Talk:18-electron rule

Error?
Hello! I may be wrong, but doesn't neutral Cobalt tetranorbornyl have 13 valence electrons rather than the 11 mentioned under "Bulky ligands"? — Preceding unsigned comment added by Livedevilslivedevil (talk • contribs) 16:26, 14 July 2012 (UTC)

To be included
M-M bonding (Mn2(CO)10, Fe3(CO)13 connected to cluster), agostic ligands, NO (linear, bent) and SO2 (planar, pyramidal). And much more.--Smokefoot 21:52, 16 May 2006 (UTC)
 * I think that most of these belong in Electron counting (how to cope with M-M bonds) or Ligand (how do SO2 and NO and NCS behave). --Dirk Beetstra 22:12, 16 May 2006 (UTC)

Examples
I've renamed 'mismatch in ligand and metal orbital energy' to 'coordination compounds' and moved it to the end, so we can discuss organometallic examples that do not obey the 18-e rule first. TiCl4 has gone into this section - it will never obey the 18-e rule, being Ti(IV) and having weak-field ligands, so is a bad example to give under lower electron counts. Perhaps a better example is needed? This section should be compounds that might obey the 18-e rule, but don't, just as the higher electron counts section is compunds that should have 18-e but actually have more. Also, I expanded a bit and tideid up some of the English.--Brichcja 08:43, 23 May 2006 (UTC)
 * I already was afraid we were going in different directions, and well, that is how the Wikipedia should work. Still I believe that TiCl4 is a good example, it should not show only compounds that might obey 18-electron rule, it should give some examples of compounds that are stable, yet do not obey 18-electron count.  I think 'coordination compounds' is not a good title either, because all the compounds discussed are (in a way) coordination compounds.  Furthermore, it should give examples of early and late TM, with an appropriate explanation.  So you could be right, maybe my choice of 'too low electron count' vs. 'too high electron count' is not good.  But to find a path inbetween, we could try and split up to some subsections giving some common reasons for not obeying 18VE, with some examples, e.g. compounds that need too many electrons to get to 18 VE, compounds with a energy-mismatch in orbital-overlap, high-spin compounds that do have a couple of unpaired electrons, thereby the orbitals are not free for overlap with two-electron donating ligands .. what do you think about that approach? --Dirk Beetstra 09:15, 23 May 2006 (UTC)

Yeah, that sounds reasonable. I kinda feel that the 18-e rule is really only meant to apply to organometallic compounds (yeah, I know, what exctly constitutes an organometallic compound?), so maybe we could split it into organometallic and non-organometallic sections, and then split the former into more and fewer than 18 and give reasons? There's no point in having a huge debate about TiCl4 when the rule wasn't meant to apply to it anyway - it says at the top of the page which classes of compounds it works well for.--Brichcja 09:28, 23 May 2006 (UTC)
 * Hmm, no, technically, all TM complexes can accomodate 18 electrons, not only organometallic compounds, just as all main-group elements can accomodate 8 electrons. So, I think it should tell why TM can accomodate 18 electrons, but that there are more exceptions from the 18 VE rule than that there are exceptions from the octet rule (exceptions from the latter should be in octet rule.  And then tell why some complexes just don't get there, or why inorganic and organometallic chemists do their best to make them deviate from 18 VE).  --Dirk Beetstra 10:20, 23 May 2006 (UTC)

No, I completely disagree with that. The 18-electron rule only applies to a relatively small percentage of TM compounds, so I see no point in introducing as a principle for all TM complexes and then pointing out that the majority of them don't obey it. To quote Elschenbroich and Salzer on the 18-e rule (p186, 2nd ed.): 'Thermodynamically stable transition metal organometallics are formed....' (my italics).--Brichcja 11:20, 23 May 2006 (UTC)
 * Hmm, we disagree quite a bit here, I think (and I do 'disagree' (bit a too strong word) with Elschenbroich and Salzer, there are also non-organometallic compounds that obey the 18 electron rule, e.g. the whole set of carbonyl compounds known, and many ammonia and water adducts). I fail to see, there are 5 d-orbitals, 3 p orbitals and 1 s orbital to fill, hence there fit 18 VE.  That compounds do not get to 18 VE, has their reason.  And I do believe that is what should be explained, there is place for 18 VE, and why do compounds exist that have way less, or a couple more?  All of them have a reasonable explanation.  --Dirk Beetstra 13:18, 23 May 2006 (UTC)
 * I really fail to see, why the hexaaqua copper(II) ion [Cu(H2O)6]2+ (21 VE) is separated from Nickelocene and cobaltocene .. just because they are organometallic? --Dirk Beetstra 13:30, 23 May 2006 (UTC)

