 |
|
|
|
|
|
|
|
|
|
|
|
|
 |
||||||||||||
 |
 |
 |
||||||||||
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
||
 |
||||||||||||
Quality Assurance Plan of the Reston Stable Isotope Laboratory
 |
 |
 |
The Reston Stable Isotope Laboratory (RSIL) is an entity that produces hydrogen, carbon, nitrogen, oxygen, and sulfur stable isotope ratio analyses of water, rock, and biological samples for the operational and research components of the U.S. Geological Survey. It is located in the John Powell National Center, Reston, Virginia.
Quality Assurance Policy Statement
The goal of the RSIL is to produce consistently the highest quality isotope ratios of hydrogen-, carbon-, nitrogen-, oxygen-, and sulfur-bearing materials, recognizing that these analytical data may need to withstand the highest scrutiny and may be used in judicial proceedings. This statement applies specifically to the primary analyses provided by the RSIL: (1) hydrogen isotope ratio analysis of water samples, (2) oxygen isotope ratio analysis of water samples, (3) carbon isotope ratio analysis of sediments and biological materials, (4) nitrogen isotope ratio analysis of sediments and biological materials, and (5) sulfur isotope ratio analysis of dissolved sulfate in water samples. However, the general principles set forth in this document are employed in all RSIL isotope ratio analyses. The purpose of this Quality Assurance Plan is to describe briefly how the RSIL achieves its quality assurance goals.
Organizational Plan
The RSIL numbers about a dozen individuals, including chemists, a hydrologist, physical science technicians, an electronics technician, and physical and biological science aids. There are no administrative subdivisions in the RSIL. The chief of the RSIL is T. B. Coplen and he is responsible for ensuring that the RSIL conducts its operating according to the guidelines set forth in the Quality Assurance Plan. Visiting scientists from other countries sometimes join the RSIL for training and cooperative development of analytical methods.
Analytical Methodology
The two key components for ensuring that a laboratory consistently provides quality data are reliable, documented analytical procedures and implementation of a laboratory information management system (LIMS). Documented analytical procedures allow another laboratory to reproduce isotope ratio measurements and should allow that laboratory to report the same isotopic composition on the same sample within analytical uncertainty.
The reliability and accuracy of isotopic data generated by a laboratory can be improved by utilizing a LIMS to (i) store information about samples, (ii) print sample labels with user's sample identification on each label, (iii) store the results of mass spectrometric isotope ratio analyses of samples, (iv) calculate analytical results using standardized algorithms stored in a database, (v) normalize stable isotopic data to international scales using isotopic reference materials, and (vi) generate multi-sheet paper templates for convenient loading of samples and isotopic reference materials in automated sample preparation manifolds of isotope ratio mass spectrometers. The database program used by the RSIL is Laboratory Information Management System (LIMS) for Light Stable Isotopes (Coplen, 2000). Major benefits of this system include (i) a dramatic improvement in quality assurance, (ii) an increase in laboratory efficiency, (iii) a reduction in workload due to the elimination or reduction of retyping of data by laboratory personnel, and (iv) a decrease in errors in data reported to sample submitters. Such a database provides a complete record of when and how often laboratory reference materials have been analyzed and provides a record of what correction factors have been used through time. It provides an audit trail for the RSIL.
When samples are submitted to a laboratory, they normally are accompanied by a paper or electronic document listing the samples and describing the analyses to be performed. In LIMS for Light Stable Isotopes (abbreviated LIMS from here on), this document is called the sample submission document. Before samples can be analyzed, they must be logged in to LIMS. The date on the sample submission document is a key for organizing the samples in LIMS. When samples arrive at the laboratory, each sample must have been assigned a unique identifier by which the sample submitter distinguishes his (her) samples, the Field ID. When samples are logged into LIMS, LIMS will assign an alphabetic Prefix and an integer, separated by a hyphen, to each sample. This is called the Our Lab ID in LIMS, and it is the unique identifier by which LIMS knows each sample. Examples are N-3345, W-332, and R-10. The prefix is determined by the type of sample submitted (for example, W for water for hydrogen and oxygen isotope ratio analysis, and S for sulfur isotopic analysis). Using this system samples are kept in numeric order and they are easy to locate in cabinets, saving laboratory personnel time. Once samples have been logged into LIMS, they can be analyzed for isotopic composition. Samples cannot be analyzed prior to logging them in to LIMS because the Our Lab ID needs to be entered into the sample ID field of each mass spectrometric analysis. Samples are prepared for isotopic analysis using established methods or procedures. When samples are analyzed for isotopic composition, LIMS requires that each mass spectrometric isotopic analysis have a unique integer value, known in LIMS as the analysis number. Once samples have been analyzed with an isotope ratio mass spectrometer, the isotopic analyses are imported into LIMS. The isotopic results are "corrected" using isotopic reference materials analyzed daily with the samples. Reports of the reproducibility of these reference materials are generated daily by LIMS and used to determine the isotopic composition of samples. These laboratory reference materials are calibrated using internationally distributed isotopic reference materials available from the National Institute of Standards and Technology (Gaithersburg, Maryland) and the International Atomic Energy Agency (Vienna, Austria). Isotopic results are reported to sample submitters on internationally accepted isotope ratio scales.
