MBL Stable Isotope Laboratory
Bulk Organic Matter, Particulates Filters, NH4 & NO3 Salts, NH4 Diffusion Filters
Both natural abundance and isotopically enriched solid samples (animal matter, vegetation, organics in soils and sediments, particulates filters, NO3 and NH4 Salts, and NH4 diffusion filters) are analyzed at the MBL Stable Isotope Laboratory for δ15N and δ13C using a continuous-flow isotope ratio mass spectrometry (CF-IRMS) system. Based on replicate analyses of isotopically homogeneous international standards, we obtain a precision (1s) of about +/- 0.1 ‰ for both δ15N and δ13C measurements. This is comparable to the precision obtained on dual inlet isotope ratio mass spectrometers. Such high precision using CF-IRMS is accomplished by setting up each analytical run with careful consideration of individual sample characteristics (sample type, N & C composition, estimated isotope enrichment), and by analyzing appropriate quality control materials within each run that: (1) allow corrections if necessary for blank, drift, linearity effects and calibration fluctuations; and (2) that provide a check on the accuracy of the results.
Description of Instrumentation
Gas source mass spectrometers measure gases and, therefore, all samples must first be converted to appropriate gases for measuring the isotope of interest. At the MBL Stable Isotope Laboratory, samples are combusted (in the presence of oxygen) to produce N2 and CO2 and to determine d15N and d13C and respectively. This conversion is done using a Europa ANCA-SL elemental analyzer with an auto-sampler that sequentially drops each sample into a quartz combustion tube held at 1000oC and filled with CuO and Cr2O3 to provide oxygen for the combustion. In addition, oxygen gas is injected into the system to insure complete combustion of the sample. The main combustion products are CO2, N2 and NOx, and H2O. These gases are carried by a helium carrier gas through a reduction tube held at 600oC and filled with Cu metal where any NOx is converted back to N2. After removal of H2O, the remaining N2 and CO2 gases are then separated by gas chromatography before introduction into the mass spectrometer.
The purified N2 and CO2 gases are introduced into a Europa 20-20 mass spectrometer via the ion source where the gas molecules are first ionized and then accelerated into the flight tube. In the flight tube, the ions pass through a magnetic field provided by a large, fixed magnet. The magnetic field separates individual ions into ion beams of different masses that are focused so that they enter the collector (composed of detectors known as faraday cups). For d15N measurements, the masses of interest are 28 (14N14N), and 29 (14N15N), and for δ13C measurements, the masses of interest are 44 (12C16O16O) and 45 (13C16O16O). As the ions impact each detector, a small electrical current is generated that is proportional to the amount of ions present. The currents generated from each mass beam are then amplified and sent to the computer for processing.
Dual-Inlet vs. Continuous-Flow Systems
In a dual-inlet system, a reference gas of known isotopic composition and the purified sample gas are alternately bled at a constant rate into the mass spectrometer and each mass beam continuously impacts the detectors producing flat-topped peaks. By measuring the heights of each mass peak, very precise isotope ratios can be determined. Systems with dual inlets commonly obtain analytical precisions of better than +/- 0.1 ‰ for both δ15N andδ13C.
In the continuous-flow system of the Europa 20-20, the sample gas signals are continuously monitored as they flow out of the gas chromatograph and through the mass spectrometer to the collectors where each mass signal slowly rises, peaks and then drops as the gas flows through the system. The area under each mass peak is then determined and the ratio calculated. These ratios are compared to those produced by in-house reference materials of known isotope composition that are analyzed in the same sample batch. Precisions of continuous flow systems are usually better than +/- 0.3 ‰ for both δ15N and δ13C, and with careful work can approach the precisions attained by dual-inlet systems.
In order to attain high precision using CF-IRMS, we analyze within each analytical run a number of blanks, one or two international standards, and numerous in-house standards of different sizes and compositions, and then put the run data through a rigorous data reduction process that corrects for instrument drift, blank addition, and possible linearity effects. In addition, we normalize the results using two well characterized, isotopically distinct in-house standards, one of natural composition and the other with an enriched composition, to ensure the accuracy and reproducibility of the data, even on enriched samples.
In setting up an analytical run, every effort is taken to minimize the effect of blank N and/or C additions to the N2 and CO2 peak signals. Running samples that are substantially larger than the blank minimizes the blank correction. We prefer analyzing samples containing at least 100 micrograms of N or C. However, we often have to analyze smaller samples where the blank effect may be significant, and sometimes we don’t know the size of the unknowns. Therefore, we routinely run a number of empty tin capsules as blanks, calculate the average blank peak size and subtract it from the standard and sample peaks before the isotope ratios are calculated. In order for this correction to be meaningful it is important that the empty tin capsules run as blanks be the same type and size as those packed with standards and unknowns. Using this method, we have obtained "reasonable" isotope results on samples with as little as 5 micrograms of N and/or C.
