Is science sometimes in danger of getting tunnel vision? Recently published ebook author, Ian Miller, looks at other possible theories arising from data that we think we understand. Can looking problems in a different light give scientists a different perspective?

Share this |

Share to Facebook Share to Twitter Share to Linked More...

Latest Posts

The usual approach to the chemical bond is to "solve the Schrödinger equation", and this is done by attempting to follow the dynamics of the electrons. As we all know, that is impossible; the equation as usually presented requires you to know the potential field in which every particle moves, and since each electron is in motion, the problem becomes insoluble. Even classical gravity has no analytical solution for the three-body problem. We all know the answer – there are various assumptions and approximations made, and as Pople noted in his Nobel lecture, validation of very similar molecules allows you to assign values to the various difficult terms and you can get quite accurate answers for similar molecules.

However, you can only be sure of that if there are suitable examples from which to validate. So, quite accurate answers are obtained, but the question remains, is the output of any value in increasing the understanding of what is going on for chemists? In other words, can they say why A behaves differently to a seemingly similar B?

There is a second issue. Because validation and the requirement to obtain results equivalent to those observed, can we be sure they are obtained the right way? As an example, in 2006 some American chemists decided to test some programs that were considered tolerable advanced and available to general chemists on some quite basic compounds. The results were quite disappointing, even to the extent of showing that benzene was non-planar. (Moran, D. and five others. 2006. J. Amer. Chem. Soc. 128: 9342-9343.)
There is a third issue, and this seems to have passed without comment amongst chemists. In the state vector formalism of quantum mechanics, it is often stated that you cannot factorise the overall wave function. That is the basis of the Schrödinger cat paradox. The whole cat is in the superposition of states that differ on whether or not the nucleus has decayed. If you can factorise the state, the paradox disappears. You may still have to open the box to see what has happened to the cat, but the cat, being a macroscopic being, has behaved classically and was either dead or alive before you opened it. This, of course, is an interpretive issue. The possible classical states are "cat alive" (that has amplitude A) and "cat dead" (which has amplitude B). According to the state vector formalism, the actual state has amplitude (A B), hence thinking that the cat is in a superposition of states. The interesting thing about this is it is impossible to prove this wrong, because any attempt to observe the state collapses it to either A or B, and the "or" is the exclusive form. Is that science or another example of the mysticism that we accuse the ancients of believing, and we laugh at them for it? Why won't the future laugh at us? In my opinion, the argument that this procedure aids calculation is also misleading; classically you would calculate the probability that the nucleus had decayed, and the probability the rest of the device worked, and you could lay bets on whether the cat was alive or dead.
Accordingly, I am happy with factorizing the wave function. Indeed, every time you talk about a p orbital interacting with . . . you have factorized the atomic state, and in my opinion chemistry would be incomprehensible unless we do this sort of thing. However, I believe we can go further. Let us take the hydrogen atom, and accept that a given state has the action equal to nh associated with any state. We can factorise that (Schiller, R. 1962. Phys Rev 125 : 1100 – 1108 ) such that
            nh  = [(nr + ½) + ( l 
+ ½)h
Here, while the quantum numbers count the action, they also count the number of radial and angular nodes respectively. What is interesting is the half quanta; why are they there? In my opinion, they have separate functions from the other quanta. For example, consider the ground state of hydrogen. We can rewrite (1) as
            h  = [( ½ ) + ( ½)]h  (2)
What does (2) actually say? First there are no nodes. The second is the state actually complies with the Uncertainty Principle. Suppose instead, we put the RHS  of (2) simply equal to 1. If we assign that to angular motion solely, we have the Bohr theory, and we know that is wrong. If we assign it to radial motion solely, we have the motion of the electron as lying on a line through the nucleus, which is actually a classical possibility. While that turns up in most text books, again I consider that to be wrong because it has zero angular uncertainty. You know the angular momentum (zero) and you know (or could know if you determined it) the orientation of the line. (The same reasoning shows why Bohr was wrong, although of course at the time he had no idea of the Uncertainty Principle.)
 There is another good point about (2): it asserts the period involves two "cycles". That is a requirement for a wave, which must have a crest and a trough. If you have no nodes separating them, you need two cycles. Now, I wonder how many people reading this (if any??) can see what happens next?
Which gets me to a final question, at least for this post: how many chemists are actually happy with what theory offers them? Comments would be appreciated.

