Silicon is a strong, brittle, tetravalent metalloid found in the group 14 series of elements, and has an electron configuration of [Ne]3s23p2. Silicon is considered relatively inert chemically and has a giant molecular lattice structure. Silicon exists in oxidation states ranging from 4 to -4, and has four electrons in its outermost shell (two s electrons and two p electrons). The most important property of silicon is its use as a semi-conductor and the element has helped to revolutionise the computer and micro-processing industries. Silicon readily donates or shares the four outer electrons in its outermost shells, and this allows it to form many different types of chemical bonds (though it is not known to form multiple bonds). In its purest form silicon also has a negative temperature coefficient of resistance, as the number of free charge carriers increases with temperature.
The most common oxide of silicon is silica (SiO2), a common acidic compound that is widespread in natural systems. Silica is the most common mineral in the earths crust (as quartz or sand), and is also prevalent as a component of the cell walls of diatoms. Oxoanions derived from silicon are called silicates - and are based on a SiO4 tetrahedral, and can also form a series of hydrides called silanes. In many compounds silicon acts like a typical non-metal, and its structural chemistry is no more complicated than that of sulphur or the carbon of carbonates. The difference between silicon and other non-metals however, is the versatility it displays in terms of its chemistry - the simple ionic compounds are only the first step of an elaborate structural chemistry that includes rings, chains, sheets and solid frameworks of interconnected silicon-oxygen groups [Krauskopf 1995]. The reason for this is that silicon, as a metalloid, has measures intermediate between those of non-metals and metals [Krauskopf 1995]. We can see, for example, the metallic type behaviors exhibited in quartz and other silica minerals (bonding strongly with oxygen in a three-dimensional framework), and the non-metallic bonding exhibited in sulphur trioxide (SO3 consists of self-contained molecules having only slight attraction to each other). This dual-capacity can be considered in a more structural sense by noting that the size of the Si4+ ion is intermediate between that of the smallest common metal ions (Ti4+ and Al4+) and the largest multivalent non-metal ions (P5+ and S5+) [Krauskopf 1995]. Thus the size of the silicon ion prohibits it from effectively polarizing oxygen ions (to form anions), or becoming an independent unit, like a metal oxide (due to it's lack of attraction to oxygen ions). Silicon is also one of the very few substances that have a higher density when in the liquid state compared to the solid state and, as a result, silicon expands when it freezes.
Silicon makes up around 28.8% of the Earth's crust when measured by mass (making it the second most abundant element in the crust after oxygen) [Sun 2010]. In spite of this silicon very rarely occurs in it's elemental form in nature, confined to volcanic intrusions and mineral deposits, and is more commonly found as various forms of silica or silicates. Silicon dioxide (silica) can be found in its purest forms as rock crystals such as quartz. In silicate, silicon is also widespread in various geological mineral formations such as those found in the feldspar group. These are commonly occurring minerals found in many granite and sandstone formations. The silicon ions in these minerals are generally always surrounded by oxygen ions at the corners of a tetrahedron. Silicates, as dominant features of the continental crust, are continuously altered by weathering processes, and studies of the silicon cycle have demonstrated that the gross input into this cycle has primarily derived from the continents (around 84%) as dissolved silicate, which in turn is dominated by river transport [Sun 2010]. Eolian transport and the submarine weathering of basalt is are additional sources of dissolved silicate and silicon, with basalt weathering primarily confined to hydrothermal areas of the sea floor [Krauskopf 1995].
The weathering effects of continental silicates have implications for not only the silicon cycle but many of the earths other biogeochemical cycles. A key component when analysing marine productivity and carbon sequestration, the silicon cycle, through the analysis of silicon isotopes can help give an understanding of both the system as a whole, and the relative concentrations of silicon in each sub-cycle. Silicon has numerous isotopes, with mass numbers ranging from 22 to 36, of which three are stable: 28Si, 29Si and 30Si. The variations in the silicon stable isotope composition have been looked at extensively in the past half-century using many different methodologies and analytical techniques. Despite this it remains a relatively poorly defined field, marked by in-complete data sets and fragmented records.
