results are published in Fracture Behavior ofa Pressure Vessel Steel in the Ductile-to-BrittleTransition Region (NISTIR 88-3099), availablefrom the National Technical Information Service,Springfield, VA 22161 for $13.95, prepaid. Orderby PB #89-189195.

CERTIFICATION OF ALUMINUM IN SERUMALBUMIN FOR FDANIST scientists recently developed methods forthe determination of aluminum in serum albumin.The serum albumin was provided by the Center forBiologics Evaluation and Research Division of theFood and Drug Administration (FDA) and will beused by FDA as a quality assurance material. Theanalytical procedures involve the use of atomic ab-sorption spectrometry with electrothermal atom-ization, molecular fluorescence, and inductivelycoupled plasma atomic emission spectrometry.

The serum albumin will be used for quality con-trol in the manufacture and analysis of parenteralmaterials. Parenteral materials enter the bodythrough means other than the digestive tract, e.g.,injections or implants. Aluminum levels in par-enteral solutions are of increasing concern becausetrace quantities of aluminum have been correlatedto diseases affecting nerve tissue and bone. One ofthe greatest impacts has been on patients with renalimpairment or patients receiving large or frequentamounts of albumin. It is expected that thisresearch will greatly aid the FDA regulatory pro-gram for parenterals.

CAC QA PROGRAM IMPROVES THEQUALITY OF MEASUREMENTS MADE INCANCER CHEMOPREVENTION STUDIESFor the past 5 years, the Center for AnalyticalChemistry has operated a quality assurance pro-gram for more than 40 laboratories that measureserum and plasma levels of selected fat-soluble vita-

379



mins as part of National Cancer Institute (NCI)supported investigations to determine the cancer-prevention benefits of selected micronutrients inhuman populations at high risk of contracting cer-tain forms of cancer.

The QA program developed at NIST to supportanalytical measurements made in these studies hasthree main components: (1) development and criti-cal evaluation of analytical methods for the deter-mination of retinol (vitamin A), a-tocopherol(vitamin E) and /3-carotene in serum; (2) interlabo-ratory measurement proficiency testing; and (3) de-velopment of serum based reference materials toserve as accuracy benchmarks.

By the spring of 1988 the first two componentswere well under way and a pilot batch of serum-based reference materials with assigned values forthe concentrations of the three fat-soluble vitaminswas developed and distributed to laboratories par-ticipating in the QA program. In 3 years prior totheir development, interlaboratory precision im-proved from 50 to -25 percent in a regular fash-ion. With reference materials available forlaboratories to validate and trouble shoot theirmethods, there has been a three-fold improvementin interlaboratory precision, to the point where theprecision today is less than 8 percent.

NIST efforts are now being channeled towardimproving measurement comparability for fivenew agents. A program is also being developedto transfer NIST measurement capabilities intwo-dimensional electrophoresis to NCI granteelaboratories for measuring the appearance/disap-pearance of selected proteins as a function of theprogress of various forms of carcinoma. This tech-nology has the potential for use as a diagnostic toolas well as a means for screening prospectivechemopreventive agents.

ATOMIC LIQUID AND SOLID BEHAVIOROBSERVED IN PLASMASNIST scientists have made the first confirmation ofthe existence of condensed states in a plasma sys-tem. Their nonneutral plasmas consisted of up to15,000 positive beryllium ions contained in an iontrap and cooled to a temperature below 10 mK. Atsufficiently high density and low temperature thesystem undergoes a transition from a gaseous ioncloud to a set of concentric shells of ions. In thisstate, ions move freely within a given shell, butmotion between shells is constrained. The system isliquid-like within a shell and solid-like betweenshells, as in a smectic liquid crystal.

There is high interest in this type of system.First, certain internal atomic transitions in thetrapped ions may be used as the basis for a veryhigh quality frequency standard. Second, thetrapped beryllium-ion plasmas at low temperaturemay serve as a model of other interesting systemsthat exhibit similar behavior but are less amenableto systematic study. Some examples are dense, ion-ized states of matter found in astrophysical objectssuch as neutron stars, conduction electrons in semi-conductors, and charged particles confined in high-energy storage rings.

NIST SERVICE FOR SETTING COMPUTERCLOCKSThe Time and Frequency Division of the Centerfor Atomic, Molecular and Optical Physics has justcompleted its first year of operation of a new ser-vice which provides computer and other digitalsystems with telephone access to NIST time ataccuracies approaching 1 ms. Features of the ser-vice include automatic compensation for tele-phone-line delay, advanced (48-day) alert forchanges to and from daylight savings time, and ad-vanced (1-month) notice of insertion of leap sec-onds. Although the main application of this systemis to set computer clocks, simple hardware can alsobe developed to set noncomputer clock systems.The NIST system, entitled Automated ComputerTime Service (ACTS), involves transmission of atime signal, which is then echoed by the user, sothat the telephone time delay can be measured byNIST. The system adjusts for this delay and ad-vances the transmission so that the signal arrives atthe user on time. User software for the service issold through the SRM program as RM8101.

PROPOSED REVISION OF STANDARDFOR POSIX: PORTABLE OPERATINGSYSTEM INTERFACE FOR COMPUTERENVIRONMENTSA proposed revision to Federal Information Pro-cessing Standard (FIPS) 151, POSIX, was an-nounced in the Federal Register. The revision willadopt IEEE Standard 1003.1-1988, Standard forPortable Operating System Interface for ComputerEnvironments which defines a C language sourceinterface to an operating system environment.FIPS 151 was adopted on an interim basis, pendingcompletion of the technical specifications by theIEEE standards committee, to enable federal agen-cies to begin specifying POSIX in procurements.The standard will assist agencies in achieving amore open software environment.

380



PRESENCE OF TOXIC S2F1 o CONFIRMED FORFIRST TIME IN SF6 EXPOSED TO DISCHARGENIST scientists have confirmed for the first timeby direct measurement that the toxic gas disulfur-decafluoride (S2F10) is formed by corona dischargesin compressed sulfur hexafluoride (SF6), a gas in-creasingly used as an electrical insulation and cur-rent-interruption medium in power transmissionand distribution equipment. The team of scientistsalso found that the observed formation rate of S2F10

is consistent with predictions of a theoretical modelfor the discharge chemistry developed by NISTscientists. When SF6 is subjected to an electricaldischarge it decomposes to form various corrosiveand/or toxic byproducts which are of concern inassessing the reliability and safety of power systemsthat use it. The suspicion, not proven until thiswork, had been that S2F10 would be among theseproducts.

NCWM TRAINING MODULE PUBLISHEDThe National Conference on Weights and Mea-sures (NCWM) has published the twelfth moduleof a planned set of training programs for state andlocal weights and measures officials that theNCWM is developing under a grant from NIST.

Module 24, Introduction to NIST Handbook 44,provides an overview of NIST Handbook 44,Specifications, Tolerances, and Other TechnicalRequirements for Weighing and MeasuringDevices, which has been adopted by all 50 states asthe basis for regulation of these devices. The mod-ule includes information on the history of the hand-book, its organization and format, and how to useit. It discusses some of the basic principles underly-ing the handbook, for example, the factors that areconsidered in establishing tolerances for weighingand measuring devices.

This module is the first to be issued by theNCWM in a self-study format. Earlier modules areintended to be taught by an instructor. Each train-ing module includes an inspector's or student'smanual and an instructor's manual or a course ad-ministrator's guide.

Copies of Module 24 have been distributed toeach state weights and measures program. TheNCWM also will have copies of the module avail-able for sale.

HIGH-SPEED METALLIZING OF FIBERSA small-scale pilot plant designed to test new ex-perimental concepts in high-speed electrodeposi-tion of metals and alloys on monofilament

high-strength fibers is now complete and in opera-tion to produce small quantities of metallized fiberfor evaluation. This technology, developed atNIST, makes use of triaxial impinging jets of elec-trolyte to increase the deposition rate. Depositionof metals such as copper, nickel, and cobalt tung-sten have been demonstrated. This experimentalapparatus has allowed cost estimates to be made,and these indicate that metal and/or metallic com-posite precursors can be produced for about$130/kg. Currently available technology is morethan an order of magnitude more expensive.

ELECTROCHEMICAL DEPOSITION OF Al-MnQUASICRYSTALS AND INTERMETALLICSElectrodeposition of aluminum alloys and inter-metallics from molten salt electrolytes offers sev-eral advantages over other metals processingtechniques. The span of operating temperaturespossible with molten salt electrolytes (120-500 'C)allows for the deposition of a variety of structures.In addition, since the solidification is isothermaland alloy composition is rigorously controlled bythe melt composition, the microstructure and com-position of electrodeposited alloys are quite uni-form and predictable. Amorphous, quasicrystallineas well as metastable and stable crystalline struc-tures have been observed in binary aluminum-man-ganese alloys electrodeposited from an electrolytecontaining AlCl3 and NaCl with controlled addi-tions of MnCl2 . Electrodeposited alloys can beformed under conditions far from equilibria wherethe deviation from equilibrium, as well as the de-gree of ordering, is defined by the concurrent pro-cesses of new layer formation and surface diffusion.

The metastable amorphous phase has beenformed by low-temperature electrodeposition. Thedirect formation of quasicrystals, having a level offree energy between that of an amorphous and astable crystalline phase, can be achieved by an in-crease in deposition temperature in a manner some-what analogous to that reported for sputterdeposition. A further increase in electrolyte tem-perature results in the formation of the stable inter-metallic phases which appear on the equilibriumphase diagram. The ability to electrodeposit thick,uniform films of the icosahedral and decagonalphases, for instance, may lead to the first mechani-cal property measurements of these very importantbinary alloys.