Of course many non-organometallic compounds have 18-electrons. However, the 18-electron rule (which is what this entry is about) is only meant to apply to organometallic compounds (including the carbonyl compounds). I understand what you're saying, but I don't think you can go extending the rule to systems that it's not meant to cover even though your point is valid. I still think discussing it wrt organometallics is best, and then putting in non-organometallic (counter)examples. Housecroft and Sharp (1st ed., p591): low oxidation state organometallic complexes tend to obey the 18-electron rule. (My italics).--Brichcja 16:03, 23 May 2006 (UTC)
 * I'll wait until some other people give their opinion. I know it is true for organometallic compounds, but it is also true for non-organometallics, the rule does not apply only to organometallics, and the quotes you give do not exclude that.  I just feel that the article in Wikipedia should be general, not specific.  --Dirk Beetstra 16:24, 23 May 2006 (UTC)
 * I recall hearing that the 18e (or EAN) rule goes way back to Sidgwick in the 1930's when the structures of metal carbonyls were unknown. Once students get past the Cp-M-CO/PR3 bit, TiCl4 is a good example. Humbling for sure because it shows the 18e rule is not-even-close.  There is Ti(neopentyl)4, which is organometallic (and apparently no agostic stuff going on either).  One theme would be to develop a set of guidelines for when 18e rule is violated: (i) bulky L's (TiNp4, Co(norbornyl)4), (ii) pi-donor L's (VCl4), (iii) late metals where d8(Ni(II) etc) and d10 (Cu(I)), and (iv) high spin.  I am traveling so I dont have access to Elschenbroich and Salzer, but I think that they enumerate these.  Item ii would be an opportunity to discussion pi-donation from L to M, related the paucity of 18e oxides, fluorides, alkoxides, nitrides - important ligands all.--Smokefoot 17:14, 23 May 2006 (UTC)


 * Dear friends - I made a big edit on the exceptions - probably blew some of my counting and wrote awkwardly. I was attempting to move the topic onward and keep the 18e rule open to all creeds, colors, ethnicities (of complexes, that is).  I will not be offended if this gets revised.  Particularly welcome would be more examples.--Smokefoot 17:54, 23 May 2006 (UTC)


 * I'm an organic chemist, so I can't claim the same expertise as Dirk and Smokefoot, but I am a Bristol graduate so I was exposed to a three years of inorganic/organometallic lectures. My impression is that the 18e rule works well for certain types of ligands (pi-acids?), including the typical ligands found in organometallics.  For related reasons it tends to work well for low oxidation states, so it works perfectly well for things like Ni(PF3)4.  To my mind, there is nothing unique about carbon that should make the rule exclusive for organometallics.    Granted, though that with strongly donating ligands and high oxidation state metals, such as (Ni(H2O)6)2+, it breaks down and you have to rely on crystal field theory.  Please forgive any mistakes of mine, it's been a lot of years since I played with these things (including making ferrocene in my attic at home, in yoghurt containers!). Hope this helps anyway, Walkerma 17:57, 23 May 2006 (UTC)

Looking much better now. I tidied up generally, correcting the English etc, but rewrote the applciations section a bit to emphasise the pi-acid and low-oxidation state sentiments expressed above. I removed hydrides and alkyls from the list of typical ligands - they're not pi-acids, and often don't give 18e complexes (e.g. [WMe6]2-, {[ReH9]2-!). I replaced them with phosphines and olefins, which more often do. In this, I think the application of the 18e rule has been shifted away from specifically organometallics to pi-acid complexes in low ox states.

I think we shpould replace Cp and Cp* with eta5-C5H5 and eta5-C5Me5, but I can't be bothered right now. And also, I don't think "High spin metal complexes have orbitals half-occupied. These orbitals could be filled by lone pairs of electrons from donor ligands, when the ligand is capable of pairing (some of the) unpaired electrons." makes much sense. Anyhow, off to bed now.--Brichcja 22:53, 23 May 2006 (UTC)

