Standard Operating Procedures

Solid Sulfur Isotope Analysis


This document outlines how to run Shrek for sulfur isotope analysis of solid materials using the Eurovector elemental analyzer. Shrek, a Finnigan MAT253, is set up as a continuous flow isotope ratio mass spectrometer for the analysis of carbon, nitrogen, and sulfur isotopes. It has two elemental analyzers (EAs), a custom low-N setup, and two peripheral / mass spectrometer interfaces (Conflo IIIs). Use the Terse Procedure section if you already have a good idea of what you are doing. Use the Exhaustive Protocol section for more thorough descriptions. This flow diagram may be helpful.


We add a combustion aid called vanadium pentoxide to each sample tin. Vanadium pentoxide is particularly hazardous and you must read and sign it's chemical SOP.

Appropriate precautions should be taken to protect yourself against the high temperatures of the elemental analyzer and the reagents used therein. Users should wear safety glasses and leather gloves when changing the insert. Preparing the insert and the combustion / reduction / desiccant columns should take place in the hood and users should wear nitrile gloves. Users are expected to know the contents of the MSDS for each of the ingredients of the columns and know what to do in case of exposure or spillage.

Terse Procedure

  1. Weigh samples and standards leaving the tins OPEN – target quantity for S is 50 µg
  2. In a fume hood, add vanadium pentoxide to each tin
  3. Still in a fume hood, close up the tins as you normally would
  4. Clean up vanadium pentoxide spills appropriately
  5. Make certain the previous run has finished
  6. Note the N2 background with the dilution off
  7. Consider if a new insert, new combustion column, or new desiccant should be installed before proceeding
  8. Load samples
  9. Complete sequence table
  10. Purge all atmospheric air out of autosampler
  11. Complete Daily Log – Shrek_DailyLog.xlsx
  12. Start your run –
  13. Use shrekS.m on matlab to reduce your data

Exhaustive Protocol

Weighing samples and standards

Samples and standards are weighed into tin capsules. Consult the microbalance method. When weighing at a micro balance, it is important to keep the tins wide open and reasonably circular at the opening to make it easier to add the combustion aid, vanadium pentoxide, in a later step. The addition of vanadium pentoxide happens in a fume hood. The target weight for Shrek sulfur is more than 50 µg of S per capsule. This target weight is optimal though the instrument can run lower and higher amounts of sulfur. To determine target weights for your sample material, divide the elemental target weight (e.g. 50 µg S) by your percent (e.g. 2 % S in decimal form). This example, then yields 50 µg S / 0.02 = 2500 µg of material.

Here is a table of our common in-house reference materials and the target weights to hit the 50 µg S mark.

StandardMaterialSulfur (%)Target (mg)
Ag2SSilver sulfide12.940.464
ZnSZinc sulfide32.910.182
BaSO4Barium sulfate13.740.437
MALMalachite Lake Sediment0.230

Vanadium Pentoxide

Vanadium pentoxide, V2O5, is a combustion aid that increases the abundance of oxygen immediately around the sample material when we want rapid and complete combustion. Given the toxicity of V2O5, we do not strictly control the amount and thus the ratio of sample/standard weight to Vanadium Pentoxide weight. Rather, we add a visually similar amount to each capsule. The V2O5 reagent does not have any sulfur in it and thus, have a standard amount is less important. The most important consideration is that oxygen is in excess and available during combustion.

Loading vanadium pentoxide
Approximate appropriate amount of vanadium pentoxide

After you have all your samples and standards weighed into tin capsules and you are ready to add the V2O5, find a clean and clear hood (or make a hood so). Don protective nitrile gloves and a lab coat. Place your tray of weighed samples and standards into the hood. Also have in the hood, a second clean tray, the vial of vanadium pentoxide, two tweezers, a micro-balance type scoop, wipes, and the vanadium pentoxide waste container. You will be working at the hood for a while and may choose to have a stool to sit on. With the hood sash lowered so as to act as face protection, open the sample tray and vanadium pentoxide. Place a scoop of vanadium pentoxide into each tin. The draft from the hood may blow some of the vanadium pentoxide into the tray or the hood area. Leave this for later cleanup. Continue adding vanadium pentoxide into each tin until you are finished. Close the vial of vanadium pentoxide and wipe the utensil clean, discarding the wipe into the waste container. Using the tweezers, grab the first tin, remove it from the tray, close it up tightly into a ball, and place it into the clean tray in the exact tray location that it was weighed into. Continue until all tins are closed and transferred to the clean tray. Close, secure, and label the tray as needed.

