Talk:Lone pair

Lone pairs vs. valence electrons
What's the difference between the two terms? --pie4all88 08:08, 24 October 2006 (UTC)
 * Valence electrons are the electrons that can participate in a reaction on an atom, in general those in the outer sphere. So a nitrogen atom has 5 valence electrons, when you construct ammonia from that atom, then three of those valence atoms are paired with three valence electrons on the hydrogen atoms (one on each hydrogen atom) giving a covalent bond. The other two valence electrons on the nitrogen are not reacted, these form a lone pair.  If you now would combine a molecule of borane (BH3) or a H+ to this ammonia, that electron pair (that lone pair) is donated to the borane respectively hydrogen cation, forming a dative covalent bond.  Hence, lone pairs are valence electrons, but valence electrons are only lone pairs, if they reside on an atom (and form a pair), but not used in a bond.  I hope this clarifies.  --Dirk Beetstra T  C 08:50, 24 October 2006 (UTC)
 * Ah yes, thank you very much, Beetstra. I'll try to update the article accordingly.  Thanks!  --pie4all88 09:11, 24 October 2006 (UTC)

Merge proposal

 * please consider merging Unshared pair (electrons) into lone pair. Identical concepts different wording thanks V8rik 22:07, 8 March 2007 (UTC)

Merge proposal

 * please consider merging electron pair into lone pair. Identical concepts different wording thanks Solejheyen (talk) 11:49, 10 November 2009 (UTC)


 * no merge! not every electron pair is a lone pair! V8rik (talk) 18:17, 10 November 2009 (UTC)

Accuracy of Article
There are serious inaccuracies on this page. For one thing, the VSEPR picture of lone pairs is completely wrong. Water does not have rabbit ears! Eugene Kwan (talk) 14:34, 17 March 2010 (UTC)


 * Hi Eugene, I can find no inaccuracies in the lead and the example section. The rabbit ears in water are modelled after Organic Chemistry John McMurry 2nd Edition page 27. Problem is you really need a graphical editor capable of depicting sp3 orbitals in 3D, can you be more precise about the inaccuracies? V8rik (talk) 18:32, 17 March 2010 (UTC)

Hi V8rik, I can understand your confusion. Take a look at the MO diagram for water. There are definitely no "sp3-hybridized" lone pairs. In fact, the lone pairs are not equivalent at all. Instead, there is one s-type and one p-type orbital. Unfortunately, a lot of textbooks are outdated and still cling to VSEPR ideas, which are now known to be incorrect. Eugene Kwan (talk) 01:45, 18 March 2010 (UTC)


 * I have had a discussion about this some time ago with user Smokefoot and this resulted:Orbital_hybridization. Current textbooks discuss hybridization and many many current scientific articles use hybridization as a simple qualitative tool. My problem is that if MO theory is the way to go why does wikipedia not have any content on the MO description of molecules like methane, benzene or water V8rik (talk) 17:19, 18 March 2010 (UTC)

Well, I don't know...maybe it's too complicated. All I'm saying is that the information here is incorrect. Eugene Kwan (talk) 05:58, 19 March 2010 (UTC)

Also, to say that hybridization has been all but abandoned by inorganic textbooks is inaccurate. It's still a very useful concept. Rather, I think what has been abandoned is the notion that an atom must form _equivalent_ bonding hybrids to all its bonding partners. This is not a question of "hybridization is qualitative," while "MO theory is quantitative." If you get right down to it, both theories can make qualitative and quantitative predictions. I can use VSEPR to make a MM force field or interpolate a bond angle, or I can use high level computations to predict the MOs of something and make a prediction about its reactivity. But that's really not at issue here. Fundamentally, VSEPR treats lone pairs incorrectly because it assumes they are formed from bonding hybrids of equal s/p composition, regardless of the local symmetry of the molecule. A simple argument is as follows. Consider the p-orbitals in formaldehyde. There's a p_x, a p_y, and a p_z. In the most basic picture, one of them gets used for the sigma bond, one of them gets used for the pi bond, leaving one p orbital unused and one s orbital unused, meaning you have a lone pair in an s-orbital and another higher energy one in a p-orbital. Of course, the s-orbital can interact favorably in both the sigma and pi bonding modes, so the bonding hybrids to carbon will certainly have some s-character in them. Necessarily, to conserve the total number of orbitals, this means the s-type lone pair will have some p-character and vice versa. But you can see there's nothing here to indicate two "rabbit ear" type, equivalent lone pairs. Similar arguments apply to the water molecule, and indeed, molecules of lower overall symmetry. There are two ways to think about it. On symmetry grounds, one cannot perturb the sigma/pi-type separation between the orbitals unless there's a reason to do so; i.e., there's a bonding interaction which gets better when an s/p hybrid is used. In all these cases, one finds that while the s-type lone pair has significant p-character, the p-type lone pair is nearly all p orbital. On energy grounds, it's more favorable to keep a lone pair in an s-orbital than a p-orbital, since s-orbitals are lower in energy. So, we don't need VSEPR at all! We just need to know the above fact, which is generalized by Bent's Rule. In the above example, you've got an oxygen. It's forming a bond to a carbon and a lone pair. Think of the lone pair being a bond to a completely electropositive atom. The bonding pair of electrons is partly shared by carbon, which is way more electronegative than the "atom" the lone pair is "bonded" to. So the oxygen will put the lone pair in an s-type orbital, and the bond with carbon gets a bunch of p-character. This is much better than the reverse, which would put more electron density in a p-heavy orbital. Anyways, these ideas are kind of confusing at first, but I think at the end of the day, you can feel a lot better about them. VSEPR always disturbed me. Why the heck are electron pairs "bigger" than bonding pairs? Why should an atom form equivalent hybrids, regardless of what it's bonding to? Just because it would separate ligands optimally in space, at say an interhybrid angle of 120 degrees, what guarantees the bonding will happen that way? One thing to realize is that the atomic orbitals themselves, and by extension their bonding hybrids, look different in different molecules. In cyclopropane, for example, there's no canonical hybridization that will get you to sixty degrees. Even if you use all p-orbitals, you still only get to 90 degrees. So, the bonds must be "bent" in some sense. OK, I talked too much already, but I just wanted to explain where I was coming from. Feel free to ask questions or complain loudly... Eugene Kwan (talk) 13:01, 19 March 2010 (UTC)


