12C Isotope Amount Ratios by

Jun 20, 2008 - A new method for the measurement of SI traceable carbon isotope amount ratios using a multicollector inductively coupled mass spectrome...
0 downloads 0 Views 158KB Size
Anal. Chem. 2008, 80, 5963–5969

Precise and Traceable 13C/12C Isotope Amount Ratios by Multicollector ICPMS Rebeca Santamaria-Fernandez,* David Carter, and Ruth Hearn LGC, Queens Road, Teddington, Middlesex, TW11 0LY, U.K. A new method for the measurement of SI traceable carbon isotope amount ratios using a multicollector inductively coupled mass spectrometer (MC-ICPMS) is reported for the first time. Carbon 13C/12C isotope amount ratios have been measured for four reference materials with carbon isotope amount ratios ranging from 0.010 659 (δ13CVPDB ) -46.6‰) to 0.011 601 (δ13CVPDB ) +37‰). Internal normalization by measuring boron 11B/10B isotope amount ratios has been used to correct for the effects of instrumental mass bias. Absolute 13C/12C ratios have been measured and corrected for instrumental mass bias and full uncertainty budgets have been calculated using the Kragten approach. Corrected 13C/12C ratios for NIST RM8545 (Lithium Carbonate LSVEC), NIST RM8573 (L-Glutamic Acid USGS40), NIST RM8542 (IAEA-CH6 Sucrose) and NIST RM8574 (L-Glutamic Acid USGS41) differed from reference values by 0.06-0.20%. Excellent linear correlation (R ) 0.9997) was obtained between corrected carbon isotope amount ratios and expected carbon isotope amount ratios of the four chosen NIST RMs. The method has proved to be linear within this range (from 13C/12C ) 0.010 659 to 13C/12C )0.011 601), and therefore, it is suitable for the measurement of carbon isotope amount ratios within the natural range of variation of organic carbon compounds, carbonates, elemental carbon, carbon monoxide, and carbon dioxide. In addition, a CO2 gas sample previously characterized in-house by conventional dual inlet isotope ratio mass spectrometry has been analyzed and excellent agreement has been found between the carbon isotope amount ratio value measured by MC-ICPMS and the IRMS measurements. Absolute values for carbon isotope amount ratios traceable to the SI are given for each NIST RM, and the combined uncertainty budget (including instrumental error and each parameter contributing to Russell expression for mass bias correction) has been found to be < 0.1% for the four materials. The advantage of the method versus conventional gas source isotope ratio mass spectrometry measurements is that carbon isotope amount ratios are measured as C+ instead of CO2+, and therefore, an oxygen 17O correction due to the presence of 12C17O16O+ is not required. Organic compounds in solution can be measured without previous derivatization, combustion steps, or * To whom correspondence should be addressed. E-mail: [email protected]. 10.1021/ac800621u CCC: $40.75  2008 American Chemical Society Published on Web 06/20/2008

both, thus making the process simple. The novel methodology opens new avenues for the measurement of absolute carbon isotope amount ratios in a wide range of samples. Differences in measured carbon isotope amount ratios (13C/ C), commonly expressed as δ13C (‰) are used in biology, geology, environmental sciences, food and drug authentication, and forensic studies.1 The most common technique to measure relative carbon isotopic abundance variations is gas source isotope ratio mass spectrometry (GS-IRMS), where carbon is measured as CO2. Carbon dioxide isotope amount ratios are then measured monitoring molecular ion masses 44, 45, and 46. The 13C/12C information becomes available through measurement of ion currents at m/z 44 and 45. However, in order to determine the 13 C/12C isotope amount ratio, a correction must be applied to the measured ion current ratios.2 This oxygen correction, commonly referred to as 17O correction accounts for the presence of 12 17 16 C O O isotopologues, which contribute 6-7% to CO2 isotopologues of m/z 45.3 To complicate matters, several correction algorithms to correct for 17O are available, which result in systematic differences of the derived 13C/12C isotope amount ratios. A review of the different options including a reformulated 17 O correction for carbon dioxide measurements has recently been published by Kaiser.3 The availability of different algorithms complicates the exact comparability of δ13C values between different laboratories,4 and the correction in itself contributes substantially to the full uncertainty of the method. Recently Valkiers et al.5 reported traceable carbon isotope ratios using the IRMM Avogadro II amount comparator spectrometer. A set of synthetic isotope mixtures of carbon dioxide were prepared and used as primary standards to the SI. The Avogadro II amount comparator is a Finnigan MAT 271 spectrometer further modified for the measurement of absolute carbon isotope amount ratios with SI traceability.6 In this work, an alternative method for the measurement of carbon isotope amount ratios using a commercially available multicollector inductively coupled mass spectrometer (MC12