You have likely spilled vanadium pentoxide into the hood, the tray, etc and have a fair bit of clean-up to do. Make sure you have the waste container with you in the hood. Use paper towels and water or ethanol or spray soap to clean all of the vanadium pentoxide from the hood surface. These paper towels should go into the waste container. All utensils should be wiped clean, disposing of the wipes into the waste container. The 96 well tray likely has loose vanadium pentoxide in it as well. DO NOT USE THE AIR GUN TO BLOW IT OUT. Try to clean the tray by laying out one or two paper towels flat on the bench surface. Squirt ethanol into the wells while hold the tray inverted and over the paper towels. Repeat until the tray is clean. Allow the paper towel to dry in the hood, then place it in the waste container. If you can not clean the tray, place it in the waste container.

We generated several runs of data using our three in-house standards analyzed with and without vanadium pentoxide in 2017. These runs seem to suggest that vanadium pentoxide is not needed. However, we still seem to use it until a more thorough data set is assembled.

Run Sequence

The goal of any run is be able to calibrate measured δ34S values to the VCDT (Vienna Canyon Diablo Troilite) scale and to calibrate SO2 peak area to sulfur quantity.

A typical tray of 49 drops could proceed qtycal_BaSO4 2x, qtycal_BaSO4 1x, qtycal_BaSO4 1/4x, qtycal_BaSO4 1/2x, 3 standards, 8 samples, 3 standards, 7 samples, 3 standards, 8 samples, 3 standards, 7 samples, 3 standards.

Previous Run

Check the Information box at the base of the IsoDat Acquisition screen to see if the previous run has finished. The information box is a play-by-play log of events that updates while the instrument is running. If the instrument was previously running and is finished, you should see a few rows indicating that the sequence finished, the last of which will have a green dot with white check mark, a time stamp, the word 'Sequencer', and the text "Sequence Finished !". It is a good idea to view the samples of the previous run as a first cut at checking for proper instrument functionality. 

Instrument Preparation

Current instrument state - Before opening the instrument up to change the insert, change columns, or load samples, make a mental note of the N2 background and the carrier flow rate. Within Isodat Acquisition, change the gas configuration to N2 by using the drop-down menu in the lower left corner of the screen. Make certain all reference gases are off by clicking any Reference gas dark gray bars in the Conflo window on the left side of the screen. Turn the dilution OFF by clicking the dilution bar in the Conflo window on the left side of the screen. Note the m/z 28 signal and then turn the dilution back ON. The N2 background should be around 20 mV. On the Eurovector EA, press any arrow key to wake up the display and the scroll to the right (right arrow) until you see Carrier. It should be about 180.

Replace the combustion column - The combustion and equilibration columns are both 45 cm quarts tubes and heated from the EA furnace. The combustion column is heated to 1020°C, and is packed with quartz chips and reduced copper (for Sulfur). The reduced copper for S is used to remove excess O2 from the Helium stream. Quartz chips at the top of the combustion column are used as a temperature buffer so the copper does not melt from the heat generated while combusting samples. The combustion column is packed from bottom to top as follows: 1 cm quartz wool, 5 cm quartz chips, 0.5 cm quartz wool, 10 cm reduced copper, 0.5 cm quartz wool, 5.5 cm quartz chips, 0.5 cm quartz wool. The combustion column should be replaced after 80 normal samples. The reduced copper is the limiting reactant. If you adjust the amount of oxygen used per sample, you should try to estimate how long the column will last.

The second equilibration column is heated to 890°C and is only packed with quartz chips. The quartz chips uniformly exchange the oxygen from the sample gas to give all combusted materials the same oxygen isotope ratio in SO2. This allows for variation in mass 66 to be interpreted as variation in δ34S.  This column does not need to be changed.

The Desiccant column is packed with Magnesium Perchlorate to remove all water from the combustion and helium stream. Its placement between the two columns is to reduce the possible influence on the oxygen isotope exchange from H2O in the equilibration column. Desiccant column should be replaced after approximately 200-250 samples have been combusted. After replacing the desiccant column, check flow rate as well as combustion timing with test samples. 

The Gas Chromatography (GC) column is packed and held at either 75°C or 80°C and is used to control the speed at which the gas enters the mass spectrometer. This column generally does not need to be changed. It should be baked out above 100 °C on occasion and, most certainly, after switching from CN. 