381



STRUCTURAL CERAMICS DATABASEDEVELOPEDA prototype structural ceramics database has beendeveloped in the Ceramics Division to provide ma-terials property data for advanced applications ofceramics. The goal of the project is to provide acritical link between the development of new mate-rials in laboratory research studies and the use ofthese materials in competitive product applications.Providing this link will accelerate the developmentof advanced technologies that require structuralceramics for higher temperature operation orgreater durability. The project is funded in part bythe Gas Research Institute, the Standard ReferenceData, and the Ceramics Division.

The prototype is designed for DOS-based per-sonal computers and emphasizes user-friendliness,flexible design, and evaluated data. Property valuesin the system are supplemented with measures ofquality, method and procedures information, andsignificant cautions. Materials are described in con-siderable detail using a modification of the ASTMCommittee E-49 recommendations, and a completebibliography of the sources of data is included.

Currently, the prototype contains thermal andmechanical property data for monolithic siliconcarbides and silicon nitrides. Preparations are nowunder way to test the prototype at selected indus-trial sites.

MAGNETIC ORDER OBSERVED INARTIFICIAL SUPERLATTICESNIST scientists and guest researchers from theUniversity of Notre Dame, have recently made thefirst neutron diffraction determination of magneticorder in artificial metallic superlattices. The speci-mens consisted of layers of dilute magnetic semi-conductor materials and were produced bymolecular beam epitaxy techniques which yielded100-200 alternate layers of MnSe and ZnSe, each ofthickness approximately 2-5 nm. The results of theneutron measurements clearly identified long-rangeantiferromagnetic order within the MnSe layers. Inthe direction perpendicular to the atomic planesthe ordering is limited in extent by the interfacewith the nonmagnetic ZnSe layer. No evidence ofmagnetic propagation between MnSe layers wasfound, a result consistent with predictions of cur-rent magnetism theory.

COLLABORATIVE RESEARCH EFFORT ONMATERIALS RESEARCH RENEWEDA major manufacturer of semiconductor devices,has begun its fourth year of collaborative research

with NIST on the study of materials used in theproduction of semiconductors. Using the neutrondepth profiling facility (NDP) operated by theCenter for Analytical Chemistry (CAC), re-searchers will continue investigations on the thinfilms that form the passivation layers on semicon-ductor devices. Elemental depth profiles are beingdetermined to better understand and control the ef-fects of processing in device fabrication. NDP, aquantitative and nondestructive technique, pro-vides a means to make repetitive analyses of a sin-gle sample at various processing stages. Othertechniques also can be used in conjunction withNDP on the same sample volume. Key experimentsare now being designed to take advantage of boththe improved detection limits for the technique andthe improved research facilities being constructedas part of the Cold Neutron Research Facility.

NY FIRM JOINS WITH NIST TOINVESTIGATE CID SENSORSA Liverpool, NY, firm has begun a twofold re-search program with researchers at NIST. One ob-jective of the program is to improve chemicalanalysis accuracy and detection limits through theuse of charge injection device (CID) sensors in an-alytical instrumentation. The second goal is toidentify and investigate the effects of various formsof nuclear irradiation on the CID sensors and theirassociated camera systems.

The CID sensor is an imaging device that offersnumerous technical advantages over relateddevices, such as the more commonly knownCharge Coupled Device (CCD). For example,CID cameras typically allow excellent exposurecontrol in low-light situations, and they also aremore tolerant to intense light, producing accurateimage detail even under extreme lighting condi-tions. This has made the devices ideal for purposessuch as missile tracking, semiconductor patternrecognition, and factory process line inspection.

Scientists from industry and NIST will use theNBSR 20-MW reactor and other intense sources ofradiation to expose the sensors and associated CIDcomponents to x-ray, gamma, beta, neutron, andcharged particle environments to determine "radia-tion hardness" of the materials. The electroniccharacteristics of the devices will be gauged be-fore, during, and after radiation exposure. The re-searchers will also explore the analyticalinstrumental uses of CIDs in a variety of spec-troscopy applications where the signal intensitymay vary greatly in magnitude during the analysisof a sample.

382



COLD NEUTRON WORKSHOP ONANALYTICAL MEASUREMENTS WITHCOLD NEUTRON BEAMSThe Center for Analytical Chemistry and the Re-actor Radiation Division hosted the third in a seriesof meetings on the Cold Neutron Research Facility(CNRF). The recent meeting was held to informthe scientific community of this national facilityand to solicit recommendations on the instrumentsbeing built. Two analytical measurement tech-niques, prompt gamma neutron activation analysis(PGNAA) and neutron depth profiling (NDP),were specifically highlighted for scientists, man-agers, and students from universities, industry, andother agencies from across the United States.Through invited talks and open discussions, the po-tential of the new instruments was explored. Theparticipants reviewed the basics of the two tech-niques and discussed applications in a variety ofareas, including semiconductors, polymers, metals,superconductors, biological materials, environmen-tal monitoring, and fundamental physics. Addition-ally, the users policy for the CNRF was described,including required procedures for submitting pro-posals to gain access to the new instruments.

HIGH EFFICIENCIES OF SOFT X-RAYMIRRORS MEASURED AT SURF IINIST scientists in collaboration with researchersfrom the University of Arizona, Goddard SpaceFlight Center, Lawrence Berkeley Laboratory, andLawrence Livermore Laboratory, have recentlymeasured the reflectivity of soft x-ray opticaldevices at angles of the incident radiation near nor-mal to the surface and at wavelengths from 150-250A at SURF II. Reflectivities up to 47 percent weremeasured. The soft x-ray optics beamline at SURFII is one of the few facilities in the world capable ofmaking these measurements. By working interac-tively with the few laboratories capable of makingefficient soft x-ray mirrors, we are helping to de-velop the emerging technology of soft x-ray optics.

This high reflectivity at normal incidence allowsscientists to make high-quality soft x-ray imagingsystems used in the remote sensing and imaging ofobjects that emit strongly in this spectral region,e.g., the Sun and stars, with unprecedented spatialresolution. Because these devices are also wave-length selective, they act as "monochromatic" fil-ters and discriminate against other wavelengths.

These wavelength selective, stigmatic mirrorsare also being used to make efficient optical cavi-ties for experimental soft x-ray laser systems. Other

applications include: soft x-ray lithography to ob-tain linewidths on semiconductor chips substan-tially smaller than now attainable and soft x-raymicroscopy to make real-time observations of liv-ing cells in the so-called "water window" between24 and 45A.

LASER ABLATION SOURCE FORSTRUCTURAL STUDIESRecently, NIST scientists completed the develop-ment of a novel molecular beam source to be usedin conjunction with a pulsed-beam Fourier trans-form microwave spectrometer for molecular struc-ture studies. This new capability now makes itpossible to examine the technologically importantrefractory materials and low vapor pressure metalsfor which very few studies have been possible. Per-haps even more exotic molecules, e.g., mixed metalcompounds and metal-molecule clusters, can be in-vestigated with this new method.

The laser ablation beam source uses a Nd-YAGlaser to vaporize solid materials into a rare gaspulsed jet, which is then examined for rotationalspectra by the Fourier transform spectrometer.This method for forming vapor phase refractorymolecules for spectroscopic studies overcomes sub-stantial limitations in the conventional method ofbrute force heating of solid samples of the startingmaterials. One of the most problematic limitationsof the conventional approach is a materials prob-lem, i.e., the ability to find high-temperature inertmaterials as a container for substances heated inexcess of 2000 'C. The laser ablation method re-quires no container and heats only a small area ofmaterial, -0.1 mm3 . The vapor is then entrained ina supersonic pulse of inert gas for delivery to thespectroscopic chamber. The initial studies with thisinstrument have included the examination of therotational spectra from SiC2, an important circum-stellar molecular species, and the oxides of the tran-sition elements Y, La, Zr, and Hf. The rotationaltransition frequencies are used to determine thebond distance between the metal and oxygen atom.Measurements of the shifts in these frequencies,when an external electric field is applied, yield themolecular dipole moment and provide importantinformation on the electronic nature of the molecu-lar bonding.

FREQUENCY TABLES FOR CARBONMONOXIDE LASERSIn collaborative work with the University of Bonn,NIST scientists have obtained new, highly

383



accurate values for the molecular constants of car-bon monoxide. This was done by measuring thefrequencies of spectral lines of a stabilized COlaser. Frequency-measured transitions are somethousand times more accurate than wavelength-measured ones. While some of the spectral lines arewell-known from earlier, isolated experiments,most of them have been improved by about a fac-tor of 1,000 in this work. The results contributesubstantially to the list of unified wave numberstandards used by the scientific community to cali-brate spectrometers. The increasing use of Michel-son interferometers has increased the need for suchstandards in all parts of the spectrum. The resultsare also useful for spectroscopic studies of impor-tant molecules in space and in the upper atmo-sphere. The measurements were in the region from1140 to 2100 wave numbers, i.e., 34-62 THz, or4.7-8.7 pm.

NEW RECORD BEAM CURRENT, 300 mA,ACCELERATED TO FULL ENERGY IN SURF-II,THE NIST ELECTRON STORAGE RINGThe performance of the Synchrotron UltravioletRadiation Facility's (SURF-II) electron storagering has been improved. Electron beams exceeding250 mA are being accelerated to operating energyon a regular basis, and a new record of 300 mA hasbeen achieved. The light intensity at all wave-lengths radiated by the relativistic electron beam isdirectly proportional to the beam current. Thehigher radiation output achieved results in im-proved signal-to-noise for the surface science,atomic and molecular physics, and optical proper-ties measurements that are investigated at the facil-ity. These investigations are done by variousgroups from NIST, other government agencies,universities, and private industry. The storage ringalso serves as a primary irradiance standard in thefar ultraviolet (4-300 nm) and this improvement hasextended the dynamic range of spectral irradiancethat is provided to groups who use the facility tomake absolute calibrations of monochromators orspectrometers.

