NIM B

Beam Interactions with Materials & Atoms

Nuclear Instruments and Methods in Physics Research B 259 (2007) 282287 www.elsevier.com/locate/nimb

Use of natural diamonds to monitor 14C AMS instrument backgrounds

a,b,* ,

John Southon

a Department of Anthropology, University of California, Riverside, CA 92354, USA Cotsen Institute of Archaeology, University of California, Los Angeles, CA 90024, USA Keck Accelerator Mass Spectrometry Laboratory, Department of Earth System Science, University of California, Irvine, CA 92697, USA

Available online 12 February 2007

Abstract To examine one component of the instrument-based background in the University of California Keck Carbon Cycle AMS spectrometer, we have obtained measurements on a set of natural diamonds pressed into sample holders. Natural diamond samples (N = 14) from dierent sources within rock formations with geological ages greatly in excess of 100 Ma yielded a range of currents (110250 lA 12C where lamentous graphite typically yields 150 lA 12C) and apparent 14C ages (64.9 0.4 ka BP [0.00031 0.00002 fm] to 80.0 1.1 ka BP [0.00005 0.00001 fm]). Six fragments cut from a single diamond exhibited essentially identical 14C values 69.3 0.5 ka70.6 0.5 ka BP. The oldest 14C age equivalents were measured on natural diamonds which exhibited the highest current yields. 2007 Elsevier B.V. All rights reserved.

1. Introduction The initial anticipation [1,2] that AMS-based systems might achieve 14C-inferred age measurements of 105 years (0.000004 fraction modern [fm]) on unknown age samples has been, to date, unrealized due to a variety of sample processing and instrument-based experimental constraints. The typical situation is well illustrated by the early work of the University of Washington AMS group. An AMS-based measurement of 69.0 1.7 ka BP was obtained on a specially prepared sample of geological graphite. However, graphite prepared from CO2 obtained from a sample of marble, which, like geologic graphite, should exhibit no 14 C activity due to its great geologic age, yielded an apparent age of 47.9 0.7 ka BP [3]. A study carried out jointly the University of California/Lawrence Livermore National Laboratory Center for Accelerator Mass Spectrometry and the University of California, Riverside Radiocarbon Laboratory obtained an average apparent age of 64.5 0.3 ka BP on samples of geologic graphite and an average
Corresponding author. Address: Department of Anthropology, University of California, Riverside, CA 92354, USA. E-mail address: [email protected] (R.E. Taylor). 0168-583X/$ - see front matter 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2007.01.239
*

(N = 19) apparent age of 52.1 0.4 ka BP on duplicate 1 mg samples of catalytically-condensed graphitic carbon prepared from carefully-pretreated wood of reportedly Pliocene age. The lowest Pliocene wood blank value achieved by the LLNL/UCR laboratory collaboration was 60.5 0.6 ka BP [46]. There are a number potential sources of a 14C background signal in an AMS-based system using catalytically-condensed graphitic carbon produced from CO2 derived from the combustion or acidication of carboniferous materials [7]. One of these categories is instrument or machine background which involves the registration of what is interpreted by the detector circuitry and/or software as a 14C-ion produced pulse when, in fact, a non-14C-ion mimics 14C or the detector counts 14C which was not originally present in the sample matrix when it was introduced into the source. To examine and monitor the level of machine background in the University of California, Irvine Keck Carbon Cycle AMS spectrometer [8], we have obtained a series of measurements on a set of natural diamonds. Because of their great geologic age, we view it as a reasonable assumption that these gem-carbon samples contain no measurable 14 C and that their unique physical characteristics signi-

R.E. Taylor, J. Southon / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 282287

