Karl Ziegler (1964) - From Triphenylmethyl to Polyethylene - Less Well-Known Facts About the Developments Leading to the Invention Made in Muelheim (German presentation)

Ladies and Gentlemen, The topic of my lecture today might lead you to expect me to talk today about the stimulation of polymerisation by free radicals. After all: the chemists among you, and I assume there are a great number of you from the younger generation, will know that you can stimulate polymerisation with radicals and that triphenylmethyl is the oldest and best-known free radical. Stimulation of polymerisations by radicals proceeds in such a way that the radical with its free electron is attracted to one side of a double bond, for example in ethylene. As a result, one electron is again freed on the other side of the ethylene, which can then repeat the process. And in this way polymerisation speeds through a large number of molecules. This is known as a radical chain mechanism. But the polymerisation of ethylene, which we discovered ten years ago in Mühlheim, does not proceed on the basis of such a radical chain mechanism and is therefore completely unrelated to it. It is, however, related to the fact that I have today the opportunity, the very welcome opportunity, to speak to you here. The context of radicals in relation to triphenylmethyl is a personal one and relates to the somewhat circuitous path that ultimately led us to an observation in 1953, from which everything else down to the present day developed. The story in which the whole thing took place is unique and may be educational for the younger generation, because it shows the roundabout routes that can sometimes lead to important findings and progress. And it begins at a time before I was born and when I was still a child. And in this context there is a remarkable circumstance namely our old master Hahn. And in his book “From Radiothorium to the Fission of Uranium”, Otto Hahn has recently written about his own early days at the Chemical Institution of the University of Marburg on the Lahn, in which he mentioned his doctoral supervisor, Theodor Zincke. And his account then goes on to portray very vividly how, after starting his career as an organic chemist, he became an inorganic chemist and radiochemist, and then achieved the pinnacle of success in those fields. My story begins in the same place but around 20 years later. And it differs in that my doctoral supervisor was no longer “old Zincke”, as we called him, but his successor, von Auwers. Moreover, I attempted, and this is another difference to Otto Hahn, to prove that you can start in organic chemistry and then achieve success in organic chemistry itself. Von Auwers, incidentally, had begun his career with August Wilhelm von Hofmann in Berlin and later worked for many years under Victor Meyer at the Heidelberg Institute before moving to Marburg via Greifswald. And for this reason, I can claim to be a scientific grandchild of August Wilhelm von Hofmann, as well as of Victor Meyer. I would first like to place the two gentlemen, the two great chemists, in context. Here you have “Old Zincke”, as I remember him and knew him, and then Mr. von Auwers, as he was when he was my doctoral supervisor. Despite this very unique personal bond to von Auwers, my story also has something to do with the old Privy Counsellor Zincke, who throughout my studies still worked assiduously in a corner of the organic chemistry room, where he had his workplace, and immediately after the First World War still had two doctoral candidates who, incidentally, remained close colleagues for life. My own relationship has a more negative slant. This has to do with the fact that Privy Counsellor Zincke, in his final period of creativity, did a lot of work on certain organic sulphur compounds. Now, sulphur compounds are, as you know, the worst toxins for sensitive catalysts. When I starting in 1917 was working in organic chemistry in Marburg and then took up my own work there in 1920, it was a standing rule that catalytic hydration never worked at the Marburg Institute. This was generally attributed to contamination of the laboratory with organic sulphur compounds, that is, the sulphur compounds from Old Zincke. Whether this really was true or whether it was simply a convenient excuse on the part of those affected, who did not always work scrupulously enough when preparing catalysts, is anyone’s guess. In any case, when I was confronted with a certain hydration problem, this situation prompted me not to even attempt catalytic hydration and to come up with something somewhat different. And this marks the real beginning of my story. Around 1920, free organic radicals of the triphenylmethyl type were a hot topic. Shortly before the war, which had only recently ended, Schlenk prepared the first really totally free monomolecular, non-associated radical, triphenylmethyl and the question regarding the causes of the stability of such radicals was the subject of lively debate. There is a