Water samples submitted for hydrogen isotope ratio
analysis:
Water samples submitted for hydrogen isotope
ratio determination are analyzed using the gaseous hydrogen equilibration
procedure published by Coplen, Wildman, and Chen (1991). As many as 50
unknowns and 10 reference water samples (Potomac River water and Antarctic
precipitation) are loaded on a sample preparation manifold. Air is evacuated
and gaseous hydrogen is added to each water sample and reference. Equilibration
is maintained at 30±0.1ºC for at least 1 hour using a platinum
catalyst; the gaseous hydrogen is expanded into the isotope ratio mass
spectrometer and analyzed. Each water samples is analyzed in duplicate.
The standard deviation is calculated by LIMS and if it is larger than
1.55 ‰, the sample is re-analyzed until the 2-sigma uncertainty of
the result is better than 2 ‰. This means that if the same sample
were submitted to the laboratory again for hydrogen isotopic analysis,
there is a 95 per cent probability that the result returned to the sample
submitter would be within ±2 ‰ of that originally reported.
The isotopic composition is reported to the sample submitter in per mill
relative to VSMOW reference water on a scale such that SLAP reference
water is -428 ‰ exactly (Coplen, 1994).
Water samples submitted for oxygen isotope ratio
analysis:
Water samples submitted for oxygen isotope
ratio determination are analyzed using the carbon dioxide-water equilibration
technique of Epstein and Mayeda (1953). As many as 48 water samples are
loaded on the equilibration introduction system of isotope ratio mass
spectrometer. Air is removed by pumping and carbon dioxide is added to
each sample. After shaking and equilibration at a temperature of 25.0±0.5ºC
for at least 7 hours, the carbon dioxide is expanded into an isotope ratio
mass spectrometer and analyzed. Every third water sample is analyzed in
duplicate. The standard deviation is calculated on duplicate analyses.
If it is larger than 0.15 ‰, the samples in the manifold are all
re-analyzed until the 2-sigma uncertainty of the result is better than
0.2 ‰. This means that if the same sample were submitted to the laboratory
again for oxygen isotopic analysis, there is a 95 per cent probability
that the result returned to the sample submitter would be within ±0.2
‰ of that originally reported. The isotopic composition is reported
to the sample submitter in per mill relative to VSMOW reference water
on a scale such that SLAP reference water is -55.5‰ exactly (Coplen,
1994).
Sediment and biological samples submitted
for carbon isotope ratio analysis:
Sediment and biological samples submitted
for carbon isotope ratio determination are analyzed using continuous flow
isotope ratio mass spectrometry after conversion of the sample into carbon
dioxide in an elemental analyzer using the method described in Qi and
others (2003). As many as 50 unknowns and reference samples are loaded
into the carousel of an elemental analyzer. Each, in turn, is converted
into gaseous carbon dioxide and swept into a continuous flow isotope ratio
mass spectrometer in a stream of helium. Every third unknown is analyzed
in duplicate. The standard deviation is calculated on duplicate analyses
using LIMS. If it is greater than 0.28 ‰, the unknowns are re-analyzed
until the 2-sigma uncertainty of the result is better than 0.4‰.
This means that if the same sample were submitted to the laboratory again
for carbon isotopic analysis, there is a 95 per cent probability that
the result returned to the sample submitter would be within ±0.4
‰ of that originally reported. The isotopic composition is reported
to the sample submitter in per mill relative to VPDB on a scale such that
L-SVEC LiCO3 carbonate and NBS 19 CaCO3, respectively, are -46.6 ‰
and +1.95 ‰ exactly.