AGREEMENTS OF OPEN SYSTEMSINTERCONNECTION (OSI) PROTOCOLSPUBLISHEDA new National Computer Systems Laboratory(NCSL) publication records working implementa-tion specification agreements of OSI protocolsamong the organizations participating in the NISTWorkshop for Implementors of OSI. Issued asNISTIR 89-4082, the document is based on the

proceedings of the NIST Workshop Plenary As-sembly held March 17, 1989. While work describedin this document is not advanced enough for use inproduct development or procurement reference,the work is intended as a basis for future stableagreements. As each protocol specification is com-pleted and becomes technically stable, it is movedfrom this working document to the stable compan-ion document, which records mature agreements.NIST Special Publication 500-162, Stable Imple-mentation Agreements for Open Systems Intercon-nection Protocols, Version 2, Edition 1, is thecurrent version of these mature agreements.

The NIST Workshop for Implementors of OSIaccepts the specifications of emerging standards forprotocols and produces agreements on the imple-mentation and testing particulars of these proto-cols. This process expedites the development ofOSI protocols and promotes interoperability of in-dependently manufactured data communicationsequipment.

NCSL RELEASES REVISED STRUCTUREDQUERY LANGUAGE (SQL) TEST SUITENational Computer Systems Laboratory (NCSL)released Version 1.2 of the SQL Test Suite whichhelps users and vendors determine compliancewith FIPS 127, Database Language SQL. The re-vised test suite adds the programming language in-terface Pascal and includes tests for features to beavailable in the next revision of FIPS 127, expectedto be approved later in 1989. Version 1.2 replacesthe original version of the SQL Test Suite releasedin August 1988. Over 30 companies presently uti-lize the SQL Test Suite.

OPTIONAL MODULE FOR INFORMATIONRESOURCE DICTIONARY SYSTEM (IRDS)STANDARD TO BE DEVELOPEDIn June 1989, ANSI's Technical Subcommittee X3approved the development of an optional modulefor the IRDS standard in the area of Naming Con-vention Verification and assigned responsibility forthe X3 Technical Report to Task Group X3H4.3,IRDS Naming Convention Verification. Approvedin March as FIPS 156, the IRDS is a software sys-tem that records, stores, and processes informationabout an organization's data and data processingresources. NCSL's Special Publication 500-149,Guide on Data Entity Naming Conventions, is be-ing used as the basis for developing the report.

384



NSF GRANT FOR NIST SENIOR RESEARCHFELLOWSHIPNational Science Foundation (NSF) has awarded agrant to the American Statistical Association forthe support of a research fellowship program atNIST in the Center for Computing and AppliedMathematics. The research fellow will be selectedcompetitively based on recommendations from anadvisory group of leading industrial engineers andstatisticians. The fellow will be a guest worker inthe Statistical Engineering Division, leading a teameffort in cross disciplinary research in process mod-eling and optimization as a basic approach to qual-ity. NSF has approved this program for 3 years.During this period, several fellows are expected tocollaborate with other NIST centers on a numberof different processing research projects.

FIRST DEMONSTRATION OF SOLITON-LIKECOMPRESSION OF PULSES FROM OPTICALFIBER LASERNIST scientists have demonstrated for the firsttime that pulses from fiber lasers have the neces-sary power and spectral purity to produce soliton-like compression in external fibers. This soliton-likereduction in pulse width has previously been ob-served only in pulses from large, high-power lasers.An application of the work is the development ofultra-high-bit-rate lightwave communications sys-tems. The team mode-locked an erbium fiber laser,formed by optically pumping various lengths oferbium-doped fiber. For a fiber cavity length of 20m, the resulting pulses are 18 ps wide at a wave-length of 1536 nm and a peak power of 0.5 W.Using a streak camera, the team then measured thewidth of pulses from the erbium laser that had tra-versed 14 km of ordinary telecommunications-grade optical fiber and found that they werecompressed to 5 ps. At the 0.5-W power level, thenonlinear component of the fiber refractive index islarge enough to cause a "chirp" or variation of op-tical frequency during the pulse. At a wavelengthof 1536 nm the group velocity dispersion in mostfibers is negative, with the result that the trailingedge of the pulse travels faster than the leadingedge, and pulse compression can result from propa-gation through a fiber.

EIA COMMITTEE ASKS NIST TO DEVELOPSTANDARDS FOR FIBER GEOMETRYElectronic Industries Association CommitteeF06.6 on Fibers and Materials has expressed con-cern over the adequacy of present methods forfiber geometry and asked NIST to develop trace-able standards. This request is a result of the com-mittee's preliminary analysis of data from aninternational comparison of relevant measurementmethods. NIST and the British Telecom ResearchLaboratories (BTRL) carried out an evaluation ofthe current state of methods for determining thegeometry of optical fiber in response to a move-ment in the lightwave industry to reduce tolerancesfor physical dimensions of optical fibers. NIST ad-ministered the comparisons among U.S. andJapanese laboratories, while BTRL interacted withEuropean laboratories. Twelve single-mode fiberswere measured for outside diameter, ovality, andconcentricity error. In total, 6 different techniqueswere evaluated, including microscopy with imageprocessing, image shearing, microscopic interfer-ometry, and Fizeau interferometry. Committees ofthe Electronic Industries Association (EIA) andthe International Telegraph and Telephone Con-sultative Committee are now studying the results,which will be the subject of a joint NIST-BTRLreport following a full analysis of the data.

GMAP SYSTEM ARCHITECTURE INSTALLEDAT THE NATIONAL PDES TESTBEDIn support of the National Product Data ExchangeSpecification (PDES) testbed project within theFactory Automation Systems Division of NIST, aguest worker from private industry, has installedthe Geometric Modeling Applications InterfaceProgram (GMAP) system architecture for support-ing product data technology on the AMRF VAX11/785. This work was done as part of the technol-ogy transfer tasks under the U.S. Air Force. Soft-ware tools to facilitate the testing of the currentPDES specification are part of the GMAP systemarchitecture and will be made available to thePDES community at large as part of the nationalPDES testbed.

The GMAP system architecture as installed atNIST supports every entity in the PDES/STEPVersion 1.0 draft proposal, available from NTIS asdocument number PB 89-144794. However, theschema upon which the system currently operates

385



can be removed and replaced with future versionsof PDES/STEP, utilizing the ability of the GMAPsystem architecture to maintain schema indepen-dence. PDES models and application programs canbe created under the GMAP system to aid in test-ing the current draft proposal.

NIST QUALITY-IN-AUTOMATION PROJECTDRAWS MACHINE-TOOL COMPANYCOLLABORATORSThe "Quality-in-Automation" project of NIST'sAutomated Manufacturing Research Facility(AMRF) has brought together senior engineersfrom a machine-tool company and a measurement-probe company to confer on a new measurementtechnique being developed within the AMRF.

As part of the AMRF project, CME researchershave developed an architecture for quality controlin automated manufacturing, one aspect of whichinvolves on-machine process-intermittent gaging,that is, high-speed automatic measurement of thedimensions of parts between the stages of the ma-chining operations. The senior engineers conferredwith CME researchers on configuring probes formachine tools to allow CME research on a generictechnique for process-intermittent gaging. Bothcompanies view the work of CME as invaluable indeveloping such advanced measurement tech-niques and demonstrating their feasibility on com-mercially available equipment such as theirs.

CBT DEVELOPS NEW DESIGN PROCEDUREFOR HEAT LOSSES FROM UNDERGROUNDPIPING SYSTEMSCBT has developed a new finite element computermodel to estimate heat loss along structural sup-ports for underground insulated steam and hot wa-ter pipes enclosed in shallow concrete trenches.Structural supports for these underground pipesare often in direct contact with the hot carrierpipes and form highly conductive heat flow pathsthrough the pipe insulation to the concrete trenchand the surrounding earth. The CBT model isbased on two-dimensional steady-state heat con-duction assuming a rectangular trench containingtwo insulated pipes with and without pipe sup-ports. Two different configurations of typicalstructural supports were modelled. Sample calcula-tions showed that heat loss per unit length in thevicinity of these structural supports could be ashigh as 17 times the value of the insulated pipealone.

CBT MONITORS THE SOURCE STRENGTHSOF VOLATILE ORGANIC COMPOUNDS IN AMAJOR NEW FEDERAL OFFICE BUILDINGVolatile organic compounds (VOC) emitted intothe air of modern office buildings that often lead to"sick building syndrome" are typically thought tobe outgas from new materials when a building isfirst constructed or newly remodelled. CBT hasdiscovered that this observation is not true for atleast one new building. Over the last 14 months,CBT researchers have been monitoring the indoorair quality in a newly constructed 10-story federaloffice building in Portland, OR, beginning with theday of occupancy. They measured VOCs alongwith CO, CO 2 , respirable particulates, formalde-hyde, and radon. In addition, they made measure-ments of air infiltration, ventilation rates, andinterzone air movement. The study indicated thatthe level of VOCs did not decrease over the moni-toring period as would be expected if the predomi-nant source had been building and interior finishmaterials. The levels remained relatively constantand could be clearly related to occupant activities.The dominant compounds found were CI0-C11branched alkanes emitted by 26 liquid process pho-tocopiers and three liquid process plotters. In addi-tion, a number of light hydrocarbons characteristicof vehicle exhaust were found and were traced tovehicles on the loading dock and in three levels ofthe underground garage. Although the concentra-tion of some of these compounds was relativelyhigh, the level of total organic carbon was threeorders of magnitude below specified occupationalthreshold limit values for exposure to these carboncompounds. The potential for the measured con-centrations of these carbon compounds to producesymptoms characteristic of the "sick building syn-drome" (i.e., mucous membrane irritation,headache, nausea, and dizziness) is unknown andrequires further study by health scientists.

