As you can see, Si is the second most abundant element in the earth's crust. Si is found chemically bound to oxygen and associated with the next six elements. Consequently, the chemical preparations of samples for Si analysis will have to contend with the presence of these elements as well in many cases. The Na, Mg, and K do not present as much difficulty as Al, Fe, and Ca.
Uses — Si is arguably the most useful element (C, H, O not considered). It is used in high-tech electronics, medicines, construction materials, and is necessary for living plant and animal organisms. Although no attempt will be made here to mention or discuss its many uses, it is important to advise the analyst to be prepared to handle a wide variety of sample types containing Si in the ppb to wt. % concentration levels.
Chemical Properties — The Group 14 elements have an electronic configuration that favors a 4-valence state. As you move to the higher atomic number elements in this group, the 2-valence state comes into play with Pb and, to a lesser extent, with Sn. In the case of Si, only the 4-valence will be considered. Si bound to oxygen is the most common chemistry encountered in the sample preparation schemes for Si.
Si Chemistry as Practiced & Observed at IV
Inorganic Ventures can and may use elemental Si0 as the starting material, with a documented purity that is claimed to be 99.999+% using ICP-MS and ICP-OES. However, we find that the handling of Si0 is plagued by contamination that cannot be avoided during the grinding and sieving required to reduce the Si0 to a workable particle size. Therefore, we almost always use fumed Si as SiO2, which readily dissolves in dilute, electronic-grade HNO3/tr HF, producing a 99.99+% solution free of metallic impurities for each lot (measured using ICP-MS and ICP-OES) and reported on the Certificate of Analysis.
Sampling and Handling
When preparing for a Si determination, it is critical to take precautions against contamination, with the task becoming increasingly difficult as the Si content decreases. In fairness, Si is not the most difficult element to sample for while avoiding contamination (this I have awarded to Na and Ca, who share this distinction); however, it is difficult at best to generate a blank that is sufficiently reliable. The problem is that Si is present in air particulate and laboratory apparatus that varies according to geographic location, weather conditions, and type/prior uses of the laboratory apparatus.
Silicon will typically be found bound to oxygen. It occurs only in this combined state in nature. It is a characteristic constituent of igneous and metamorphic rocks and of the sediments derived from them.
For details on sampling and subsampling, see Chapter 3 of the Inorganic Ventures Trace Analysis Guide.
The Metal, Oxides, Minerals, and Zeolites
Metal — One gram of high-purity Si0 can be dissolved in 30 mL of HNO3/HF/H2O (1:1:1). The more crystalline the Si0 and smaller the surface area, the slower the dissolution. Warming is OK, but vigorous heating is not recommended (without validation — see below), due to the loss of H2SiF6↑, which is roughly as volatile as acetone. It is important to note that the volatile fluoride does not form in the presence of water. This volatile form of silicon fluoride forms readily when HF is added to concentrated, hot H2SO4 and/or HClO4.
The addition of H2O2 is effective in the dissolution of Si alloys. The use of H2O2 is only recommended if the intent is to avoid HNO3 altogether. It is important to remember that Si and its alloys are dissolved in HNO3/HF, and H2O discourages the formation of volatile silicon fluorides. The confirmation of "no Si losses" should always be validated by preparing a sample of pure Si or SiO2 using your sample preparation procedure.
Oxides, Minerals, and Zeolites by Fusion — I have found few, if any, sample types that cannot be dissolved using either HF in combination with other mineral acids, or a Na2CO3 fusion.
I have used the carbonate fusion for the sample types that defy acid digestion, such as SiC. The following is a general guide for carbonate fusions:
- Make certain that the sample is well mixed with the sodium carbonate.
- A 5-9's pure sodium carbonate is recommended and available from E.M. Science.
- Mix the sample with the flux at no more than a 1:20 ratio. Typical sample-to-flux ratios are in the 1:10 area.
- If organic matter is present, either the sample is mixed with the flux initially and heated slowly to 500 °C for ~2 hours before bringing up to full temperature, or the sample can be pre-ashed at 500 °C and then the ash mixed with the flux.
- Use Pt as the crucible container material.
- Perform the fusion at 1000 °C in a muffle furnace. Avoid flames, since this fusion is difficult to perform in a flame due to the high melting point of the sodium carbonate.
- Most fusions are complete in 15 minutes, and some require up to 45 minutes.
- Dissolve the fuseate in dilute HCl (1:1).
Oxides, Minerals, and Zeolites by HF Acid Digestion — The work that Inorganic Ventures has done in the sample preparation of Si by HF acid digestion has been applied to a wide variety of sample types. This guide, available from the Guides and Papers section of the IV website, details this relatively fast sample preparation method.
Testing Methods
Measuring Si by ICP-OES and ICP-MS can be quite challenging without the proper preparation or knowledge of how your samples were prepared. It is common for HF to be present in samples containing Si, which can cause contamination from glass components of ICP sample introduction systems. This type of contamination is detailed in Chapter 6 of the Inorganic Ventures ICP Operations Guide. HF-resistant sample introduction components are commercially available and can significantly reduce this issue. HDPE bottles can also be a source of Si contamination and should be avoided if measuring low concentrations of Si. Analysts should run Si-free blanks to establish instrument baselines, and closely monitor washout (even the HF-resistant sample introduction system parts can cause Si washout problems at high concentrations).
We recommend testing for Si on ICP by diluting samples down under 10 µg/mL, as this will not require any HF to be added to create a stable sample dilution. See Chapter 1 of the Inorganic Ventures ICP Operations Guide for more information on elemental and matrix compatibility.
Si is generally not a sensitive element when tested by ICP-MS; however, we have found that measuring 28Si while utilizing a H2 reaction gas can produce linear calibration curves down to 1 µg/L.