still recent, really new - back then really new - monograph by Schmidlin on the field, and on the last page he suggested that researchers might consider searching for a trivinylmethyl, which would contain three monounsaturated residues, the bottom formula here. The implication was that researchers should conduct a search here based on the analogy between aromatic and unsaturated aliphatic substituents. The first success of my independent scientific research lay along these lines, and in 1923 I was able to synthesize the first radical with an unsaturated substituent, namely tetraphenylallyl, which you can see at the top left. And soon afterward, I managed to synthesise an analogous substance containing two unsaturated substituents, pentaphenyl cyclopentadienyl, in 1925 (top right). Both products showed a marked tendency to associate, and they truly were free radicals. Of course I then set about to run a cross-check and see what would become of such radicals after hydration of the double bond. In the process, hydration should ... So, what I was trying to do was make a substance like the one expressed by the bottom formula. That involves nothing more than hydrating this tetraphenylallyl at the double bond. The hydration should not be performed on the free radical itself, because it was feared that the free electron would also enter into the hydration reaction, but rather on a precursor. On a precursor, namely the corresponding ether, which you see up here. And because, as I mentioned, I didn’t expect much from catalytic hydration, I immediately chose a different path. Around ten years before, Wilhelm Schlenk namely had found that you can add alkali metals to certain double bonds, particularly those containing aromatic substituents. And so I conceived the pathway shown in these two formulas. I wanted to add sodium to this ether and then, by way of conventional hydrolysis, replace those two sodium atoms with hydrogen. That was the starting point and the concept. I was very fortunate that things didn’t go quite to plan and that I was given the opportunity to discover my first new reaction. No alkali metal attached to the double bond; instead, the two sodium atoms caused the ether to cleave at this bond, as you can see here in this diagram, to form an organosodium compound plus sodium alcoholate. Thus, on the basis of this very complicated example, a new, very simple method was discovered to produce organo compounds of the alkali metals. And, of course, I then established what structural requirements such ethers must possess for them to be cleanly cleaved by alkali metals. It was found that ethers of tertiary alcohols containing at least one aromatic residue on the tertiary carbon atom undergo cleavage particularly easily. The simplest ether that worked was phenylisopropyl methyl ether, a tertiary ether with an aromatic substituent that undergoes cleavage extraordinarily easily, especially by potassium, in accordance with this reaction. In this way, phenylisopropyl potassium was produced, the substance at the bottom left, which at the time was the most readily accessible and highly reactive organopotassium compound. I would like to take this opportunity to reassure you that in my later chemical career I managed to carry out catalytic hydration with ease. So I’ve rectified this deficiency from my early days. But at the time I had every reason to attribute this to the old Privy Counsellor Zincke and his sulphur compounds and actually suspect – I didn’t even attempt the experiment Because that is how I came to study organometallic compounds, to which I owe a whole series of new discoveries and successes right down to the present day. The next link in the chain of development was a rumour that I heard shortly after starting work in Heidelberg in 1925. The word was that Wilhelm Schlenk in Berlin had produced addition products of alkali metals, for example on stilbene the symmetrical isomer of 1,2-diphenylethylene, whose configuration was said to correspond to that of diastereomers. So it was thought that when lithium is added to stilbene, shown as a stereoformula, as a projection formula, such an addition product results, and when sodium is added, an analogous but different product results. Schlenk published these results later. I heard at the time about it from …, as it happens, from rumour. Schlenk later published his findings, and it still remains doubtful whether the published data would have withstood close scrutiny. A lot of other things attributed to the Berlin laboratory at the time were not true. But I began to wonder whether you could also make such diasteriomers with the same metal if you produced the addition products using two different methods. One such method would be to attach the metal directly to the double bond, and another, for example, an indirect route via a reaction in which the metal is exchanged by way of a reactive organometallic compound. Based on these considerations, I reacted stilbene - that was in Heidelberg, where