Posted by Ian Miller on Oct 22, 2017 9:45 PM BST
Following the Alternative interpretations theme, I shall write a series of posts about the chemical bond. As to why, and I hope to suggest that there is somewhat more to the chemical bond than we now consider. I suspect the chemical bond is something almost all chemists "know" what it is, but most would have trouble articulating it. We can calculate its properties, or at least we believe we can, but do we understand what it is? I think part of the problem here is that not very many people actually think about what quantum mechanics implies.

In the August Chemistry World it was stated that to understand molecules, all you have to do is to solve the Schrödinger equation for all the particles that are present. However, supposing this were possible, would you actually understand what is going on? How many chemists can claim to understand quantum mechanics, at least to some degree? We know there is something called "wave particle duality" but what does that mean? There are a number of interpretations of quantum mechanics, but to my mind the first question is, is there actually a wave? There are only two answers to such a discrete question: yes or no. De Broglie and Bohm said yes, and developed what they call the pilot wave theory. I agree with them, but I have made a couple of alterations, so I call my modification the guidance wave. The standard theory would answer no. There is no wave, and everything is calculated on the basis of a mathematical formalism.

Each of these answers raises its own problems. The problem with there being a wave piloting or guiding the particle is that there is no physical evidence for the wave. There is absolutely no evidence so far that can be attributed solely to the wave because all we ever detect is the particle. The "empty wave" cannot be detected, and there have been efforts to find it. Of course just because you cannot find something does not mean it is not there; it merely means it is not detectable with whatever tool you are using, or it is not where you are looking. For my guidance wave, the problem is somewhat worse in some ways, although better in others. My guidance wave transmits energy, which is what waves do. This arises because the phase velocity of a wave equals E/p, where E is the energy and p the momentum. The problem is, while the momentum is unambiguous (the momentum of the particle) what is the energy? Bohm had a quantum potential, but the problem with this is it is not assignable because his relationship for it did not lead to a definable value. I have argued that to make the two slit experiment work, the phase velocity should equal the particle velocity, so that both arrive at the slits at the same time, and that is one of the two differences between my guidance wave and the pilot wave. The problem with that is, it puts the energy of the system at twice the particle kinetic energy. The question then is, why cannot we detect the energy in the wave? My answer probably requires another dimension. The wave function is known to be complex; if you try to make it real, e.g. represent it as a sine wave, quantum mechanics does not work.

However, the "non-real" wave has its problems. If there is actually nothing there, how does the wave make the two-slit experiment work? The answer that the "particle" goes through both slits is demonstrably wrong, although there has been a lot of arm-waving to preserve this option. For example, if you shine light on electrons in the two slit experiment, it is clear the electron only goes through one slit. What we then see is claims that this procedure "collapsed the wave function", and herein lies a problem with such physics: if it is mysterious enough, there is always an escape clause. However, weak measurements have shown that photons go though only one slit, and the diffraction pattern still arises, exactly according to Bohm's calculations (Kocsis, S. and 6 others. 2011. Observing the Average Trajectories of Single Photons in a Two-Slit Interferometer Science 332: 1170 – 1173.) There is another issue. If the wave has zero energy, the energy of the particle is known, and following Heisenberg, the phase velocity of the wave is half that of the particle. That implies everything happens, then the wave catches up and sorts things out. That seems to me to be bizarre in the extreme.

So, you may ask, what has all this to do with the chemical bond? Well, my guidance wave approach actually leads to a dramatic simplification because if the waves transmit energy that equals the particle energy, then the stationary state can now be reduced to a wave problem. As an example of what I mean, think of the sound coming from a church organ pipe. In principle you could calculate it from the turbulent motion of all the air particles, and you could derive equations to statistically account for all the motion. Alternatively, you could argue that there will be sound, and it must form a standing wave in the pipe, so the sound frequency is defined by the dimensions of the pipe. That is somewhat easier, and also, in my opinion, it conveys more information.

