InChI metal-reconnected layer

While I wasn’t looking, the InChI folks implemented the metal-reconnected layer. Isn’t it nice? I discovered it quite by chance thanks to the Beilstein-Institut ChemInfo Labs page. You can see how it works on InChI Web Demo.

Consider ferrocyanide (a):

(a)

[Fe(CN)6]4− hexacyanidoferrate(4−) (additive) ferrocyanide (trivial)

Its standard InChI is:

InChI=1S/6CN.Fe/c6*1-2;/q;;;;;;-4

(1)

The main layer contains two types of entities: 6CN (i.e. six CN molecules) and Fe (one iron atom). If we try to convert (1) back to structure, using the same Web Demo tool, we get six free-floating CN^• radicals and a separate Fe⁴⁻ anion. Ew. But if we tick the “Include Bonds to Metal” box in the Web Demo tool, we have

InChI=1/6CN.Fe/c6*1-2;/q;;;;;;-4/rC6FeN6/c8-1-7(2-9,3-10,4-11,5-12)6-13/q-4

(2)

where the metal-reconnected layer (/r) appears. It looks like an alternative InChI added directly after the standard one, with its own connectivity (/c) and charge (/q) sublayers. In this layer, there is only one entity: C6FeN6, i.e. [Fe(CN)₆]. The string (2) is correctly converted back to the structure (a).

Now let’s look at the structure of a salt known as Prussian Blue (b):

(b)

Fe4[Fe(CN)6]4− iron(3+) hexacyanidoferrate(4−) (additive) ferric ferrocyanide (trivial) Prussian Blue (trivial)

Its standard InChI is:

InChI=1S/18CN.7Fe/c18*1-2;;;;;;;/q;;;;;;;;;;;;;;;;;;3*-4;4*+3

(3)

The main layer contains two types of entities: 18CN (i.e. 18 CN molecules) and 7Fe (seven iron atoms). Converting (3) to structure brings about a horrible mess. With “bonds to metal”, however, we get

InChI=1/18CN.7Fe/c18*1-2;;;;;;;/q;;;;;;;;;;;;;;;;;;3*-4;4*+3/r3C6FeN6.4Fe/c3*8-1-7(2-9,3-10,4-11,5-12)6-13;;;;/q3*-4;4*+3

(4)

In the metal-reconnected layer (/r) we see two different types of entities: 3C6FeN6 (i.e. three [Fe(CN)₆]) and 4Fe. The string (4) is correctly converted back to the structure (b).

Years ago, I was complaining (to the universe) about different InChIs for the same molecular entity, viz. chromate (c—e). I’ve revisited it with the new version of InChI.

(c)	(d)	(e)

[CrO4]2− chromate (trivial) tetraoxidochromate(2−) (additive) tetraoxidochromate(VI) (additive)

Alas, the standard InChIs for the representations (c), (d) and (e) remain different. Try to convert them back to structures: they also are all different and all wrong (all have extra hydrons). Nevertheless, I see a sign of progress: the metal-reconnected layers for the corresponding strings (5), (6) and (7) are identical!

(c)	InChI=1/Cr.4O/q;;;2*-1/rCrO4/c2-1(3,4)5/q-2	(5)
(d)	InChI=1/Cr.4O/q-2;;;;/rCrO4/c2-1(3,4)5/q-2	(6)
(e)	InChI=1/Cr.4O/q+2;4*-1/rCrO4/c2-1(3,4)5/q-2	(7)

Moreover, all three strings, (5), (6) and (7), are converted back to the structure (c).

What about our old friend ferrocene? Depends how you draw it. I’ll stick to the ChEBI’s decacoordinate-iron representation (f):

(f)

bis(η5-cyclopentadienyl)iron (additive) ferrocene (trivial)

The standard InChI for ferrocene is:

InChI=1S/2C5H5.Fe/c2*1-2-4-5-3-1;/h2*1-5H;

(8)

Converting (8) to structure results in two standalone cyclopentadienyl radicals and a neutral iron atom. With “bonds to metal”:

InChI=1/2C5H5.Fe/c2*1-2-4-5-3-1;/h2*1-5H;/rC10H10Fe/c1-2-4-5-3(1)11(1,2,4,5)6-7(11)9(11)10(11)8(6)11/h1-10H

(9)

In the /r layer we see a single entity, C10H10Fe, i.e. [Fe(C₅H₅)₂]. The string (9) is correctly converted back to the structure (f).