I took up work in 1926 - with phenylisopropyl potassium, which had just been discovered in Marburg, in the expectation that a dipotassium compound would form along the lines of such an exchange. I then wanted to determine if it was identical to the compound obtained by direct attachment of potassium to Stilbene and the product obtained by combining the two remaining residues of phenylisopropyl potassium. But again things didn’t go to plan. The potassium compound simply added to the ethylene in accordance with this reaction: Here is potassium on the one side, the predicted phenylisopropyl group on the other. And in this way I discovered my second new reaction of organoalkali metal compounds, and namely one that went significantly beyond the organometallic compounds known up to then. Because the Grignard magnesium compounds, which dominated the field at the time, added only to CO double bonds and not to CC double bonds of olefin hydrocarbons. In the course of events, those observations I made in Heidelberg may have played a role. The Baden Aniline and Soda Factory, which was working on synthetic rubber, was nearby. And I was familiar with the well-known polymerisation method of butadiene by sodium, which later led to the general name of Buna rubber. I clearly recall the evening we succeeded in adding phenylisopropyl potassium to stilbene and definitively established that this reaction pathway had taken place, and that even on the way home from the institute, I realized that this reaction might hold the key to understanding the polymerisation of butadiene by alkali metals. If one assumes that such an addition product is formed from butadiene and sodium, you would have a disodium compound that would certainly correspond to this phenylisopropyl potassium in the nature of its bonds. And if it adds here and on both sides of the butadiene molecule, the result could be a large molecule of this type. Now, based on the example of this phenylisopropyl potassium, it was possible to show that this type of reaction actually is possible. An addition reaction of this kind does actually occur between phenylisopropyl and butadiene, and – as I discovered in Heidelberg – you can repeat it. That is the first reaction step. And then, in the same way, you can add the second, the third, the fourth, and so on. I discovered in other words and demonstrated a reaction type that I called stepwise organometallic synthesis, and then later, in a completely different form, namely with aluminium as the carrier metal of the organometallic compound and using ethylene instead of butadiene, developed it into an industrial process. It’s clear that after the discovery of such a new reaction, researchers would search the entire field for reactions of a similar nature in order to explore the boundaries within which such reactions are possible. So I tested many combinations of organometallic compounds and olefin carbohydrates available at the time and found that not all unsaturated carbohydrates are suitable for addition, not all alkali metal compounds as well. And when you come up against such highly variable reactivity, it’s logical, to draw exact comparisons, to take kinetic measurements of the reaction. Now, for such measurements phenylisopropyl potassium is very unsuitable, because, insofar as it reacts at all, it always does so at an extraordinary speed, making it impossible to carry out comparative rate measurements. We then included the lithium alkyls discovered by Wilhelm Schlenk in 1917 in these investigations. At the time, the method for producing them was very elaborate, requiring mercury-silver alkyls to be heated with lithium shavings in organic solvents. We were also able to add these substances to suitable unsaturated carbohydrates, and usually the reaction did not occur instantly but required some time. We therefore had ideal objects for carrying out comparative kinetic measurements. In our kinetic experiments we were faced with the task of distinguishing the addition products of lithium alkyls ourselves by means of a suitable reaction in order to follow the progress of addition over time. If, for example, you add lithium methyl, a colourless substance that is readily soluble in organic solvents, to a suitable hydrocarbon, for example this diphenylethylene, addition takes place, and the addition product is deep orange-red in colour. We then found that this red instantly disappears after the addition of organohalogen compounds, especially butyl chloride. In other words, this type of organometallic compound in proximity to aromatic substituents reacts extraordinarily rapidly with the halogenated alkyl. And we also found that the starting product, the purely aliphatic lithium methyl, does not react at all with the halogen, with the organohalogen, for example butyl chloride or ethyl chloride, within a reasonable time. So that we were able to carry out our kinetic experiments very nicely by removing the lithium from the more reactive, coloured species