X-RAY METHOD REMOVES AIDS VIRUSFROM EVIDENCEResearchers at NIST, working with the FederalBureau of Investigation and the National CancerInstitute, have tested an x-ray technique that wipesout the AIDS virus on criminal evidence withoutdestroying important biological components.Workers in crime laboratories face a potential riskof accidental AIDS infection from exposure tocriminal evidence-bloodstained clothing, forexample-that may contain body fluids contami-nated with the AIDS virus. Though steam

386



sterilization and related procedures can effectivelyinactivate the virus, these methods also can destroyimportant biological components, lessening or evencanceling the value of the samples as criminal evi-dence. The researchers used an industrial radio-graphic instrument to produce the x rays andbombarded virus-tainted samples with varying in-tensities of x-ray radiation to determine the smallestamount needed to inactivate the AIDS virus at thelowest detectable concentration of the virus. Theyarrived at a dose that is about 25,000 times largerthan a typical chest x ray but concluded that higherviral concentrations would require even larger ra-diation doses. The FBI is incorporating the newtechnique into its forensic labs, and other law en-forcement agencies will likely follow suit.

BALDRIGE QUALITY AWARD PROGRAMSEEKS EXAMINERSThe Malcolm Baldrige National Quality AwardProgram currently seeks applications from individ-uals who can qualify as examiners. The award isoffered annually to American companies thatdemonstrate the highest levels of total quality man-agement. Information on the program and applica-tions to serve as an examiner are available from theMalcolm Baldrige National Quality Award, NIST,A1123 Administration Bldg., Gaithersburg, MD20899. As an examiner, the individual is responsiblefor reviewing and evaluating applications submit-ted for the award. Those chosen meet the higheststandards of qualification and peer recognition andmust participate in a preparation course based onthe examination items, the scoring criteria, and theexamination process. The Board of Examiners nownumbers over 130 leading quality experts selectedfrom industry, professional and trade organiza-tions, and universities.

INTERNATIONAL DIRECTORY OFSTANDARDS GROUPS UPDATEDThe newly revised Directory of International andRegional Organizations Conducting Standards-Related Activities (NIST SP 767) provides infor-mation on 338 groups that conduct standardization,certification, and laboratory accreditation activi-ties. The directory is designed to serve the federalagencies, standards writers, manufacturers, ex-porters, and others concerned with U.S. participa-tion in international standards. With more than 60new listings, the volume is one in a series toprovide information on multinational standards-re-lated endeavors. International and regional organi-

zations are listed in alphabetical order. Informationincludes acronyms, national affiliations, U.S. par-ticipants, membership restrictions, scope of inter-est, and availability of standards in English. Copiesof NIST SP 767 are available from the Superinten-dent of Documents, U.S. Government Printing Of-fice, Washington, DC 20402. Order by stock no.003-003-02937-8 for $22 prepaid. For a list of otherNIST standards-related and certification directo-ries, contact: Office of Standards Code and Infor-mation, NIST, A629 Administration Bldg.,Gaithersburg, MD 20899; telephone: 301/975-4031.

EQUATION DEVELOPED FOR PREDICTINGSTEEL BEHAVIOR IN FIRENIST scientists have developed an equation forpredicting the performance of ASTM A36 struc-tural steel during and after a fire. Designed for per-sonal computers (PCs), the equation predicts theperformance of the most commonly used steel inthe United States when the material is exposed tofire for periods from 2 min to 8 h over a tempera-ture range from 350 to 650'C. To assess deforma-tion, an analysis of the various stresses,temperatures, and times involved are required inorder to calculate the resulting strains. Each typeof strain-elastic, recoverable; plastic, time-inde-pendent; and creep, time-dependent-has its ownformula. The equation also can be used to con-struct "deformation mechanism" maps to show thedominant behavior of steel at any given tempera-ture, time, and stress, and to compare the perfor-mance of A36 to foreign steels. A report, ElevatedTemperature Deformation of Structural Steel(NISTIR 88-3899), with the new equation and in-formation on the correlations between measuredand predicted strains for ASTM A36, AustralianAS A149, and Japanese SS41 structural steel steels,is available from the National Technical Informa-tion Service, Springfield, VA 22161. Order byPB# 89-172621/AS for $21.95 prepaid.

HIGH-QUALITY AMORPHOUS SILICONFILMS PRODUCEDGovernment, industry, and academia are workingto develop an inexpensive source of solar electric-ity-amorphous silicon photovoltaic solar cells.NIST and University of Colorado researchers atthe Joint Institute for Laboratory Astrophysicshave produced high-quality hydrogenated amor-phous silicon films by decomposing silane gas on avery hot surface and depositing it on a nearby rela-tively cool 210 'C substrate. The technique uses a

387



low-pressure feed gas that offers great potential forhigh deposition rates, above 1 nm/s. The resultsobtained in this work demonstrate a low incidenceof defects, which interfere with the photovoltaicefficiency and electron transport mechanisms.While high deposition rates were not the goal ofthe research, a low-pressure feed gas can be effi-ciently dissociated into a large flux of depositingradicals so that the technique looks promising formore efficient production of solar cells. The lowpressure also avoids the undesirable production ofdust which introduces defects into the depositedlayers. A paper describing the process is availablefrom Jo Emery, NIST, Division 104, Boulder, CO80303; telephone: 303/497-3237.

NON-TOXIC ACIDS ARE BASIS FOR NEWCLASS OF CEMENTSA new family of nonaqueous dental cements havebeen developed based on nontoxic dimer andtrimer acids. They can be used in temporary fill-ings, liners, and as a foundation for other restora-tive materials. The new patented formulations,developed by a NIST scientist, use a dimer and/ora trimer acid derived from the partial polymeriza-tion of simple unsaturated fatty acids. These liquidscan be mixed with a variety of basic powders usedin restorative materials such as zinc oxide and cal-cium hydroxide. When mixed with powders, thebulky, flexible, hydrophobic nature of the dimerand trimer acids with their relatively low car-boxylic acid content produce cements that aretough, low shrinking, water resistant, hydrolyti-cally stable, and nonirritating. These new materialscan be replacements for the eugenol cements thatare mechanically and hydrolytically weak. For in-formation on the new nonaqueous dental cements,contact: Dr. Joseph M. Antonucci, NIST, A143Polymers Bldg., Gaithersburg, MD 20899; tele-phone: 301/975-6794.

QUICK WAY TO MEASURE HEAT CONDUCUIONResearchers from the NIST Center for BuildingTechnology and the Virginia Polytechnic Instituteand State University have investigated a quick tech-nique for measuring on-site the heat-conductingproperties of building insulation materials. Gener-ally, this information is obtained through labora-tory testing. Once in place, however, the materialcan settle or accumulate moisture and lose some ofits insulation value. The new technique uses a ther-mistor, a semiconductor device, as a probe alongwith a computer which controls the system and

gathers data. Once inserted into the insulation, theprobe acts both as a heater and a sensor to measurehow the heat is being dissipated through the insula-tion over time. While the technique does have ad-vantages, it also has limitations-primarily, a lackof suitable calibration materials. A report,Development of an Automated Probe for ThermalConductivity Measurements (NIST-GCR-89-563),is available from the National Technical Informa-tion Service, Springfield, VA 22161. Order byPB# 89-209324 for $21.95 prepaid.

IS THERE A ROBOT IN YOUR FUTURE?If you work in a federal building, there might be.The use of robotics in factories has exploded withinthe past decade with over 30,000 robots in use inthe United States and more than 200,000 world-wide. Outside factories, using robots to performtasks such as building operations and maintenanceis receiving increased attention. In a study for theGeneral Services Administration (GSA), whichconstructs, operates, and maintains federal build-ings, the NIST Center for Building Technologyconducted a state-of-the-art survey of robotic tech-nology. The NIST researchers also identified po-tential barriers to using service robots in GSAbuildings and suggested some changes to GSA'shandbook of facility standards to accommodaterobotic technology. Assessment of Robotics forImproved Building Operations and Maintenance(NISTIR 88-4006) is available from the NationalTechnical Information Service, Springfield, VA22161. Order by PB# 89-189237/AS for $15.95prepaid.

PROPERTIES OF METHANE DEVELOPEDThe thermophysical properties of methane havebeen tabulated by NIST for a large range of fluidstates based on recently formulated correlations.The use of improved correlations is important eco-nomically because methane is the major constituentof natural gas. The tables include: thermodynamicproperties at temperatures from 91 to 600 K andpressures less than 100 MPa, viscosity at tempera-tures from 91 to 400 K and pressures to 55 MPa,and thermal conductivity from 91 to 600 K andpressures to 100 MPa. Algebraic expressions andassociated tables of coefficients are given to permitadditional property calculations. A FORTRANprogram is listed for the evaluation of methanethermophysical properties using the Schmidt-Wag-ner equation of state with separate correlations forthe viscosity and thermal conductivity. Copies ofTables for the Thermophysical Properties of

388



Methane (TN 1325) are available from the Superin-tendent of Documents, U.S. Government PrintingOffice, Washington, DC 20402. Order by stocknumber 003-003-02947-5 for $23.00 prepaid.