One-electron carbon—carbon bond

What is a covalent bond? We learn in school that it is a chemical bond formed by shared pairs of electrons between atoms. The Gold Book provides a bit more careful definition:

A region of relatively high electron density between nuclei which arises at least partly from sharing of electrons and gives rise to an attractive force and characteristic internuclear distance.

Well, now the textbook definition will have to be changed [1]. In the study published this week in Nature [2], Shimajiri and co-authors

report the isolation of a compound with a one-electron σ-bond between carbon atoms by means of the one-electron oxidation of a hydrocarbon with an elongated C—C single bond. The presence of the C•C one-electron σ-bond (2.921(3) Å at 100 K) was confirmed experimentally by single-crystal X-ray diffraction analysis and Raman spectroscopy, and theoretically by density functional theory calculations.

Cf. the length of single C—C bond in diamond: 1.54 Å. In 2018, Ishigaki et al. [3] reported much longer two-electron C—C bond of 1.806(2) Å in a polycyclic hydrocarbon, dispiro[(dibenzo[a,d]cycloheptatriene)-5,1′-(1′,2′-dihydropyracylene)-2′,5″-(dibenzo[a,d]cycloheptatriene)]^* (10c):

Now the same team took the compound 10c of [3] and crystallised it with iodine. In the resulting stable salt (10c^•+)(I₃⁻), the C1—C2 bond lost one electron to I₃.

This is not the first one-electron bond observed (see [2] and references 1—4 therein) but the first involving carbon atoms.

The crystal structures of (10c^•+)(I₃⁻) are deposited with CCDC, entries 2301032 through 2301039.

*

This is the name given to the compound by Shimajiri [4]. ‘Pyracylene’ is a trivial name of cyclopent[fg]acenaphthylene. Saturating the bond of the cyclopentane ring, we get 1,2-dihydrocyclopent[fg]acenaphthylene. In spiro nomenclature, the locants of the second component are primed (and of the third component doubly primed, etc.), thus 1′,2′-dihydrocyclopent[fg]acenaphthylene. Therefore, the fully systematic name should be dispiro[(dibenzo[a,d]cycloheptatriene)-5,1′-(1′,2′-dihydrocyclopent[fg]acenaphthylene)-2′,5″-(dibenzo[a,d]cycloheptatriene)].

References

1. Bourzac, K. (2024) Carbon bond that uses only one electron seen for first time: ‘It will be in the textbooks’. Nature, online ahead of print.
2. Shimajiri, T., Kawaguchi, S., Suzuki, T. and Ishigaki, Y. (2024) Direct evidence for a carbon—carbon one-electron σ-bond. Nature, online ahead of print.
3. Ishigaki, Y., Shimajiri, T., Takeda, T., Katoono, R. and Suzuki, T. (2018) Longest C—C single bond among neutral hydrocarbons with a bond length beyond 1.8 Å. Chem 4, 795—806.
4. Shimajiri, T. (2022) The nature of ultralong C—C bonds: Demonstration of the longest Csp³—Csp³ single bond beyond 1.8 Å and discovery of flexible covalent bonds. Doctoral dissertation, Hokkaido University.

Boron hydride nomenclature

Can we expand the parent hydride naming philosophy much beyond organic chemistry? Not going too far, let’s have a peek at carbon’s immediate neighbour in the periodic table, boron.

(a)

BH3 borane (preselected name) boron trihydride (binary) trihydridoboron (additive)

The mononuclear hydride (a) is systematically named ‘borane’ while neutral boron hydrides as a class are called boranes.

von Hofmann’s footnote

Systematic name formation in chemistry typically happens through compounding, derivation, or mix of both. The semantic modification of a combining form through umlaut-like vowel change as seen in alkanes/alkenes/alkynes appears to be unique. Its origin could be traced to the 1866 publication of August Wilhelm von Hofmann [1]; I probably would never know about it if not for an illuminating blog post by Joe Dixon [2].