(1) Benson, S.; Lennard, C.; Maynard, P.; Roux, C. Forensic Sci. Int. 2006, 157, 1–22. (2) Craig, H. Geochim. Cosmochim. Acta 1957, 12, 133–149. (3) Kaiser, J. Geochim. Cosmochim. Acta 2008, 72, 1312–1334. (4) Werner, R. A.; Brand, W. A. Rapid Commun. Mass Spectrom. 2001, 15, 501–519. (5) Valkiers, S.; Varlam, M.; Ruβe, K.; Berglund, M.; Taylor, P.; Wang, J.; Milton, M. J. T.; De Bievre, P. Int. J. Mass Spectrom. 2007, 264, 10–21. (6) De Bievre, P.; Lenaers, G.; Murphy, T. J.; Peiser, H. S.; Valkiers, S. Metrologia 1995, 32, 103–110.

Analytical Chemistry, Vol. 80, No. 15, August 1, 2008

5963

ICPMS) is described. ICPMS has rarely been used for measurements of carbon isotope amount ratios mainly due to the poor ionization of carbon in the ICP. Carbon’s first ionization potential is 11.26 eV; thus, typically 0.05% on ratios close to 1 and precision worsens for small or large ratios.8–12 The use of MC-ICPMS instruments for high-precision measurements of isotope amount ratios is now widespread for a large range of elements, and recent publications have demonstrated that MC-ICPMS can produce accurate isotope amount ratio measurements with typical precision between 0.05 and 0.002% relative standard deviation (RSD).13–18 To the authors’ knowledge, this is the first time the performance of a MC-ICPMS has been evaluated for carbon isotope amount ratio measurements at natural isotopic abundance. Vogl and Heumann19 quantified dissolved organic carbon by quadrupole ICP-IDMS, and Luong and Houk determined carbon isotope amount ratios in proteins and amino acids using a home-built twin quadrupole ICPMS.20 In both cases, the precision of carbon isotope amount ratios expressed as RSD was between 1 and 3%. In this work, the use of a MC-ICPMS improves the precision of those measurements (typical RSD of 0.04%) given that measurements are performed under optimum conditions and the concentration of carbon in the solution is ∼300 µg g-1 (leading to a signal intensity of ∼7 V for 12C+). The use of a multicollector ICPMS has several advantages over alternative methods in the context of this work. First, the precision of the ratio measurements is considerably superior to those made by scanning ICP. Additionally, the carbon isotope amount ratios are measured as atomic C+ ions instead of molecular ions, and therefore, a correction for oxygen is not required. Since samples (7) Houk, R. S. Anal. Chem. 1986, 58, 97A105A. (8) Kim, C. S.; Kim, C. K.; Martin, P.; Sansone, U. J. Anal. At. Spectrom. 2007, 22, 827–841. (9) Dreyfus, S.; Pecheyran, C.; Lienemann, C. P.; Magnier, C.; Prinzhofer, A.; Donard, O. F. X. J. Anal. At. Spectrom. 2007, 22, 351–360. (10) Ruiz Encinar, J.; Garcia Alonso, J. I.; Sanz-Medel, A.; Main, S.; Turner, P. J. J. Anal. At. Spectrom. 2001, 16, 315–321. (11) Begley, I. S.; Sharp, B. L. J. Anal. At. Spectrom. 1994, 9, 171–176. (12) Prohaska, T.; Latkoczy, C.; Stingeder, G. J. Anal. At. Spectrom. 1999, 14, 1501–1504. (13) Santamaria-Fernandez, R.; Evans, P.; Wolff-Briche, C. S. J.; Hearn, R. J. Anal. At. Spectrom. 2006, 21, 413–421. (14) Santamaria-Fernandez, R.; Hearn, R. Rapid Commun. Mass Spectrom. 2008, 22, 401–408. (15) Makishima, A.; Nagender, N.; Nakamura, E. J. Anal. At. Spectrom. 2007, 22, 407–410. (16) Malinovsky, D.; Stenberg, A.; Rodushkin, I.; Andren, H.; Ingri, J.; Ohlander, B.; Baxter, D. C. J. Anal. At. Spectrom. 2003, 18, 687–695. (17) Balcaen, L.; Schrijver, I. D.; Moens, L.; Vanhaecke, F. Int. J. Mass Spectrom. 2005, 242, 251–255. (18) Barbaste, M.; Robinson, K.; Guilfoyle, S.; Medina, B.; Lobinski, R. J. Anal. At. Spectrom. 2002, 17, 135–137. (19) Vogl, J.; Heumann, K. G. Anal. Chem. 1998, 70, 2038–2043. (20) Luong, E. T.; Houk, R. S. J. Am. Soc. Mass Spectrom. 2003, 14, 295–301.