283

cantly reduce or eliminate exogenous contamination from more recent carbon sources. On this basis, we propose that we have eliminated the major sources of mass 14 ion with the exception of that contributed from various components of instrument or machine background signal and perhaps that contributed from the sample holder itself. Other backgrounds, such as 13CH ! 13C+* ! 13C+, from charge exchange and scattering, can also contribute to producing 14C peaks if no dE/dx measurements are performed. However, since UCI backgrounds for processed (graphitized) samples are comparable with those on larger machines where 13C can be separated using dE/dx, 14C from the actual sample is probably the dominant component of the routine background. 2. Sources of AMS background Almost three decades of experience with AMS technology has yielded an appreciation of the diculty of achieving the initial optimistic projections of the potential of achieving meaningful 14C-inferred ages to as much as 105 years. A review of the literature generated by AMS laboratories dealing with background issues and our own experience reveal a large number of factors that have been identied or suspected as sources of 14C from blanks or pseudo-14C-signals in the various instrumental systems developed for AMS 14C work [917]. Table 1 summarizes seven major potential sources of pseudo-14C signal in AMS-based 14C measurements using catalytically-condensed graphitic carbon produced from CO2 from combustion/acidication of carboniferous material. These factors operate singly or in various combinations to yield the apparent nite ages exhibited in 14Cfree or presumed 14C-free sample materials used as backgrounds or blanks in AMS 14C systems. To monitor one component of our instrument background, ion source memory, which we view as the largest component of the machine background in the UCI AMS system, we have taken advantage of several unique physical properties of naturally occurring diamonds. The most important of these properties is the reported ability of diamonds to exclude almost all molecular species from adhering to its surface and/or penetrating below the surface. We wished to test the hypothesis that the surface of a natural diamond would eliminate or signicantly reduce the adhesion of exogenous CO2 and other-carbon-containing molecules. 3. Natural diamonds as AMS targets Except for trace impurities such as boron, oxygen, sulfur and nitrogen, diamond, a natural crystalline mineral substance, is the transparent species of a concentrated form of pure carbon. Its surface property of extreme harness the hardest surface known with a resistance to physical abrasion more than four times that of the next most resistant natural material is the result of covalent chemical

Table 1 Potential sources of pseudo-14C signal in AMS-based

14

C measurements

(1) Pseudo C-free sample background: C is present in carboniferous material that should not contain 14C because of its geologic age. (1) Non-14C-free sample: sample material erroneously assumed to contain no detectable 14C. (2) Contaminated sample: 14C introduced into the 14Cfree sample material during chemical or physical pretreatment. (2) Combustion/acidication background: 14C introduced during production of CO2 from sample. (1) Materials contamination: 14C introduced from materials in combustion/acidication tube other than sample (e.g. adsorbed CO2 on oxidizer). (2) Tube contamination: 14C introduced from walls of combustion/acidication tube (e.g. adsorbed CO2). (3) Tube leakage: 14C introduced from leakage of atmospheric CO2 into combustion/acidication tube. (3) Graphitization background: 14C introduced during graphitization process. (1) Materials contamination: 14C introduced from materials in reaction tube (e.g. from Fe or Co powder as graphitization catalyst). (2) Tube contamination: 14C introduced from walls of reaction tube. (3) Vacuum line contamination: 14C introduced during CO2 transfer in vacuum line. (4) Transfer background: 14C introduced during graphite transfer to sample holder. (1) Target/cathode contamination: 14C introduced from sample holder. (2) Manipulation contamination: 14C introduced during physical transfer/packing of graphite into sample holder. (6) Storage background: 14C introduced at any point in the sample processing sequence from any containment vessel. (1) Particulate contamination: 14C physically introduced from carbon-containing particulates derived from storage containers. (2) Atmospheric contamination: 14C introduced from CO2 in air or from out gasing from storage container. (7) Instrument background: 14C signal registered in detector circuitry when 14C-ion not present. (1) Detector anomaly: pseudo-14C pulse registered when no ion of mass 14 is present at the detector (e.g. electronic noise). (2) Ion identication anomaly: mass 14 particle which is not 14C reaches detector and is misidentied as 14C. (3) Beam line anomaly: 14C internally derived from some component of beam line reaches detector.

bonding forming a tetrahedral structural unit. This characteristic makes diamond highly resistant to chemical corrosion. This is related to the ability of diamonds to repel and exclude water from adhering to its surface, a very unusual property for a mineral [18,19]. It was this unique physical characteristic of diamond that was the basis of our hypothesis that this surface would eliminate or signicant reduce the adhesion of carbon or carbon-containing molecules