Minerals formed at different temperatures show varying isotopic signals with regards to silicon - for example those enriched in the lighter stable isotopes (28Si) are produced as a result of decreasing crystallization temperatures during formation. Based on these theoretical calculations, scientists have predicted a positive correlation between 30Si enrichment and increasing silicon content in rocks [Sun 2010 - quoting Grant 1954], and isotope analysis is an important component in looking at silicon measurements as a whole. There are very small natural variations in isotopic signatures of silicon, and as a result of this the analytical methods for determining these ratios need to be robust in terms of both accuracy and reproducibility. Many of the current stable isotope ratio studies involving silicon are primarily concerned with low temperature processes [Sun 2010] - biogenic opal formation [Rocha et al 1998], clay formation [Janvier 2006], and the chemical weathering processes found in groundwaters and river systems [George et al 2009].
As mentioned above, silicon can be released in large quantities during weathering of silicate minerals. When the cations are set free during the weathering process however, the Si-O frameworks of the original silicate minerals are in part decomposed and part reconstituted into the structures of clay minerals, so that only a fraction of the silicon is able to find its way into solution [Sun 2010]. As a result of this, silicon released during the weathering of silicate-rich minerals is retrieved in soil solution through clay formation, adsoroption onto secondary oxides, and plant uptake [Opfergelt 2010]. In order to measure this, the stable silicon isotopes found in these matrixes are analysed and measured in terms of the changes in the isotope signature from one sample to the next.
Silicon is also widely found in biological samples, and serves as an important nutrient in the biology and metabolic functionality of both plants and animals. It is a key component of epidermal cell walls and is beneficial for plants growing under biotic or abiotic stress conditions [Terreza et al. 2006]. As an essential element in nutritional growth, silicon fertilizes the seas by stimulating the production of marine siliceous phytoplankton such as diatoms. This group of phytoplankton organisms has an absolute requirement for silicon for cell-wall growth, and contributes up to 75% of the oceans total primary productivity by fuelling both pelagic and benthic food webs [Fripiat 2009]. As a result of this diatoms are a dominant factor in the biological carbon sequestration in the oceans [Sun 2010]. The analysis of silicon isotopes can help reveal both present and past patterns of silicic acid, primarily by diatoms, in surface waters of the ocean [Demarest 2009].
Analytical difficulties in testing silicon isotopes has resulted in a fairly limited amount of available data recorded, however it is becoming a more and more important branch of scientific analysis.
The isotope analysis of silicon analysis is generally free from matrix effects. As a result of this, the isotopic signature of silicon concentrations has the potential to be incorporated into a definitive analytical approach that can provide reference values for concentrations in physiological and pathological conditions [Aggarwal 1994].
Thus, the isotope dilution analysis via mass spectrometry results in data sets that are fundamentally free from constraints associated with the quantitative recovery of the analyte. This is an essential requirement in other analytical techniques that is difficult to achieve with complex biological samples [Aggarwal 1994]. As a result of this Mass spectrometry is a very useful tool in looking at silicon concentrations, being a powerful analytical method in quantified isotope chemistry.
IRMS allows the analysis of most types of bulk sample (whether rock, diatoms or water), and can be used to generate isotopic signatures for the stable silicon isotopes in these matrices. One of the advantages of this method is that the analysis of Si isotopes using fluorination can be conducted alongside oxygen isotope measurements – allowing for cross-referencing of results. Leng  described this procedure of simultaneous measurement using biogenic silica. The pure silica is disassociated into the oxygen and silicon compounds using fluorination and cryogenic separation using cold traps. The results gave a yield of around 70-80% for biogenic silica, and 97-99% for quartz [Leng 2008].
