15 August 2011

Data File: 15082011 water.pdf, 15082011 water.csv

This was an equilibration run with the new NTW. The experiment was designed to test the functioning and reproducibility of data obtained from the CO2-H2O equilibration unit prior to a series of calibration and laboratory intercomparison runs using IAEA VSMOW, IAEA GISP, USGS 67400, USGS 64444 standards and IAEA inter laboratory comparison samples 01 to 04. The data files can be downloaded by clicking on either the .pdf or .csv versions.

The data are plotted as a function of run number in the plot below. The average of all the data is -7.228 +/- 0.08 per mille (n=40). Analysing each bank (A, B, C and D) separately gives average and precision respectively as:

Bank A: -7.204 +/- 0.069 (n=12)
Bank B: -7.297 +/- 0.072 (n=6)
Bank C: -7.179 +/- 0.086 (n=10)
Bank D: -7.257 +/- 0.050 (n=12)

The histogram shows that, with the exception of one outlier, the data are approximately normally distributed. The origin of the lower precision for Bank C is not clear and maybe simply an artefact of too small a sample population.

Overall each bank in the equilibration unit is giving the same measured d18O composition for water with no signs of systematic bias.


17 August 2011

This sample batch includes two USGS standards, 67400 and 64444 and the new NTW internal lab standard. USGS 67400 has a composition of d18O(vsmow) = -1.97 per mille, d2H(vsmow) = +1.25 per mille, and USGS64444 has composition of d18O(vsmow) = -51.14 per mille, and d2H(vsmow) = -399.1 per mille. These are close to VSMOW and SLAP and provide an opportunity for a preliminary calibration of NTW. In addition, the 50 per mille range in composition between USGS 67400 and 64444 will allow an assessment of any stretch/compression in the isotope scale that we see on the equilibration system, as well as an assessment of inter sample memory effects. Banks B, C and D contain a variety of samples (wine, seawater and rainwater) together with NTW.


The results for this run for the 3 standards are:

USGS 67400: d18O(vsmow) = -1.970 +/- 0.023 per mille
USGS 64444: d18O(vsmow) = -51.257 +/- 0.03 per mille
NTW 201108: d18O(vsmow) = -7.250 +/- 0.033 per mille

A comparison of the result for USGS64444, which is close in composition to SLAP, shows it to be about 0.1 per mille depleted in 18O compared to the accepted value of -51.14 per mille. Using the measured difference between USGS67400 and USGS64444, and the accepted difference a small scale correction factor of 0.997626 is calculated. i.e. the scale is stretched by ca. 0.24%. The corrections for samples within 20 per mille of VSMOW are small and on the order of the measurement precision (+/- 0.050 per mille) and can usually be ignored. The correction only becomes important for samples that are strongly depleted in 18O.

Examination of the figure above shows that on passing from NTW to USGS64444 there is a small memory effect. This is on the order of 1.4% of the difference in d18O between adjacent samples in the analysis batch. For samples that are close in composition (e.g. a suite of groundwater samples from the same region) the memory effect is not important. Where samples that are widely different in composition it is advisable to run 3 or 4 samples ins sequence so that the effect can be eliminated.

18 August 2011

IAEA-TEL-2011-01 proficiency test on the determination of stable isotopes in water (H,O)

The Terrestrial Environment laboratory of the IAEA, Siebersdorf, Vienna is co-ordinating a laboratory proficiency test for the analysis of oxygen and hydrogen isotopes in water samples of unknown (to the participating laboratories) isotope composition. There are 4 waters in this test: IAEA-1 to 4. The water descriptions are:

IAEA-1: The water sample is a mixture of water originally collected in 1999 for the preparation of VSMOW2. It mainly consistes of lake water and groundwater samples. The sample was stored in a stainless steel container and was analyzed multiple times with different instrumentation.

IAEA-2: Water from lake Balaton, Hungary, sampled in 2006.

IAEA-3: Shallow groundwater sample from Strebersdorf, Vienna, sampled in 2007.

IAEA-4: Melted snow sample from Huttschlag, Austrian Alps, sampled in 2007.

The SIL is analysing these water samples using two different methods. The first is mass spectrometery using the traditional CO2-H2O equilibration method for oxygen isotope composition. The hydrogen isotope analysis by mass spectrometry is done using high temperature pyrolysis and mass spectrometer analysis of the resultant hydrogen gas using the Delta XP TC/EA analysis system. The second method is Cavity Ring Down Infra Red (CRD-IR) spectroscopy using a Picarro liquid water isotope ratio analyser.

A) Mass spectrometric analysis of oxygen isotope composition

The raw data files and results of our analysis for oxygen isotope composition are in the pdf and csv files (20110818). To our best estimate the test sample compositions are:

IAEA-1: d18O(vsmow-slap) = +0.41 +/- 0.12 per mille
IAEA-2: d18O(vsmow-slap) = -5.27 +/- 0.08 per mille
IAEA-3: d18O(vsmow-slap) = -10.02 +/- 0.08 per mille
IAEA-4: d18O(vsmow-slap) = -15.34 +/- 0.08 per mille

To check data quality we ran 3 other standard samples with certified values. The results are (certified values in brackets):

USGS67400: d18O(vsmow-slap) = -1.95 +/- 0.08 per mille (-1.97)
USGS64444: d18O(vsmow-slap) = -51.14 +/- 0.08 per mille (-51.14)
GISP: d18O(vsmow-slap) = -24.73 +/- 0.08 per mille (-24.76)

The data are plotted below as a function of their run position in the equilibration unit.

B) Mass spectrometric measurement of hydrogen isotope composition

C) CRD-IR absorption spectroscopy measurement of oxygen and hydrogen isotope analysis