Is silicon a covalent or ionic compound

Lecture: Silicate Chemistry

The comparison of the compound formation of silicon and its lighter homologue, carbon, with non-metals shows that due to the binding energies of carbon, C-C or C-H bonds are preferred, whereas in silicon the Si-O bonds are preferred.

In most cases 4, less often also 6 occur. The tetrahedral coordination sphere, which a sp3-Hybrid (with proportions of π-π bonds between O-p and Sip and very limited also SidConditions, see below). A distance is characteristic of this coordination number dSi-O from 162 pm. The coordination number 6 is also found less frequently, i.e. octahedral environment (hypervalence as in SF6) with a typical distance dSi-O of 177 pm (pressure-distance paradox!). This coordination occurs with Si-O high pressure phases, with Si phosphate, some Si complexes (e.g. with catechol as a ligand) and in SiF62- (i.e. with strongly electronegative ligands).
Due to the general structural principles, the linking of SiO4Tetrahedra with different numbers of corners, is coordinated in silicates by either two (bridging) or only one (terminal) Si atom.

For the Si-O bond (ionic formulation according to Si4+ and O2- versus covalent formulation) results from Pauling E.g. for quartz an ionic character 1-e1/4 (χA.B.) of 15%.
For the in silicates results

  • assuming a purely ionic bond from the sum of the ionic radii according to Shannon a value of 176 pm (r(O2-) = 140 pm and r(Si4+) = 36 pm).
  • The compute Si-O bond length, which is pure assuming covalent bond from the sum of the covalent radii of 66 pm for O and 117 pm for Si, is even 183 pm.
  • Observed however, a value of approx. 162 pm, i.e. a lower value than calculated according to all concepts.
These short bond distances dSi-O can only be explained by a considerable proportion of double bonds (see below).

The Si-O distances within the silicates do not vary very much, but there are clear tendencies: Gibbs and Brown (1969) already have a correlation between the coordination number of oxygen (calculated with Si and all others A.Cations) and the Si-O bond length, which is shown in Fig. 1.2.1: According to this, the Si-O distance increases with the coordination number of oxygen, in SiO2 the distance is shorter than e.g. in orthosilicates, where the O atoms have high overall coordination numbers.

Fig. 1.2.1. Correlation of the mean Si-O distance with the total coordination number of oxygen (according to Brown and Gibbs)‣SVG
If one only looks at the coordination of O against Si, a clear tendency also becomes clear: Terminals (CNO= 1) Si-O distances are always slightly shorter than distances to bridged oxygen atoms.
The observed averages 140 O, with the range between approximately 110 and 180O fluctuates. It should be noted, however, that the Si-O-Si angles, which are mostly determined by X-ray, are falsified in not a few cases by the special positions of the oxygen atoms! The double bond portion of the Si-O bond can again be used to explain this O-Si-O angle, which is significantly widened compared to the tetrahedral angle.

The discussion of oxido-ortho anions (orthosilicate [SiO4]4- but also phosphate, sulfate, perchlorate, etc.) on the basis of current theoretical calculations now gives a reasonably reliable picture, although depending on the calculation method, the proportions of π-π bonds between O-p and Sip or Sid-States is discussed (which also affects the 'allowed' spellings for the valence line formulas of the anions). This is shown in Figure 1.2.2. schematically shows the MO scheme of the orthosilicate ion.

Fig. 1.2.2. Schematic MO scheme of the orthosilicate ion‣SVG
The involved atomic orbitals of the central Si atom are Sis (a1, in red) and Sip (t2, in blue) states that are occupied by four electrons. The d-Orbitals form an e in the tetrahedron (i.e.z2- and dx2-y2-Orbitals, in green) and one (t2, in blue) Orbital set. The s-Orbitals of the four oxygen ligands combine to form a totally symmetric a1 and a t2-Sentence. These orbitals are so deep in energy that they are not involved in the bond (high s-p-Separation at O). The 12 p-Orbitals of the four O atoms, which are occupied by 20 electrons (4 + 4 = 16 + 4 for the charge), form SALCs of the irreducible representations a1, e, t1 and 2 x t2.
The resulting molecular orbitals of the anion are:
  • The deep a1- and t2-Sentences of the O-s States are non-binding (8 electrons).
  • The a1- and t2-Molecular orbitals (from O-p and Sis/p-AO) form the σ-bond between Si and O (8 electrons).
  • The e and t2-Molecular orbitals arise from linear combinations of the O-p-States (main components) and Sip- (t2, in blue) or Sid- Atomic orbitals (small proportions). The lowering of these two molecular orbitals compared to the SALCs of the O ligands on the right represent the π components of the Si-O bond.
  • The nonbinding orbital set t1 from O-p-SALCs form the highest occupied state (6 electrons).
It goes without saying that such a local molecular orbital model (if only because of the considerable charge, practically impossible to calculate!) Cannot correctly reproduce the bonding relationships in a solid.