In an extended footnote, Hofmann proposed to call the first ten alkanes as follows: methane, ethane, propane, quartane, quintane, sextane, septane, octane, nonane and decane.

Chains and rings

After hours spent looking in my books and searching the internet, I came to the conclusion that chemists talk about chains and rings without explaining what they mean. The only definition I found so far, viz. that of Gold Book, is specific for polymers and seems to be too complex to be used in general chemical nomenclature:

The whole or part of a macromolecule, an oligomer molecule or a block, comprising a linear or branched sequence of constitutional units between two boundary constitutional units, each of which may be either an end-group, a branch point or an otherwise-designated characteristic feature of the macromolecule.

(1)

On the other hand, general dictionary definitions of (chemical) chains are not precise enough. For example, Collins English Dictionary defines chain (chemistry) as

two or more atoms or groups bonded together so that the configuration of the resulting molecule, ion, or radical resembles a chain.

(2)

whereas Merriam-Webster says that it is

a number of atoms or chemical groups united like links in a chain.

(3)

So chain (chemistry) is like a chain. Is it?

Binary pnictogen halides

Here’s a tricky (trick?) question from the final exam of MITx course Introduction to Solid State Chemistry:

Which compound has the higher boiling point, phosphorus trifluoride (PF₃) or phosphorus pentafluoride (PF₅)?

Naturally, you are supposed to figure this out from the first principles, or rather, from some principles taught in this course, not from Wikipedia (or “by googling”, as some put it).

(a)	(b)

The problem is, the “right” answer, PF₃, is actually, factually wrong. Even though this question cost only two points (of 150), a rather animated debate followed the exam.

Those who defended the “right” but factually wrong answer (a) were proposing that what the problem was testing our thinking rather than actual knowledge, and our thinking should have been along the lines of VSEPR model. VSEPR rules correctly predict PF₃ to be trigonal pyramidal and PF₅ to be trigonal bipyramidal. The dipole moment of a polar molecule PF₃ should make it less volatile than apolar PF₅. Those who chose the “wrong” but factually correct answer (b) were arguing that polarisability of larger PF₅ is higher than that of PF₃ and therefore the London dispersion forces in PF₅ would beat dipole-dipole interactions in PF₃. The (a) party were saying that making the answer you’d get by applying principles different to the one you’d get by “googling” is a good protection against cheating. The (b) party were retorting that this is a silly way of protection, that the question asked was what has the higher boiling point, not what could be expected to have the higher boiling point, and that expecting students to come up with the factually wrong answer is not exactly pedagogical.

Truth to be told, the methods of estimating boiling or melting points of materials were simply not a part of this course. The only thing one could do was to determine whether the molecule has a non-zero dipole moment. But there is no way to figure out which effect will be stronger, the increase in dispersion forces or dipole-dipole interactions.

One would think that the physical properties of such simple compounds as binary halides of Group 15 elements (pnictogens) are studied well and long ago. Not really. I tried to compile a table of dipole moments and melting/boiling points for pnictogen tri- and pentahalides, MX₃ and MX₅, using various resources [1—5]. As you can see, there are still many gaps.

Trihalide	μ (D)	mp (°C)	bp (°C)	Pentahalide	mp (°C)	bp (°C)
NF3	0.234	–207	–129	—
PF3	1.03	–151.5	–101.8	PF5	–93.7	–84.5
AsF3	2.59	–6.0	62.8	AsF5	–79.8	–52.8
SbF3	?	290	345	SbF5	8.3	141
BiF3	?	649	900	BiF5	154.4	230
NCl3	0.6	–40	71	—
PCl3	0.97	–93.6	76.1	PCl5	167	160s
AsCl3	2.15	–16.0	130.8	AsCl5	–50	?
SbCl3	2.75	73.4	223	SbCl5	4	140
BiCl3	4.6	233.5	441	BiCl5	?	?
NBr3	?	?	?	—
PBr3	?	–41.5	173.2	PBr5	<100d	106d
AsBr3	1.66	31	221	AsBr5	?	?
SbBr3	2.47	96	288	SbBr5	?	?
BiBr3	3.6	219	462	BiBr5	?	?
NI3	?	–20s	?	—
PI3	~0	61.1	227	PI5	41	?
AsI3	?	141	400	AsI5	?	?
SbI3	1.58	170.5	401	SbI5	79	401
BiI3	?	408.6	~542	BiI5	?	?

d, decomposition
s, sublimation

What, if any, trends can we see?

