Introduction

Inductively Coupled Plasma Optical Emission Spectroscopy, ICP-OES, is a method to perform elemental analysis on a sample in solution (ODP Technical Note 29). This manual covers hard rock sample preparation for ICP-OES analysis. Hard rock samples are prepared via the 'flux fusion' approach. This technique ensures complete dissolution of sample allowing a full elemental analysis. Solutions are stable which allows further analysis, and involves no HF making. It is a safe and ideal method for shipboard analyses (ODP Technical Note 29).

Rock samples are crushed to a fine, talc-like powder using multiple cleaning, crushing and grinding procedures. After grinding, loss on ignition (LOI) is determined. Ignited material is fused with flux powder forming a glass sample bead. Fused beads are then dissolved in 10% HNO_{3 .} This is further diluted and the resulting solution is processed by the ICP-OES machine.

The complete process (from sample table to ICP-OES machine) takes at least 4 days in the very best scenario:

• Day 1 involves: sampling, polishing on the Diamond Wheel, cleaning (sonic bath), and drying samples overnight.

• Day 2: Crushing in the X-Press, grinding in the Shatterbox, a pre-ignition (for LOI) weight, and placing samples in the muffle furnace overnight.

• Day 3-4: Taking a post-ignition (for LOI) weight, giving ignited samples to the chemistry technician to add sample to the pre-weighed flux, and fusing the sample bead. The beads are then handed back to the chemistry technicians to continue ICP analysis. 

Wait until there are about six or more samples so you can prepare a 'batch' of six or more at a time.

Note: LOI determines the amount of alteration of the sample, generally, scientists use this to determine the 'freshness' of the ingenious rock being analyzed.  Typically we use unignited powder in the bead preparation as that is what correlates to the certified values.  However it is not always the case. Please speak to the geochemists regarding their preference of ignited or unignited powders in the bead preparation. There are many standards (both unignited and ignited) available for the geochemists to include in their dataset.

Apparatus, Reagents, & Materials

Laboratory Apparatus

General Laboratory Equipment

• Compensated Dual Analytical Balance System (Mettler Toledo balances)

• Drying ovens at 110°C and 60°C

• Muffle furnace

• Sonicating bath

Hard Rock Processing

• Splitting room saw

• Buehler grinder/polisher with 70 µm grit diamond grinding wheel

• Sonicator (with small water bath)

• X-Press crusher 

• Spex Shatterbox with tungsten carbide (WC) grinding vessel 

• Spex Mixer Mill with tungsten carbide or alumina canisters 

• Retsch MM400 Mixer Mill

LOI/Bead-Making

• Fisher Ashing muffle Furnace 

• Sample Bead Maker

Reagents

• 0.172 mM LiBr wetting agent (0.15 mg ultrapure LiBr in 10 mL DI water)

• 400mg of drew sighed lithium metabolite flux (weighed on shore)

• 10% nitric acid (143 mL concentrated nitric acid/L of solution). Caution! always add acid to water.

• Isopropyl alcohol, laboratory grade

• Deionized water (18.2 MΩ/cm² laboratory water obtained from Chemistry Lab)

Materials

Grinding Samples

• Beakers

• Glass cleaner

• Tweezers

• Teflon spatula

• X-Press aluminum die

• Core liner pieces

• Delrin plugs

• Sample vials or jars

• Weighing paper, 6 x 6

• Kimwipes

LOI/Sample Bead

• Quartz or alumina ceramic crucibles

• Tongs

• Vials containing 400 mg lithium metaborate flux (preweighed on shore)

• Milligram calibration weighing set

• Weighing paper, 4 x 4

• Vials for excess ignited powder

• Agate mortar and pestle

• Pt-Ag crucibles

Preparing Rock Samples

Rock samples are prepared for ICP analysis using the following procedures on each sample:

