The zigzag is short by its gap
Worth reading first: A nitrogen's two bonds cancel on an angular ring · An angular ring rescales what lies beyond it.
A heteroatom placed on one carbon of a chain of fused benzene rings changes the bond orders of every ring, and the change falls off with distance along the chain. That fall is the reach of a local perturbation through a delocalised system, and it has been measured here in one model throughout: Hückel’s, with each bond’s coupling allowed to relax with its own bond order and a stagger of site energies holding the chain’s gap open so that its length does not shrink the gap as the chain grows. The floor was in the bookkeeping removed an artefact that stopped the response falling after five rings; the reach is the molecule’s found the decay the same from the end and from the middle; and in the straight chain of twelve rings the decay length is 0.719 rings.
Angular fusions complicate it. An angular ring rescales what lies beyond it found each one multiplying the response past it by a factor without changing the decay much; on an angular ring the carbon decides found that an angular ring has only two CH carbons, and a heteroatom on either of them sends the far chain about half what it sends from a straight ring; and a nitrogen’s two bonds cancel traced that half to the angular ring’s strong bond between its two CH carbons, which sends the far chain a response almost opposite to the one from the carbon’s junction bond.
That essay ended with a prediction about the chain that is angular everywhere. In the zigzag, every interior ring has its own strong bond between its CH carbons, and a response travelling along the chain must cross every one. If each takes back part of what reaches it, as the heteroatom’s own strong bond did, the zigzag’s reach should fall faster than the straight chain’s for a reason that has nothing to do with its gap — the opposite of the ring currents of the acenes, whose reach was the gap and nothing else.
The zigzag does reach less far. The prediction about why is wrong.
One dial, two curves
The comparison that would show a reason beyond the gap needs the gap held fixed, and the stagger is the dial that sets it. Turning it up on the straight chain opens its gap and shortens its reach, along one curve. Turning it up on the zigzag does the same along another. If the strong bonds cost the zigzag something its gap does not, the zigzag’s curve lies below the straight chain’s at every gap.
It lies above. At narrow gaps the two curves run together: the zigzag at a gap of 1.21 reaches 0.711 rings, the straight chain at 1.21 reaches 0.719. As the gap opens they part, and the zigzag keeps more: at a gap near 2.0 it reaches 0.547 rings where the straight chain reaches about 0.49; near 2.5, 0.469 against about 0.41. The chain with one kink, two adjacent angular rings in the middle, sits between them, nearly on the straight chain’s curve.
The published comparison used one stagger for every chain, and at that stagger the zigzag’s gap is not the straight chain’s. With the stagger at 0.6 the straight chain’s relaxed gap is 1.206 and the zigzag’s 1.492: the angular fusions open the gap by a quarter on their own, as phenacenes’ gaps are wider than acenes’. The zigzag’s shorter reach at a common stagger is a wider gap seen at a common dial setting.
Where the zigzag’s extra gap comes from
The two chains differ in gap at a common stagger by an amount that depends on the stagger, and the dependence says what the gap is made of. At a stagger of 0.3 the straight chain’s relaxed gap is 0.612 and the zigzag’s 1.099 — nearly twice as wide. At 0.6 they are 1.206 and 1.492; at 1.4, 2.802 and 2.926. The stagger opens both gaps at about the same rate, two hopping units of gap for every unit of stagger, and the zigzag carries an extra gap of its own on top that shrinks, relative to the total, as the stagger grows.
That extra gap is the topology’s. A straight chain of fused rings, an acene, has a gap that closes as it lengthens; a zigzag chain, a phenacene, keeps a wide one, because its angular fusions localise the π bonding into the strong CH–CH bonds and break the chain into nearly separate double bonds. With no stagger at all the relaxed straight chain of twelve rings has a gap of 0.131 and the zigzag one of 0.944, seven times wider; the kink, with two angular rings, has 0.348. The stagger was introduced to hold the straight chain’s gap open so that a decay length could be read at all, and at the stagger that does that adequately for the straight chain, the zigzag’s own gap still adds a quarter.
So the comparison at a common stagger was never a comparison at a common gap, and it was the stagger, not the gap, that the earlier essays held fixed. The decay length read at a common stagger is 0.638 rings for the zigzag against 0.719 for the straight chain, a ratio of 0.887; read at a common gap it is 1.042. The two numbers describe the same pair of molecules, and only the second isolates what the angular rings do. The difference between them is not small or subtle: it changes the sign of the conclusion, from “the zigzag passes a perturbation on less well” to “the zigzag passes it on slightly better”, and it does so through a quantity — the gap — that every one of the earlier comparisons computed and none of them held fixed.