after some time by adding halogen compounds and then decomposing the whole thing, and finally filtering out the halogen. With regard to these experiments – I’d like to describe all the circumstances to you today they were simply some type of experiments. They did not lead to far-reaching new findings. But they did make us realize that simple aliphatic lithium compounds with halogenated alkyls do not react rapidly at all, that they can be stored indefinitely. This had naturally a very specific consequence, because from this realisation to the discovery of a new, simple reaction for producing lithium alkyls was just a small step. While I was still a student, the dogma still prevailed that organic halogen compounds only enter into the so-called Wurtz reaction with alkali metals. That is this reaction which you can see written down here. I note, however, that although the diagram is rather old, something is missing: a 2 should appear here, otherwise it’s incomprehensible. So the Wurtz reaction, where a metal acts on halogen compounds, simply welds together two such residues, and the metal goes over to the halogen compound. While I was a young assistant, Schlubach was able to show, with the help of a special trick, that organoalkali metal compounds are produced as intermediate products in the process. He was also able to show that the reaction proceeds in two steps. Firstly the first step: Formation (ME simply stands for metal here), formation of the organometallic compound and metal halogenide. And then, in the second step: the halogen compound reacts with the metallic compound to form the product of Wurtz synthesis. But, as Schlubach observed at the time, these two reactions proceed so rapidly, one after the other, that they couldn’t be used, or at least it was believed that they couldn’t be used, to produce the organoalkali metal compounds themselves. Today we have long since overturned that view, and we know that, using a suitable technique, even organosodium compounds can be easily produced from halogen alkyls. But at the time we had no idea of this. And so we realised for the first time in the course of our kinetic experiments that you must be able to produce lithium alkyls directly from lithium and halogen alkyls. Because as I said, the reaction that we observed to be very slow, was nothing other than reaction 2 on this diagram. And that actually proved to be the case, and as a result in 1930 we were able to transfer the known Grignard reaction, as it were, to lithium, so that a whole series of organic lithium compounds became as readily accessible as the Grignard magnesium compounds. And these findings subsequently opened the way, first by us and then by many colleagues around the world, to a broad chemical field of organolithium compounds. I do not want to pursue this, but I would like to mention that in the following period we explored the reactions between lithium alkyl, particularly between lithium butyl and butadiene, very carefully. That which was originally possible with the potassium compound, we were now able to demonstrate for lithium compounds as well. Initially, a single addition product forms and then one butadiene after another links up, and products of this general formula are formed. This reaction would have been of interest for the synthesis of long, straight-chain aliphatic compounds if it uniformly proceeded according to the principle of 1,4 addition. In fact, however, these reactions were subject to two different principles simultaneously, namely 1,4 addition products resulted with formation of the straight chain, as well as 1,2 addition products with formation of the branched chain. If the next butadiene attaches there at the site where the lithium is, it is clear that this group here will form a side chain, and you get a branched product. And the more frequently that can happen and the longer, meaning the greater the number of butadiene molecules that attach, the more complicated will be the mixtures of the reaction products obtained. In fact, using carbon dioxide, carbonic acids can be produced from these substances. These are however always complicated mixtures of straight-chain and branched products. And it’s not a neat synthesis of uniform substances. So it was theoretically very interesting, but had no practical application whatsoever. For the further course of events, attention to an extremely simple problem was then crucial: Schlenk had in 1917 described lithium ethyl as a crystalline, colourless substance that is soluble in hydrocarbons and mentioned incidentally that lithium ethyl partly sublimates in conventional tubes used for melting-point determination. Now that I had large quantities of lithium alkyls available and could easily produce, I wanted, to complete our knowledge, to check in general whether you can distil lithium alkyls under suitable conditions. Because we knew that sodium and potassium alkyls are evidently purely heteropolar salts that cannot be distilled without undergoing