Reactivity
Compounds deviating from 18 VE indeed tend to be more reactive, nickelocene is a perfectly stable compound, as long as you don't feed it anything to react with .. just as with ferrocene, ferrocene will also react with oxygen, but it needs some activation.
 * So you agree with what I said. Nickelocene and Cobaltocene react spontaneously with oxygen, but Fc needs activation. --Brichcja 11:23, 23 May 2006 (UTC)
 * Yes, I do agree to the fact that they react without activation to other (in most cases undefined) products, but they are stable compounds, I have seen undergrad students seen making gram quantities of the compounds. So what I mean is, that their deviation of 18 VE makes them more reactive .., but also ferrocene is very easily oxidised to a 17 VE compound (ferrocinium), which is just a stable compound.  --Dirk Beetstra 13:05, 23 May 2006 (UTC)
 * Responding to some comments above. Actually nickelocene is rather slow to react with air - "spontaneous" has a special meaning that eludes me now - but to kineticists it is not about speed, its about order of reaction I think. BTW, it's WMe6 neutral (trigonal prismatic by the way).  [ReH9]2- is 18e, so that's a nifty example.  Somewhere above reference is made to the [Ti(CO)7]2-, which is not on, but some dmpe analogues are known by J. Ellis et al.  The article is nice - a real collaboration.  But we need more examples.--Smokefoot 03:09, 24 May 2006 (UTC)

The next step
Indeed, I think we are getting there more and more, we need some more examples (we have quite a number, already, I guess we need some 'difficult' ones (where the ligands are participating, e.g. bipy vs it's radical anion and dianion, or maybe the pyridine diimine ligand, also important in the electron counting article to show that counting is sometimes difficult or even ambiguous!!), and we need some references.

One remark, there is a statement "late transition metals often violate the 18 VE rule", because they have a d-orbital which is high in energy. I still believe that the same is true, if not more true, for ETM. These compounds tend to be happy with 16 VE or less, some examples:
 * Group 3 compounds behave very ionic, d-orbitals hard to reach.
 * V(III) compounds are d2, if they are 10, 12 or 14 VE, they are often paramagnetic (triplet state, a rare example may be the dinitrogen bridged R3V-N2-VR3 (10 VE), but that may be due to antiferromagnetic coupling between the metal centres, or write the compound as R3V=N-N=VR3, though that is an incorrect notation, if I am correct), but many, if not all of the 16 VE compounds of V(III) are diamagnetic and very stable, because the last d-orbital is quite high in energy, higher than the pairing energy of the electrons (hé, we missed that effect somewhere, that NEEDS a section at least, maybe an article of its own, any true inorganic chemists around?). Guess we need to find a reference for this for sure.
 * In olefin polymerisation with ETM it is often stated, that the compound has to be a) cationic (makes it want an incoming olefin more), b) have a metal-carbon bond (to insert the ethene into), c) it has to have 14 VE or less, and d) have a vacant orbital cis to the M-C. That 14 VE does suggest me again, that it needs an orbital with appropriate energy, and that the other free orbital is (often) not accessible (i.e. of wrong energy) for the incoming monomer, though it is a excessive free orbital ..

It may be that I see things here the wrong way, but well. That's why it is on the talk-page ..

I'll have a look around in my personal library for some nifty references, though I think most will be about electron deficient compounds .. --Dirk Beetstra 07:58, 24 May 2006 (UTC)

As Brichcja already mentioned, the description of making high-spin compounds low spin compounds (I changed it .. but) is not correct. What I mean is, that, e.g., coordinating CO to certain paramagnetic compounds makes all the energies of the ligands lower, resulting in spin pairing and hence diamagnetic (or better, low spin) compounds, I must confess, I don't know how the mechanism here goes (I guess the CO has to dock first, and hence, there has to be room for that, before the orbitals go to low spin, can somebody please try to rewrite that into something more scientifically correct?? --Dirk Beetstra 09:26, 24 May 2006 (UTC)

I think what we need is a spectrochemical series page - there's link to a ligand-field page that also doesn't exist, and high vs low spin and pairing energy etc could go there Compounds that obey the 18-electron rule are all low-spin!. Apologies for [ReH9]2- - I obviously can't count!--Brichcja 10:16, 24 May 2006 (UTC)

I agree to that, but maybe we could start that from this page, or from the ligand page (or start the ligand-field page with that list). Make a list of ligands sorted by field strength, and put in a small explanation why that list is important. The pages low_spin and high_spin do exist .. oh, wait, they redirect to ligand field theory .. that may be what we are looking for, then the spectroscopic series is better on the ligand-page (I'll make a redirect from ligand field to ligand field theory) --Dirk Beetstra 10:32, 24 May 2006 (UTC)

have a look at the crystal field theory page - there's a lot there already.--Brichcja 19:03, 24 May 2006 (UTC)