284

R.E. Taylor, J. Southon / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 282287

from the ion source of an AMS spectrometer that would contribute to a trace memory or sample cross talk eect. 4. Materials and methods The natural diamonds employed in this study were obtained from alluvial deposits in the state of Minas Gerais in Brazil [20]. The youngest geological contexts of Brazilian alluvial diamonds are early Paleozoic, thus we have assumed an age for these materials greatly in excess of 100 my. The diamonds obtained were less than one caret in size (1 caret = 200 mg), but of gem quality. Nine diamonds were selected for measurement with one being cut into six segments. Under magnication, each diamond fragment was appropriately cut, sized and pressed into a 1.6 mm hole drilled into the center of aluminum cathode sample holder. Possible surface contamination introduced by the cutting and manipulation required to mount a diamond in a sample holder was monitored by observing the initial count rates coming from samples. All diamond samples exhibited somewhat higher initial 14 C count rates from surface contamination which then stabilized and plateaued as the surface was penetrated by cesium sputtering.

Fig. 1 is a microphotograph of one of the mounted samples. In an eort to increase the thermal conductivity of ve of the diamonds (UCIAMS-15443 to UCIAMS-15457) with respect to the metal surface of each of their cathodes, silver powder was placed in the bottom of the target hole prior to the mounting of the diamond fragment. We saw no increase in the maximum current exhibited by these samples, but they showed slightly higher 14C content than UCIAMS-9638 to 9640. Previous tests showed that 14C count rates from silver powder cathodes were comparable with those from diamonds. Because of this and because the silver packing was largely shielded from the Cs beam by the diamonds themselves, the excess 14C was therefore probably due to dierences between diamonds or run to run changes in the spectrometer, not from carbon in the silver powder. 5. Results Table 2 lists the results we have obtained on the six segments cut from the same diamond (Section A) and the measurement on eight other individual natural diamonds (Section B). In addition to the 14C content, we report also the maximum 12C currents achieved on each sample and

Fig. 1. Microphotograph of UCIAMS-15453 diamond mounted in cathode. Diameter of diamond = 1.6 mm. Microphotograph taken by Guaciara M. Santos (UCI).

Table 2 Carbon isotope and 12C current measurements on (A) six splits from of a single natural diamond and (B) individual natural diamonds (N = 8) with associated analysis of Ceylon graphite made during diamond measurements Sample no. Sample materiala Maximum current (lA12C) d13Cb (&) 0.3 5.6 4.2 19.3 3.2 0.5 4.9 11.0 14.1 15.4 20.5 23.1 19.8 20.5 3.1 4.7 11.5 10.7 9.4 3.1 0.2
14

C Content fraction modern (fm)

Conventional 69.4 0.5 69.6 0.8 70,0 0.5 70.3 0.5 70.6 0.7 70.6 0.6 66.5 0.6 68.3 1.1. 68.1 1.2c 69.4 1.0c 70.9 1.0c 71.7 1.0c 3.3 1.6c 75.7 0.8d 78.4 0.9d 80.0 1.1d 58.4 0.5 64.0 0.5 64.7 0.8 70.1 0.8e 67.0 0.7f

14

R.E. Taylor, J. Southon / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 282287

C age (ka yrs. BP)

A. Splits from single natural diamond and associated Ceylon geological graphite UCIAMS-12677 D 131 UCIAMS-12678 D 127 UCIAMS-12676 D 145 UCIAMS-12679 D 168 UCIAMS-12674 D 122 UCIAMS-12675 D 184 UCIAMS-12680 UCIAMS-12681 G G 81 70

0.00018 0.00001 0.00017 0.00002 0.00017 0.00001 0.00016 0.00001 0.00015 0.00001 0.00015 0.00001 0.00025 0.00002 0.00020 0.00003 0.00021 0.00003 0.00018 0.00002 0.00015 0.00002 0.00013 0.00002 0.00011 0.00002 0.00008 0.00001 0.00006 0.00001 0.00005 0.00001 0.00069 0.0004 0.00035 0.00004 0.00032 0.00003 0.00024 0.00002 0.00024 0.00002

B. Individual natural diamonds and associated Ceylon geological graphite UCIAMS-15445 D 136 UCIAMS-15444 D 125 UCIAMS-15443 D 125 UCIAMS-15446 D 127 UCIAMS-15447 D 127 UCIAMS-9638 D 250 UCIAMS-9640 D 240 UCIAMS-9639 D 197 UCIAMS-15440 UCIAMS-15441 UCIAMS-15442 UCIAMS-9641 UCIAMS-9642
a b c d e f

