Analyzing Blanks & Standards (EA)

The final step in evaluating the system performance prior to analyzing samples is to establish that the blank levels are low and to analyze suitable standards

Elemental and Stable isotope standards

It is essential that standards, both isotopic and elemental (if percent composition is important) are analyzed prior to (and during) sample analysis.

For the most robust data, the principle of identical treatment of the standards and samples should be followed as closely as possible in that both should be analyzed in the same way. The isotopic standards must be homogeneous and ideally be similar in matrix to samples to be analyzed. Two or more standards should be analyzed, which bracket the isotopic range of the samples to provide at least a 2-point isotopic calibration.

Analyzing an EA standard with known isotopic compositions can serve to evaluate peak shapes, sensitivity, linearity, accuracy and also provide a mechanism to calibrate sample measurements. A vast array of stable isotope standard materials are readily available from the International Atomic Energy Agency (IAEA), National Institute of Standards and Technology (NIST) and the United States Geological Survey (USGS). Arndt Schimmelmann from Indiana University also provides a wide range of stable isotope standards for EA and GC.

A cost effective approach is to also obtain in-house standards; here the objective is to obtain a suite of pure, isotopically homogeneous compounds with varying values that have been calibrated using the IAEA/NIST/USGS standards. This can be achieved by analyzing a range of potential in-house standards via EA-IRMS; homogeneity of the potential standards may be improved by first grinding with a pestle and mortar.

If, as well as isotopic values, the elemental compositions are also required from the analysis, then a standard of known composition is necessary (e.g. sulfanilamide, acetanilide, etc.); these are provided with the instrument and are available directly from Elementar.

EA methods

Two different methods are required when running EA IRMS, the lyticOS method and the EA method.

A range of default lyticOS EA methods called ‘EA Analysis {type}’ are pre-installed within the software and can be accessed in the same way as the ‘stability’ and ‘linearity’ methods discussed above, i.e. under the ‘tasks’ tab, select ‘methods’ in the ‘manage’ section. The input parameters for ‘EA Analysis NCS’ as shown in Figure 6-23 illustrate that EA keywords from the EA software have been selected and that the EA method selected is ‘2mg70sIRMS’.  The vial number (sample position) and sample weight (mg) are set to 1 by default and the source inlet position is set to EA and no dilution is specified. The EA acquisition delay (the time the EA starts after the acquisition has started), the monitoring gas pulse delay and duration are all set to 30s (‘00:00:30’). The ‘EA Analysis NCS’ method involves the measurement of three species (‘S1’, ‘S2’ and ‘S3’). ‘S1’ is set to N₂ and the tuning specified, in this example, as ‘N2_600uA’, one monitoring gas pulse is required and the time window (‘S1 peak duration’) for the N₂ sample peak is set to 00:03:30. Similar parameters are then specified for CO₂ (‘S2’) and SO₂ (‘S3’). Similar to before these parameters can be adjusted within the task list if the ‘editable in the task list’ checkbox has been ticked.

A range of default EA methods are included as part of the EA software and can be accessed by selecting (‘Options’ > ‘Settings’ > ‘Methods...’). The method parameters include, oxygen dosing time (for combustion modes only), autozero delay times (defines the time after the TCD is set to zero before integration), peak anticipation times for each gas species and the desorption temperatures for the APT columns (Figure 6-24). The default methods include the typical sample size that should be analyzed, the time for oxygen dosing and whether it is an IRMS or EA method within its name, e.g. 2mg70sIRMS, i.e. 2mg typical sample weight, 70s of oxygen dosing and it is an IRMS method. The IRMS method differs from the EA in that they have longer peak anticipation times to allow for monitoring gas pulses to be included without co-eluting with sample peaks. The methods labeled as EA should only be used when operating the EA in standalone mode, for instructions, please refer to the EA manual.

Analyzing blanks and standards

Before the analysis of samples it is necessary to evaluate the blank levels and the performance of known standards.

It is useful to determine the blank of the system with and without a capsule that way the ‘system’ blank and the contribution solely from the capsule can be assessed. This is achieved by running a task with, in this case, the ‘EA Analysis NCS’ method selected in the task list (Figure 6-25). Although all the input parameters are pre-determined within the method, they can also be altered within the task list; in the example discussed here, the ‘EA keyword’, ‘EA method’, sample ‘weight’, ‘vial number’ or sample position and the ‘dilution type’ can all be overwritten within the task list.

The ‘2mg70sIRMS’ EA method was selected in the example in Figure 6-25, however if the blanks are high and the quality of the oxygen gas is suspected as the source, then running blanks with and without the introduction of oxygen can determine if the oxygen gas is the problem (‘Blank with O’ and ‘Blank without O’ EA method); please refer to the EA manual and software for more details.

The ‘EA keyword’ column allows for the EA keywords such as  ‘blnk’, ‘RunIn’ or ‘standard name’ from the EA software to be included. These keywords are essential when measuring the elemental composition from the sample runs, however they can be assigned to runs post-analysis within the EA software. Please refer to the EA manual and software for more details.