The results of a DFT band structure calculation (fully relative FP-LAPW method) by BaSiO3 (a synthetic chain silicate) show the considerable Sid-Contents (highlighted in gray) below the Fermi level, however, also very clearly (see tDOS (above) and pDOS of Si and O (below) in Figure 1.2.3.).

Fig. 1.2.3. Total and partial density of states in BaSiO3‣SVG
The illustration of the electron density from this calculation (Fig. 1.2.4.) Shows that the bond-critical point (saddle point of the electron density) is very close to silicon and the value for the electron density at this point is only very low, the ionic bond component is very clear after this calculation.
Fig. 1.2.4. Valence electron density in BaSiO3
The most important cations in silicates and their usual coordination numbers (CN) with respect to oxygen are listed in Table 1.2.1.
cation CN versus O
Be2+ 4
Li+ 4, (6)
Al3+ 4,6
Mg2+ (4), 6
Fe2+, Ti4+ 6
N / A+, Approx2+ 6, (8)
K+ 6-12
Tab. 1.2.1. Cations in silicates

Thereafter, the cations of Be2+, Li+ and Al3+ Replace silicon on the tetrahedral sites (so-called isomorphic replacement). Aluminum has a double role, it either acts as a 'real cation' (aluminum silicate) or it replaces silicon in its places (aluminum silicate). This dual role of Al makes the interpretation of the formulas of the corresponding Si / Al compounds difficult. The cations have an inductive effect on the Si-O bond, i.e. the cations, as electron donors, reinforce the character of the Si-O double bond. As a result, the bridge bonds become longer (mean value: 163 pm), while terminal (terminal) Si-O-Si bonds become shorter (mean value: 158 pm) - corresponding to a greater proportion of double bonds. The following empirical rule describes the corresponding bond length variation:
dSi-O = 157.9 + 0.15 * CNO

water is almost always built into silicates as real crystal water. There are only a few hydroxy silicates with [SiOx(OH)y] Tetrahedra are known. This fact, too, greatly facilitates the understanding of the structural chemistry of (hydrated) silicates.

Due to the relatively high proportion of ionic bonds, the structures of the silicates can A.x[SiyOz] also often as ion crystals from O2-Anions and Si4+- and A.n +-Cations are described (concept of, KKPs). In this case, those apply to ion crystals, which in a simplified manner reflect the minimization of the potential energy in ion crystals. The Pauling rules almost always apply to silicates. (See also chapter 4.2. of the lecture Inorganic Structural Chemistry on the Pauling rules).
  1. The coordination number in ion crystals is determined by the ratio of the radii of the ions involved. The distances themselves follow the sum of the ionic radii. More precisely: After the introduction of polyhedra around the cations (definition of the coordination number) and when assigning ionic radii to the individual particles, the coordination number (CN) is determined by the radius ratio, the atomic distances are determined by the value of the ionic radii.
    A coordination polyhedron is formed around each cation. The distance between cation and anion is determined by the sum of the ionic radii, while the coordination number is determined by the radius ratio.

    Examples: The best-known example is the sequence of structure types (ZnS, NaCl, CsCl) for simple salts. In the case of silicates, due to the ionic radii of Si4+ and O2- directly the coordination number 4 for silicon versus oxygen. The values ​​for other cations correspond to those given in Table 1.2.1 above and are shown graphically in Fig. 1.2.5. shown.