FIBER LASER RESEARCH PROGRESSINGNIST researchers are investigating a laser consist-ing of a 90-cm length of optical fiber made of glassdoped with erbium. The fiber, which has an 1 l-mcore, is pumped with green continuous-wave (CW)argon laser light at 528 nm. It produces CW in-frared laser light at 1540 nm, suitable for transmis-sion through optical fiber networks. NIST's workis aimed at reducing the linewidth of the infraredoutput to about 1 MHz, from the current 5 to 10GHz, to realize a wavelength standard in this spec-tral range. A narrow bandwidth fiber laser couldbe used as a reference, or possibly the communica-tion source, in a coherent communications systemwith increased channel capacity. Conventionalpulse modulated communications over opticalfibers require that the signal be amplified and re-conditioned by repeaters at intervals of 30 to 60km. Coherent communications, on the other hand,do not use wide bandwidth pulses to carry informa-tion but can use other techniques, such as closelyspaced frequency modulation (FM) channels.Coherent communication signals are less affectedby dispersion and do not need reconditioning asoften. This means the repeaters could be spacedfarther apart, which would reduce the cost of thenetwork.

HAZARD I RELEASED TO HELP REDUCEFIRE DEATHS, COSTNIST released HAZARD I, a unique analysismethod and computer program that promises torevolutionize fire safety practices and lead tomarked reductions in fire losses and costs. HAZ-ARD I can be used on a personal computer to pre-dict the hazards to occupants in a burning buildingand the relative contribution of products, such asfurniture, to those hazards. It is the first such com-prehensive use of fire modeling in the world.HAZARD I can be used in a variety of ways infields ranging from fire investigation and recon-struction and fire protection engineering to thoseconcerned with product development, manufactur-ing, and marketing and architectural design. Also,it can be used to help building and fire codes offi-cials evaluate new materials or design and con-struction techniques. A three-volume report and aset of computer disks are available for $225 fromthe National Fire Protection Association, One Stop

Data Shop, Batterymarch Park, Quincy, MA 02269or the National Technical Information Service,Springfield, VA 22161. Use PB# 89-215404 whenordering from NTIS.

NIST LAUNCHES CONSORTIUM TODEVELOP FUTURISTIC LABWith the goal of pooling the resources of industry,government, and academia to develop a totally au-tomated analytical chemistry laboratory, NIST hasinstituted the Consortium on Automated Analyti-cal Laboratory Systems (CAALS). Industry willlikely welcome this futuristic laboratory, where au-tomated devices-including robots-will performthe entire analytical procedure under the control ofa sophisticated computer system. Analytical labo-ratory work often is labor intensive, performed byhigh-cost employees using expensive equipment.This makes the millions of yearly chemical analysesquite costly to industry and underscores the poten-tial benefits of an automated laboratory. CAALSwill pool the resources of a variety of experts tocreate a world-class team that will produce a viablerobotics system that can be freely adapted. NISTwelcomes consortium participants. For more infor-mation, contact Dr. H.M. (Skip) Kingston, NIST,A353 Chemistry Bldg., Gaithersburg, MD 20899;telephone: 301/975-4136.

INDUSTRY TO STUDY WEAR OF CERAMICSAT NISTPrivate industry is sponsoring a Research Associ-ate Program at NIST to study the tribologicalcharacteristics of ceramics and ceramic compositesunder various temperatures. Researchers will use aNIST high-temperature wear test facility and otherspecial equipment to conduct studies on the fric-tion, wear, and mechanical behavior of advancedceramic materials. NIST currently is conducting atribology research program that includes advancedceramics, coatings, and composites as well as thelubrication requirements of these materials. Forfurther information, contact: Dr. Said Jahanmir,NIST, A215 Metrology Bldg., Gaithersburg, MD20899; telephone: 301/975-3671.

EVALUATION OF HIGH-FREQUENCYPOWER METERSAccurate power measurements are fundamental toassessing the performance of almost all radiofre-quency, microwave, and millimeter wave equip-ment. Power considerations are also crucial for thedesign of efficient, cost-effective, and safe systems.

389



Measurements at these higher frequencies are af-fected by many factors. Impedance mismatch, in-terference, leakage, nonlinear effects, and othersources of error must be assessed and minimized. ANIST publication, Performance Evaluation ofRadiofrequency, Microwave, and Millimeter WavePower Meters (NIST TN 1310), describes mea-surement techniques for evaluating the electricalperformance of certain commercially availablepower meters that use bolometric sensors and oper-ate typically from 10 MHz to 26.5 GHz. TN 1310 isavailable from the Superintendent of Documents,U.S. Government Printing Office, Washington,DC 20402. Order by stock number 003-003-02931-9for $8.50 prepaid.

PORTLAND CEMENT CLINKER MATERIALSAVAILABLEProducers of Portland cement have three new ce-ment clinker reference materials (RMs) for assess-ing the quality of materials during processing.Portland cements produced in the United Statesare evaluated by ASTM C150, "Standard Specifi-cation for Portland Cement" comprising chemicalcomposition, fineness, and performance in physicaltests. The materials selected for the RMs differwidely in phase abundance, crystal sizes, and distri-bution crystals. The phase abundances in weightpercent, in RMs 8486, 8487, and 8488, respectively,are 58.47, 73.39, and 64.97 alite; 23.18, 7.75, and18.51 belite; 1.15, 12.09, and 4.34 aluminate; and13.68, 3.27, and 12.12 ferrite; along with concentra-tions of free calcium/oxide, periclase, and alkalisulfate. The phases were determined by pointcounting using reflected light microscopy. Theother properties were determined by chemicalanalysis. Each RM consists of three hermeticallysealed glass vials with approximately 10 grams ofcrushed clinker. For information on RMs 8486,8487, and 8488, available for $96 each, contact theOffice of Standard Reference Materials, NIST,B311 Chemistry Bldg., Gaithersburg, MD 20899;telephone: 301/975-OSRM (6776).

OH, SAY CAN YOU SEE?Lighting can be soft or harsh or too dim or toobright or even just right. But, no matter how it isperceived, lighting greatly influences not only ourability to see but also our feelings and even ourhealth. In an effort to understand how lightingaffects human performance, researchers at NISTand the Lighting Research Institute studied energyconsumption, brightness, and peoples' perceptions

of their lighting systems. The researchers used datafrom 912 workstations in 13 office buildings.Among the findings: About 70 percent of the occu-pants surveyed were satisfied with their lighting.Of those who were dissatisfied, 37 percent hadfurniture-integrated task lighting combined with anindirect ambient lighting system. Energy consump-tion also was higher for this system. Dissatisfactionwas related to the patterns of brightness-worksta-tions with lower brightness were less successful.Copies of the report, Evaluating Office LightingEnvironments: Second Level Analysis (NISTIR89-4069), are available from the National TechnicalInformation Service, Springfield, VA 22161. Orderby PB# 89-189153 for $21.95 prepaid.

NEW NIST STANDARD REFERENCEDATABASENIST Standard Reference Database 17-NISTChemical Kinetics, released earlier of this year,provides rapid access to kinetics data for gas phasereactions. The database contains more than 8,000data items on 2,000 separate reactions. Searchableby reactants or by reference, retrievals provide sur-veys of the literature on a particular reaction, all orsubsets of the reactions of a specific species, andthe data available from a given paper. NIST Chem-ical Kinetics, which can function on any MS-DOSor PC-DOS, has rapidly become a best-seller.

CERAMIC POWDERS SYNTHESIS-MAGNETIC MEDIAThe Ceramics and Metallurgy Divisions are exam-ining a new class of magnetic composite materials.These materials consist of ultrafine magnetic parti-cles of less than 20 nm, and offer considerable po-tential for ultradense magnetic recording media.

Information transmission, storage, and retrievaltechnologies have experienced phenomenal growthwithin the past two decades. Thus, there exists acontinuing need for new or improved approachesto support this expanding field. For example, thearea of magnetic recording is highly dependent onmagnetic particulate media.

Magnetic composite materials are generated us-ing sol-gel techniques appropriate to the formationof diphasic iron-silica xerogels. These compositesare synthesized by room temperature polymeriza-tion and curing of gels derived from solutions oftetraethylorthosilicate, ferric nitrate, and water.The hydrofluoric acid is used as a catalyst. Theresulting composite is a homogeneous dispersion ofultrafine iron-containing particles in a silica-likematrix.

390



The chemical and magnetic states of the iron inthese composites can be established in situ by ex-posure to gaseous reducing or oxidizing environ-ments at temperatures less than 400 'C. Magneticmeasurements and Mbssbauer spectroscopy onthese systems have established that these nanocom-posites have magnetic regions with diameters<20 nm. Furthermore, the bulk materials can becontrollably engineered to display paramagnetic,superparamagnetic, ferrimagnetic, or ferromag-netic behavior.

EXPERT SYSTEM FOR MATERIALSELECTION ADVICE FOR THE PETROLEUMINDUSTRYA knowledge base expert system has been devel-oped in the corrosion data center at NIST in coop-eration with the New Zealand Department ofScientific and Industrial Research (DSIR) and theNational Association of Corrosion Engineers.

The program operates on personal computersand provides material selection advice for compo-nents of pumps used in the downhole environment(sucker rod pumps). Industry standards, togetherwith corrosion advice and information obtainedfrom oil and gas production companies and equip-ment manufacturers, provide a basis for assessingcurrent industry practice and key environmentalfactors. The expert system considers a broad rangeof user-supplied chemical and physical parametersto characterize the anticipated corrosivity of com-plex downhole environments and the relative suit-ability of corrosion resistant engineering materials.The system defines the downhole environment byapplying rules and inference procedures structuredfrom the domain expertise derived from the infor-mation sources. The knowledge base encompassesover 600 rules and provides options of materialsselection consultation, review of related domain in-formation, and addition of custom rules to high-light user-specified in-house requirements.

SANDIA TO PARTICIPATE IN COLDNEUTRON FACILITY INSTRUMENTDEVELOPMENTThe Organic and Electronic Materials Departmentof Sandia National Laboratories (SNL) agreed toparticipate with NIST in the development and useof two high-resolution instruments at the ColdNeutron Research Facility (CNRF). This is thefirst such facility in the United States.