Loading Samples

Costech autosampler 3-way valve venting
Autosampler 3-way valve venting
Costech isolation valve
Isolation valve closed

Preparing the carousel - With the isolation valve closed (arrow pointing towards you), vent the autosampler by pointing the three-way valve on the lid towards the vent (away from vacuum). Loosen the three nuts that hold the lid in place (you may need pliers) until the nut assembly hinges and falls down. Open the autosampler lid. Check the carousel for debris and blow out as needed. If you are unable to clean the carousel with it in place, remove the carousel by pulling straight up while grasping the center portion. You can remove the three screws located on the underside of the carousel to remove the base plate. Use wipes with ethanol to wipe the carousel pieces clean. Wipe the body of the autosampler in the same way. If it is still dirty, you can soak the carousel in soapy water and scrub as needed. Reassemble the carousel when dry and replace into the autosampler body.

Costech autosampler shaft and guide pin
Autosampler shaft and guide pin
Costech carousel shaft and guide-pin holes
Carousel shaft and guide-pin holes
Costech carousel that is not properly seated
Carousel not properly seated
Costech carousel properly seated
Carousel properly seated
The pin in the center as well as the guide pin located underneath the carousel must line up and have a tight tolerance. You can manually rotate the carousel in either direction until the pin underneath lines up properly.

Autosampler alignment - Whether you remove the carousel for cleaning or not, you should check the alignment to ensure proper dropping of samples. Ideally the carousel hole is lined up exactly with the base plate such that you can not see the base plate. It is always a good idea to advance the autosampler manually with the Costech control box to ensure proper alignment. Press "MANUAL ADVANCE" on the costech control box and the autosampler will advance one position. If you wish to advance multiple positions, wait 2-3 seconds between each push of "MANUAL ADVANCE". If the carousel is not exactly aligned, you will need to make fine scale adjustments with the "jog" switch. Toggling the Jog switch up (Jog +) will very slowly move the carousel in the positive direction (same direction as the manual advance button), while toggling the jog switch down (Jog -) will very slowly move the carousel backwards. YOU MUST ALWAYS FINISH YOUR JOG ADJUSTMENTS IN THE POSITIVE DIRECTION. If you move the jog switch at all, always finish these fine scale adjustments with one last "MANUAL ADVANCE" to ensure proper alignment.

Rear view of Costech zero-blank autosampler control box
Autosampler control box - rear view
Front view of the Costech zero-blank autosampler control box
Autosampler control box - front view

Load your samples - Using tweezers of your choice, carefully place samples and standards sequentially in the autosampler starting to the right of the hole (which may be position 50; e.g. start loading at position 1). Continue until all samples and standards are loaded. Now is a good time to make final adjustments on your sample capsules if they are very full. If they are touching multiple sides of the autosampler hole, you may want to squeeze them to be thinner and taller.

Close the lid - Wipe the o-ring with your a wipe or your finger to remove any lint or debris. Close the lid and secure the bolts into place, making them finger tight. Rotate the lid valve towards vacuum to pump out the atmospheric air from the autosampler. As it is pumping, you will need to keep tightening the lid nuts, again, to finger tightness. IT IS NOT NECESSARY TO USE A TOOL TO MAKE THE NUTS TIGHTER THAN FINGER TIGHT. Evacuate autosampler for 5 minutes.

Costech autosampler 3-way valve evacuating
Autosampler 3-way valve evacuating
Proceed to next step while waiting.

Purging autosampler

Costech isolation valve
Isolation valve open
Costech helium purge valve
Helium purge valve closed
Costech helium purge valve
Helium purge valve open
Costech autosampler 3-way valve closed
Autosampler 3-way valve closed

After the 5 minutes of evacuation, close the three-way valve. It should now be pointing towards the white board. Open the helium purge valve. Vent helium by gradually turning the three-way valve clockwise and then closing it after the hissing of helium stops. Repeat three-ish times. End with the three-way valve closed and then close the helium purge valve. Open the isolation valve SLOWLY. Wait for helium flow to stabilize (it should be 130-134 mm on the ball flow meter on the north side of Shrek). Change the flow meter / mass spec valve to point towards mass spec. Watch the m/z 28 signal and if it stabilizes below 50 mV turn the dilution off. Keep venting helium using the three-way valve on top of the autosampler (leaving the helium purge valve closed) and watching m/z 28 signal until it stabilizes very near 20 mV.

Eurovector Program

The traditional processing program on the machine is Method 7. This method describes the default pressures of the EA and program settings. The EA should already be in method 7 for regular samples. Check the EA screen using the right hand arrow keys. 

Sequence Table

On the left side of the screen, find the Sequence tab under the Browser window bar to open the appropriate sequence.