Drift in isotope measurements during the course of an automated twelve-hour analytical run is not unusual and we monitor drift by the frequent placement of similarly sized reference standards in each run. Usually, most of the drift occurs at the beginning of the run and in order to minimize the correction, we analyze a number of test standards and blanks, prior to running samples.
If the mass spectrometer source parameters are incorrectly set, linearity effects may occur where the result will vary depending on the sample size. We carefully tune the source to reduce or hopefully remove this effect. However, we always run a sequence of standards of different sizes to quantify any effect remaining to allow for corrections to be applied to results. The best way to minimize any linearity effect is to run standards and samples with similar amounts of N and/or C. It is for this reason that we ask for additional information (%N, [N], %C, [C]) about submitted samples so that we can weigh out samples and standards of the same approximate size.
Finally, in order to monitor any day-to-day calibration fluctuations, we normalize the blank, drift, and linearity corrected data based on two well-characterized, isotopically distinct in-house test standards that are included in each analytical run. This normalization ensures consistent, accurate and reproducible results as documented by the results on replicate analyses of international and in-house standards (see below).
Accuracy and Precision of Isotope Data
Accepted MBL Stable Isotope Lab Results
Standard Material Values Average (n) Std Dev (1σ)
δ15N vs. AIR
International δ15N Standards
N-1 (NH4)2SO4 + 0.4 ‰ + 0.45 ‰ (23) +/- 0.10 ‰
N-2 (NH4)2SO4 + 20.3 ‰ + 20.34 ‰ (3) +/- 0.08 ‰
In-House δ15N Reference Materials
Glycine + 10.7 ‰ + 10.73 ‰ (127) +/- 0.13 ‰
Citrus Leaves + 4.8 ‰ + 4.83 ‰ (38) +/- 0.16 ‰
Peptone + 7.4 ‰ + 7.33 ‰ (17) +/- 0.17 ‰
(NH4)2SO4 +193.5 ‰ +193.47 ‰ (43) +/- 0.33 ‰
δ13C vs. PDB
International δ13C Standards
NBS-21 Graphite - 28.1 ‰ - 28.19 ‰ (13) +/- 0.11 ‰
In-House δ13C Reference Materials
Glycine - 34.0 ‰ - 34.02 ‰ (70) +/- 0.12 ‰
Citrus Leaves - 27.3 ‰ - 27.32 ‰ (26) +/- 0.14 ‰
Peptone - 14.7 ‰ - 14.67 ‰ (8) +/- 0.19 ‰
Dextrose - 9.8 ‰ - 9.80 ‰ (24) +/- 0.09 ‰
The analytical uncertainty on %N, [N], %C, and [C] measurements is approximately +/- 1% Relative.
When enriched samples are analyzed there is always the possibility of significant memory or carryover effects especially when there is a substantial jump in isotopic composition between two consecutive samples. There are a number of ways to deal with this problem. Firstly, we avoid, if possible, running enriched and natural composition samples in the same run. Secondly, if estimates of isotope enrichment are available, analyzing the samples in increasing or decreasing order of enrichment can minimize any memory effect. Finally, it is possible to quantify the memory effect by running a number of samples in triplicate and documenting the magnitude of the effect for different isotopic jumps, both up and down, and then correct the results if the effect is significant; however, this is a last resort.
Quality Control Limits
We closely monitor the magnitude of all of these corrections, and if any of the corrections become significant (i.e. > 1 ‰ or so), or if there is any other indication of a poor run, for instance poor peak shapes, but especially poor standard results (i.e. > +/- 0.5 ‰ of the accepted values listed in Table 1), we re-analyze the samples if there is sufficient material available.
Additionally, we routinely analyze 1 in 10 samples in duplicate (at no extra charge), and if these duplicates do not reproduce within 1 per mil, we analyze a third aliquot. Poor duplicates most often reflect poorly homogenized samples, so it is important that the researcher provides properly homogenized sample powders.
Finally, we are more than willing to analyze the researcher’s own quality control standard materials at no charge, but please let us know how we did!
It is emphasized that the accuracy, precision and ultimate reliability of the isotope results are not so much a feature of the instrumentation, but are largely due to care in sample preparation and analysis. Don’t forget that the key to obtaining good results is discussing your samples with the analyst in advance of submittal, and providing as much information about them as possible. Please contact him or visit our website the lab manager for additional information.
NH4 - NO3
We extract dissolved NH4 and/or NO3 from water samples off-line using the NH4-diffusion method of…. For this we need good estimates of concentration and d15N composition.
O2-Ar Triple Oxygen Productivity Determinations
AMBIENT GASES IN AIR