That is worth saying because the same trap is open wherever two chains of different topology are compared at one set of model parameters. A stagger, a bond alternation, a heteroatom strength: each is a dial that sets the gap jointly with the topology, and a comparison at a common dial setting compares two gaps as well as two topologies.
Matched gap by gap
The curves read the comparison by eye; the matched comparison makes it a number. For each stagger of the zigzag, the straight chain’s stagger is found that gives exactly the same relaxed gap — by the secant method, to a part in — and the two decay lengths are divided.
At the stagger used throughout the argument, the zigzag reaches 1.042 times as far as the straight chain at its own gap: 0.638 rings against 0.612. Across the nine staggers the ratio runs from 0.99 at the narrowest gaps to 1.17 at a gap of 2.5, and back to 1.15 at the widest. The relaxation strength barely moves it: with the bond-order feedback halved, to 0.2, the ratio runs from 1.00 to 1.18; increased to 0.6, from 0.95 to 1.16. The smallest value anywhere is 0.950, at the narrowest gap and the strongest relaxation — the one corner where a response crossing many strong bonds might have been expected to lose the most, and where it loses five per cent.
The relaxation strength is itself a test of the mechanism. λ is what makes a strong bond strong: it is the feedback by which a bond with a high order shortens and couples more, so tripling it from 0.2 to 0.6 deepens the alternation of every angular ring’s bonds. If those bonds were taking back the response, a stronger feedback should cost the zigzag more. At the published stagger the matched ratio is 1.018 at λ = 0.2, 1.042 at 0.4 and 1.044 at 0.6 — it rises, slightly. Only at the narrowest gaps does a stronger feedback lower it, and there by five per cent.
The kink sits within six per cent of the straight chain at every matched gap, between 0.99 and 1.06. Two angular rings in a chain of twelve change the reach by about what the counting of angular rings found they change it by.
At no gap does a run of ten strong bonds cost the zigzag the reach the prediction asked for. The prediction’s mechanism would have shown up as a ratio falling below one and falling further the more bonds were crossed; what the numbers show is a ratio above one that grows as the gap opens.
Ring by ring
The fitted length is a summary, and it would be possible for each strong bond to take back part of the response and for something else to make it up. The profiles show which.
Set against the straight chain at the same stagger, the zigzag’s response falls faster ring after ring, and that was the observation the prediction set out to explain. Set against the straight chain at the same gap, it tracks it and finishes above it: on the last ring the zigzag’s response is against the matched straight chain’s , twice as large, though on the first ring past the heteroatom’s the zigzag’s was the larger by only a ninth, against .
The zigzag’s profile is not smooth, and that is where the strong bonds are.
From one ring to the next the straight chain keeps between 0.16 and 0.23 of its response, varying gently. The zigzag keeps 0.15 at one step and 0.20 at the next near the heteroatom, and further along it swings: 0.29, 0.13, 0.42, 0.14, 0.41, 0.17. Every other passage loses much more than the straight chain’s and every other one much less. That is a strong bond taking back part of what reaches it, as the prediction said — and the next ring giving it back. The geometric means over the ten steps are 0.206 for the zigzag and 0.194 for the straight chain at the same gap, which is the whole of the difference in their decay lengths.
One reading of the alternation is geometric. In the zigzag the strong bonds alternate in their position relative to the chain’s axis, one ring’s strong bond on one side and the next ring’s on the other, and a bond-order perturbation crossing a ring whose strong bond is on its far side meets the strong bond’s opposition head on, while one crossing a ring whose strong bond is on its near side passes it before the response has spread. What the model shows is that the losses and gains come in pairs and very nearly cancel; why the gains slightly exceed the losses is not something this calculation separates.
From the middle of the chain
Every response above is to a heteroatom on the end ring, which is not angular in any of these chains. The prediction was about a heteroatom on an angular ring’s CH carbon, with the strong bond at its own site, so it is worth testing from there too.
From a CH carbon of ring three — an angular ring in the zigzag, a straight one in the straight chain — the response towards the far end decays with a length of 0.679 rings in the zigzag and 0.588 in the straight chain at the same gap: the zigzag reaches 1.155 times as far. The heteroatom’s own strong bond cuts the zigzag’s response near the heteroatom, as the two bonds’ cancellation found it does on a single angular ring; the far decay is longer, and the alternation reappears as soon as the response leaves the heteroatom’s ring.