The analysis of plant tissue also can also be measured using the IRMS method – one particular study using plant uptake in rice species in isotopic signatures [Koster 2009]. Rice is known to accumulate silicon, with the uptake of silicon carried out by ‘transporter' genes (Lsi1 - the influx gene, and Lsi2 - the efflux gene). The deposition of silicon in the plant tissue of rice discriminates against the heavier isotopes 29Si and 30Si, resulting in isotope fractionation within the plant, and can be measured using IRMS [Koster 2009]. The IRMS used in Kosters study looked at the 29Si isotope signals to establish the silicon content in the husk and straw to outline the translocation and deposition processes that take place within the plant [Koster 2009]. Following on from this a new preperation method has been formulated (involving Cs2SiF6) has improved IRMS - the drawbacks of which are its complicated and hazardous nature [Brzezinski 2006].
The stable isotope of 29Si has also been studied as a means of analysing silicon signals within natural systems. A 2010 study by Sun et al. looks at the δ29Si signal and the precision achievable using a single focusing multiple collector inductively coupled plasma mass spectrometer (MC-ICP-MS) equipped with a hexapole gas-collision cell [Sun 2010], to enable measurements of this δ29Si at reduced levels. The MC-ICP-MS analytical method has created a wide range of possibilities in terms of developing the measurements of silicon-isotope signals, and has a level of precision that is comparable with the IRMS method. The Sun study in 2010 produced a solid and useful data set using the new MC-ICP-MS method. The study took place over a 4-year period (which is a fairly long-term project when analysing this sort of chemical signal) and showed that the MC-ICP-MS instrument and methodology was capable of providing rapid analysis and offered a dependable long-term analytical method for the analysis of d29Si in purified samples with low silicon concentrations (18 µM Si) [Sun 2010]. The major advantages with ICP source mass spectrometry are that the sample preparation is simplified and less hazardous (even on the bulk samples), and it requires shorter analytical time and smaller sample size compared to gas-source IRMS [Sun 2010]. IRMS measurements also have problems within the sample preparation, where fractionation may occur at several steps. The one drawback of the MC-ICP-MS method used in Sun's 2010 study is that the determination of the d30Si measurements is not possible. Only measurements of the lighter 28Si and 29Si are considered reliable, due to the large interface of 14N16O on mass 30Si. d30Si has to be calculated from d29Si with an empirical relation [Sun 2010].
Looking into more detail with regards to the MC-ICP-MS analysis of silicon-isotopes, the sample preparation methods often include a cation-exchange purification, and previously studies have eluded to the fact that this step was sufficient for geological materials as the occasional enrichment of anionic species would not compromise silicon-isotope analysis [Sander 2009]. Some reports, however, suggest that there are major offsets in the silicon-isotopes data caused by increased sulphur levels in groundwater samples, and in silicified rock samples, where alteration had been accompanied by some sulphur enrichment [Sander 2009]. Sander's study proposed an additional purification step to negate this problem, in removing sulphur deposits from solid samples - suggesting that further development was required in the MC-ICP-MS methodology as chemical purification prior to analysis seems to be necessary to remove potential interferences that could drastically alter the results.
In oceanic systems, the silicon cycle can be analysed by looking at the silicon-isotopic compositions of seawater and sediments. One of the main isotopes looked at is 30Si, and the concentrations of this can be measured through many techniques. In looking at the simultaneous determination of the rates of production and dissolution of biogenic silica in the marine environment papers, in 2009 and 2010 used a new method involving the measurements of the 30Si isotope. Using a dilution technique with a high-resolution sector field inductively coupled plasma mass spectrometer (HR-SF-ICP-MS) [Fripiat 2009 and 2010] looked at 30Si incubation experiments, and natural silicon isotopic measurements. The new ICP-MS method uses a variable amount of seawater (usually 6 litres collected in the euphotic layer), and the measurements of silicon addition of a purification step by cation exchange chromatography) [Fripiat 2009]. The Fripiat papers claim that this new method is quicker and simpler than the thermal ionization-quadrupole mass spectrometry (TIMS) or isotope ratio production and dissolution rates (conventionally this has not been well assessed). The HR-SF-ICP-MS method is tested and illustrated for two contrasting situations. This included the assessment of silicon isotope signatures in waters with low silicon content (which required the mass spectrometry – IRMS discussed earlier). It was found that the sensitivity of this method was more than one order of magnitude better than if the same analysis had been carried out using TIMS.