If elemental composition of the samples are required, the analysis of a standard with known elemental composition is required to be analyzed. Figure 6-26 lists tasks where ‘sulfanilamide’ is analyzed; the ‘EA keyword’ was left blank, this can be added within the EA software. Please refer to the EA manual and software for more details.

When running isotopic standards, two or more should be analyzed for each element with values that span the expected range of the samples. In NCS mode, the example in Figure 6-27 lists two glutamic acids (USGS40 and 41) and two silver sulphides (IAEA-S1 and S-2); the USSG40 and 41 have known d¹⁵N and d¹³C values, while IAEA-S1 and S2 have known d³⁴S values.

When evaluating the system with standards it is advisable to analyze them several times to determine the level of precision that can be achieved. To investigate sample linearity, analyze different amounts of the same standard.

For information on how to wrap standards (or samples) within tin or silver capsules, please refer to the EA manual. Typically, tin capsules are used for combusting samples, while silver is more suited to pyrolysis. Tin burns easily and can elevate the local temperatures to 1800°C, thus aiding the combustion reaction. Silver capsules typically have a much lower oxygen content than tin and are therefore used when measuring in pyrolysis mode.

Evaluation of data

Within the system page the raw data and the chromatograms from each run can be seen in the ‘data analysis’ and ‘data view’ windows, respectively,  however to collectively view (and process) all the runs together it is advisable to group them within a batch.

To add task runs to a batch, simply highlight the runs and under the ‘Tasks’ tab, in the ‘Batch’ section, click on the ‘Add to Batch’ icon (Figure 6-28) where the ‘batch selector’ window will open (Figure 6-28).

Within ‘batch selector’ there is an option to add the tasks to pre-existing batches or to create a new one by clicking on the ‘create new batch’ option in the top right corner. When creating a new batch, the user is prompted to select what type, i.e.  ‘GC’, ‘EA’ or ‘General’ and create a name for the batch (Figure 6-30). The ‘GC’ and ‘EA’ batches are for processing GC and EA data, while a ‘general’ batch is for grouping stability, linearity, background scans, etc. 

Figure 6-31 displays an EA batch called ‘EA standard calibration’, which contains 16 x EA standard runs with d¹⁵N, d¹³C and d³⁴S values relative to the monitoring gas. Further runs can be added from the task list like before or by pressing the ‘Add’ batch icon under the ‘EA’ tab.

To obtain average values and standard deviations for each of the standards, the analyst can group the replicates by highlighting the runs and selecting ‘Set Group’ from the toolbar (Figure 6-32).

The runs that were added to the batch were all standards and these can be assigned as such within the batch to enable the software to compare their measured values with their known.

Firstly the standard data, i.e. known isotopic values, are required to be made available to the batch; this is done by pressing the ‘standards’ icon under the ‘EA’ tab (Figure 6-33). If no standards are listed then they can be included by pressing the ‘+’ icon and typing in the known stable isotope values.

However, it is recommend to store the standard data within the global settings of lyticOS and these can be simply imported by pressing the ‘import global’ icon and selecting the required standards (Figure 6-34).

Figure 6-35 displays the standards and their known stable isotope values. To create or import standards into the global settings, please go to the lyticOS Home Screen > Settings > Global Settings > Processing > Standards.

Once the required standards are available within the batch then each standard run is assigned as their corresponding standard. This done by highlighting the required runs, right-clicking and selecting ‘Set Sample Type…’ from the menu (Figure 6-36).

The ‘sample type’ options will open and the standard can be selected.

The averages and standard deviations for the standard runs are calculated automatically within the batch, however to further evaluate the quality of the data we can compare the measured stable isotope values for the individual components with their expected values using the ‘Calibration & Corrections wizard’ (Figure 6-38).

After pressing the wizard, the analyst is presented with options to either perform a ‘blank subtraction’, ‘drift correction’ and ‘calibration’; for this evaluation, select the ‘calibration’ checkbox only (Figure 6-39) followed by next.

You should see that ‘only create a single calibration group’ is selected (Figure 6-40).

After pressing ‘finish’, the software will automatically compare the measured stable isotope values with the known values to produce calibrated d¹⁵N, d¹³C and d³⁴S values relative to Air, VPDB and VCDT, respectively (Figure 6-41). By hovering over the cursor over the ‘(C1)’ labels in Figure 6-41, calibration curves for N₂ (Figure 6-42), CO₂ (Figure 6-43) and SO₂ (Figure 6-44) can be displayed along with their gradients (slopes), their intercept (or constant) and if there are more than two points, the R squared value (statistical measure of how close the data are to the line). Ideally the gradients should range from 0.95 to 1.08 and the R squared value greater than 0.99.

To establish if the system is ready to analyze samples is very much dependent on the analyst and their specific application. If the answer to the following questions is yes, then the instrument is ready to analyze samples:

• Is the level of sensitivity good enough for the samples / application?
• Is the precision good enough?
• Is the linearity good enough?
• Is the calibration curve as expected?

If not, please refer to the troubleshooting section of this manual.