Sediment and biological samples submitted for nitrogen isotope ratio analysis:
Sediment and biological samples submitted
for nitrogen isotope ratio determination are analyzed using continuous
flow isotope ratio mass spectrometry after conversion of the sample into
gaseous nitrogen in an elemental analyzer using the method described in
Qi and others (2003). As many as 50 unknowns and reference samples are
loaded into the carousel of an elemental analyzer. Each, in turn, is converted
into gaseous nitrogen and swept into a continuous flow isotope ratio mass
spectrometer in a stream of helium. Every third unknown is analyzed in
duplicate. The standard deviation is calculated on duplicate analyses
using LIMS. If it is greater than 0.28 ‰, the unknowns are re-analyzed
until the 2-sigma uncertainty of the result is better than 0.4 ‰.
This means that if the same sample were submitted to the laboratory again
for nitrogen isotopic analysis, there is a 95 per cent probability that
the result returned to the sample submitter would be within ±0.4
‰ of that originally reported. The isotopic composition is reported
to the sample submitter in per mill relative to nitrogen in air. The nitrogen
isotopic composition of various internationally distributed reference
materials is given by Böhlke and Coplen (1995).
Nitrate samples submitted for nitrogen (and oxygen)
isotope ratio analysis:
Dissolved nitrate and nitrate in soil submitted for nitrogen (and oxygen)
isotope ratio determination are analyzed using continuous flow isotope
ratio mass spectrometry after bacterial denitrification of the sample
into nitrous oxide (Sigman and others, 2001; Casciotti and others, 2002;
Coplen and others, 2004). As many as 32 unknowns and reference samples
(Böhlke and others, 2003) are analyzed daily with this technique.
Nitrate is converted into gaseous nitrous oxide and swept into a continuous
flow isotope ratio mass spectrometer in a stream of helium. Samples typically
are analyzed in duplicate using the nitrate concentrations provided by
the sample submitter, and these concentration must be accurate to within
±10 percent. For samples with nitrate concentrations of at least
0.06 mg/kg as N, the standard deviation is calculated on duplicate analyses
using LIMS. If the 1-sigma of nitrogen isotope ratios is greater than
0.25 ‰ (or the 1-sigma of oxygen isotope ratios is greater than
0.5 ‰), the unknowns are re-analyzed until the 1-sigma uncertainty
of the result is better than 0.25 ‰ for nitrogen isotope ratios
(or 0.5 ‰ for oxygen isotope ratios). This means that if the same
sample were submitted to the laboratory again for nitrogen isotopic analysis,
there is a 95 per cent probability that the result returned to the sample
submitter would be within ±0.5 ‰ of that originally reported.
The uncertainty for nitrate samples with concentrations less than 0.06
mg/kg as N is twice that indicated above. The isotopic composition is
reported to the sample submitter in per mill relative to nitrogen in air.
The nitrogen isotopic composition of various internationally distributed
reference materials is given by Böhlke and Coplen (1995). The oxygen
isotopic composition is reported to the sample submitter in per mill relative
to VSMOW reference water on a scale such that SLAP reference water is
-55.5 ‰ exactly (Coplen, 1994).
Water samples submitted for sulfur isotope ratio
analysis of dissolved sulfate:
Water samples submitted for sulfur isotope
ratio determination of dissolved sulfate are analyzed using continuous
flow isotope ratio mass spectrometry after precipitation of the sample
as barium sulfate (Carmody and others, 1997) and conversion of the barium
carbonate into gaseous sulfur dioxide in an elemental analyzer (Glesemann
and others, 1994). As many as 50 unknowns and reference samples are loaded
into the carousel of an elemental analyzer. Each, in turn, is converted
into gaseous sulfur dioxide and swept into a continuous flow isotope ratio
mass spectrometer in a stream of helium. Every third unknown is analyzed
in duplicate. The standard deviation is calculated on duplicate analyses
using LIMS. If it is greater than 0.28 ‰, the unknowns are re-analyzed
until the 2-sigma uncertainty of the result is better than 0.4 ‰.
This means that if the same sample were submitted to the laboratory again
for sulfur isotopic analysis, there is a 95 per cent probability that
the result returned to the sample submitter would be within ±0.4‰
of that originally reported. The isotopic composition is reported to the
sample submitter in per mill relative to VCDT defined by adopting a value
of -0.3 ‰ for IAEA-S-1 silver sulfide (Coplen and Krouse, 1998).
External Review
In addition to locally produced laboratory
reference materials that are interspersed with samples submitted for determination
of stable hydrogen and oxygen isotopic composition, the RSIL analyzes
unknown water samples provided by the National Water Quality Laboratory
(NWQL). Approximately 1 out of every 50 samples submitted by the NWQL
is an unknown. The isotopic results of these unknowns are available from
the NWQL. To date, this external review program has confirmed that the
RSIL is meeting its Quality Assurance objectives.
References