All of which is all very well, but where does it take us? I hope to offer some food for thought in the posts that will follow.
Posted by Ian Miller on Aug 28, 2017 12:19 AM BST
In a recent Chemistry World there was an item on chemistry in India, and one of the things that struck me was that Indian chemists seemed to be criticized because they published a very low proportion of the papers in journals such as JACS and Angewandte Chemie. The implication was, only the "best" stuff gets published there, hence the Indian chemists were not good enough. The question I want to raise is, do you think that reasoning is valid?
One answer might be that these journals (but not exclusively) publish the leading material, i.e. they lead the way that chemistry is taking in the future. When I started my career, these high profile journals were a "must read" because they were where papers that at least editors felt was likely to be of general interest or of practical interest to the widest number of chemists were published.
But these days, these sort of papers do not turn up. There may be new reactions, but they are starting to involve difficult to obtain reagents, and chemical theory has descended into the production of computational output. These prestige journals have moved on to new academic fields, which is becoming increasingly specialized, which increasingly needs expensive equipment, and which also needs a school that has been going for some time, so that the background experience is well embedded. There are exceptions, but they do not last, thus graphene was quite novel, but not for long. There are still publications involving graphene, but chemists working there have to have experience in the area to make headway. More importantly, unless the chemist is actually working in the area, (s)he will never touch something like graphene. I am certainly not criticizing this approach by the journals. Rather I am suggesting the nature of chemical research is changing, but I feel that in countries where the funding is not there to the same extent, chemists may well feel they might be more productive not trying to keep up with the Joneses.
Another issue is, by implication it is claimed that work published in the elite journals is more important. Who says? Obviously, the group who publish there, and the editorial board will, but is this so? There may well be work that is more immediately important, but to a modest sized subset of chemists working in a specific area. Now the chemist should publish in the journal that that subset will read.
My view is that chemistry has expanded into so many sub-fields that no chemist can keep up with everything. When I started research, organic chemists tended not to be especially interested in inorganic or physical chemistry, not because they were not important, but simply because they did not have the time. Now it has got much worse. I doubt there is much we can do about that, but I think it is wrong to argue that some chemistry that can only be done in very richly funded Universities is "better" or more important than a lot of other work that gets published in specialized journals. What do you think?
Posted by Ian Miller on Jul 3, 2017 3:23 AM BST
Some time ago now I published an ebook "Planetary Formation and Biogenesis", which started with a review including over 600 references, following which I tried analyzing their conclusions and tried to put them together to make a coherent whole. This ended up with a series of conclusions and predictions on what we might find elsewhere. It was in light of this I saw the article in the May edition of "Chemistry World". That article put up reasons to back some of the various thoughts as to where life started, but I found it interesting that people formed their views based on their chemical experience, and they tended to carry out experiments to support that hypothesis. That, of course, is fair enough, but it still misses what I believe to bee the key point, and that is, what is the most critical problem to overcome to get life started, and how hard is it to do?

The hardest thing, in my opinion, is not to make polymers. I know that driving condensation reactions forwards in water is difficult, but as Deamer pointed out in the article, if you can get a lipid equivalent, it is by no means impossible. No, in my opinion, the hardest thing to do is to make phosphate esters. Exactly how do you make a phosphate ester? As Stanley Miller once remarked, you don't start with phosphoryl chloride in the ocean. The simplest way is to heat a phosphate and an alcohol to about 200 degrees C.  Of course, water will hydrolyse phosphate esters at 200 degrees C, so unless you drive off the water, which is difficult to do in an ocean, high temperature is not your friend because the concentration of water in the ocean always exceeds the concentration of phosphate or alcohol. You simply cannot do that around black smokers.

The next problem is, why did nature choose ribose? Ribose is not the only sugar that permits the formation of a duplex when suitably phosphated and bound to a nucleotide. Almost all other pentoses do it. So the question remains, why ribose? The phosphate ester is an important solubilizing agent for a number of biochemicals necessary for life but it invariably occurs bound to a ribose, which in turn is usually bound to adenine. The question then is, is this a clue? If so, why is it largely unnoticed? My conclusion was, ribose alone can form a phosphate ester on a primary alcohol group in solution because only ribose naturally has reasonable concentrations of itself in the furanose form.

