What makes a good working standard?
The first concern is that the working standard has the elements of interest and is in a form that is useful for your analytical techniques. For example, a fine powder for EA work would be great. Isotopic homogeneity is probably the most important characteristic. If a batch of material is not isotopically homogenous, then its measured isotope ratio will vary with each measurement. Hence, the way to check a working standard is to make multiple measurements of it using a variety of sample sizes. If it is sufficiently homogenous, then you should get the same isotope ratio within acceptable precision over multiple measurements. As a rule of thumb, mixed liquids are homogenous. Hence, if you can dissolve the working standard in a solution for your analyses (for example making a large batch of standard for the TOC analyzer) then you can be sure that the homogeneity is ok.
More detailed information on the working standards for waters, bulk C and N, and total organic/inorganic C are provided below.
Working standards for EA and TOC analyses of Carbon and Nitrogen:
Sucrose and KHP for δ13C on the TOC analyzer and the EA.
Glycine (from Aldrich) for δ13C and d15N (referred to as "Glycine Low")
Glycine (from Aldrich) spiked with 2-13C-glycine and 15N-glycine (both also from Aldrich) for δ13C and δ15N. (referred to as "Glycine Mid")
Glycine (from Aldrich) spiked with more 2-13C-glycine and 15N-glycine (both also from Aldrich) for δ13C and δ15N. (referred to as "Glycine High")
NIST 2710, Montana soil for δ13C of low wt %C samples
NIST 2711, Montana soil for δ13C of low wt %C samples
Carbon calibrations for the glycine and NIST working standards were performed using USGS 40, USGS 41, NBS 18, and NBS 19. The KHP and sucrose carbon standards were calibrated with USGS 40 and USGS 41. Nitrogen numbers for the NIST standard were provided by Rick Doucette. All other nitrogen numbers were measured with USGS 40, USGS 41, and IAEA-N-2.
Note that the international standards that were used to calibrate these working standards have a standard deviation of 0.1 ‰. Although we report our numbers with two digits past the decimal, the accuracy of the final number for any measurement made in the lab cannot be known to within better than +/- 0.2‰ (+/- 2s) of the actual value.
Working standards for δ2H and δ18O measurements of waters:
18O nearzero: Prepared from one gallon of Trader Joe's Pure New Zealand water spiked with 10% 18O labelled water to get it close to zero.
2H nearzero: Prepared from one gallon of Trader Joe's Pure New Zealand spiked with D2O.
PNZ: Prepared from Trader Joe's Pure New Zealand bottled water.
Dasani: Prepared from Dasani bottled water.
MSW: Prepared from Trader Joe's Mountain Spring Water.
DMSW: Prepared from distilled Trader Joe's Mountain Spring Water.
DSW-ANT: Prepared from a blend of distilled Trader Joe's Mountain Spring Water and Antarctic ice water.
ANT: Prepared from Antarctic ice water graciously given to us by Tyler Coplen's group at USGS.
Working Standards for δ18O measurements:
*n = 7 to 9
The raw data for the 18O working standard calibrations is here in excel format. No drift was observed here so no drift correction was performed on this data. If you use the raw data in the second worksheet then you can recalculate the values for the standards and check them against the ones shown in the table above.
Working Standards for δ2H measurements:
*n = 7 to 9
All working standards are calibrated against VSMOW and SLAP. GISP is run with the calibration as an accuracy check. The raw data for the 2H working standard calibrations is here in excel format. It seems that we didn't do so great a job sealing some of these vials as several apparently had no H2 in them. A drift correction was used here but drift was insignificant in the end. The precision of the measurements was slightly improved with the drift correction, however, it did not change the actual final measured values. If you use the raw data in the second worksheet then you can recalculate the values for the standards and check them against the ones shown in the table above.
Working standards are stored in amber 240mL Qorpak bottles with the polyseal cap. When a new working standard bottle is opened for use, the entire contents is transferred to 5mL glass ampoules in 5mL aliquots and flame sealed. For any given analysis multiple working standards are used (two to bracket the samples and at least one with a midway value as a check of measurement accuracy and precision). The working standards are spread throughout the sequence so as to minimize effects of instrumental drift.