How the chains were compared
Each chain has twelve fused rings, built from its list of angular rings with the angular rings read back from the geometry: none for the straight chain, rings one to ten for the zigzag, rings five and six for the kink. A stagger of ±δ on alternate sublattice sites holds the gap open. The plain chain’s couplings are relaxed to self-consistency with their bond orders at strength λ, the chain is relaxed again with one carbon’s site energy raised by one hopping unit and its couplings measured from the plain chain’s mean bond order, and each ring’s response is the mean absolute change of the bond orders on its six bonds. The heteroatom is on the end ring’s outermost carbon, or on the CH carbon of ring three nearest the far end. The decay length is the negative reciprocal of the slope of the logarithm of the response against ring number, fitted from the first ring past the heteroatom for as long as the response keeps falling — eleven rings from the end, eight from ring three.
The gap is that of the relaxed plain chain. The straight chain’s matched stagger is found by the secant method from a bracket around half the target gap, converging to in the gap.
The published numbers come back: the straight chain’s 0.719 rings and the zigzag’s 0.638 at a stagger of 0.6. Every chain’s reach shortens monotonically as its stagger opens the gap. And the refusal is the prediction itself made specific: a response losing part of itself at every strong bond would put the zigzag’s ratio below one by a margin growing with the gap, and the ratio must instead stay above 0.94 everywhere and above one from the published stagger upwards.
What a relaxed Hückel chain cannot settle
One electron and one relaxation law. The bond-order feedback is linear with a single strength, and the π electrons do not interact. In a real phenacene the bonds alternate because of the σ framework and electron correlation as well as the π bond orders, and the strong CH–CH bonds are stronger than any value of λ here makes them. Whether a larger alternation tips the balance between the losses and the gains along the zigzag is not something a linear feedback can say.
A stagger is not a molecule. The gap is held open by alternating site energies, which no hydrocarbon has. That device is what makes the matched comparison possible — no real pair of molecules differs in gap and in nothing else — and it is also why the matched straight chain is a model object rather than a compound.
The decay length is a fit. Its value depends on where the fit starts and how far the response keeps falling. The ring-by-ring passage shows the alternation that the fitted length averages away, and a fit over an odd number of steps would weight one kind of passage more than the other; over eleven rings the effect is a few per cent.
The gap was the account after all
The angular rings matter to the zigzag’s response in a way the straight chain cannot show: they make it alternate, ring by ring, between keeping much less and keeping much more. What they do not do is shorten it. The zigzag’s reach at a common stagger is shorter because its gap is wider, and at a common gap it is as long as the straight chain’s or longer. The same statement the ring currents made about the acenes — that a reach is set by the gap — holds for the bond-order response of a chain angular everywhere, once the gap is the thing held fixed.
Still open: why the gains exceed the losses, and a branch
The obvious open question is the small excess. The zigzag’s passages alternate, and their geometric mean exceeds the matched straight chain’s by about five per cent at the published stagger and by more as the gap opens. A two-ring unit cell — the zigzag’s actual period — has its own band structure, and the decay of a perturbation through a periodic chain is set by the complex wave number at the energy of the heteroatom’s level. Computing that for the zigzag’s two-ring cell and the straight chain’s one-ring cell at equal gaps would say whether the excess is a property of the periodic chain or of these twelve rings.
The nearer question is the one this sequence has deferred three times: a ring fused on three sides, triphenylene’s central ring, where a response leaving one outer ring has two routes to the other two. In the zigzag the strong bonds’ losses and gains came in pairs along one route; with two routes, a loss on one could meet a gain on the other at the same ring, and whether they add or cancel there is a sharper question now that the alternation is known.
What links here
Computed from the collection rather than written here: the essays that point at this one.
Reads more easily once this is understood
Essays that name this one as worth reading first.
Shares its objects with
Essays naming at least two of the same things, that neither author linked.
- A bend is not an end — both name bond order, delocalisation, hückel theory, reference state
- An anomaly that is not the first of a series — both name bond order, delocalisation, hückel theory, reference state
- Six electrons in a ring that is not all carbon — both name bond order, delocalisation, hückel theory, reference state
- The frame that was allowed to relax — both name bond order, delocalisation, hückel theory, reference state
- The same ring, three charges — both name bond order, delocalisation, hückel theory, reference state
- Which numbers carry a frame — both name bond order, delocalisation, hückel theory, reference state
Named objects
A dashed tag is an object no other essay names yet.
Band gapBond orderDelocalisationHückel theoryReference state