δ18O and δ13C Isotope Analysis of Carbonate Rock
Dry and finely ground carbonate mineral samples and standards are weighed into clean exetainers, which are sealed with rubber septa. The exetainers are loaded in the gas bench at 72°C and all subsequent operations are carried out using the PAL A200S autosampler. Air is removed by replacement with ultrapure helium using the autosampler, then the sample is acidified with 100% phosphoric acid. After an hour the CO2 produced by the reaction is sampled and sent to the mass spectrometer in helium, via a water trap and GC column in the gas bench. Analysis is in continuous flow mode, with CO2 gas of known isotopic composition used as the reference gas. Fractionation is calculated by multiple analyses of internal standards that have been calibrated against international standards NBS 18 and 19.
Carbonate samples should preferably be submitted as powders in a small glass vial (plastic causes static problems). If powdering is not possible, small chips can be sent. Approximately 200 micrograms is analysed of pure carbonates; the weight goes up for samples with low percentage of carbonate content. The submitted samples should be accompanied by a list that includes the percentage of carbonate in the rock, and the (approximate) proportions of carbonate minerals in the case of a mixture (calcite, dolomite, magnesite, hydromagnesite, siderite, ankerite, etc.).
Please ship the samples to Janet Gabites, and also email the list to firstname.lastname@example.org.
Dissolved Inorganic Carbon
DIC is analysed on the gas bench by adding 600 microlitres of the sample water to a clean exetainer sealed with a rubber septum. The exetainer is flushed with helium to replace air in the headspace and to sparge CO2 dissolved in the water. Phosphoric acid is added, the sample is shaken then left to equilibrate for one hour. The CO2 produced is introduced into the mass spectrometer in continuous flow mode.
Non-Traditional Stable Isotopes
A wide range of non-traditional stable isotope systems are currently analysed at PCIGR. For these elements, the natural mass dependent isotopic fractionation is measured; it is therefore crucial to accurately and precisely correct for the instrumental mass fractionation without eradicating this natural signature. Data are also sensitive to both spectral and non-spectral matrix effects (Barling et al, 2006) and thus every effort is made to remove all matrix. Every element system has its own issues in this regard.
Analytical protocols at PCIGR have been set up for Li, Fe, Cu, Zn, Mo and Cd. Where the element of interest is the focus of graduate research, the student, under supervision, is responsible for establishing the protocol for the element. The first step is to establish conditions for precise and accurate measurement of a single standard (i.e. measurement of del = 0). To do this a number of factors are investigated. These include, but are not limited to: length of analysis, need for on-peak-zeros, automatrix effects, optimal analyte element/external normalizing element ratio. After this a secondary standard is introduced to determine that accurate and precise non-zero dels can be measured. Finally, samples are measured. At this point unsuspected matrix effects may be detected due either to a specific element causing a spectral interference or to a high level of matrix causing non-spectral matrix effects. If such matrix effects are detected then sample purification protocols are fine tuned either to eliminate specific elements and/or for specific types of matrix e.g. geological v. biological matrices (Barling et al., 2006, Shiel et al. 2009).
The range of analytical issues covered and the rigour needed in order to establish measurement protocols for analysis of a non-traditional stable isotope system means that once a PCIGR student has done this they are fully capable of applying what they have learned to the isotopic analysis of any other element of interest on the Nu Plasma or indeed on any MC-ICP-MS.
Lithium has no similar mass element that could be used to monitor instrumental mass fractionation externally and only two isotopes. For Li therefore, instrumental mass fractionation can only be compensated for by assuming a stable drift in instrument behaviour and samples are therefore measured by simple sample standard bracketing of measured ratios. The existence of only two isotopes means that Li is vulnerable to unidentified spectral and non-spectral matrix effects since these cannot be identified in the data. However, the low mass of Li means that few elemental and isobaric interferences are present, although 14N2+, 12C2+ can occur (Tomascak, 2004). Non-spectral matrix effects are mitigated by thorough calibration of the Li separation chemistry.
At PCIGR we have adopted the approach of Jeffcoate et al. (2004). Using this method we are currently able to measure d7Li/6Li in samples to within ± 1‰ (2SD).
Li isotope systematics are currently being applied to pegmatite geochemistry (Barnes et al., 2008a, 2008b).
Mo, Cd, Zn isotopes
In contrast to Li, elements like Mo and Cd, have multiple isotopes and a choice of possible elements to use to normalize the data externally for instrumental mass fractionation. For these elements data quality can be assured by agreement of results from different data reduction methods (e.g. simple sample standard bracketing of measured ratios, sample standard bracketing of externally normalized data and graphical data reduction) and by internal consistency of the del/amu of multiple isotope ratios for individual runs. One has therefore several means of identifying spectral and non-spectral matrix effects for these elements.