• The dipole moment of MX₃ grows larger down the group of the central atom M, e.g. μ(NF₃) < μ(PF₃) < μ(AsF₃), and grows smaller down the group of ligand atom X, e.g. μ(SbCl₃) > μ(SbBr₃) > μ(SbI₃).
• As the sizes of both central atom and ligands go up, so do the melting and boiling points.
• As dipole moments go up, so do the melting and boiling points.

Something curious happens, though, when one crosses the phosphorus—arsenic borderline. AsF₃ has a dipole moment of 2.59 debye. As expected, both mp and bp of AsF₃ are, respectively, higher than those of AsF₅. PF₃, however, has much lower moment of 1.03 D. Both mp and bp of PF₃ are, respectively, lower than those of PF₅. Similarly, mp of AsCl₃ is higher than mp of AsCl₅, whereas mp of PCl₃ is lower than mp of PCl₅. Similarly... but no, there are too many gaps in the table “down there”.

Which shows, by the way, that “googling” does not help if the data is not available. For the future, the course authors may consider asking a very similar question about a pair of compounds from “down there”. Thus the whole conflict between the (as yet unknown) “truth” and “expected answer” could be easily avoided.

Since the electronegativities decrease down the group for both M and L, the most polar M—L bond must be Bi—F bond and BiF₃ should have the largest dipole moment. Well I couldn’t find its value. But it is known that bismuth trifluoride has ionic structure, and has the highest melting (649 °C) and boiling (900 °C) points of all binary pnictogen halides. On the other side of the spectrum, we have extremely sensitive nitrogen triiodide. A feather tickle, a loud noise and, I suppose, any attempt to measure its dipole moment will set off an explosive decomposition (see the video below):

2 NI₃ → N₂ + 3 I₂

References

1. Earnshaw, A. and Greenwood, N. (1997) Chemistry of the Elements, 2^nd Edition. Butterworth-Heinemann, Oxford.
2. Nelson, R.D., Jr., Lide, D.R., Jr. and Maryott, A.A. (1967) Selected values of electric dipole moments for molecules in the gas phase. National Standard Reference Data Series — National Bureau of Standards 10, Washington, DC.
3. Cotton, S. (2001) Nitrogen triiodide. Molecule of the Month collection, University of Bristol.
4. WebElements
5. atomistry.com

Visualising hexabenzocoronene

A few years ago, I wrote that we do not know how to draw ferrocene or a nitro group. (Still true.) Is the situation with polycyclic aromatic hydrocarbons any better?

Take hexabenzo[bc,ef,hi,kl,no,qr]coronene, one of the subjects of the single-molecule visualisation study published last week in Science [1]. One way to draw it shown in diagram (a):

(a)

I chose this one (out of many other possible Kekulé representations) because I can reproduce it on a paper napkin (beermat, Post-it note, you name it). If you look carefully, you will notice that the central ring and the six outermost rings are connected with single bonds.

(b)

Continuing the paper-napkin-doodle argument, it is even easier to draw a circle inside of each ring as in all-delocalised representation (b). However, that would not be a preferred diagram from IUPAC point of view [2, GR-6.5]: for example, benzene ⏣ is acceptable but ⌬ is preferred. Moreover, “it is generally not acceptable to use curves in two adjacent fused rings”. Still, I’d stick with circles.

The question is, do I have to draw a circle within each ring? Of course not. If I draw seven aromatic rings and connect the with single bonds as shown in (c), the resulting structure will be the same. In this way, I can even save some ink (graphite, chalk, etc.)

(c)

Without the circles, the six rings that surround the central ring in (c) start to look, well, more empty. Using the noncontact atomic force microscopy (NC-AFM), the team behind the study [1] were able to show (and in this case “to show” really means “to show”), that those rings are indeed slightly larger. The C—C bonds in the central ring (i-bonds, 1.417 Å) are 0.03 Å shorter than the bonds connecting that ring with the six outermost rings (j-bonds, 1.447 Å).