The instruments are the high-resolution neutrontime-of-flight spectrometer and the back-reflection

spectrometer, both of which are already in the pre-liminary design phase at NIST. Both instrumentsare for inelastic neutron scattering studies in whichthe dynamics and interactions of atoms andmolecules are elucidated through the observationof their characteristic energy states and diffusivemotions. This type of measurement has alreadyproven highly important in studies of dynamic pro-cesses in a wide variety of materials and chemicalsystems; for example, as a probe of catalysis mecha-nisms, hydrogen interactions with metals, phasetransformations, magnetic materials, and in studiesof intermolecular potentials and rotational motionsin molecules and macromolecules. The two spec-trometers being developed at the CNRF will beequal to or better than similar instruments any-where in the world.

INDUSTRY SUPPORTS RESEARCH ON THEDEVELOPMENT OF OPTICAL WAVEGUIDESENSORSA consortium of biotechnology companies hassigned a 3-year cooperative research agreementwith NIST to support research in the Organic Ana-lytical Research Division of the Center for Analyt-ical Chemistry. The investigation will entail thedesign, fabrication, and evaluation of planar andfiber optic waveguide phase-sensitive analyticalsensors for real-time monitoring. The device willconsist of a Mach Zehnder interferometer ontowhich analyte-specific antibodies have been immo-bilized. Analyte quantitation will be achieved bymeasurement of the enthalpy (heat) of the antigen-antibody binding reaction. The heat generated atthe sensing arm of the interferometer will perturbthe phase of the guided laser light with respect tothe reference arm. Interferences common to botharms will be rejected.

The analyte chosen to evaluate the efficacy ofthe interferometric immunoassay system is BTtoxin, a natural insecticide (entomopathogen) pro-duced biotechnologically from a spore-formingbacteria. This mammalian-harmless toxin is usedagainst larval gypsy moths, as well as mosquitoesand blackflies. To maximize the production oftoxin, the sporulation cycle must be terminated be-fore sporulation occurs and this requires a real-time, BT toxin-sensitive sensor. By the appropriatechoice of antibody, this approach should be appli-cable to a broad range of analytes.

391



FEDERAL INFORMATION PROCESSINGSTANDARD (FIPS) PROPOSED FOR THEUSER INTERFACE COMPONENT OF THEAPPLICATIONS PORTABILITYPROFILE (APP)A Federal Register notice announced and re-quested comments on a proposed FIPS whichwould adopt the X Protocol, Xlib Interface, XTIntrinsics, and Bitmap Distribution Format specifi-cations of the X Window System, Version 11, Re-lease 3 (X Window System is a trademark of MIT).The standard covers the data stream encoding,data stream interface, and subroutine foundationlayers of the reference model for network-basedbit-mapped graphic user interface standards, whichis detailed in the appendix to the proposed stan-dard.

The proposed standard is part of a series ofspecifications required for applications portability,the ability to move or port an application from oneoperating system environment to another. NCSLhas proposed an application portability profile thatprovides an architectural approach to applicationsportability and consists of a group of standard ele-ments including database management, data inter-change, network services, user interfaces, andprogramming languages.

OPEN SYSTEMS INTERCONNECTION/INTEGRATED SERVICES DIGITALNETWORK (OSI/ISDN) TRIAL HELDTo demonstrate the use of ISDN as a lower layertechnology for OSI applications, NCSL organizedthe OSI/ISDN trial as a necessary first step forincluding ISDN as a lower layer technology in theGovernment Open Systems Interconnection Pro-file (GOSIP) Version 2. Approved as Federal In-formation Processing Standard (FIPS) 146 in 1988,GOSIP specifies a set of OSI protocols for com-puter networking for use by government agenciesin the procurement of products and services.GOSIP Version 1 supports the message handlingsystems and file transfer, access, and managementapplications; GOSIP Version 2 proposes additionalfunctionality including Virtual Terminal Service,Office Document Architecture, and ISDN.

The OSI/ISDN trial was held June 20-21, 1989,at Mather Air Force Base, CA. Trial objectiveswere to support the inclusion of ISDN transportcapabilities in the GOSIP Version 2, to demon-strate the operation of OSI applications usingISDN transport capabilities, and to demonstrate

the interworking of ISDN and existing transporttechnologies. Among the ISDN functions testedwere CLNP over ISDN X.25, X.25 D-Channel,and X.25 Preallocated B-Channel.

The trial successfully demonstrated two ISDNcapabilities: ISDN as a transit subnetwork along anend-to-end communication path where neithersource nor the destination are directly attached tothe ISDN itself and the transit properties of anISDN as an OSI subnetwork; and the interconnec-tivity capability of an ISDN subnetwork withother types of subnetworks and end-to-end com-munication across multiple interconnected subnet-works where one of the subnetworks is an ISDN.

NIST BREAK-JUNCTION MEASUREMENTPROVIDES FIRST DETERMINATION OF SISTUNNELING GAP OF A THALLIUM-BASEDHIGH-TEMPERATURE SUPERCONDUCTORA NIST scientist has carried out the first mea-surements of the superconductor-insulator-super-conductor (SIS) tunneling energy gap in thallium-based high-temperature superconductors. Usingthe break-junction method he pioneered, the scien-tist measured crystals grown and characterized atSandia National Laboratories. Energy gap is a fun-damental characteristic of a superconductor and inthis case is important to the ultimate understandingof the superconducting behavior in thallium-basedand other high-transition-temperature materials.The measured gap value of 30 meV is consistentwith the Bardeen-Cooper-Schrieffer theory of su-perconductivity, applied to the configuration of thebreak junction. The scientist was able to detect asupercurrent when the break junction was oper-ated in a point-contact mode at temperatures ashigh as 95 K.

NIST TO PARTICIPATE IN CONSORTIUMORGANIZED TO STUDY S2F1 oNIST is a participant in a new consortium forstudying the formation and detection of disulfurde-cafluoride (S2F10 ) in electrical power systems. Acooperative investigation led by NIST staff re-cently confirmed for the first time the presence ofthis extremely toxic gas (threshold limit value of 10parts per billion) in sulfur hexafluoride (SF6) ex-posed to electrical discharge. Because SF6 is usedwidely as an insulating gas in high-voltage electricpower transmission systems, serious concern existsabout the hazards associated with exposure to de-composed SF6 during maintenance and repair ofelectric power equipment such as circuit breakersand gas-insulated transmission lines.

392



NIST MINIATURE BROADBAND ELECTRIC-FIELD PROBE DESIGN COMMERCIALIZEDTwo U.S. organizations are offering commercialversions of the 8-mm electric field probe developedby NIST staff that have an isotropic response ofbetter than ±0.3 dB from 100 kHz to 18 GHz. TheNIST probe can measure fields between 1 and 1600V/m from 1 MHz to 15 GHz, flat to ±2 dB.Probes are available from the Electro-MechanicsCompany of Austin, TX and Denver ResearchInstitute (DRI), affiliated with Denver University;the Italian aerospace manufacturer Aeritalia hasalso indicated its intention to market the probe.The probe's performance is achieved through theuse of thin-film dipole elements having a preciselytailored resistive profile.

Calibration Services

NEW AUTOMATED GAGE BLOCKCALIBRATION SYSTEM AT NISTA new calibration system has been successfully im-plemented by the NIST gage block laboratory.The new system allows NIST laboratory personnelto implement routinely a measurement assuranceprogram for all NIST reference standard mastergage blocks. With the new measurement design,typical estimates for the uncertainty of the calibra-tions by mechanical intercomparison have alreadybeen substantially reduced from between 2.0 and4.0 gin (0.05 to 0.10 gim) to between 1.0 and 2.0 gin(0.02 to 0.05 gim) or less for gage blocks under 2 in(50 mm). Although uncertainty estimates for thelonger gage block sizes have remained unchanged,improved and frequent surveillance of NIST gageblock calibration history are expected to reducepresently reported uncertainty estimates for thesesizes in the future. The new system includes a data-base and analysis software to record and monitorthe measurement history of customers' gage blocksand, thereby, reduce risks of reporting inaccuratemeasurements and improving overall measurementconfidence.

393



Calendar

January 16-18, 1990HYPERTEXT STANDARDIZATION

WORKSHOP

Location: National Institute ofStandards and TechnologyGaithersburg, MD

This workshop will disseminate technical informa-tion on Hypertext to users and designers in indus-try, government, and academia. Workshop goalsare to consider Hypertext system definitions, toidentify viable approaches for pursuing standards,to seek commonality among alternatives whereverpossible, and to make progress towards a coordi-nated plan for standards development, i.e., a Hy-pertext reference model. Hypertext technology isan approach to information management in whichdata are stored in a network of nodes connected bylinks.

Contact: Jean Baronas, B266 TechnologyBuilding, NIST, Gaithersburg, MD 20899, 301/975-3338.

June 11-13, 199050th ANNUAL CONFERENCE ON

PHYSICAL ELECTRONICS


The conference will focus on the physics andchemistry of solid surfaces and interfaces. Topicsof interest include electronic, chemical and crystal-lographic properties of surfaces and interfaces aswell as kinetic and dynamic mechanisms of physi-cal and chemical reactions, phase transitions, andadsorption. The properties of surfaces that areclean or that incorporate foreign atoms are of inter-est, as are new methods of surface analysis and

characterization. Emphasis will be placed on thedescription of surface and interface properties at afundamental atomic and molecular level. Spon-sored by the American Physical Society, NIST,and the University of Maryland.

Contact: Richard R. Cavanagh, B248 ChemistryBuilding, NIST, Gaithersburg, MD 20899, 301/975-2368.

