SHRIMP analysis types
Stable and radiogenic isotopes and trace elements can be measured using SHRIMP. The ion abundances are converted into actual concentrations and isotopic ratios using calibration procedures.
Stable isotope analysis
Kinetic processes affect abundances of stable isotopes such that they may be different between two coexisting minerals, or between mineral and fluids, etc. These fractionations are also temperature dependent. Kinetic mass fractionation is mass dependent and so lighter isotope systems have greater potential because of the larger fractional difference in mass. Common systems for analysis are C, O, and S.
The stable isotopes generally emit strongly as negative ions requiring use of the Cs primary beam and charge neutralization in the case of the analysis of insulators. Precisions around 0.1 permil and accuracy of the order of 0.3 permil can be obtained on SHRIMP in O and S isotopic analysis.
Instrumental effects easily fractionate light isotope ratios measured on SHRIMP. One such is the steering required for maximum beam transport through the source slit, which is a function of the distance the sample sits away from the extraction plate. To be sure that each of the unattended analyses hits the correct location and is in focus a post analysis image or snap (shot) is taken. Below the out of focus sample is an obvious outlier and can be rejected on the basis of the post analysis snap. The average 2 standard deviation of the remaining bracketing Amelia albite standards is much better than 0.3‰ with an overall session statistics of δ18O/16O of 10.7‰ ± 0.34‰ (2 standard deviations).
16 micron spots for δ18O to avoid
alteration & overgrowths & create
high resolution records of past
environments (Smith et al. 18; Science Advances)
Trace and volatile element analysis
Trace and volatile element abundances offer important constraints on petrogenetic processes. These elements are measured through a suitable isotope (either the most abundant isotope or one free from interference). The isotope ratios are ratioed to the isotope of a major element (e.g. Ca, Si) whose concentration can be determined through electron microprobe anlaysis. The ion ratios are calibrated to a standard for which the trace or volatile element concentrations are independently known. For example:
[Ci] = SFi*[Ca]*(ci+/44Ca+)
Ci is the concentration of the trace or volatile element i
SFi is the sensitivity factor of element i relating the yield of the element to the yield of the normalizing isotope (44Ca) and the Ca concentration of a standard.
[Ca] is the concentration of Ca in the unknown determined by electron microprobe
(ci+/44Ca+) is the measured ion intensity of the relevant isotope to 44Ca+.
SHRIMP II and SHRIMP RG have been calibrated for rare earth and trace elements in many materials. SHRIMP SI has been calibrated for hydrogen and sulfur abundances in some materials.
Trace element abundance pattern in Archean komatiite glass inclusions showing systematic incompatible element depletion. Zr and Ti are also depleted because these elements have fractionated into clinopyroxene. (McDonough and Ireland 1993).
SHRIMP U-Pb analysis consists of two separate measurement types.
1) The isotopic composition of the Pb isotopes is taken as measured. Molecular interferences are well resolved (in zircon, 180Hf28Si from 208Pb requires 5,500). Hydride contributions are generally less than 0.1%. Isotopic mass fractionation is taken as negligible but could be as high as 0.2% per amu.
2) The measurement of the U-Pb ratio (206Pb/238U) is an interelement ratio and so is affected by the different ionization yields of Pb and U. Moreover, the ratio is variable from standard materials with known ratios. It has been found that changes in the Pb/U ratio are reflected in the abundances of U, UO, and UO2 ion abundances. Hence, measured U-Pb isotope ratios must be calibrated to a standard to get age information. The first measurements on SHRIMP used U/UO covariance to monitor U/Pb fractionation.
The measured ion ratios of Pb/U covary with the concurrent measurement of UO/U. This allows the regression of the standard data to define a standard point for the session (B). Treating the unknown sample in the same way and regressing to the same UO/U will yield another point (A). For calibration to the 1099 Ma FC1 standard, the age of the unknown will be
Age(A) = (Pb/U)A/(Pb/U)B*1099