1. Cut to size (see Cutting Samples to Size below)

2. Polish (see Polishing Samples on Diamond Wheel below)

3. Clean (see Cleaning Samples below)

4. Dry (see Drying Samples below)

5. Crush (see Crushing Samples in the X-Press in XRD Sample Preparation Hard Rock )

6. Grind (see Grinding Samples in the Shatterbox in XRD Sample Preparation Hard Rock)

Cutting Samples to Size

To cut samples for the X-Press, use the splitting room rock saws (located in the Core Deck) following these guidelines:

• Cut samples to ~1–2 cm in length and width. Avoid cutting irregular pieces; ideal samples are cubes

• Avoid veins, infilled vugs, etc

• Remove as much contaminated material as possible

• Contact the petrologist(s) if cutting reveals unexpected features

Notes about altered samples:

• It may be desirable to hand-pick and separate alteration material such as vesicles and/or veins from whole-rock basalt

• Sometimes veins/alterations do not become apparent until after cleaning and drying.

• Speak to the petrologist about that method if alteration is visible

Polishing Samples on Diamond Wheel

Polishing the samples remove contamination caused by drill bit, saw blade, or other unwanted material. Grind each surface (each outer side) on a high-speed, diamond disc.

Apparatus and Materials

• Buehler Grinder/Polisher

• Diamond grinding disc

• Sample Beaker

Each sample will correspond to one beaker; collect as many beakers as needed. Beakers need to be cleaned (DI water and isopropyl alcohol) and labeled (ex. 1, A, or sample label).
Next start the grinding process using the Buehler grinder located in the X-Ray preparation area of the Thin Section Lab (Figure 1).

Note: The diamond disc is attached to a magnetic disc which is then placed on the wheel plate. Diamond disks are located on the stud shelf in the ICP preparation area of the Thin Section Lab. Use the silver (125 µm mesh) or purple (220 µm mesh) diamond disc.

To start the polisher flip the 'On' switch in the back (Figure 1A). The power indicator light (Figure 1B) should illuminate. Press the timer on/off button (Figure 1C) to get continuous disc rotation. Turn the water on by flipping the 'Water' switch to the 'On' position (Figure 1D). The water flow can be adjusted by turning the knob shown in Figure 1E.

The rotation speed is controlled by the dial shown in Figure 1F. The range is 0 rpm to 500 rpm. 150 rpm is a good starting position. Adjust the speed if needed: faster for hard rocks and slower for softer rocks.
With the water on and the RPM adjusted press the 'Run' button (Figure 1G). Move the sample back and forth across the disc to prevent making a groove. If there is an emergency press the 'Emergency Stop' button (Figure 1H) to stop rotation and cut off the water. To enable the wheel again twist the knob until it pops back out.

Polish the rocks with a Buehler diamond disc (Figure 2) until they're completely smooth and round on all edges. The purpose of grinding on a diamond disc is to remove any possible contaminates caused by the drilling mud (or drill bit) or the rock saws in the splitting room. There should be no pits or jagged corners. Put the rock into a labeled beaker (Figure 3). Before polishing the next sample, clean the diamond disc with a dressing stone. Dressing stones are found in the Splitting Room of the Core Lab. Do this for all samples and then move on to 'Cleaning Samples'.

Open

Cleaning Samples

To remove contamination (oil, skin, and residue from the diamond wheel) wash the polished samples in 70% isopropyl alcohol and DI water. From this point onward, wear gloves when handling samples to avoid reintroduction of contaminants.

Apparatus and Materials

• Sonicator with basket

• Beakers

• 70% Isopropyl Alcohol

• Tray(s)

Place the basket in the sonic bath and fill to just above the wholes in the basket with DI water. Pour enough isopropyl alcohol (70%) into each beaker to cover the sample. DI water can also be used if the scientists prefer. Place the beakers in the sonic bath basket. There should be enough liquid in the beaker to keep the sample from floating in the basket (Sonic Bath without basket Figure 4).