decomposition, the question regarding the distillability of lithium alkyls was of some interest, because, according their properties, these substances occupy an intermediate position with respect to the purely homeopolar types of the zinc diethyl type. Shortly before leaving Heidelberg in 1936, I ran a few experiments along these lines, which, however, did not produce a clear result, so that the question regarding the distillability of lithium alkyls remained unanswered for the time being. I then moved to Halle and concerned myself with a series of other problems that are of no interest to us here. And finally, in 1943 I landed in Mühlheim an der Ruhr and considered what productive work I could do there at the Coal Research Institute. Sure there were back then many who believed – and you still encounter similar opinions today – that the director of a coal research institute should of course pursue research into coal. Under this assumption I had however not come to Mühlheim. I was generously granted instead free rein to explore the entire field of the chemistry of carbon compounds without regard to whether the work was related to coal. At first I had very few coworkers. Mainly, the work by my predecessor at the institute was still ongoing. So the thought occurred to me that I could revisit ground that I had not explored since 1936, and which would require little effort. So I began with this question regarding the distillability of lithium alkyls. I was then able to show that these substances can be distilled without decomposing if their molecular weight is not too high, and if the distilling takes place in an extremely high vacuum and along a very short distillation path. But much more important than the thing we were looking for, and which, by way of exception, we actually found this time around, was a side reaction that we observed. We found, namely, that lithium alkyls readily split to form lithium hydride and olefins at temperatures above around 100 degrees. That was not very exciting in itself. American authors had already demonstrated that several years earlier for sodium ethyl . And we ourselves had already shown around 1930 in Heidelberg a tendency for lithium hydride to split off from certain complex organolithium compounds. But we also found that the decomposition of lithium methyl gives rise not only to ethylene, but also to butylene, a hydrocarbon containing four carbon atoms. This means that a synthesis had occurred, and we were then able to show in suitably designed experiments, in which lithium alkyls were heated with excess ethylene under pressure, that a stepwise organometallic synthesis is possible between lithium alkyls and ethylene. The next image shows all these possibilities. Up here we have lithium methyl, which can split into lithium hydride and ethylene. And in the further reaction with ethylene under pressure it can be converted first into lithium butyl and then with another ethylene to lithium hexyl, and so forth in general. In a manner similar to that which I showed before for butadiene, it was possible to synthesize straight-chain aliphatic compounds in this way. In retrospect, you might ask today: Why didn’t you discover this very simple reaction much earlier and why did it take such a roundabout way to find this primitive reaction? The answer is as follows. Firstly: when we discovered the general reaction between organic and alkali metal compounds and olefin unsaturated hydrocarbons, it seemed that it would only be possible with olefin double bonds that were activated somehow by a suitable substituent, for example by a phenyl. Earlier, I presented diphylethylene as a reaction product, and styrene, also known as phenylethylene, also undergoes the reaction or activates via a second double bond, as in butadiene. In Heidelberg we had no ethylene available. We had tested other simple olefins, for example cyclohexene, and found that these did not result in addition. In fact, we now know that ethylene and ethylenes with substituted saturated hydrocarbon residues must not be equated in terms of their reactivity. Ethylene is by far the most reactive. Finally, before we discovered our new synthesis, the organolithium compounds were accessible only via mercury-silver alkyls, so with great difficultly. And in the long course of our research we have repeatedly observed that a field experiences a great boost only after it has become possible to produce the necessary starting materials easily in the laboratory in quantities of several 100 g or, better, in kilogram quantities. Only in Mühlheim were all the requirements met. Above all, ethylene was native to the Ruhr region from coking plants. All of this conspired in such a way that in the first few years at Mühlheim we made important advances, achieving the transfer of these reactions from butadiene to ethylene. From the moment of the discovery of this very simple way to synthesize straight-chain aliphatic compounds from lithium alkyls, the course of research at Mühlheim