 * I did see the page, and I noticed the spectrochemical series was on both pages. As it is on spectrochemical series, by the way.  It makes sense to me that ligand has a table with ligands with more info in it, and the spectrochemical series is a neat way to sort them.  We'll have to see how crystal field theory and [spectroscopical series]] (the latter is also a logical place for the same table, but I would not go to look for ligands on that page, I would type 'ligand' in the Wikipedia search engine) develop.  By the way, Brichcja, you are right, the page is very good, though there is more to tell.  But it is a perfect starting point, at first glance I have no objections, maybe only that I miss some pictures, it would clarify it even more.  The examples given are nice starting points as well, see if we can come up with some more difficult examples as well (what about ferrocene). --Dirk Beetstra 22:41, 24 May 2006 (UTC)

Field strength
I have made a start with a table on Ligand containing ligands, sorted by Field strength. I don't have much time during the day, I will try to go on with it, please feel free to do some as well, I will save regularly. --Dirk Beetstra 13:27, 24 May 2006 (UTC)

32 electron rule
I propose removing the paragraph about the 32 electron rule near the top. There's no such thing - the only hits for this on Google come from this article (ie, whoever wrote it made it up), whereas you get millions for the 18 electron rule. I don't even think the concept is valid, as lanthanide ions (especially) show little if any covalency in their bonding, so they never have any electrons in their outermost s or d orbitals. If anything, it would be a 14 electron rule as the only valence orbitals (in the ions) are the f-orbitals. I challenge anybody to show me a compound that obeys this rule. Any thoughts?? —The preceding unsigned comment was added by Brichcja (talk • contribs).
 * I can agree partially to that .. they in general have no effect, except for a couple of elements of the actinide series (uranium, thorium). Maybe we should drop the '32 electron rule' (which as such does not make much sense), but I think there should be some 'disclaimer' about the lanthanides and actinides, which do fill up to 32 electrons.  --Dirk Beetstra T  C 20:33, 26 September 2006 (UTC)
 * No, they don't fill up to 32 electrons. Give me an example!--Chris 20:46, 26 September 2006 (UTC)
 * Err .. you are right .. but Radon and Uuo (can't write that one full out, though) do .. But indeed, it does not make a difference (it may be when g starts filling up, but we are not there (yet)). OK .. remove the sentence .. never looked at it that way .. Good thinking (don't know if I should feel ashamed, now ..)!  --Dirk Beetstra T  C 21:13, 26 September 2006 (UTC)
 * Rewrote the paragraph, I think it is not it yet, certainly needs some tweaking. Could you have a look?  --Dirk Beetstra T  C 21:30, 26 September 2006 (UTC)
 * OK. I moved it to the section on applications of the 18-e rule, and took out the bit on the lanthanide contraction because it's not really relevant here.--Chris 09:09, 27 September 2006 (UTC)

This is expected to be a thing, but only in the early actinides around uranium where the 5f orbitals are relativistically destabilised to the point that they participate on the same footing as 6d, 7s, and 7p. Double sharp (talk) 15:39, 14 July 2016 (UTC)

Some references suggested by Gerard Parkin
Prof. Ged Parkin sent me an email with some references which, he thinks, could be used to upgrade this article. Please have a look:


 * J. Chem. Educ., 2006, Vol 83, p 791
 * A chapter in 'Classification of Organometallic Compounds', 1.01, 2007
 * Polyhedron 2004, 23, 2879-2900.

(He is co-author in all three of them, it seems best that I suggest them here). --Dirk Beetstra T C 15:22, 1 February 2008 (UTC)
 * The numbers of publications on this concept are numerous. Lots of textbooks cover this material, and such sources seem more helpful and accessible to the general reader. J. Chem. Ed. is more difficult than most journals to access on-line.  I welcome his including a self-citation, esp from an insightful scientist such as Parkin.--Smokefoot (talk) 15:02, 2 February 2008 (UTC)

Comment removed
I removed this hidden comment from the article: In practice, of course, orbitals cannot directly accept electrons, otherwise one would encounter ions such as Fe10− and Pt8−. However,There are also some higher-energy anti-bonding orbitals). The complete filling of these nine lowest-energy orbitals with electrons, whether those electrons originate from the metal or from any ligands, is the basis of the 18-electron rule. If it is correct, perhaps it should be included in the article. Biscuittin (talk) 17:53, 29 December 2011 (UTC)