 * Thanks, yes the simple picture has its limitations. I am going to find out what the current textbooks have to say on lone pairs and rabbits ears and I will get back on it V8rik (talk) 23:47, 19 March 2010 (UTC)
 * I have checked with some of the general textbooks, they all use VB theory. See citations in my edit. Typically, several textbooks also offer MO theory but only with respect to energies but not geometry. The section may be simplistic but is not inaccurate. V8rik (talk) 21:07, 22 March 2010 (UTC)

Respectfully, a lot of current textbooks are incorrect, as they are based on older textbooks. For an up to date view, I suggest you read Valence and Bonding (Weinhold and Landis). A more classic choice would be Coulson's Valence. A dramatic demonstration of how incorrect the lone pairs is can be found in JACS 1984 106 1018-1025. There, you can see the regions of hydrogen bonding do not correspond to rabbit ears at all. The valence bond model is not wrong, it just needs modification to account for inequivalent bonding hybrids. Eugene Kwan (talk) 21:54, 22 March 2010 (UTC)


 * Thanks for your comment. I think the way lone pairs are described in the current textbooks is a good starting point We then have to move forward to modifications and the more complex models V8rik (talk) 22:18, 22 March 2010 (UTC)

Here is an insightful discussion on the topic. Among other things it shows that the photoelectron spectra of water shows inequivalent lone pair electrons. Laing, M. J. Chem. Ed. 1987, 64, 124. What can I say? Textbooks are wrong. VSEPR theory just happens to work well. It's more or less like putting balloons on balls and sticks. It's bound to get things wrong some of the time. Eugene Kwan (talk) 04:23, 7 April 2010 (UTC)


 * Thanks I will get hold of it, I am updating the MO diagram article and will eventually arrive at water. V8rik (talk) 17:27, 7 April 2010 (UTC)

A few examples in MO diagram theory
I've been looking through info on non-bonding orbitals (n orbitals). According to traditional Valence bond theory and VSEPR theory, hydrogen fluoride (HF) should have 3 equivalent electron lone pairs on the F atom, water (H2O) should have 2 equivalent lone pairs on the O atom (I assume those are the rabbit ears User:E kwan refers to), and formaldehyde (H2CO) should have 2 lone pairs on the O atom, which should essentially occupy n orbitals in Molecular orbital diagrams. Instead, I found MO diagrams which showed HF has only 2 pairs of n electrons, formaldehyde has only 1 pair of n electrons, and water has an unusual mixture of MO orbitals in which there are not 2 equivalent pairs of n electrons. Maybe sometime when I have more time, I will write something more about this. H Padleckas (talk) 11:06, 12 January 2011 (UTC)
 * the water rabbit ears make an appearance in MO diagram, HF is also mentioned V8rik (talk) 21:53, 12 January 2011 (UTC)

...but WHY? (Introduction)
I gather from reading other Wikipedia entries - Lewis acids and bases, for example - that the concept of the lone pair is pretty important in chemistry, at least in some formulations. But I wouldn't get that at all from the intro to this piece, which tells me exactly what they are (in very abstract terms) but gives me no clue what that means in terms of its implications for chemistry and so on. Anyone up for expanding on this? --Oolong (talk) 12:39, 2 November 2013 (UTC)