It was not always unnoticed. There is a clearly plausible route, substantiated by experiment (Ponnamperuma, C., Sagan, C., Mariner, R., 1963. Synthesis of adenosine triphosphate under possible primitive earth conditions. Nature 199: 222-226.) that shows the way. What was shown here was that if you have a mixture of adenine, ribose and phosphate, and shine UV light that can be absorbed by the adenine, you make adenosine, and then phosphate esters, mainly at the 5 position of the furanose form, so you can end up with ATP, a chemical still used by life today. Why is that work neglected? Could it be that nobody these days goes back and reads the literature from 1963?
Why does this synthesis work? My explanation is this. You do not have to get to 200 degrees to form a phosphate ester. What you have to do is provide an impact between the alcohol group and phosphate equivalent to that expected at 200 degrees. If we think about the experiment described above, there is no way an excited electronic state of adenine can be delocalized into the ribose, so why is the light necessary?

My conclusion was that the excited state of the adenine can decay so that quite a considerable amount of vibrational energy is generated. That will help form the adenosine, but after that the vibrational energy will spread through the sugar. Now we see the advantage of the furanose: it is relatively floppy, and it will vibrate well, and even better, the vibrational waves will focus at C-5. That is how the phosphate ester is formed, and why ribose is critical. The pyranose forms are simply too rigid to focus the mechanical vibrations. Once you get adenosine phosphate, in the above experiment the process continued to make polyphosphates, but if  some adenosine was also close by, it would start to form the polymer chain. Now, if that is true, then life must have started on the surface, either of the sea or on land. My view is the sea is more probable, because on land it is difficult to see where further biochemicals can come from.
Posted by Ian Miller on May 28, 2017 11:45 PM BST
One issue that has puzzled me is what role, if any, does theory play in modern chemistry, other than having a number of people writing papers. Of course some people are carrying out computations, but does any of their work influence other chemists in any way? Are they busy talking to themselves? The reason why this has struck me is that the latest "Chemistry World" has an article "Do hydrogen bonds have covalent character?" Immediately below is the explanation, "Scientists wrangle over disagreement between charge transfer measurements." My immediate reaction was, what exactly is meant by "covalent character" and "charge transfer"?  I know what I think a covalent bond is, which is a bond formed by two electrons from two atoms pairing to form a wave function component with half the periodic time of the waves on the original free atom. I also accept the dative covalent bond, such as that in the BH3NH3 molecule, where two electrons come from the same atom, and where the resultant bond has a strength and length as if the two electrons originated from separate atoms. That is clearly not what is meant for the hydrogen bond, but the saviour is that word "character".  What does that imply?
 What puzzles me here is that on reading the article, there are no charge transfer measurements. What we have, instead, are various calculations based on models, and the argument is whether the model involves transfer of electrons. However, as far as I can make out, there is no observational evidence at all. In the BH3NH3 molecule, obviously the two electrons for the bond start from the nitrogen atom, but the resultant dipole moment does not indicate a whole electron is transferred, although we could say it is, and then sent back to form the bond. However, in that molecule we have a dipole moment of over 6 Debye units. What is the change of dipole moment in forming the hydrogen bond? If we want to argue for charge transfer, we should at least know that.
From my point of view, the hydrogen bond is essentially very weak, and is at least an order of magnitude less strong than similar covalent bonds. This would suggest that if there were charge transfer, it is relatively minor. Why would such a small effect not be simply due to polarization? With the molecule BH3NH3 it is generally accepted that the lone pair on the ammonia enters the orbital structure of the boron system, with both being tetrahedral in structure, more or less. The dipole moment is about 6 Debye units, which does not correspond to one electron fully transferring to the boron system. There is clear charge transfer and the bond is effectively covalent.
Now, if we then look at ammonia, do we expect the lone pair on the nitrogen to transfer itself to the hydrogen atom of another ammonia molecule to form this hydrogen bond? If it corresponded to the boron example, then we would expect a change of at least several Debye units but as far as I know, there is no such change of dipole moment that is not explicable in terms of it being a condensed system. The article states there are experimental data to support charge transfer, but what is it?