Open

 Sonicate the samples for 15 minutes. You should notice the isopropyl or water becoming cloudy from residue being shaken off the samples. Then follow the wash sequence below:

1. Decant as much of the liquid as possible

2. Sonicate again with DI water for 10 min

3. Decant liquid into the sink

Repeat the rinse cycle until the water is clear. If the samples are soft and/or clay rich, they will not reach the "clear water" state. Continuing to sonicate will only dissolve the sample. If after 3–4 washings, the water still isn't clear, go to the next step. After the final rinse, decant as much water from the beaker as possible.

Drying Samples

This step requires at least 12 hours and therefore should be done towards the end of your shift. Turn oven on (Figure 5) and adjust the dial to about 110°C (this is marked on oven, or you will have to turn on oven a couple hours before and use the thermostat inside to get the temperature correct. It is important to not overheat the sample as it may affect some minerals. Temperatures less than 110°C is ok, but it may take longer than 12 hours for the sample to dry.

Apparatus and Materials

• ICP Oven

• Samples

Note: The ICP oven should be kept clean at all times, as samples are left open and are susceptible to contamination. The ICP oven should only be used for ICP samples. If the oven shows any sign of rusting, please notify the ALO as a new oven will need to be ordered.

Place the beakers into the ICP Oven at 110°C for at least 12 hours (Figure 5). You can cover the beakers with a larger clean beaker if there is any possibility of rust or debris falling onto the sample. 

Open

After 12 hours the samples should be dry, remove the beakers from the oven and place them inside the desiccator (Figure 6) while you prepare the X-press station.

Crushing Samples in the X-Press

See XRD Sample Preparation Hard Rock 

Grinding Samples in the Shatterbox

See XRD Sample Preparation Hard Rock

Determining LOI

Loss on Ignition, or 'LOI', compares a mass measurement taken before and after a sample is subjected to extreme heat. Petrologists use LOI as an indication of degree of alteration. Low LOI values suggest relatively fresh, unaltered basalt; whereas high LOI numbers suggest alteration (clay, alteration minerals, etc.). LOI is determined by weighing a small amount of the sample (~5 g) before and after ignition. Samples typically lose weight as water is driven off, though an iron-rich, water-poor sample may gain weight. 5g of sample is suggested to reduce the %LOI error to around 1% or less. Advise the scientists if they want less ignited.

LOI is not required for all types of ICP Preparation. Check with the science party to determine if LOI is a desired measurement. If the science party does not want an LOI measurement move on to the section Making the Sample Bead.

Loss on Ignition

Determining a sample LOI comprises three procedures:

• Pre-ignition weighing

• Igniting samples

• Post-ignition weighing

Advice on LOI Procedures (from Exp. 366 Methods, from Exp 393/Jeff Ryan)

Shipboard sample preparation and LOI determination procedures described in Murray (2000) and updated in recent IODP Proceedings volumes (e.g., Reagan et al. [2015] for Expedition 352) are appropriate for a range of sediment and rock compositions, but some care must be taken with unusual sample matrixes. As an example, attempting sample ignitions on carbonate-rich materials can lead to spurious results and issues with contamination if quartz crucibles are used for sample ignitions because carbonates will react with quartz upon heating to both devitrify and decompose the crucible. Alumina ceramic crucibles may be better for carbonates but risk contamination for Al and potentially other elements due to spallation over time. Putting carbonate rich samples or sediment in quartz crucibles ruins the crucibles and the LOI determination (consulted with Jeff Ryan on this, Exp 393). Consult with the scientists to be determine what crucible will be needed.

Maximum ignition temperatures of 1000°C and higher are appropriate for ultramafic and some mafic igneous materials but may result in sample sintering and/or sticking to some Si- or Ca-rich materials. Ignition temperatures of <850°C are inadequate to decompose carbonate minerals in sediment samples, even if samples are held at temperature for several hours. In general, igniting samples to at least 900°C as a maximum temperature is advisable to decompose all volatile-bearing phases and obtain reliable measures of LOI.