has been described and published many times in rather great detail, so that I can end my story here. Combining our knowledge about the tendency of lithium hydride to split off from lithium alkyls and the ability of lithium hydride to undergo addition reactions with ethylene, we began to suspect that you should be able to convert ethylenes into higher alpha olefins by using lithium hydride as a catalyst. That is shown in this reaction diagram. If you consider each reaction stage to be an equilibrium reaction involving the scission and re-deposition of lithium hydride, you can easily combine them, so that when you simply heat lithium hydride together with ethylene, you would expect all these products to appear again catalytically on the right side. Thus, butylene, hexene, octylene and the higher homologues should form simply by combining these two reactions. We found this to be true in principle, but the reaction does not run smoothly. It is complicated by side reactions and follow-on reactions. One is shown at the top. A sort of substitution reaction occurs. For example, when lithium methyl reacts with butylene, a lithium compound of butylene like this is able to form. And if lithium hydride is removed, you have butadiene, and from there it proceeds to polymerisation and the like. Those are complications that you cannot necessarily predict, but which nevertheless prevent what we were looking for to occur straightforwardly. Subsequently – and as I said, this has been described many times that we could achieve what we were aiming for very nicely with lithium hydride alone if we used the lithium aluminium hydride complex, which had been discovered in the US a short time before. And it was just after the war that we became aware of the substance. This observation, that it works very well with lithium aluminium hydride, despite working poorly or not at all with lithium hydride, this observation opened the way to proceed from the lithium alkyls to the aluminium alkyls. Surprisingly, with these it was possible to fully realize everything we had expected from the lithium alkyls and which the latter only exhibited weakly or poorly. In this context, we were particularly intrigued by the problem of synthesising straight chains from aluminium alkyls and ethylene, a process we referred to as a growth reaction, which proved highly suitable for the synthesis of aliphatic compounds, particularly alcohols with a chain length of 10 to 20 atoms, especially if you combined it with subsequent oxidation. I’ve again written down the principle of this growth reaction again for aluminium alkyl. It’s simply the addition of this Al-C bond to the ethylene. It can be written like this. The reaction can be repeated, and this chain then grows on the aluminium by attaching ethylene consecutively at this site so that ultimately this is the final product. The result must, of course, be a straight chain, because no possibility other than straight growth is possible using ethylene as a building block. These substances can enter two possible reactions. You can oxidise them with oxygen at this site. The result is then what I mentioned before: a long-chain alcohol with a terminal OH group, a higher fatty alcohol. These compounds also exhibit the property – and I’ll come to speak about that in a moment – of simply releasing aluminium hydride under certain conditions. That results in this long chain, an olefin. This could provide a synthesis pathway for such olefins, exactly what we had previously sought with lithium hydride. We did not succeed in attaching as many ethylenes to an aluminium ethyl as would be required to form a true plastic-like polyethylene. You would have to add at least around 1000 ethylene molecules, whereas the best we managed was 50 to 100, because cleavage of olefins As a result, growth could not continue indefinitely. Before that could happen, relatively short chains of around 50 to 100 ethylene molecules split off. Clearly, when we discovered this reaction of aluminium alkyls with ethylene, this growth reaction, we realized of course that this could be related to the formation of a high-molecular-weight plastic. In truth – and contrary perhaps to reports in the daily press But, of course, this drew our attention to the possibility. It looked as if it was not achievable with aluminium alkyls alone and that at best you could only approach molecular sizes corresponding to fatty acids and the like. In fact, we had already given up on this problem of achieving a true plastic. But then serendipity once again lent a helping hand. One day, during an experiment in which we were attempting to synthesize non-straight chains from aluminium propyl, a trace amount of metallic nickel found its way into the reaction unbeknownst to us. The result was that the entire growth or synthesis reaction, as we called it, no longer took place. So instead of these nice large aluminium alkyls that we had already made a dozen times, suddenly we got: first propylene from