Back to my original problem with computational chemistry: what role, if any, does theory play in modern chemistry? In this article we see a statement such as the NBO method falls foul of "basis set superposition error". What exactly does that mean, and how many chemists appreciate exactly what it means? We have a disagreement where one is accused of focusing on energies, while they focus on charge density shifts.  At least energies are measurable. What bothers me is that such arguments on whether different people use the same terminology differently is a bit like arguing about how many angels can dance on the head of a pin.  What we need from theory is a reasonably clear statement of what it means, and a clear statement of what assumptions are made, and what part validation plays in the computations.
Posted by Ian Miller on Apr 24, 2017 12:54 AM BST
An interesting thing happened for planetary science recently: two papers (Nature, vol 541 (Dauphas, pp 521 – 524; Fischer-Gödde and Kleine, pp 525 – 527) showed that much of how we think planets accreted is wrong. The papers showed that the Earth/Moon system has isotope distributions across a number of elements exactly the same as that found in enstatite chondrites, and that distribution applied over most of the accretion. The timing was based on the premise that different elements would be extracted into the core at different rates, and some not at all. Further, the isotope distributions of these elements are known to vary according to distance to the star, thus Earth is different from Mars, which in turn is clearly different from the asteroid belt. Exactly why they have this radial variation is an interesting question in itself, but for the moment, it is an established fact. If we assume this variation in isotope distribution follows a continuous function, then the variations we know about have sufficient magnitude that we can say that Earth accreted from material confined to a narrow zone.
 Enstatite chondrites are highly reduced, their iron content tends to be as the metal or as a sulphide rather than as an oxide, and they may even contain small amounts of silicon as a silicide. They are also extremely dry, and it is assumed that they were formed at a very hot part of the accretion disk because they contain less forsterite and additionally you need very high temperatures to form silicides.
In my mind, the significance of these papers is two-fold. The first is, the standard explanation that Earth's water and biogenetic material came from carbonaceous chondrites must be wrong. The ruthenium isotope analysis falsifies the theory that so much water arrived from such chondrites. If they did, the ruthenium on our surface would be different. The second is the standard theory of planetary formation, in which dust accreted to planetesimals, these collided to form embryos, which in turn formed oligarchs or protoplanets (Mars sized objects) and these collided to form planets must be wrong. The reason is that if they did collide like that, they would do a lot of bouncing around and everything would get well-mixed. Standard computer simulations argue that Earth would have formed from a distribution of matter from further out than Mars to inside Mercury's orbit. The fact that the isotope ratios are so equivalent to enstatite chondrites shows the material that formed Earth came from a relatively narrow zone that at some stage had been very strongly heated. That, of course, is why Earth has such a large iron core, and Mars does not. At Mars, much of the iron remained as the oxide.
In my mind, this work shows that such oligarchic growth is wrong and that the alternative, monarchic growth, which has been largely abandoned, is in fact correct. But that raises the question, why are the planets where they are, and why are there such large gaps? My answer is simple: the initial accretion was chemically based, and certain temperature zones favoured specific reactions. It was only in these zones that accretion occurred at a sufficient rate to form large bodies. That, in turn, is why the various planets have different compositions, and why Earth has so much water and is the biggest rocky planet: it was in a zone that was favourable to the formation of a cement, and water from the disk gases set it. If anyone is interested, my ebook "Planetary Formation and Biogenesis" explains this in more detail, and a review of over 600 references explains why. As far as I am aware, the theory outlined there is the only one that requires the results of those papers. So, every now and again, something good happens! It feels good to know you could actually be correct where others are not.
So, will these two papers cause a change of thinking. In my opinion, it may not change anything because scientists not directly involved probably do not care, and scientists deeply involved are not going to change their beliefs. Why do I think that? Well, there was a more convincing paper back in 2002 (Drake and Righter, Nature 416
: 39-44) that came to exactly the same conclusions. Instead of ruthenium isotopes, it used osmium isotopes, but you see the point. I doubt these two papers will be the straw that broke the camel's back, but I could be wrong. However, experience in this field shows that scientists prefer to ignore evidence that falsifies their cherished beliefs than change their minds. As a further example, neither of these papers cited the Drake and Righter paper. They did not want to admit they were confirming a previous conclusion, which is perhaps indicative they really do not wish to change people's minds, let alone acknowledge previous work that is directly relevant.
Posted by Ian Miller on Feb 5, 2017 9:41 PM GMT