Pre-ignition Weighing

Apparatus and Materials

• Mettler Toledo Dual Balance

• Acid Washed Quartz or Alumina Ceramic Crucibles

• 4x4 Weigh Paper

• Reference Weights

• Thermolyne Muffle Furnace

Clean the balance area including the balance plates inside the balance. Any dust or particles on the plate could throw off the weight measurements. Before beginning make sure there are enough acid washed crucibles for your samples. The crucibles are located in the X-Ray lab in a plastic container labeled 'Acid Washed Crucibles'. Wear gloves while handling the crucibles.

Place a large sheet of paper in front of the balances and place supplies here. For each sample you need weigh paper (Figure 28A), a scoopula (Figure 28C), and a crucible set (Figure 28B). Clean the scoopula with isopropyl alcohol in between each sample as it has direct contact with the sample powder.

Setting up the Mettler Toledo Balances

Samples are weighed on the Mettler-Toledo Dual Balance. The Dual balance uses two weighing stations to compensate for shipboard motion: one a 'known' reference weight (Figure 29A) and the other an 'unknown' sample weight (Figure 29B). The balance takes a series of measurements and uses the average value as the final weight (for a more in-depth guide refer to the Balance User Guide on Cumulus). Each balance has a control panel plate, which constantly records weight. These plates communicate with the "Mettler Balances" program.

Open the Mettler Balances program (Figure 30). There are multiple panes and parameters that are set before we start measuring.

Open

Graphical Measurement Window: Shows a line graph of the live-time measurement weight and the running average weight.
Final Weight Panel: Displays the Final Weight after all measurement counts have been made.
Statistics Panel: Shows the average weight adjusting with time.
Counter Weight: Enter in the reference weight

Counts: The elapsed amount of measurements
Tare: Shows and applies the tare or 'zero' value.
Sample ID: Name the sample being measured.
Commands Panel: Executable commands
Weigh: Starts measurement
Tare: Determines the 'zero' weight. This value is applied to the final weight.
Halt: Stops a measurement before it has gone through all counts
History Panel: Shows statistics on all measurements taken. This file can be exported into an excel file by using the 'Export' button. Note: The 'Export CSV' file does not work.
Options Panel: Editable measurement parameters. We measure using the 'Counts' feature. 'Counts' is active when the dot is blue. Change the number of counts or measurements the balance takes here.
At the beginning of a series of measurements, tare the balances. To do this first make sure that the 'Counter Weight' field is set to '0' and then set the 'Count' value. The 'Count' is dependent of the sea state: 600 for calm waters and 1000 counts for rough waters. If seas are too rough than wait until the weather settles before continuing to measure.
A rule of thumb is that the measurement of a known reference mass shouldn't have a larger deviation than the accuracy desired. For example, our accuracy is +/- 0.05 grams; weigh a reference mass in the unknown balance that is close to the masses you are measuring (e.g., 25 grams) and perform the measurement with the appropriate counterbalance mass in the reference balance pan. You should get a final mass of 24.95—25.05 grams.
Once parameters are set, select the 'Tare' button. When the tare is complete the 'Final Weight' Section turns orange and the 'History' Section updates (Figure 31).

Put in a reference weight into the "Reference' balance. With the tweezers, select the 20g weight and place it in the center of the 'Reference' balance (Figure 32). To have a more accurate measurement, the reference weight should be close to the expected 'Unknown' sample weight (roughly ~20g). Enter this reference weight into 'Counter Weight' in the Statistics panel.

Open

Weighing Crucibles

 The quartz crucibles have three sections: an outer (or large) crucible, an inner (or small) crucible, and a lid (Figure 33). If alumina ceramic crucibles are needed, as for carbonate rich samples or sediment samples, replace the inner quartz crucible with the ceramic crucible.

Open

The inner crucible holds the sample material and is the only piece that is weighed. They are assembled as seen below with the inner crucible inside of the outer crucible and the lid sitting over the entire unit (Figure 34). Crucible sets (large, small and lid) are engraved and lettered and should be kept as a set. For example, Crucible lid 'A' should always be run with large and small crucible 'A'. If a crucible is unlabeled use a diamond-tipped pen to etch in an unused lid.

