IAEA-TEL-2011-01 proficiency test

The International Atomic Energy Agency (IAEA) is running a proficiency test (IAEA-TEL-2011-01) for the determination of stable isotopes (H, O) in water samples. The aim of the test is to evaluate the performance of participating laboratories for the measurement of oxygen and hydrogen stable isotope composition of four unknown water samples (IAEA-1 to 4). The samples consist of two lake waters, one groundwater and one melted snow sample. Results of the analyses are to be submitted to the IAEA by 31 August 2011.

In the SIL we have analysed all four samples, together with five standards (VSMOW, USGS64700, USGS64444, GISP and NTW201108). The analyses were carried out using the three routine methods we have available for water isotopes:

  1. CO2-H2O equilibration followed by mass spectrometric analysis for 18O. 2.4mL aliquots of samples were run in quadruplicate on the UEA auto-equilibration system followed by analysis on the Europa-SIRA series II dual inlet IRMS.
  2. High temperature pyrolysis of water followed by mass spectrometric analysis for 2H. 0.1micro-litre sample volumes were injected into a ThermoFinnigan TC/EA pyrolysis system and analysed on a Delta Plus mass spectrometer in continuous flow mode. Samples were injected 10 times and the first three analyes ignored to eliminate memory effects.
  3. Both oxygen and hydrogen isotopes were determined by infra-red spectroscopy using the Picarro Cavity Ring Down (CRD-IRS) water vapour analyser equipped with a liquid water vapourisation interface. 2.2 microlitre sample volumes were used for the analysis resulting in a cell vapour concentration of approximately 20,000 ppm. Samples were injected 10 times and the first four injections ignored to eliminate possible memory effects.

All results are normalised to VSMOW = 0, corrected for scale compression/stretch using the certified composition of USGS64444 (d18O = -51.14 per mille, d2H = -399.1 per mille). Results are reported on the VSMOW-SLAP scale.

The quoted precisions represent our best estimate of the one sigma standard deviation of analyses based on a propagation of all contributing errors.

The results are summarised in the folowing table. All raw data are available here.


Sample
certified value
oxygen isotope data - d18O(vsmow-slap)
hydrogen isotope data - d2H(vsmow-slap)


CRD-IRS
precision
CO2-H2O/ms
precision
CRD-IRS
precision
pyrolysis/ms
precision










NTW201108

-7.26
0.12
-7.25
0.08
-48.72
0.40
-47.90
1.40
USGS67400
(-1.97, +1.25)
-2.05
0.12
-1.95
0.08
0.00
0.40
1.60
1.40
VSMOW
(0, 0)
0.00
0.12
0.00
0.08
0.00
0.40
0.00
1.40
IAEA-1

0.39
0.12
0.41
0.08
-0.93
0.40
0.57
1.40
IAEA-2

-5.35
0.12
-5.27
0.08
-42.93
0.40
-41.90
1.40
IAEA-3

-10.05
0.12
-10.02
0.08
-72.57
0.40
-72.40
1.40
IAEA-4

-15.45
0.12
-15.34
0.08
-113.39
0.40
-112.50
1.40
GISP
(-24.76, -189.5)
-24.80
0.12
-24.73
0.08
-188.99
0.40
-189.30
1.40
USGS64444
(-51.14, -399.1)
-51.14
0.12
-51.14
0.08
-399.09
0.40
-399.10
1.40

Until laboratories are told of the certified composition of samples IAEA-1 to 4 it is not possible to draw conclusions concerning the quality of our analyses. At a first pass the agreement between our measured and certified compositions for GISP and USGS67400 gives us a degree of confidence that our performance with respect to the unknown water samples is good.

Moreover, within measurement error the different techniques for both the oxygen and hydrogen isotope analyses give identical sample compositions. The one to one correspondence between mass spectrometric and CRD-IRS analysis is further illustrated in the next 2 charts which plot the CRD-IRS results versus the mass spectrometer analysed data. The gradient of the linear regressions between the two sets of data for both oxygen and hydrogen are unity with all residuals significantly less than the estimated analytical precision.

Picarro_versus_ms_oxygen.jpg

Picarro_versus_ms_hydrogen.jpg