G G G G G

131 122 95 106 111

D = Diamond, G = Ceylon geological graphite. d13C Values measured using the AMS spectrometer. These values can dier, typically by 13&, from that measured on a conventional mass spectrometer. Associated Ceylon geologic graphite measurements: UCIAMS-15440, 15441 and 14440. Associated Ceylon geologic graphite measurements: UCIAMS-9641 and 9642. Graphite baked in air prior to analyses. Graphite baked in hydrogen prior to analyses.

285

286

R.E. Taylor, J. Southon / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 282287

the d13C values measured during the AMS analysis. For comparison, the 12C currents and apparent age exhibited by Ceylon geological graphite measured at the same time as the diamonds are also reported. Our natural diamond samples yielded a range of currents (120250 lA 12C where catalytically-condensed graphitic carbon in this instrument typically yields 150 lA 12C) and apparent 14C ages (64.9 0.4 ka BP [0.00031 0.00002 fm] to 80.0 1.1 ka BP [0.00005 0.00001 fm]). The six fragments cut from a single diamond exhibited essentially identical 14C values 69.4 0.6 ka to 70.6 0.5 ka BP. The oldest 14C age equivalents were measured on natural diamonds which exhibited the highest current yields. Conditions varied somewhat from run to run but, in general, the currents exhibited in the diamonds came up as rapidly as those in the Ceylon geologic graphite samples and were stable within 10% for periods of 1.5 3 h. Earlier investigators have reported wide variations in d13C values in diamonds including variations within single stones [2123]. As indicated in Table 2, we have observed similar d13C variations. Our measurements have conrmed our hypothesis that diamonds represent a much cleaner surface with respect to adhesion of carbon-containing molecules from the ion source that contribute to a trace memory or sample cross talk eect. At this time, it is not clear to us what factors might be involved in the greater variability in the apparent 14 C concentrations exhibited in individual diamonds (Section B) as opposed to splits from a single natural diamond (Section A). Possible factors suggested to us are greater variability in the orientation of the crystal facies and microfractures in individual diamonds.

require a means of synthesizing or activating in gas phase carbon-containing precursor molecules. This is currently accomplished by thermal (e.g. hot lament), electric discharge (e.g. DC, RF or microwave), or in a combustion environment (e.g. as in an oxyacetylene torch). Typically, the precursor gas (usually CH4,but the use of CO2 is currently being explored by several groups) is diluted in an excess of hydrogen and the temperature of the substrate (e.g. Si) is usually greater than 700 C to avoid the formation of amorphous carbon rather than diamond. Various experimental arrangements have been rened to optimize for various desired outcomes such as high growth rate or large area deposition and a large scientic and commercial literature now exists which describes the varying applications [25]. It is not clear whether appropriate experimental arrangements necessary to convert CO2 to a diamond lm can be designed that will permit ecient and reproducible deposition on a appropriate substrate while avoiding signicant contamination processes with respect to 14C. One of the rst future experiments that will need to be undertaken is to determine the degree of variability in 12C currents exhibited by thin diamond lms in contrast to natural diamonds. We intend to carry out these experiments in the near future. Acknowledgements The authors thank the Dean of Physical Sciences and Vice Chancellor for Research, University of California, Irvine, the National Science Foundation, and the Gabrielle O. Vierra Memorial Fund for nancial support and Guaciara M. Santos (UCI) for the photomicrograph. We also zel acknowledge the assistance of Emanuel Jacobson, O Fine Jewelry, Redlands, California who obtained the diamonds and cut and mounted them into our sample holders and the helpful comments of reviewers. References
[1] R.A. Muller, Science 196 (1977) 489. [2] R.A. Muller, Physics Today 32 (1979) 23. [3] F.H. Schmidt, D.R. Balsley, D.D. Leach, in: H.E. Gove et al. (Eds.), Accelerator Mass Spectrometry, Amsterdam, 1987. [4] D. Kirner, R.E. Taylor, J.R. Southon, Radiocarbon 37 (1995) 697. [5] D. Kirner, J.R. Southon, P.E. Hare, R.E. Taylor, in: M.V. Orna (Ed.), Archaeological Chemistry Organic, Inorganic and Biochemical Analysis, Washington, DC 1996. [6] D. Kirner, R. Burky, R.E. Taylor, J.R. Southon, Nucl. Instr. and Meth. B 123 (1997) 214. [7] R.E. Taylor, D.J. Donahue, T.H. Sable, P.E. Damon, A.T.J. Jull, in: J.B. Lambert (Ed.), Archaeological Chemistry III, Washington, DC 1984. [8] J. Southon, G. Santos, K. Druel-Rodriguez, E. Druel, S. Trumbore, X. Su, S. Grin, S. Ali, M. Mazon, Radiocarbon 46 (2004) 41. [9] R. Gillespie, R.E.M. Hedges, Nucl. Instr. and Meth. B 5 (1984) 294. [10] J.S. Vogel, D.E. Nelson, J.R. Southon, Radiocarbon 29 (1987) 323. [11] R.P. Beukens, in: R.E. Taylor et al. (Eds.), Radiocarbon After Four Decades: An Interdisciplinary Perspective, New York, 1992. [12] P.M. Gulliksen, M.S. Thomsen, Radiocarbon 34 (1992) 312. [13] R.P. Beukens, Nucl. Instr. and Meth. B 79 (1993) 620.