Open

Weight measurements are recorded in an excel spreadsheet which will be uploaded to LIMS at the end of an expedition (Figure 35). Open the excel spreadsheet titled 'LOI Template' found in Local Disk > DATA and save the spreadsheet in Local Desktop > XRD-ICP Prep Documents >LOI spreadsheets.

Open

The spreadsheet has multiple columns to fill in.'SITE', TEXT ID', 'CORE/SECT/INTERVAL', 'CRUCIBLE ID', 'CRUCIBLE WT', 'CRU+FRESH SAMPLE WT', and 'CRU+IGN SAMPLE WEIGHT'. 'SAMPLE WEIGHT', POST IGNITION LOSS', and '%LOI' are calculated values based on the weights entered in columns E – G.
The first measurement taken will be the initial weight of an empty crucible. To complete this measurement, open the side door and place an empty inner crucible in the center of the 'Unknown' balance (Figure 36). Record the number or letter etched onto the crucible in the excel spreadsheet under 'Crucible ID.'

Close the door and click 'Weigh'. Wait for the counts to finish and then record the 'Final Weight' in the spreadsheet under 'Crucible Wt'.

Weighing Sample

Weigh out 5 grams of sample powder into the crucible within +/- 0.05g (Figure 37). The total weight should be the crucible weight + 5 grams within +/- 0.05 grams. For example, a crucible weighs 14.32g, thus the total weight plus the sample will be between 19.27 – 19.37g. We use 5g because it is a good representation of the sample, to decrease the %LOI error to about 1%, and it fills the crucible appropriately. You will find that you will loose some of the sample due to it sticking to the crucible when you are finished with the LOI.

 Note: If there is only a small amount of material, you can use less but the %LOI error will be larger. 

Open

When the sample is close to this range click 'Weigh.' Press 'Halt' to stop the measurement and either add or remove sample if needed, and then click on 'Weigh' again to take a new measurement. 

When a sample's final weight is within the allowable range, record the 'final weight' value into the spreadsheet under 'CRU + FRESH SAMPLE WT'.

Carefully remove your sample from the balance. Place your crucible into the larger quartz holder and cover with a lid (Figure 38). Repeat this process for all samples. After all samples have been weighed and recorded, take samples from the desiccator and bring over to the muffle furnace in the Chemistry Lab.

Igniting Samples

Samples are ignited in the Thermolyne Muffle furnace located in the Chemistry Laboratory. The entire ignition cycle takes approximately 20 hours to complete and cool down to ~200°C. After ignition, samples need to be taken out when they come down to ~ 50°C-200°C. If the samples sit for too long they will reabsorb moisture and the 'Post-Ignition Weight' will be inaccurate. Time this accordingly.

Using the Muffle Furnace

Bring crucibles over to the Muffle Furnace (Figure 39). There is a wooden tray to assist with the transfer of crucibles from one lab to the other. Turn the power switch on and the control panel will illuminate.

Confirm with scientists, what temperature and how long samples should be run for. Below is a quick reference guide:

A common program is an increase in temperature of 3°C/min to a target temperature of 900°C and a hold of one hour. Then ramp up at a rate of 3°C/min to a target temperature of 1025°C and hold for four hours before cooling. This ramp cycle is already programmed into the furnace and corresponds to 'Program 1'.

Oven Programs: (Note: Once a program has finished, the temperature setting returns to room temperature, ~25°C.)

• Program 1: Ramp up to 900°C at 3°C/min and hold for 1 hour, then continue to 1025°C at 3°C/min and hold for 4 hours.

• Program 2: Ramp up to 950°C at 3°C/min and hold for 4 hours.

• Program 3: Ramp up to 550°C at 3°C/min and hold for 1 hour.

• Program 4: Ramp up to 900°C at 3°C/min and hold for 1 hour, then continue to 1025°C at 3°C/min, no hold.

To check or edit a program see additional guides attached to the furnace itself. It is also possible to run the furnace manually without a ramp up cycle. Discuss with scientists their preference.

Either enter the desired temperature manually or select your program. If you are running the furnace manually, enter the desired temperature; no other buttons or steps are needed. If selecting a program, press and hold 'Run' (Figure 40).

