An oxidation number is an idealization of charge distribution on a molecule. Usually, the oxidation number does remain within bounds of +8 and -8. Good and bad approximations of reality are possible, since an oxidation number does not capture covalent bond characteristics. While a given element has a variety of oxidation states possible under the right energetic conditions, there are patterns among most common oxidation numbers observed in halogens and transition metals. Last, the oxidation number of single-element atoms in a molecule, such as sulfur (S8) or nitrogen (N2), is defined as zero.
When describing compounds and predicting reactions, quantifying the charge on each atom is necessary. An oxidation number assigns each atom an “oxidation state” in chemical equations. When combined with the charge conservation requirement, the oxidation number predicts charge on a given atom. The oxidation state is determined by an ideal situation of entirely ionic chemical bonds.
Consider common salt, sodium chloride (NaCl). In terms of oxidation state, the sodium (Na) atom has a +1 charge, while chloride (Cl) has a -1 oxidation state. For sodium chloride, this is very close to reality, since the NaCl bond is nearly 100 percent ionic.
By contrast, water (H2O) is an example of an oxidation number being a helpful but ultimately poor correlation to the reality of atomic charge distribution. Each hydrogen (H) atom has +1, and oxygen has a -2 oxidation state. Since water is a polar-covalent compound, there is a slight positive charge on the hydrogen atoms and a negative charge on the oxygen atom. However, ionic character is smaller in magnitude than +1 or -2 if the isolated electron charge is taken as one unit of charge magnitude.
Poor Approximation Explained
It is not correct to say that each of the hydrogen and oxygen atoms contains or is missing fractions of an electron. Rather, the electron cloud is “thicker” at the oxygen atom than at the hydrogen atoms. Therefore, there is net negative charge at the oxygen end and net positive charge at the hydrogen ends.
In most molecules, halogen elements have an oxidation number of -1. This is close to reality in bonding with nonhalogen atoms. With a -1 oxidation number, valence electron orbitals on halogen atoms achieve full and therefore the most stable state.
The oxidation number on transition metals is observed to vary more than in other element groups. For example, iron (Fe) has common oxidation numbers +2 and +3, corresponding to an ideal subtraction of two and three electrons, respectively. Less common oxidation states like +6 can also occur. Once again, multiple oxidation states happen in other element groups under the right conditions. However, without specialized temperature and/or pressure conditions, transition metals display this characteristic most frequently.
When atoms of one element are bonded to each other, as in fluorine (F2), the oxidation number on each atom is zero. If the oxidation state is maintained at -1, it would lead the whole molecule to have a net -2 charge. Such oxidation number assignment is inaccurate, since fluorine molecules are observed to be electrically neutral. Furthermore, in a symmetrical arrangement of identical atoms, it is difficult to justify asymmetrical electron cloud density that would lead to a nonzero charge on a given atom.