    Fig. 1.2.5. Radius ratio rule and typical cation coordination‣SVG
    If different cations are present in a compound, then it may not be possible for all of them to find ideal relationships. In this case, the cations with the lowest charge and the largest radius switch to other coordination numbers. For example, sodium is found in the sodalite Na8[Si6Al6O24] Cl2 not as usual with the coordination number 6, but only with a coordination number of 4, while Si (higher charge!) remains in CN 4.
  2. For each cation i with the charge Z and the coordination number CN, the 'electrostatic bond strength' S resultsi to:
    S.i=(Z e)/CN
    A stable ion lattice exists when the charge X e of the anions corresponds to the sum of the bond strengths of the coordinating cations:
    X = Σi si
    The summation takes place via the i-cations around the respective anion.

    Or as text:

    The valence of an anion in a stable ionic structure tries to compensate for the strength of the electrostatic bonds of the surrounding cations (and vice versa).


    1. Perovskite CaTiO3:
      The following applies to the individual cations:
      • Ca: Z = +2; CN = 12, i.e. Z/CN = 1/6
      • Ti: Z = +4; CN = 6 i.e. Z/CN = 2/3
      Since oxygen (Z = -2) is coordinated by two Ti and four Ca atoms, S = 4 * 1/6 + 2 * 2/3 = 2. That means that the charge of the O2- is exactly balanced.
    2. Spinel: MgAl2O4
      The following applies again to the cations:
      • Mg: Z = +2; CN = 4, i.e. Z/CN = 1/2
      • Al: Z = +3; CN = 6, i.e. Z/CN = 1/2
      Since oxygen is coordinated by one Mg and three aluminum ions, the charge is also here from O2- balanced ((1 + 3) * 1/2 = 2)
    3. Grenade (A.32+B.23+[SiO4]3):
      Based on the coordination numbers, the following results for the cations (using the grossular example):
      • Ca (A): Z = +2; CN = 8, i.e. Z/CN = 1/4
      • Al (B): Z = +3; CN = 6, i.e. Z/CN = 1/2
      • Si: Z = +4; CN = 4, i.e. Z/CN = 1/1
      Since the oxide ions are surrounded by two Ca, one Al and one Si cations, it follows for these anions:
      2 * 1/4 + 1 * 1/2 + 1 * 1 = 2 (q.e.d.)
    4. In the case of the silicates (further calculations will follow for some structures) there are maximum deviations from the electrostatic valence rule of only 1/6. In the case of silicates, this rule can be used to determine the Al positions in aluminosilicates. It is also used to distinguish between O2-, OH- and H2O positions and to determine the oxidation states of ions (e.g. Fe2+/3+).
  3. In the case of salts with several cations, those with a high charge are installed as far away from each other as possible so that the cations can be shielded from each other as well as possible. This means that the cation coordination polyhedra should have as few polyhedron elements as possible in common.

    Or as text:

    Splitting of edges and especially of surfaces between coordination polyhedra reduces the stability of a structure. This effect is particularly pronounced for cations of high valence and low coordination number.


    1. In the case of silicates, no examples of common surfaces between tetrahedra and no examples of common edges between tetrahedra are known. The SiO4-Tetrahedra are always either isolated or connected to one another via common corners.
    2. The comparison of the series Silicates [SiO4]4- -> phosphates [PO4]3- -> sulfates [SO4]2- shows that because of the increasing charge of the cation in this series, phosphates usually only appear with a few common corners (often isolated, as dimers or at most chains) and that in the case of sulfates there is no tendency for the [SO4] Tetrahedron is observed.
    As an extension of the 3rd Pauling rule, several cations with a high charge should be spatially separated from one another, i.e. they should have as few polyhedral elements as possible in common.

    Or in the text

    In a structure with multiple cations, cations with high charges evade parts of components.

    The consequence for aluminosilicates is the so-called Löwenstein rule, according to which two aluminum atoms are never built into the tetrahedral structure next to one another.

  4. Overall, as few coordinations as possible are implemented.

    In the text:

    The number of different components in a crystal structure is small.
In general, it can be seen that three-dimensional associations are more stable than two-dimensional ones and that these in turn are more stable than isolated assemblies.

As with other ion crystals, instead of the concept of KKPs, a consideration as (oxide ions), in which the Si and A.-Cations occupy the tetrahedral and octahedral spaces, very useful. This applies in particular to compounds such as orthosilicates, where the covalent description is based on differently condensed SiO4-Tetrahedron does not interfere.