July 23-26, 1990CONFERENCE ON ADVANCES IN

CEMENTITIOUS MATERIALS


The purpose of the conference is to advance under-standing of cementitious materials so as to providea rational approach to the design of new cementi-tious materials and composites. The conferencewill address the fundamental science underlyingthe behavior of all cementitious materials, whethermass concrete or chemically bonded ceramics.However, to limit the scope of the conference,degradation processes will not be included. Theconference should be of interest to all who are con-cerned about enhancing the performance of con-crete and making it a more predictable material.Holding the conference at the National Institute ofStandards and Technology will make it possible forparticipants to visit the laboratories of the NISTCenter for Building Technology. Session topics in-clude characterization of particulate starting mate-rials and of resultant microstructure; physical andchemical phenomena in the flow of cementitiousmaterials prior to hardening; relationships betweenmicrostructure and physical and mechanical behav-ior; interactions of water with the hardened cementpaste matrix; the microstructure and micromechan-ics of cementitious composites, including the

395



effects of interfaces; and new cements and novelcement matrices.

Contact: Geoffrey Frohnsdorff, B368 BuildingResearch Building, NIST, Gaithersburg, MD20899, 301/975-6706.

August 16-21, 1990INTERNATIONAL CONFERENCE ON THE

CHEMISTRY OF ELECTRONICCERAMIC MATERIALS

Location: Grand TetonNational Park, WY

Materials chemistry has evolved from the moreidentifiable disciplines of inorganic and solid statechemistry, ceramics, and materials science. It is in-herently interdisciplinary, attracting scientists frommany different backgrounds, such as materials sci-ence, physical, inorganic, and organo-metallicchemistry, physics, ceramics, and mineralogy. Thisconference will bring together major national andinternational researchers to bridge the gap betweenthose primarily interested in the pure chemistry ofinorganic solids and those interested in the physicaland electronic properties of ceramics. One of theprimary goals of the conference will be to evaluateour current understanding of the chemistry of elec-tronic ceramic materials and to assess the state of afield that has be come one of the most importantareas of advanced materials research. Sponsored byNIST, the Office of Naval Research, and otherscientific organizations-federal, university, andindustry.

Contact: Robert Roth, B126 Materials Building,NIST, Gaithersburg, MD 20899, 301/975-6116.

September 18-20, 1990ELECTRONIC PUBLISHING '90


Electronic Publishing '90 (EP90) will be the thirdin a series of international conferences establishedto bring together researchers in all areas of elec-tronic publishing systems. EP86, held in Notting-ham, England, was sponsored by the BritishComputer Society. EP88, held in Nice, France,was sponsored by INRIA. The proceedings of pa-pers presented at the earlier sessions were pub-lished by Cambridge University Press and havereceived wide dissemination. The proceedings ofEP90 will also be published in book form and willbe available at the conference. EP90 will adopt abroad definition of "electronic publishing." Elec-tronic publishing will be taken to encompass all as-pects of computer-assisted preparation, presen-tation, transmittal, storage, and retrieval of docu-ments. The scope of the conference also includesthe design of the related computer systems, the de-sign of their components, and the theory that un-derlies such systems. Both linear and non-lineardocuments are appropriate subjects for discussion.Sponsored by NIST.

Contact: Lawrence A. Welsch, B252 Technol-ogy Building, NIST, Gaithersburg, MD 20899,301/975-3345.

396



Subject Index to Volume 94Numbers in parenthesis after the page number are the issue numbers

No.No.No.No.No.No.

1 January-February

2 March-April3 May-June4 July-August5 September-October6 November-December

A

absolute ratio..................analytical chemistry............assay .........................atomic weight.................automation ....................

347(6),

347(6),. . .. ...

357(6)25(1)

347(6)357(6)311(5)

B

becquerel ............................Bergman kernel. ......................biological material ....................biotechnology ........................botanical material. ....................

363(6)117(2)215(4)59(1)

215(4)

C

conformal mapping ...................consensus values......................Consultative Committee on Electricity . .Consultative Committee on Standards

for the Measurement of IonizingRadiations .........................

convergence proof......copper oxides ..........cotinine................cotinine perchlorate.....coupling ...............cryogenic engineering. ..crystal structure ........crystallographic databaseCRYSTDAT...........cubic splines ...........

CAD .....calibrationcapacitors.CBRIS ...CCE .....109Cd .....ceramics ...................

. .. .. .. .. .

. .. .. .. .. .

. . . .. .. .. .

. .. .. .. .. .

. . .. . .. .. .

. .. .. .. .. .

.... 37(1),characteristic branching ratios

of ionic substructures ................

chemical thermodynamics data .........chemical thermodynamics.............CIPM ...............................

coaxial air line........................collisionally-activated dissociation.......compact domains .....................comparative modeling.................comparisons of activity measurements ...components of variance (within-and

between-groups) ....................computer systems.....................computer communication networks .....computerized database.................computers .

281(5)179(3)179(3)281(5)95(2)

363(6)147(3)

281(5)21(1)21(1)95(2)

117(2)281(5)65(1)79(1)

363(6)

197(3)59(1)59(1)37(1)

................... ...25(1), 311(5)

D

data banks ....................data files......................database ......................

database management systems ...delay .........................diffusion coefficients ...........digital systems. ................dimethylglyoxime ..............dissipation factor...............

E

electric power........................electrical standards...................electron diffraction...................electronic structure ...................electrostatics .........................energy levels.........................enthalpy .............................error sources.........................evaluated data........................evaluation ....... .

397

117(2)197(3)95(2)

363(6)197(3)147(3)305(5)305(5)117(2)147(3)147(3)

. 49(1)1), 49(1)

117(2)

.............. 25(1),. . .. .... . .. .... . .. .... . .. .... . .. .... . .. ...

85(1)21(1)

281(5)59(1)

311(5)105(2)311(5)347(6)179(3)

179(3)179(3)15(1)

147(3)79(1)

221(4)21(1)

117(2)21(1)25(1)

...........

...........

...........

...........

...........

...........

...........

... ........

..........

.... I ......

...

...

...

...

...

...

...

...

9(

...

..........................

.............

.............

.............

.............



medical informatics .modeling .

259(4) molecular biology ...311(5) molecular graphics..

MS/MS .

... 59(1)

... 65(1)59(1), 85(1)... 85(1)... 281(5)

GN

GC-MS..............................

H

heat capacity .......................a helix .............................high Tc superconductors .............hydrophobicity .......................

305(5)

21(1)65(1)49(1)79(1)

National Library of Medicine (U.S.).....neutron scattering....................nickel ..................... 347(6),NIST services.......................NIST-EPA International Round Robin..nuclear magnetic resonance ............numeric databases ........ ... 9(1), 15(l1

I 0

i25.i

identification .............industrial use of numeric databases......infrared spectrum ...................International Committee of Weights and

Measures ..........................International Reference System for

radionuclides .......................International System of Units...........iodine ...............................

ion-molecule kinetics.................isotopic abundance ............. 347(6),

J

JANAF Thermochemical Tables .......Josephson effect.....................Josephson frequency-to-voltage quotient.

L

laser cooling .........lattice matching ......

363(6)15(1)9(1)

25(1)

ohm.................on-line searching......optical molasses ......overview ............

95(2)

363(6)95(2)

215(4)281(5)357(6)

95(2)

9(1)373(6)147(3)

P

parameters ...........passive smoking ......periodic surface profileperiodic waveform....perovskites..........phase characterization.phase transitions.piston gage.

21(1) platinum.95(2) prediction of ferroelectricity 95(2) pressures.. ..

pressure measurement .pressure metrology.pressure standards .

373(6) principal mode.9(1) propagation delay.

protein structures .......... 6!

........ 221(4)........ 305(5)........ 367(6)........ 367(6)........ .147(3)

...... ... 15(1)21(1)

........ 343(6)

........ .113(2)9(1)

113(2), 343(6)........ 343(6)........ 343(6)........ 343(6)........ .117(2)........ 311(5)5(1), 79(1), 85(1)

M

manometer ...........................mass spectrometry ...... 215(4), 347(6mass spectrum ......................material properties ..................materials science ............ 9(1measurement contour.................measurement precision ...............measurement protocol................measurement science ..................

343(6)), 357(6)

25(1)37(1)

), 147(3)117(2)117(2)

281(5)147(3)

Q

quantized Hall resistance..............quantum Hall effect..................

R

radioactivity standards ...............random error........................resistance ............................

95(2)95(2)

363(6)367(6)113(2)

398

F

fluids ........frequency ....

................ .....113(2),

59(1)259(4)357(6)179(3)281(5)25(1)

), 21(1)

. . . . . . . . .. . . . . . . . .. . . . . . . . .. . . . . . . . .. . . . . . . . .



rheology..................root-mean-square value.....

S

SANS..........................

scattering parameters. ............75se ... .. .. .. .. .. .. .. .. .. .. .. .. .sequence alignment ..............sequence profiles. ................shear ...........................SI..............................

side stream smoke................skin effect.......................spectra .........................

spectral library ..................SRM 986.......................Standard Reference Materials ...standardization .................standards .......................Structural Ceramics Database .....structure prediction ..............superconductivity ................supercritical fluid chromatography.surface roughness..............synchronization, time.............

W259(4)

367(6)

wavelengths........................weighted least squares regression .......weighted average.....................

