I do some guesswork remembering my inorganic chemistry study.

We used to characterize the ionicity/covanlency/metal-like behavior of bonds by the so-called electronegatvity triangle (you can see several illustrations in Google). I would say an oxide is metallic if, say the sum X value is less than 1.8 and the delta x value is larger than 0.75. This is, however, only the first approximation, as the electronegativity depends on the oxidation number. http://en.wikipedia.org/wiki/Electronegativity#Variation_of_electronegativity_with_oxidation_number

I would consider Cr2O3 a metal oxide but CrO3 a non-metal oxide.

When you state "metal" vs "non-metal, are you referring to the conductivity? In general, the elements on the left-hand side of the Periodic Table are metals (they conduct) while those on the right-hand side (as I am sure you already know) are non-metals. The electronic criterion is whether the filled shell energy overlap the conduction band of the solid. See my book "Luminescence and the Solid State". Also there are some alkaline earth oxides that exhibit metal conductivity at romm temperature "Enclyclopedia of the Alkaline Earth Compounds" - RC Ropp

Cn you refine the question? Are you asking for the difference between oxides of metal (metal-oxide) and oxides of non-metals? You are excluding non-solid oxides. So, probably electrical conductivity is not the criterion you take. Your example is interesting. We used to debate whether to include Si as metal or as non-metal in metallurgist's parlance. Some more clarity on your question will be useful.

The elements along the "ladder" B, Si, Ge, As, Sb, are usually considered to be "semi.metals". Sn and even Pb in their high oxidation state, as well as Cr. Mn in high oxidation states can be considered as non-metals.

Thanks everybody for comments so far.

We all know that solid silicon is not a metal: it is a semiconductor (resistivity vs. temperature, for example).

However, once silicon atom (or another atom, which in solid state is classified as non-metal) belongs to the "compound with oxygen SiO2", is there any general property that distinguishes that compound with respect to other compounds where you have a metal, instead of Si?

In other words, does the classification "metal" or "non-metal" oxide exactly correspond to the classification "metal" or "non-metal" of the atom bonded to oxygen in the oxide?

As a matter of fact, in the literature you frequently found SiO2 classified as "metal" oxide: is this correct?

Up there was an answer to your question concerning electronegativity and I think this is a very good way to solve the problem. At least we have quantitative criterion to draw a line between metals and non-metals.

From my side I may suggest another criterion: to find out what correspond to this particular oxide - acid or base. if acid, then it is non-metal oxide, if base, then it's metal oxide, if something in between - amphoteric oxide. With this criterion SiO2 is definitely non-metal oxide. And, on the other hand, TiO2 is metal oxide, for example.

I´m agree with professor Gyorgy Banhegyi. But I must admit that the Pauling`s electronegaty concept (and method*) based on the binding energies determined thermodynamically despite its simple and intuitive appearance is one of the most difficult to understand and even harder to use. This would be the motive of some

inaccuracies or difficulties to establish clear limit .The electronegativity concept is defined for each atom involved in a bond inside a molecule. And there are always more than one component in each bond that's depends of the complete molecule structure and its particular atomic arrangement. By example the Si-O bond in pure silica is not the same than in polisiloxane. That is, there are not only ionic and/or covalent contributions. So, each bond are like as a unity of opposites which can not be clearly defined because they are not "pure" at all. However, in my opinion still it is the best criterion to classify the oxides. (By the way "Silica" has more than one crystallographic structure)

*I reccomend you the text "Advanced Inorganic Chemistry" by F. A. Cotton & G. Wilkinson

The problem with electronegativity concept is not that it is complicated but that it is strictly quantitative. There is no qualitative gap between metals and non-metals in this model, it only can be established voluntary and therefore problems with transitional cases are inevitable.

By the way, I've stated so affirmatively that titania is metal oxide as it corresponds to base and only few hours later found old bottle marked "Titansäure". It's all so perplexed!

The acid/base property was mentioned earlier. If you have a M-O-H bond the question is whether it dissociates as M+ and OH- or as MO- and H+ in aqueous environment. In principle such calculations can be perforemd, but only of the hydrate layer and the aqueous medium is taken into account. If the charge of M is high and the ionic radius is small, it will dissociate as MO- and H+ (i.e. as an acid), while if M has low charge and lareg radius, it will dissociate as M+ and OH- (as a base). Amophoterics (semi-metals) will do both, depending on the pH. Electronegativity can be defined in sevearl way, one of them involves ionic radii and chartges - that is the closest to the above environment. I do not thing that the name (metal oxide or non-metal oxide) is important by iteslf if one knows what he/she speaks about.

Emanuele Barborini has brought more clarity to his question. He is asking whether the oxide of Si, i.e. SiO2, can be considered as a metal oxide? Or, should it be considered as a non-metal oxide? i.e. oxide of a non-metal? The answer to this question can be straight. Oxides of metals are basic in nature. The oxides of non-metals are acidic in nature. In metallurgical slags, for example, SiO2 is the most acidic oxide. Hence, SiO2 should be considered as an oxide of a non-metal. The same conclusion can be arrived at if chemical reactions of the oxides in aqueous environments are considered. Hence, it is appropriate to consider SiO2 to be an oxide of a non-metal.

I think that answer of Alexander Vanatsev is the best - bacause it is the simplest. It is the first (Lavoisier's) definition, and as metal/nematal is the concept of the 18th century, this definition is correct. In modren chemistry there is no strict distinction between metals and nemetals, and therefore it is not possible to find a parameter

which would strictly distinguish them.

The simple answer is, as one can also deduce from discussions below, there are no rigorous definitions.

Thank you everybody for useful comments.

Metallic oxides are basic oxides, whereas non-metallic oxides are acidic oxides.

However, for acids and bases, Oxides that react with water giving acids are acids such as B2O3. On the other hand, bases are oxides that react with water giving bases such as Na2O. Other oxides which do not react with water such as SiO2 do react with NaOH, thus showing their acidic properties. Therefore, SiiO2 is considered as an acid and it belongs to non-metal oxides.

We can investigate the problem also form the Lewis concept: oxides with electron pair deficiency are Lewis acids, those with electron pair excess are Lewis bases.

FOLLOW