Peak Center - The first column is the peak center column and it should have a check in every row. A peak center is accomplished by increasing the speed of ions exiting the source forcing them to sweep from one side of the faraday cup to the other and then by decreasing the speed of the ions forcing them to sweep past the faraday cup in the reverse direction. The software sets the speed of the ions (i.e. high voltage) such that the ions are hitting the middle of the faraday cup.

Amount - Enter sample mass under the Amount column heading. This will allow for the C and N content calculation.

Identifier 1 - Enter the appropriate standard and sample names under the Identifier 1 column heading. If you intend to use the 'shrek' matlab script, make sure to type the reference materials in appropriately and identically (e.g. GA1, GA2, SA).

Identifier 2 / Preparation / Comment - Use these columns to enter ancillary information you wish to stay with each sample or standard.

Method - Make sure the method is what you want it to be. Methods with a 'dilution' tag will typically dilute the CO2 signal by 10. This is typical for modern organic samples so that both N2 and CO2 have reasonable signals. Your choice of diluting should be the same for the entire run. The dilution process fractionates. Treat all your samples and standards the same.

Zeros - Its your choice to include zeros or not. One reason to include them is if you have high backgrounds due to a fresh insert, don't have time to let the backgrounds drop, but want to see the first drop combustion. Include 5-10 zeros after the first drop. Another reason to include zeros is to ensure the mass spectrometer is functioning properly. Here you would include 3-5 zeros near the beginning of the run (probably after the first drop) and again at the end of the run.

Purging autosampler

After the 5 minutes from above, close vacuum, open the helium purge slider valve, vent helium 3 times, open isolation valve, wait for helium flow to stabilize (should be 130-134 mm), change valve to IRMS Conflo, watch the m/z 28 signal and if it stabilizes below 50 mV turn dilution off, keep venting helium from top of autosampler and watching m/z 28 signal until it stabilizes very near 20 mV.

Daily log

Once autosampler is completely purged with helium, complete a row in Shrek_DailyLog.xlsx and save the file. Make sure all the cylinders are on and have sufficient pressure of at least 100 psi except for SO2. This method uses O2, SO2, and He. (They should already be on). On the computer in the acquisition window, switch the gas to N2  and remember the baseline values. Use pre-defined gas configurations for backgrounds.


Highlight rows containing samples and standards if carousel is not full. Click the Start button. Folder Name should be your sample set ID. Leave File Name alone. Export Format should be .csv and File Name should be sample set ID. Click OK.


Move ONLY the .csv file from C:\Thermo\Isodat NT\Global\User\Conflo II Interface\Results\* to S:\data\projects\*\raw\. * is appropriate directory for your run and your project. Create a folder called ‘reduced’ at the same level as ‘raw’. Open matlab and type ‘shrekS’. Type comments, look at figures, etc. A reduced data file was placed in the ‘reduced’ folder.


  1. Computer Crash – If the IsoDat software and / or computer crash, quit everything and reboot the computer. Open Acquisition and turn the source on by clicking the red sun icon in the upper left portion of the window.
  2. High backgrounds – Use the mini gas leak detector to check for leaks. A helium leak will show as red LEDs. If the lid is leaking, repeat steps 6 – 10 for cleaning the lid and o-ring. You can also change to mass 40 (on the middle cup of CO2…as above) and then use the Argon cylinder to spray connections and watch for increases in mass 40. Try to assess if the high backgrounds are due to an atmospheric leak or are due to something internal. High N2 and Ar might indicate an atmospheric leak. High H2O indicates a wet GC column or a saturated Desiccant Column.
  3. Unusually low backgrounds – If you are getting values of twice as low or more in Nitrogen and Argon from the historical values seen in Shrek_DailyLog.xls, there may be a leak large enough to not allow gas to pass into the IRMS. For example, Mass 40 has a 20-30 mV signal when it usually is ~100 mV.
  4. Peak tailing – If your samples having smaller amplitudes and longer scan widths than normal, then they are tailing. This can occur from a full insert, a void in the column, improper combustion, low flow rate or a combination of these ways. The first check would be to see if the insert is filled past the two sets of slits. The second check would be to remove the column and inspect the packing and or possible voids. The third check is checking the combustion of test samples with standards. The last option would be to change the flow rate of the EA.
  5. Poor combustion – This can be seen from peak-tailing or a “memory effect” on samples. This occurs when a sample is not receiving enough oxygen for complete combustion. Peak tailing occurs when a sample sits and smolders leading to a long signal. If a sample does not completely combust, it will finish combusting with the next sample and will affect the resulting values.

Suggested Reading