Trends in the isotopic composition of oceanic silicon in the upper ocean are driven by the production of biogenic silica by diatoms. Demarest 2009 looked into this fractionation during the dissolution of biochemical silica associated with diatoms that were suspended in seawater under closed conditions. Analysis of the δ30Si value using the HR-SF-ICP-MS method was utilized and helped to reveal that the dissolution of biogenic silica produces dissolved silicon with reduced δ30Si values when compared to the parent silica [Demearest 2009].
In biological terms, silicon measurements in plant life are important in looking at nutrient cycling and overall patterns found in biological systems. When looking at the methodology for the analysis and quantification of biologically associated silicon there are several different types of analysis available, each with its own set of advantages and disadvantages. The traditional method of establishing silicon-isotope ratios is by the fluorination of silicon and introduction of the resulting SiF4 gas into a gas-source isotope ratio mass spectrometer (IRMS).
In looking at measuring silicon concentrations in geological systems (terrestrial, riverine, and biological) Opfergelts  looked at the measurements pf silicon released during weathering - in terms of clay formation, adsorption onto secondary oxides, and plant uptake- to study the contribution of biogenic silicon in a soil–plant system involving basaltic ash soils differing in weathering degree under intensive banana cropping [Opfergelts 2010]. The δ30Si signal was determined in various matrices (bulk soils, sand, silt, amorphous silicon and clay) using an MC-ICP-MS Nu Plasma. This can also be further quantified using the germanium/silicon ratios of the dissolved silicon that finds its way to riverine systems. This Ge/Si ratio was calculated after determination of the germanium and silicon concentrations using HR-ICP-MS and ICP-AES respectively [Opfergelts 2010]. In order to quantify this data the ASi fractions were analysed with microscopic counting of phytoliths, diatoms and ashes, which helped to correlate the data sets observed [Opfergelts 2010].
Some recent studies have started to look into the potential of the radioactive Silicon-32 isotope as a tool for paleontological studies. 32Si is produced by the of cosmic ray spallation of argon, and falls to the Earth's surface as precipitation. It is now being considered by many researchers, that the methods of analysing concentrations of 32Si may be used to date siliceous sediments and sponges, groundwater and glacial ice. With a half-life of approximately 140 years, the Silicon-32 isotope is ideally suited to provide paleontological information in the range of 50-1000 years, and may have the potential to fill the dating gap that lies between the chronologies based on the shorter-lived isotopes of 3H and 210Pb, and those studies based on the longer-lived isotopes of 14C [Fifield 2009]. Measurement of 32Si concentrations in these archives is, however, not straightforward. The Fifield study used two methods for analysing 32Si: radioactive decay counting of the activity of the daughter nucleus, 32P, and accelerator mass spectrometry (AMS) [Fifield 2009]. Of the newer methods, one of the most interesting applications could be the analysis of silicon isotopes by AMS, as this AMS method is suitable for analysing ice-core samples, and requires only ca. 1kg of water for measurement. The recent use and application of scintillation spectronomy for the detection of 32Si in sediments, shows the use of extremely selective radiochemical purification procedures. Moving forward it could be considered that the 32Si concentrations are now a standard tool for sediment dating. Despite these advantages the Fifield study in 2009 does seem to highlight some of the limitations associated with the use of this method in using 32Si as a dating mechanism. In both instances it was found that the detection of pushed the boundaries of the technique and highlighted the lack of precision in the 32Si record. There are extremely low natural concentrations in isotopic ratios and the half-life is still not known to a precision of better than +/-10% [Fifield 2009].
Secondary Ion Mass Spectrometry (SIMS) can also be used as an analytical tool for measuring silicon content in samples, however it cannot be used on water samples and is not capable of performing analysis on bulk samples. Instead, this analysis is restricted to the analysis of individual grains. In this respect this methodology for measuring silicon grains allows a more indepth assessment of the different types of grains that constitute a single rock specimen, whereas IRMS and MC-ICP-MS are only capable of providing an overall mean value. The weaknesses with this method are the restrictions in sample that can be used for the analysis, and a reduction in the analytical precision of the method.