5964

Analytical Chemistry, Vol. 80, No. 15, August 1, 2008

do not need combustion to CO2 gas, aqueous solutions can be introduced to the plasma without a previous derivatization or combustion step and therefore the time of sample preparation and analysis is reduced considerably. Overall, the method offers potential for a wide range of applications such as food authenticity, pharmaceutical counterfeit detection, and forensic sciences and most importantly for the measurement of SI traceable carbon isotope amount ratios. Currently carbon isotope amount ratios are reported as δ13CVPDB values (in parts per thousand units; ‰) within the Vienna Pee Dee Belemnite scale.21 The first international scale for 13C/ 12 C ratios used a carbonate laboratory standard reported by Urey et al.;22 this Cretaceous Belemnite material collected from the Pee Dee formation in South Carolina was then proposed as the international reference material (PDB) for the carbon and oxygen isotope amount ratio scales by Craig in 1957.2 Carbon isotope amount ratios were measured in carbon dioxide generated after reaction of PDB with H3PO4, and Craig2 reported 0.011 237 2 as the measured 13C/12C carbon isotope amount ratio. This material no longer exists, and therefore, another carbonate (NIST RM8544, NBS-19) has been assigned an exact δ13C value relative to PDB.23 The new adopted scale is called Vienna Pee Dee Belemnite (VPDB) in recognition of the role that the International Atomic Energy Agency (IAEA) in Vienna has played in redefining the scale. Since then it is recommended that all data for carbon isotope composition should be reported as δ13CVPDB values related to VPDB by assigning a value of +1.95‰ exactly to NIST RM8544 (NBS 19 calcium carbonate; 13C/12C ) 0.011 202)4 and -46.6‰ to NIST RM8545 (LSVEC Lithium carbonate)23 according to eq 1.

δ13CsampleVPDB (‰) ) [(13C⁄ 12C)sample - (13C⁄ 12C)VPDB] ⁄

(13C⁄12C)VPDB × 1000 (1) There are several δ13CVPDB standards covering a range from +38 to -47‰ and ranging in composition from carbonates to CO2 gaseous references. The approach therefore relies on reference samples. In this paper, the authors have measured carbon isotope amount ratios as C+ and corrected for instrumental mass bias effects monitoring boron isotope amount ratios. Then the corrected ratios can be compared to those of four NIST RMs previously analyzed for δ13CVPDB. Carbon isotope amount ratios are not given for those reference materials and therefore the ratio of VPDB (13C/12CVPDB) has to be used together with the reference δ13CSampleVPDB value so that the expected carbon isotope amount ratio for the sample (13C/12CSample) can be extracted from eq 1. The absolute 13C/ 12 CVPDB isotope amount ratio used in eq 1 was 0.011 802 ± 0.000 028, based upon previously reported values.4,24 (21) Coplen, T. B.; Brand, W. A.; Gehre, M.; Groning, M.; Meijer, H. A.; Toman, B.; Verkouteren, R. M. Anal. Chem. 2006, 78, 2439–2441. (22) Urey, H. C.; Lowenstam, H. A.; Epstein, S.; McKinney, C. R. Geol. Soc. Am. Bull. 1951, 62, 399–416. (23) Coplen, T. B. Pure Appl. Chem. 1994, 66, 273–276. (24) Kuder, T.; Wilson, J. T.; Kaiser, P.; Kolhatkar, R.; Philp, P.; Allen, J. Environ. Sci. Technol. 2005, 39, 213–220.