6. Future studies The demonstration that some natural diamonds can be used to characterize one component of machine background in an AMS system would be of limited general applicability unless there was the interest in exploiting this observation as a means of extending the 14C time frame. To take advantage of the lowered background blanks and thus the potential of an extended AMS 14C dating range, for routine AMS 14C measurements it would be necessary to demonstrate a routine capability to synthesize a diamond matrix through some intermediate chemical form from CO2 derived from the combustion or acidication of various types of carbonaceous samples. Beginning in the late 1960s, studies were initiated to accomplish this directly from a gas phase at relatively low pressures and temperatures. After several decades of experimentation, several types of chemical vapor deposition (CVD) methods were demonstrated to be practical and eective in producing diamond lms [19,24]. CVD techniques involve a gas phase chemical reaction occurring above a solid which causes a diamond lm deposition on the solid substrate. Contemporary techniques

R.E. Taylor, J. Southon / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 282287 [14] M. Schleicher, M. Grootes, M.J. Nadeau, A. Schoon, Radiocarbon 40 (1998) 85. [15] M.J. Madeau, P.M. Grootes, A. Voelker, F. Bruhn, A. Duhr, A. Oriwall, Radiocarbon 43 (2001) 169. [16] S. Freeman, P. Bishop, C. Bryant, G. Cook, D. Dugans, T. Ertunc, A. Fallick, R. Ganeshram, C. Maden, P. Naysmith, C. Schnabel, M. Scott, M. Summereld, S. Xu, Nucl. Instr. and Meth. B, these Proceedings, doi:10.1016/j.nimb.2007.01.312. [17] A. Galindo-Uribarri, J. Gomez del Campo, P. Hausladen, J. Beene, B. Fuentes, F. Liang, Y. Liu, E. Padilla, D. Stracener, A.E. Litherland, J. Doupe, Nucl. Instr. and Meth. B, these Proceedings, doi:10.1016/j.nimb.2007.01.295.

287

[18] T. Stachel, G.P. Brey, J.W. Harris, Elements 1 (2005) 73. [19] J.E. Field, The Properties of Natural and Synthetic Diamond, New York, 1992. zel Fine Jewelry, Redlands, Calif., per. comm. 2004. [20] E. Jacobson, O [21] M.B. Kirkley, J.J. Burney, M.L. Otter, S.J. Hill, L.R. Daniels, Appl. Geochem. 6 (1991) 477. [22] P. Cartigny, J.W. Harris, M. Javoy, Earth Planet. Sci. Lett. 185 (2001) 85. [23] P. Cartigny, Elements 1 (2005) 79. [24] B. Dishler, C. Wild (Eds.), Low-Pressure Synthetic Diamond, Amsterdam, 1998. [25] P. May, Phil. Trans R. Soc. Lond. A 358 (2000) 473.