λ-convention

The whole edifice of substitutive nomenclature is based on concept of parent structures, most importantly parent hydrides. Implicit in parent hydrides are the valencies, or bonding numbers, of non-hydrogen atoms. The standard bonding numbers of neutral atoms in parent hydrides are given thus [1]:

Bonding number n	3*	4	3	2	1
	B	C	N	O	F
	Al	Si	P	S	Cl
	Ga	Ge	As	Se	Br
	In	Sn	Sb	Te	I
	Tl	Pb	Bi	Po	At

These bonding numbers correspond to the number of hydrogen atoms in mononuclear hydrides for elements of Group 13 to Group 17.

Ah, if we only had to name compounds containing just carbon, hydrogen and oxygen. As soon as we move beyond, the trouble starts. Consider phosphorus. We know from school chemistry that this element has valencies of 3 and 5. For trivalent phosphorus compounds we can create substitutive names based on the parent hydride PH₃, phosphane (a). But how to deal with pentavalent phosphorus?

(a)	(b)

phosphane (preselected) phosphine (retained) λ5-phosphane (preselected) phosphorane (retained)

Well, one can create yet another parent hydride name where bonding number differs from standard. For example, the (hypothetical) molecule PH₅ (b) is known as phosphorane. This does not seem to be such a great idea though as there’s a lot of elements that have variable valencies. According to the Blue Book [2, P-21.1.2.1],

The names ‘phosphorane’ for PH₅, ‘arsorane’ for AsH₅, and ‘stiborane’ for SbH₅, are retained for use in general nomenclature. However, the names ‘sulfurane’ for SH₄, ‘selenurane’ for SeH₄, ‘iodinane’ for IH₃, ‘persulfurane’ for SH₆, and ‘periodinane’ for IH₅, which have been used in recent literature, are not recommended.

More general approach, known as “lambda convention”^†, is to modify the existing name of the hydride with the symbol λⁿ, where ‘n’ is the bonding number [1]. So the structure (b) is named λ⁵-phosphane.

Now let’s have a look at some organic molecules. Consider the stucture (c). As its additive name, benzonitrile oxide, suggests, it is a derivative of benzonitrile.

(c)

benzonitrile oxide (substitutive + additive) (benzylidyneammoniumyl)oxidanide (substitutive) benzylidyne(oxo)-λ5-azane (substitutive)

Can we name it using purely substitutive nomenclature? Yes we can, using oxidane as parent hydride, but the resulting name, (benzylidyneammoniumyl)oxidanide, is rather cumbersome. And it is not just its length. Here, benzylidyne, ammoniumyl and oxidanide refer to the ≡CPh group, –N⁺≡ group, and HO⁻ anion, respectively. So the whole name sounds as that of an anion, while (c) is a neutral molecule. One can argue that it simply reflects the way the structure (c) is drawn.

Very well. Let’s redraw the structure in such a way that the charges disappear^‡:

(c′)

By applying λ-convention, we can come with a different substitutive name, viz. benzylidyne(oxo)-λ⁵-azane, where benzylidyne, oxo and λ⁵-azane correspond to the ≡CPh group, =O group, and hypothetical NH₅ parent hydride, respectively.

The same principle applies to heterocycles containing heteroatoms with nonstandard bonding numbers:

(d)	(e)

thiophene 1-oxide (additive) 1H-λ4-thiophen-1-one (substitutive, PIN) 1-oxo-1H-1λ4-thiophene (substitutive) 1-hydroxy-1λ3,2-benziodoxol-3(1H)-one 1-oxide (Hantzsch-Widman + fused ring + substitutive + additive) 1-hydroxy-1-oxo-1λ5,2-benziodoxol-3-one (H-W + fused ring + substitutive) 1-hydroxy-1λ5,2-benziodoxole-1,3-dione (H-W + fused ring + substitutive)

For the structure (d), I prefer the additive name, thiophene oxide, to the unwieldy 1H-λ⁴-thiophen-1-one, just because the former is shorter. On the other hand, for the structure (e) I would avoid 1-hydroxy-1λ³,2-benziodoxol-3(1H)-one 1-oxide because it mixes substitutive and additive nomenclatures: one =O group is named ‘one’ and another ‘oxide’. On top of that, λ³ descriptor for clearly pentavalent iodine looks weird. Similarly, in a substitutive name 1-hydroxy-1-oxo-1λ⁵,2-benziodoxol-3-one the first =O group is ‘oxo’ and another ‘one’. The name 1-hydroxy-1λ⁵,2-benziodoxole-1,3-dione is the most compact (and logical) one.

Simetimes the introduction of λⁿ descriptors may look like an unnecessary trick. For instance, 1,3,5,2,4,6-triazatriphosphinine (f) cannot be considered a parent hydride of apholate (g) for a simple reason that (f) does not have any hydrogen atoms. Solution? A pair of hydrogens is added to each phosphorus atom to generate a parent hydride, 1,3,5,2λ⁵,4λ⁵,6λ⁵-triazatriphosphinine, only to be taken away when substituted by aziridinyl groups, giving 2,2,4,4,6,6-hexakis(aziridin-1-yl)-1,3,5,2λ⁵,4λ⁵,6λ⁵-triazatriphosphinine. Dropping those unpronounceable “lambdas” from that name won’t result in any ambiguity: 2,2,4,4,6,6-hexakis(aziridin-1-yl)-1,3,5,2,4,6-triazatriphosphinine is a name that a coordination chemist would come up with anyway. Still a mouthful, but every little helps.

(f)	(g)

1,3,5,2,4,6-triazatriphosphinine (H-W) 1,3,5-triaza-2,4,6-triphosphacyclohexa-1,3,5-triene (replacement) apholate (trivial) 2,2,4,4,6,6-hexakis(aziridin-1-yl)-1,3,5,2λ5,4λ5,6λ5-triazatriphosphinine (H-W + substitutive)

*	In the original publication on lambda convention [1, Lm-1.2], the standard bonding number ‘3’ was given to boron but not to the rest of Group 13; I allowed myself to complete the table.
†	It was never explained in the original publication [1] why the Greek letter λ (lambda) was chosen.
‡	Whether one could (or should) draw pentavalent nitrogen was discussed years ago on this blog in relation to nitro group.

References

1. Powell, W.H. (1984) Treatment of variable valence in organic nomenclature (lambda convention) (Recommendations 1983). Pure and Applied Chemistry 56, 769—778.
2. Favre, H.A. and Powell, W.H. Nomenclature of Organic Chemistry: IUPAC Recommendations 2013 and Preferred IUPAC Names. Royal Society of Chemistry, Cambridge, 2014.