the aluminium propyl (the small “Al” is simply meant to denote a third large Al). And furthermore butylene. We also found that we no longer obtained higher products with aluminium methyl either, it was as if it were jinxed, but evidently only by catalysis. The aluminium triethyl acted as a catalyst for the smooth conversion of ethylene to butylene. The cause in principle was, of course, immediately clear. Something had entered the reaction, some trace catalyst, that strongly catalysed this reaction, which I’ve written down here, in which an olefin was displaced from this chain by ethylene. In that case, it is, of course, clear that when an ethylene attaches to the aluminium ethyl and you therefore get a chain of four carbon atoms, this displacement occurs much more rapidly than the synthesis, due to the additional catalysis, and only the first synthesis stage is fixed, forming butylene and nothing else. After we had determined this and also discovered the reason – it took us around three months to realize what was going on, because the quantity of nickel required to trigger the effect is minute, so that the cause eluded us for some time – it was a logical step, since trace metal catalysis was involved, for us to expect that the previous attempts to synthesise really extremely long chains on aluminium had failed due to the presence of similar trace catalysts and that we would have to aim to work as aseptically as possible in the future. For that reason, we began to examine what other substances might exist that act similarly to nickel and exclude them in the expectation that we could approach around 1000 molecules of ethylene by way of this simple synthesis reaction. As you will be aware, from this systematic examination something very positive, new and surprising again emerged. Instead of a large number of other substances analogous to nickel, we found that there are entirely novel, up till then unknown combinations, especially organoaluminium compounds, with, for example, the chlorides of titanium, zirconium, vanadium, chromium and others, that were completely novel, highly active polymerisation catalysts, with the help of which it is possible to produce high-quality polyethylene at atmospheric pressure. Just how surprising this effect must have seemed at the time is reflected in the fact that among the many unsaturated compounds that exist, a whole series had been known for years that very readily undergo polymerisation - including products such as Plexiglas, polystyrene and polyvinylchloride. However ethylene, the parent substance, had resisted all polymerisation attempts and it was not until 1937 that researchers in England succeeded in achieving this under extreme conditions, namely at 1000 atmospheres and 200 degrees… For the gentlemen from the press, please note: not 1000 degrees and 200 atmospheres. The figures are often reversed, resulting in complete nonsense, because such substances do not withstand 1000 degrees. Anyway, that was discovered in 1937. And at the end of 1953 – 16 years later – we then discovered that the entire reaction easily proceeds with ethylene at one atmosphere with these highly active polymerisation catalysts. We called these combinations of certain heavy metal compounds with organometallic compounds of aluminium and other metals organometallic mixed catalysts. And it was found that they can be used to polymerise not only ethylene, but also may other substances to produce a large number of novel plastics. In particular, using our catalysts, others – not us – succeeded in producing a substance that is truly identical to natural latex. I have compared the entire development we set in motion ten years ago with an explosion of scientific knowledge and technical development. Let me explain to you why this comparison is not entirely unjustified with the help of two maps I will give you (the next image please). The first is a map of Europe and the second a map of the world, on which the factories that produce high-molecular plastics, whether polyethylene or others, using Mühlheim catalysts are marked in red. Among them, as you see, are three factories in Russia. Now, those marked in green by contrast are plants in which organometallic compounds, especially aluminium compounds, are produced and processed. The field has also developed industrially, but not to the same extent. And please now the map of the world. The same appears again on the world. Because of the small scale, the individual points are difficult to make out, so I have written figures that express the number of plants at each relevant location on the world map. As you see, there are three in Japan. So, that is the end of the development and it may be of interest … or perhaps the current state of development, probably not the end. It is perhaps interesting to see how it all looked ten years previously at the Mühlheim Institute. I would therefore like to describe to you in detail the small apparatus which we used to study the normal-pressure