...... 259(4)

...... .117(2)

...... 363(6)

...... .79(1)

...... ..65(1)

...... 259(4)

...... .95(2)

...... 305(5)

..... .117(2)

...... 221(4)

...... 281(5)

...... 357(6)

215(4), 357(6)

...... 281(5)

e ... 305(5)

....... .37(1)

65(l), 85(l)

9(1), 147(3)

...... 105(2)

117(2), 367(6)

...... 311(5)

221(4)

197(3)197(3)

x

XQQ instruments (QQQ, BEQQ, etc.) ... 281(5)

T

tandem mass spectrometers ............. 281(5)Taylor series ................. ........ 197(3)telegraphist equations .................. 117(2)telephone ............................ 311(5)templates ............................ 65(1)terrestrial samples ..................... 357(6)thermal conductivity ......... ......... 113(2)tobacco .............................. 305(5)trace analysis ................. ........ 215(4)transient hot-wire ..................... 113(2)tungsten ............................. 221(4),3 turn ............................... 65(1)

U

uncertainty.....user-friendly ...

367(6)

37(l)

V

vapor cell...........vapor pressure.......volt ................voltage transformers .

373(6)21(1)95(2)

179(3)

399

..........................

.............

.............



Author Index to Volume 94Numbers in parenthesis after the volume number are the issue numbers

No. 1

No. 2No. 3

No. 4No. 5No. 6

A

January-FebruaryMarch-AprilMay-JuneJuly-AugustSeptember-OctoberNovember-December

Bruno, T. J.

Allan, D. W.

The NIST Automated Computer Time Ser-vice. Levine, J., Weiss, M., Davis, D. D., Allan,D. W., and Sullivan, D. B. 94(5), 311 (1989).

Anderson, W. E.

Calibration of Voltage Transformers and High-Voltage Capacitors at NIST. Anderson, W. E.94(3), 179 (1989).

B

Barnes, I. L.

Absolute Isotopic Abundance Ratios andAtomic Weight of a Reference Sample ofNickel. Gramlich, J. W., Machlan, L. A. Barnes,I. L., and Paulsen, P. J. 94(6), 347 (1989).

The Absolute Isotopic Composition and AtomicWeight of Terrestrial Nickel. Gramlich, J. W.,Beary, E. S. Machlan, L. A., and Barnes,I. L. 94(6), 357 (1989).

Bean, V. E.

The Reduction of Uncertainties for Absolute Pis-ton Gage Pressure Measurements in the Atmo-spheric Pressure Range. Welch, B. E., Edsinger,R. E., Bean, V. E., and Ehrlich, C. D. 94(6), 343(1989).

Beary, E. S.


Bennett, L. H.

A Brief Review of Recent SuperconductivityResearch at NIST. Lundy, D. R., Swartzendruber,L. J., and Bennett, L. H. 94(3), 147 (1989).

A Supercritical Fluid Chromatograph forPhysicochemical Studies. Bruno, T. J. 94(3), 105(1989).

Byrd, G. D.

A Cotinine in Freeze-Dried Urine ReferenceMaterial. Sander, L. C., and Byrd, G. D. 94(5),305 (1989).

CCabeza, M. I.

The Spectrum of Doubly Ionized Tungsten(W iii). Iglesias, L., Cabeza, M. I., Rico, F. R.,Garcia-Riquelme, O., and Kaufman, V. 94(4), 221(1989).

Carr, M. J.

NIST/Sandia/ICDD Electron Diffraction Data-base: A Database for Phase Identification by Elec-tron Diffraction. Carr, M. J., Chambers, W. F.,Melgaard, D., Himes, V. L., Stalick, J. K., andMighell, A. D. 94(1), 15 (1989).

Chambers, W. F.


Chase, M. W.

Numeric Databases in Chemical Thermodynam-ics at the National Institute of Standards and Tech-nology. Chase, M. W. 94(1), 21 (1989).

D

Davis, D. D.


401



E

Edsinger, R. E.

The Reduction of Uncertainties for AbsolutePiston Gage Pressure Measurements in the Atmo-spheric Pressure Range. Welch, B. E., Edsinger,R. E., Bean, V. E., and Ehrlich, C. D. 94(6), 343(1989).

Ehrlich, C. D.


G

Garcia-Riquelme, 0.

The Spectrum of Doubly Ionized Tungsten(W IIi). Iglesias, L., Cabeza, M. I., Rico, F. R.,Garcia-Riquelme, O., and Kaufman, V. 94(4), 221(1989).

Gramlich, J. W.

Determination of Trace Level Iodine in Biologi-cal and Botanical Reference Materials by IsotopeDilution Mass Spectrometry. Gramlich, J. W., andMurphy, T. J. 94(4), 215 (1989).


The Absolute Isotopic Composition andWeight of Terrestrial Nickel. Gramlich,Beary, E. S. Machlan, L. A., andI. L. 94(4), 357 (1989).

AtomicJ. W.,

Barnes,

HHimes, V. L.


Holt, D. R.

Scattering Parameters Representing Imperfec-tions in Precision Coaxial Air Lines. Holt,D. R. 94(2), 117 (1989).

Erratum: Scattering Parameters RepresentingImperfections in Precision Coaxial AirLines. Holt, D. R. 94(5), 323 (1989).

Hoppes, D. D.

Special Report on Standards for Radioactivity-Report on the 1989 Meeting of the RadionuclideMeasurements Section of the Consultative Com-mittee on Standards for the Measurement of Ioniz-ing Radiations. Hoppes, D. D. 94(6), 363 (1989).

Hubbard, C. R.

The Structural Ceramics Database: TechnicalFoundations. Munro, R. G., Hwang, F. Y., andHubbard, C. R. 94(1), 37 (1989).

Hwang, F. Y.


IIglesias, L.

The Spectrum of Doubly Ionized Tungsten(W iii). Iglesias, L., Cabeza, M. I., Rico, F. R.,Garcia-Riquelme, O., and Kaufman, V. 94(4), 221(1989).

K

Kaufman, V.

The Spectrum of Doubly Ionized Tungsten(W III). Iglesias, L., Cabeza, M. I., Rico, F. R.,Garcia-Riquelme, O., and Kaufman, V. 94(4), 221(1989).

LLesk, A. M.

The Computational Analysis of Protein Struc-tures: Sources, Methods, Systems, and Re-sults. Lesk, A. M., and Tramontano, A. 94(1), 85(1989).

Levine, J.


Lias, S. G.

Numeric Databases for Chemical Analysis.Lias, S. G. 94(1), 25 (1989).

402



Lundy, D. R.


MMachlan, L. A.



Mandel, J.

Consensus Values, Regressions, and WeightingFactors. Paule, R. C., and Mandel, J. 94(3), 197(1989).

Martinez, R. I.

Instrument-Independent MS/MS Database forXQQ Instruments: A Kinetics-Based MeasurementProtocol. Martinez, R. I. 94(5), 281 (1989).

Matula, R. A.

The Importance of Numeric Databases to Mate-rials Science. Matula, R. A. 94(1), 9 (1989).

Masys, D. R.

New Directions in Bioinformatics. Masys,D. R. 94(1), 59 (1989).

Melgaard, D.


Migdall, A. L.

A Search for Optical Molasses in a Vapor Cell:General Analysis and Experimental At-tempt. Migdall, A. L. 94(6), 373 (1989).

Mighell, A. D.


Moult, J.

Comparative Modeling of Protein Structure-Progress and Prospects. Moult, J. 94(1), 79 (1989).

Munro, R. G.


Murphy, T. J.

Determination of Trace Level Iodine in Biologi-cal and Botanical Reference Materials by IsotopeDilution Mass Spectrometry. Gramlich, J. W., andMurphy, T. J. 94(4), 215 (1989).

PPaule, R. C.

Consensus Values, Regressions, and WeightingFactors. Paule, R. C., and Mandel, J. 94(3), 197(1989).

Paulsen, P. J.


Perkins, R. A.

Relation Between Wire Resistance and FluidPressure in the Transient Hot-Wire Method.Roder, H. M., and Perkins, R. A. 94(2), 113(1989).

RReid, L. S.

The Use of Structural Templates in ProteinBackbone Modeling. Reid, L. S. 94(1), 65 (1989).

403



Rico, F. R.

The Spectrum of Doubly Ionized Tungsten(W III). Iglesias, L., Cabeza, M. I., Rico, F. R.,Garcia-Riquelme, O., and Kaufman, V. 94(4), 221(1989).

Roder, H. M.

Relation Between Wirp Resistance and FluidPressure in the Transient Hot-Wire Method.Roder, H. M., and Perkins, R. A. 94(2), 113(1989).

SSander, L. C.

A Cotinine inMaterial. Sander,305 (1989).

Siegrist, T.

Applications ofAnalysis Systemence. Siegrist, T.

Freeze-Dried Urine ReferenceL. C., and Byrd, G. D. 94(5),

the Crystallographic Search andCRYSTDAT in Materials Sci-94(1), 49 (1989).

Stalick, J. K.


Straty, G. C.

Apparatus for Neutron Scattering Measurementson Sheared Fluids, Straty, G. C. 94(4), 259 (1989).

Sullivan, D. B.

The NIST Automated Computer Time Ser-vice. Levine, J.,Weiss, M., Davis, D. D., Allan,D. W., and Sullivan, D. B. 94(5), 311 (1989).

Swartzendruber, L. J.


TTaylor, B. N.

Special Report on Electrical Standards-New In-ternationally Adopted Reference Standards ofVoltage and Resistance. Taylor, B. N. 94(2), 95(1989).

Teague, E. C.

On Measuring the Root-Mean-Square Value of aFinite Record Length Periodic Waveform.Teague, E. C. 94(6), 367 (1989).

Tramontano, A.

The Computational Analysis of Protein Struc-tures: Sources, Methods, Systems, and Re-sults. Lesk, A. M., and Tramontano, A. 94(1), 85(1989).

WWeiss, M.


Welch, B. E.


404

Co ntents · 2003-10-02 · Co ntents Articles The Reduction of Uncertainties for Absolute B. E....

Documents

Transcript of Co ntents · 2003-10-02 · Co ntents Articles The Reduction of Uncertainties for Absolute B. E....