How Do You Tell When A Compound Will Form A Coordinate Bond?

[Priyadarshini]

How can you tell when a compound will form a covalent bond or a coordinate bond? I know that a coordinate bond is a special type of covalent bond and if during covalent bonding, if the elements taking part do not obtain a noble gas configuration, they for coordinate bonds. But take for example, boron trifluoride. Why does this not form a coordinate bond? Boron is electron deficient and has not obtained a noble gas configuration, shouldn'y it form a coordinate bond with fluorine?

 

[tommyxu3]

I think the point is not whether the element has obtained a noble gas configuration, but how the electrons provided. For your example, the bonds between $B$ and $F$ are all provided by they both. A coordinate bond's two electrons should be both provided by one of the two elements.

 

[Priyadarshini]

I think the point is not whether the element has obtained a noble gas configuration, but how the electrons provided. For your example, the bonds between $B$ and $F$ are all provided by they both. A coordinate bond's two electrons should be both provided by one of the two elements.

But if a compound is given to me and I have to draw its bonding, how would I know if it'll form a covalent bond or a coordinate bond?

 

[cgk]

Priyadarshini, this is a tough question, and you might not find a fully satisfactory answer anywhere. The types of bonding which atoms can do is really not quite as diverse as it is often claimed---in the end what makes atoms attract to each other can really only come from the following effects:

• covalent-type bonding (this has something to do with the kinetic energy, which in quantum systems is lower when electrons are spread out over more space; in covalent bonds this kinetic energy win is large enough to offset the potential energy loss due to electrons getting away further from the cores).
  
• electrostatic interactions (i.e., electrostatic interactions due to atoms/molecules having electronic densities in space which result in net attraction)
• electrodynamic interactions (e.g., the London-dispersion force, which also occurs between atoms and molecules without permanent multipole moments)

In the end, all other kinds of "formal" bonding (e.g., coordinate bonds, metal bonds, hydrogen bonds, halogen bonds etc.) are either variants or mixtures of these basic bonding motives (especially the first two---the last one is much weaker). Coordinate complexes can make a classification especially difficult, since they form a broad spectrum of mixtures of covalent and electrostatic type (and, additionally, they often have covalent bonding motives which are rarely seen elsewhere, e.g., all kinds of strange 3-center bonds). This can make it very hard to predict how the bonding in a complex really looks like without an explicit first principles calculation, and even first principles calculations may not make the result obvious.

In the end the question of whether or not something is a coordinate bond or a covalent bond comes down a lot to interpretation. For example, recently a lot of progress as been made in interpreting various centers normally considered to lie in the realm of organic chemistry (e.g., carbenes, phosphorous, etc) as coordination complexes. For example, see:
http://dx.doi.org/10.1039/C0SC00388C
http://dx.doi.org/10.1002/anie.200701632
http://dx.doi.org/10.1021/ja510558d

 

[Priyadarshini]

Priyadarshini, this is a tough question, and you might not find a fully satisfactory answer anywhere. The types of bonding which atoms can do is really not quite as diverse as it is often claimed---in the end what makes atoms attract to each other can really only come from the following effects:

• covalent-type bonding (this has something to do with the kinetic energy, which in quantum systems is lower when electrons are spread out over more space; in covalent bonds this kinetic energy win is large enough to offset the potential energy loss due to electrons getting away further from the cores).
  
• electrostatic interactions (i.e., electrostatic interactions due to atoms/molecules having electronic densities in space which result in net attraction)
• electrodynamic interactions (e.g., the London-dispersion force, which also occurs between atoms and molecules without permanent multipole moments)

In the end, all other kinds of "formal" bonding (e.g., coordinate bonds, metal bonds, hydrogen bonds, halogen bonds etc.) are either variants or mixtures of these basic bonding motives (especially the first two---the last one is much weaker). Coordinate complexes can make a classification especially difficult, since they form a broad spectrum of mixtures of covalent and electrostatic type (and, additionally, they often have covalent bonding motives which are rarely seen elsewhere, e.g., all kinds of strange 3-center bonds). This can make it very hard to predict how the bonding in a complex really looks like without an explicit first principles calculation, and even first principles calculations may not make the result obvious.

In the end the question of whether or not something is a coordinate bond or a covalent bond comes down a lot to interpretation. For example, recently a lot of progress as been made in interpreting various centers normally considered to lie in the realm of organic chemistry (e.g., carbenes, phosphorous, etc) as coordination complexes. For example, see:
http://dx.doi.org/10.1039/C0SC00388C
http://dx.doi.org/10.1002/anie.200701632
http://dx.doi.org/10.1021/ja510558d

Thank you for your reply!
However, what is a first principle calculation? Also, how can a bond be a "mixture" of these basic bonds? How can a bond both share electrons (as in a covalent bond) as well as transfer electron to form ion which attract (as in ionic bonds)?