1. Gross, L., Mohn, F., Moll, N., Schuler, B., Criado, A., Guitián, E., Peña, D., Gourdon, A. and Meyer, G. (2012) Bond-order discrimination by atomic force microscopy. Science 337, 1326—1329.
2. Brecher, J. (2008) Graphical representation standards for chemical structure diagrams (IUPAC Recommendations 2008). Pure Appl. Chem. 80, 277—410.

Carbon—carbon quadruple bond

Quadruple and higher order metal—metal bonds are known for transition metals, lanthanoids and actinoids. But for main group elements? Using four different computational methods, Shaik et al. [1] show that

C₂ and its isoelectronic molecules CN⁺, BN and CB⁻ (each having eight valence electrons) are bound by a quadruple bond. The bonding comprises not only one σ- and two π-bonds, but also one weak ‘inverted’ bond, which can be characterized by the interaction of electrons in two outwardly pointing sp hybrid orbitals.

According to Shaik, the existence of the fourth bond in C₂ suggests that it is not really diradical C₂^2• [2]:

If C₂ were a diradical it would immediately form higher clusters. I think the fact that you can isolate C₂ tells you it has a barrier, small as it may be, to prevent that.

1. Shaik, S., Danovich, D., Wu, W., Su, P., Rzepa, H.S. and Hiberty, P.C. Quadruple bonding in C₂ and analogous eight-valence electron species. Nature Chemistry 4, 195—200.
2. Extance, A. Calculations reveal carbon-carbon quadruple bond. Chemistry World, 29 January 2012.

Sulfimide bond in collagen IV

The recent paper in Science describes the sulfilimine (sulfimide, in IUPACese) bond, “not previously found in biomolecules”, identified in collagen. The bond (>S=N–) cross-links methionine and hydroxylysine residues of adjoining protomers.

Charge-shift bonding

Here’s something one doesn’t see in a chemistry textbook. The recent perspective paper in Nature Chemistry deals with a distinct class of electron-pair bonding called “charge-shift” (CS) bonding, which exists alongside classical covalent and ionic bonding. And in not-so exotic molecules.

<A> striking example is the difference between H₂ and F₂; two homonuclear bonds that by all criteria should be classified as covalent bonds, but exhibit fundamental differences. Consider the energy curves (Fig. 1) of the two bonds calculated recently. Figure 1a shows that the H—H bond is indeed covalent; its covalent structure accounts for most of the bonding energy (relative to the ‘exact’ curve). By contrast, for the F—F bond in Fig. 1b, the covalent structure is entirely repulsive, and what determines the bonding energy and the equilibrium distance is the covalent–ionic mixing. This mixing leads to a resonance energy stabilization, which we have termed the ‘charge-shift resonance energy’ (RE_CS). Thus, despite their apparent similarity, the two bonds are very different; whereas the H—H bond is a true covalent bond, the F—F bond is a CS bond that is completely determined by the RE_CS quantity.

No less striking example is so-called inverted C—C bond in [1.1.1]propellane (described in a paper from the same group of authors), which “closely resembles the single bond of difluorine”.

Aurophile, argentophile...

One can expect that these terms have something to do with alchemy. Wrong. Apparently, aurophilic bond is just a weak Au—Au bond, and argentophilic bond is a Ag—Ag bond. On the other hand, Merriam-Webster’s Medical Dictionary defines argentophilic (argyrophilic) as

having an affinity for silver — used of certain cells, structures, or tissues that selectively reduce silver salts to metallic silver.

“Cuprophilic” has been used in both senses, viz. Cu—Cu bond (as, for example, here) and “having an affinity for copper” (as in here). Similarly, “metallophilic” has been used to describe both “generic” metal—metal bond and for “metallophilic cells”. I find the use of this terminology in its former (more restrictive) sense both confusing and unnecessary. For example, this paper describes “Hg(II)···Pd(II) metallophilic interactions”. It could as well be named simply “Hg(II)—Pd(II) interactions”.