For Mo and Zn all the necessary isotopes can be measured in a single static cycle, with typical 2SD errors on d/amu that are better than ± 0.05‰. In contrast a dynamic approach, using two cycles is required in order to measure Cd isotopes. This is due to the range of masses measured; from 107Ag (external normalization element) to 118Sn (isobaric interference). The dynamic approach can introduce additional uncertainty to the measurements due to the inherent instability of a plasma source, even so typical 2SD errors on d/amu for two cycle dynamic measurements is better than ± 0.1‰.
Cd, Zn and Pb isotope systematics are currently being applied to the sourcing of metals in bivalves (Shiel et al., 2008).
Fe isotopes present their own special set of difficulties due to major polyatomic interferences inherent to the ICP source: ArN+ (mass 54), ArO+ (mass 56) and ArOH+ (mass 57). A number of approaches can be used to minimize the significance of these interferences, however, they generally rely on increasing the Fe signal relative to the interferences and thus require µg size Fe samples. In order to measure smaller (<300ng) Fe samples at PCIGR we have been investigating various means of reducing argide production and transmission without sacrificing instrument sensitivity for Fe (Aimoz et al., 2008).
Cartoon illustrating parameters investigated by Aimoz et al., 2008.
Radiogenic Isotopes (by MC-ICP-MS)
Nd, Hf and Pb isotope ratios are measured at PCIGR (e.g. Weis et al. 2005, 2006, 2007). Each analysis takes approximately 13 minutes and consumes 100, 75 and 40 ng respectively of Nd, Hf and Pb with regular settings. With the ES interface, the analyte size can be reduced to 25, 20 and 10 ng respectively for Nd, Hf and Pb.
Annual external reproducibility (2SD) for standards are currently: 43 ppm on 143Nd/144Nd, 27 ppm on 176Hf/177Hf and 91 ppm on 206Pb/206Pb. External reproducibility during a single measurement session is better than this by a factor of two or more. (See individual element links below). For sample runs, standards bracket every two samples and data are normalized to these bracketing standards offline. After offline normalization, typical 2SD external reproducibility of replicate samples is better than 40 ppm for 144Nd/143Nd, 70 ppm for 176Hf/177Hf and 200 ppm 206Pb/204Pb.
For Nd isotope ratio measurements, masses 150, 148, 146, 145, 144, 143, 142 are measured together with monitoring of Sm at mass 147 and Ce at mass 140, which allows interference corrections to be applied to masses 150, 148, 144 and 142. Nd isotope measurements are normalized internally for instrumental mass fractionation to a 146Nd/144Nd ratio of 0.7219 using an exponential correction. Interferences on Nd are calculated using natural abundances for the interfering element and adjusting them for instrumental mass fractionation as monitored by the normalizing ratio used to correct the Nd. (Annual values and reproducibility for the JNDI Nd standards in 2014.)
Nd isotopes can be measured by both MC-ICP-MS and TIMS at PCIGR. MC-ICP-MS is usually preferred due to its faster output (~13 min/sample instead of ~1 hour) for similar precision.
The configuration used to measure Hf isotopes enables simultaneous collection of Hf (180, 179, 178, 177, 176 and 174) together with monitoring of Lu at mass 175 and Yb at mass 172, which allows interference corrections to be applied to masses 174 and 176. Hf isotope measurements are normalized internally to a 179Hf/177Hf ratio of 0.7325 using an exponential correction. (Annual values and reproducibility for the JMC Hf standard in 2014.)
For Pb isotopic analyses the instrument is configured for simultaneous collection of Pb (208, 207, 206 and 204) together with Tl (205 and 203), which is used to monitor and correct for instrumental mass discrimination and Hg potential interference (202), which is used to correct mass 204 for the presence of 204Hg. Mercury levels are always below 0.7 mV and more typically are less than 0.2 mV of 202, corresponding to a correction of less than 0.18 (0.05) mV on the 204 peak. To improve the reproducibility of the analytical conditions for the Pb isotopic analyses, and thus the precision, all sample solutions are analyzed with the same Pb/Tl ratios as the NIST SRM 981 standards which requires determination of the exact Pb content after column separation. (Annual values and reproducibility for the NIST SRM 981 standard in 2014.)
A particular concern at PCIGR is that our MC-ICP-MS Pb isotope data achieve the best levels of precision and accuracy required for investigation of isotope systematics in mantle geochemistry (e.g. Weis et al., 2011). We therefore continue to investigate the roles of residual matrix, leaching and sample purification on the ultimate precision and accuracy of MC-ICP-MS Pb data (Barling & Weis 2008, 2012; Hanano et al. 2009; Nobre Silva et al. 2009).
Sr isotopes (by TIMS)
For Sr isotopic analyses, both TIMS instruments are configured for simultaneous collection of Sr Sr (84, 86, 87, 88) and Rb at mass 85, which allows for interference corrections to be applied to 87Sr for the potential presence of 87Rb. Each analysis takes approximately 1 hour. Sr isotope measurements are normalized internally to an 86Sr/88Sr ratio of 0.1194 using an exponential correction. For sample runs, data are normalized to the average value of the standard SRM NBS 987 (typically 3-4 standards analysed for 17-18 samples).