Open

When the furnace finishes it's cycle and cools down to ~50°–200°C, remove the crucibles with tongs or the padded gloves (if cool enough). Put samples back into the wooden tray and store tray in the desiccator. Keep samples in the desiccator and remove one at a time while weighing. It is very important for the samples to not reabsorb moisture so begin weighing as soon as possible.

Post-Ignition Weighing

Post-ignition measurements should be taken soon (within an hour) after removing crucibles from the furnace. Failure to do so will disrupt the LOI values. Reweigh the crucible plus the ignited sample to determine how much weight was gained or lost. Follow the same weighing procedure as in Pre-ignition Weighing.

1. Record the final weight in the excel spreadsheet under 'CRUCIBLE + IGN SAMPLE WT'. The spreadsheet will populate the columns 'Post Ignition Loss' and '%LOI'.

2. The formula used to calculate LOI is:

%LOI = 100 x (weight change during ignition)/ (fresh sample weight).

Note: By convention, weight lost during ignition is recorded as a positive LOI value; whereas weight gained is recorded as a negative LOI value. Report the results to 2 decimal places.

Uploading LOI Data To LIMS

Open the Excel File LOI_Spreadsheet in Local Disk > Documents > XRD > Sample and ICP Prep > LOI > LOI Spreadsheets > LOI_ExpXXX.xlsx

The spreadsheet has three sheets:

• The first sheet is as follows:

• The second sheet is for repeats.

• The third sheet is the format for uploading the data using the spreadsheet uploader (Figure 41).

Put your LOI information into the spreadsheet following the example format. Fill in the following:

• Text ID: e.g., WDGE11258831 (Note: capital letter matters)

• Analysis: LOI

• Replicate: leave blank, unless true replication (not just reanalyzing) of a sample, then begin replicate numbering with '0', then '1', then '2'.

• crucible_number: Crucible ID (e.g., A, B, F...)

• crucible_preignition_mass: crucible mass only (small inner crucible) (component unit: grams)

• crucible_and_preignition_sample_mass: inner crucible with fresh sample before ignition (component unit: grams)

• crucible_and_postignition_sample_mass: inner crucible with ignited sample after ignition (component unit: grams)

• loi_percent: LOI in % (component unit: none; the value is already a percentage)

• reference_mass: sample mass (calculated difference) (component unit: grams)

• comment: specify temperature and time used for LOI (e.g., 950 deg C for 4 hours; used during Exp 391)

Open up the program Spreadsheet Uploader Pinned to the Taskbar (Figure 42). 

Copy and paste your spreadsheet into that uploader. Click the 'Edit' button and 'Validate Sheet'. This checks and highlights any errors that need to be fixed. When the spreadsheet comes up clean, click 'Lims' and 'Upload'. The sheet will turn green when the measurements are successfully uploaded. The data is now in LIMS under Chemistry > ICP-AES Solids >Expanded LOI.

Cleaning the Quartz or Alumina Ceramic Crucibles

1. Wash the crucibles with DI water and a small piece of a scouring pad (no soap).

2. Rinse several times with DI water.

3. Place crucibles in a 10% HNO₃ bath for 12 hr. If in urgent need, the quartz crucibles can be soaked a minimum of 2 hours, but extra flux blanks will need to be made and added to batches that use these crucibles.

4. Rinse the quartz crucibles 3 times with DI water after the acid bath. For the ceramic crucibles, rinse 3 times with DI and then soak them in a large beaker of DI water for 4-6 hours to dilute the acid absorbed by the crucibles.

5. Dry the either type of crucibles in the oven at a maximum temperature of 60°C. For the ceramic crucibles, heat them to 800°C in the muffle furnace for 2 hours and allow to cool to room temperature before removing.

Some crucibles will develop a thin white or red cloudy film, become spotted, or start flaking. If any of these things happen discard the crucible in the sharps container. When the crucible undergoes one of those changes the quartz has started to react at high temperatures, and could start contaminating the samples.