 

[cgk]

Priyadarshini, a "first principles" calculation in this context is a calculation which actually attempts to solve (approximately) the many-body electronic Schrödinger equation for the molecule in question. I.e., which determines the electronic structure based on the "first principles" of quantum mechanics instead of applying empirical rules. In practice this refers to both various kinds of wave function methods (e.g., Hartree-Fock, MSCSF, and correlation methods) and, maybe more often, Kohn-Sham Density Functional Theory methods. These methods can calculate the electronic structure of molecules, and there are ways of extracting empirical bonding pictures of the thus obtained wave functions.

Regarding mixtures of bonds: Basically, a hybrid between a standard covalent bond and an ionic bond is what you get if an electron pair is not shared equally between the involved partners. For example, in the HF molecule, the bonding electron pair is strongly localized on the F atom, which gives this molecule both ionic (different net charges on H and F) and covalent (shared electron pairs) character at the same time. This is not at all rare: Many bonds are polar (unless the electronegativities of the involved atoms are very similar), and the question of whether you call something an ionic bond or a covalent bond is often a matter of degree more than a matter of yes/no.

 

[cgk]

btw, I probably should have added this: A "first principles" calculation is not something you do with pen and paper. This is done with quantum chemistry software packages, which are complex scientific applications with often several 100k lines of code, and developed by many people over the range of decades.

 

[Paul Janson]

Here is a simple thing which can help. but it does not guarantee correct . But can help most of the time . If you have a compound like NO₃ ^- . First put the valence electrons of the center atom then the charge . and the add the valency of the ligands one by one . if it exceeds 8 then there may form coordinate bond . for the elements in the second period it should be a coordinate bond for the others it may not.
e.g. NO₃^-
valence of central atom = 5
charge = -1
valence of ligands * 3 = 2 *3
5+(-1)+2+2+2=(5+1+2)+2
a octet + dative bond

 

[Priyadarshini]

Priyadarshini, a "first principles" calculation in this context is a calculation which actually attempts to solve (approximately) the many-body electronic Schrödinger equation for the molecule in question. I.e., which determines the electronic structure based on the "first principles" of quantum mechanics instead of applying empirical rules. In practice this refers to both various kinds of wave function methods (e.g., Hartree-Fock, MSCSF, and correlation methods) and, maybe more often, Kohn-Sham Density Functional Theory methods. These methods can calculate the electronic structure of molecules, and there are ways of extracting empirical bonding pictures of the thus obtained wave functions.
Regarding mixtures of bonds: Basically, a hybrid between a standard covalent bond and an ionic bond is what you get if an electron pair is not shared equally between the involved partners. For example, in the HF molecule, the bonding electron pair is strongly localized on the F atom, which gives this molecule both ionic (different net charges on H and F) and covalent (shared electron pairs) character at the same time. This is not at all rare: Many bonds are polar (unless the electronegativities of the involved atoms are very similar), and the question of whether you call something an ionic bond or a covalent bond is often a matter of degree more than a matter of yes/no.

Then are non polar bonds purely covalent bonds? Is there anyway to tell if the compound is more ionic or more covalent? The properties of covalent compounds and ionic compounds are almost opposites (like one is soluble the other is not), so how does it affect their properties? Like with polar molecules, if it is partially covalent and partially ionic, will it possesses the properties of both covalent and ionic compounds?
Wow! Quantum mechanisms software packages!

 

[Priyadarshini]

Here is a simple thing which can help. but it does not guarantee correct . But can help most of the time . If you have a compound like NO₃ ^- . First put the valence electrons of the center atom then the charge . and the add the valency of the ligands one by one . if it exceeds 8 then there may form coordinate bond . for the elements in the second period it should be a coordinate bond for the others it may not.
e.g. NO₃^-
valence of central atom = 5
charge = -1
valence of ligands * 3 = 2 *3
5+(-1)+2+2+2=(5+1+2)+2
a octet + dative bond

What is a valence of ligands?

 

[Paul Janson]

What is a valence of ligands?

Valency mean what it can get or loose to get octet. For oxygen it's 2 and for chlorine 1 just like that

 

[Priyadarshini]

Valency mean what it can get or loose to get octet. For oxygen it's 2 and for chlorine 1 just like that

Sorry, I meant to ask "what is a ligand?"

 

[Borek]

Sorry, I meant to ask "what is a ligand?"

Come on, you are asking about coordinate bonds and you don't know what a ligand is?

https://en.wikipedia.org/wiki/Ligand