In comparing the analytical techniques, the tables below give an overview of the main methods used and the isotopes associated with each method, which shows the domination of the mass spectrometer methodologies, and the confinement of the SIMS analytical method to individual grains associated with thin sections and polished grains.
(1) Douthitt (1982); (2a) Ding et al. (1996); (2b) Ding et al. (2004) 5; (3a) De La Rocha et al. (1996); (3b) De La Rocha et al. (1997); (3c) De La Rocha et al. (1998); (3d) De La Rocha et al. (2000); (4) De La Rocha et al. (2002), De La Rocha et al. (2003); (5) Cardinal et al. (2003) [Taken from www.nature.com/nature/journal/v433/n7024/.../nature03217-s1.doc]
C. L. De La Rocha, M. A. Brzezinski, M. J. DeNiro and A. Shemesh  Silicon-isotope composition of diatoms as an indicator of past oceanic change: Nature; Vol 395, pp 680-683
Demarest, M.S, Brzezinski, M.A and Beucher, C.P  Fractionation of silicon isotopes during biogenic silica dissolution: Geochimica et Cosmochimica Acta; Vol 73, pp 5572-5583
Fifiled, K and Morgenstern, U  Silicon-32 as a tool for dating the recent past: Quaterny Geochronolgy, Vol 4, pp 400-405
Fifiled, K  Isotopic approaches in the silicon cycle: The Southern Ocean case study
Fripiat, F, Corvairier, R, Navez, J, Elskens, M, Schoemann, C, Leblanc, K, Andre, L and Cardinal, D  Measuring Production-dissolution rates of marine biogenic silica by 30Si-isotope dilution using a high-resolution sector field inductively coupled plasma mass spectrometer: Limnology and Oceanography: Methods; Vol 7, pp 470-478
F. S. Grant , Geochim. Cosmochim. Acta: Vol 5, pp 225–242
Janvier, P  Palaeontology: Modern look for ancient lamprey: Nature: Vol 443, pp 921-924
Krauskopf, K and Bird, D  Introduction to Geochemistry: New York, McGraw-Hill
Opfergelt, S, Cardinal, S, Andre, L, Delvigne, C, Bremond, L and Delvaux, B  Variations of δ30Si and Ge/Si with weathering and biogenic input in tropical basaltic ash soils under monoculture: Geochemica et Cosmochimica Acta; Vol 74, pp 225-240
Sun, X, Andersson, P, Land, M, Humborg, C and Mörth, C  Stable silicon isotope analysis on nanomole quantities using MC-ICP-MS with a hexapole gas collision cell: Journal of Analytical Atomic Spectrometry; Vol 25, pp 156-162
Terraza, P, Castillo, G.A, Gonzalez, R. Koashi-Shibata, J, Garza, M and Lopez, C  Comparison of three methods of silicon analysis in a cucumber leaf tissue:
R.B. Georg, A.J. West, A.R. Basu and A.N. Halliday  Silicon fluxes and isotope composition of direct groundwater discharge into the Bay of Bengal and the effect on the global ocean silicon isotope budget: Earth and Planetary Science Letters; Vol 283, pp 67-74
Mark S. Demarest, Mark A. Brzezinski, and Charlotte P. Beucher  Fractionation of silicon isotopes during biogenic silica dissolution Geochimica et Cosmochimica Acta
Vol 73 pp 5572
Melanie J. Leng , Hilary J. Sloane  Combined oxygen and silicon isotope analysis of biogenic silica: Journal of Quaternary Science: Vol 23, pp 313 – 319
Jan Reent Köster, Roland Bol, Melanie J. Leng, Adrian G. Parker, Hilary J. Sloane, Jian F. Ma  Effects of active silicon uptake by rice on 29Si fractionation in various plant parts: Rapid Communications in Mass Spectrometry: Vol 23, pp 2398 - 2402
Aggaral, S.K, Kinter, M, Fitzgerald, R.L, Herold, D.A  Mass Spectrometry of trace elements in biological samples: Crit Rev Clin Lab Sci; Vol 31, pp 35-87