Table 1. Operating Conditions and Data Acquisition Parameters for the Neptune MC-ICPMS Instrument Settings rf power cool gas flow auxiliary gas flow sample gas flow sampler and skimmer cones extraction voltage

1230 W 14 L min-1 Ar 1.1 L min-1 Ar 1.4 L min-1 Ar Ni -1800 V

USN/Desolvator Settings 0.85 L min-1 160 °C -3 °C 160 °C

Ar sweep heater 1 condenser heater 2

Data Acquisition Parameters collection mode cup configuration

dynamic L3 L2 C H2 H4 low

resolution mode

main: C 12C 12.55 13C

secondary: B 10B 10.13 11B

continuous sample introduction acquisition method for solutions acquisition method for gas sample

3 blocks, 10 cycles, 4.194-s integration each, 3-s idle 1-min washout time with HNO3 1% (v/v) 10 blocks, 10 cycles, 4.194-s integration each, 3-s idle

EXPERIMENTAL SECTION Multicollector ICPMS. All measurements were performed using a MC-ICPMS (Neptune MC-ICPMS, ThermoFinnigan, Bremen, Germany). The Neptune has an argon inductively coupled plasma ion source, forward Nier-Johnson geometry analyzer, and nine Faraday detectors. The axial Faraday can be interchanged with a secondary electron multiplier enabling detection below the normal operating range of the Faraday cups. Further details of the instrument design may be found elsewhere.25 Instrumental operating conditions and data acquisition parameters including the cup configuration used for measurements are given in Table 1. The mass difference between 10B and 13C is ∼30%, beyond the normal Faraday detector array of the Neptune MC-ICPMS instrument used in this study. Thus, each measurement comprised two parts: 12C and 13C (4.194-s dwell time) were measured simultaneously on the “low 3” and “high 2” Faraday cup detectors, respectively. Subsequently, 10B and 11B were measured (4.194-s dwell time) on the “low 2” and “high 4” Faraday cup detectors. A 3-s delay was used between carbon and boron measurements to allow the instrument to completely stabilize. Hence, the difference between each carbon and boron measurement cycle was minimized to 7 s. Each acquisition for liquid samples comprised 30 cycles (divided in (25) Weyer, S.; Schwieters, J. Int. J. Mass Spectrom. 2003, 226, 355–368.

3 blocks). The total acquisition time for each sample was 7 min. Measurements for the gas cylinder comprised 50 cycles (divided in 10 blocks), and the total acquisition time was 12 min. Ni sample and skimmer cones were used in this study, and deposition of solids was not observed. Gas samples were introduced to the plasma directly; the gas cylinder was connected to the ICPMS via PTFE tubing and a oneflow restrictor (Kuhnke, Hants, UK) that allows regulating the flow without using a mass flow controller and therefore avoids potential isotopic fractionation. Liquid samples were introduced to the plasma after ultrasonic nebulization and desolvation (USN CETAC U-6000AT+, CETAC Technologies, Omaha, NE). The ultrasonic nebulizer/ membrane desolvator provides enhanced desolvation for aqueous and organic samples. The membrane desolvator effectively removes solvent vapor thus increasing plasma stability and potentially minimizing the 12C1H+ interference on 13C+. Sensitivity when not using the USN was 1 order of magnitude lower. The argon sweep flow was optimized for maximum sensitivity and optimum peak shape. Background levels were determined by measuring the corresponding signal intensity (V) of pure water (Milli-Q) in which samples were dissolved. Blank measurements lead to a signal intensity of < 0.01 V for 12C+ (