polymerisation of ethylene and then present a short film. You can easily image what a setup would like to run an experiment at 1000 atmospheres. The apparatus is a bit unusual. The glass vessel in the middle is a conventional five-litre jar. It is fitted with a stirring device. At the top is the drive; the motor is behind the paper, and to the right is the ethylene steel bottle with a reduction valve and manometer. Here is a wash bottle, which can be used to estimate the flow rate by counting the bubbles. Then on the other side is again … you can see how the ethylene escapes unchanged, there’s a thermometer in it, and so forth, which is what such an apparatus looks like. Here are a few coloured pictures showing how the experiment was run. Afterwards you will see the same thing in the film. Unfortunately, it’s an amateur film we made ourselves. It’s a bit fast, and it’s not always possible to explain what’s going on in detail. Therefore I’m going to show you parts of the apparatus again beforehand in static pictures. Here the stirring has begun. You can see that it’s foaming here. Here is the liquid inside, and the foam has risen. This is nothing more than a petrol-like hydrocarbon serving as a suspension medium. And here we have added a mixture of diethylaluminium chloride and titanium tetrachloride. A brown precipitate forms. That is the polymerisation catalyst. It’s 9:05. It is still very thin. No ethylene has been added yet. It’s just the catalyst alone. Perhaps a gram or so of it is suspended in the mixture. In the next stage, about an hour later, you see that the brown catalyst is still there. But you can already see a thick mass forming. Up here some of it is already adhering. So polyethylene has been formed by introducing, say, per hour 200 litres of ethylene. This is very rapidly absorbed, at the same time the substance precipitates, the temperature rises, and you have to cool the setup and hold it at around 60 or 70 degrees. And the catalyst is still in it. Now you can decompose it. Up to now the reaction has to be carried in the absence of air. You can decompose it by adding alcohol and admitting air. The thing then looks snow-white. This is somewhat later. You can then filter the thick slurry (the next picture). This is in the same experiment. The decomposition process has not become so prominent. You can see that brown catalyst is still present. After it has been exposed to air for a short time, it turns snow-white. Here you see a thick slurry being put on the filter above the vacuum flask. And the end result is a heap of white powder on the table. Here is a picture from the very early days. That’s the first tube which we late in 1953 fabricated ourselves from the material, and we first formed a solid by hot pressing and then drilled out the cylinder. We then conducted a pressure test with it. I believe the inside diameter was around 12 mm and the walls were perhaps 4 mm thick. And we pressed water into it, and the tube burst at 180 atmospheres. So we had a material with very remarkable properties, after all such performance had not been possible with the earlier high-pressure polyethylenes. It had a somewhat different structure for specific reasons. You can see that here too: If you press two objects made of the conventional product and the new material, in this case two beakers, you can see that the one is quite floppy and soft, while the other retains its shape pretty well. Using this simple test, you can always tell very easily where the products come from. However, the test is no longer perfect, because meanwhile high-pressure polyethylenes have been improved so that they are now much stiffer. Anyway, that was the state of affairs at the time. Nowadays, the result wouldn’t be quite the same. But there has always been a notable difference And how it today ... Those are the small objects from the early days. Now I’d like to demonstrate to you on the maps what progress has been made by showing you the largest vessels that can be made from the material today, which of course has excellent properties in terms of resistance to aggressive agents. Those are the largest containers with a capacity of several cubic metres. And the fact that they are very light can be seen in the picture, which shows how just a few people can easily lift such a large container. Incidentally, there’s no need to construct demonstration objects of this material for lectures, because all you have to do is stroll through the streets of the city. Right here in Lindau I saw a display window yesterday filled with objects that can be made from this material. And I also want to show you … all this in the time-lapse film, which unfortunately But now that you’ve seen the individual components, it’ll be easier for you to understand the film And here it is now …, that’s the apparatus again, similar, here ethylene is fed through, here then the catalyst is produced. One component. What’s more, time 0 is registered on the clock. Now comes … that’s probably the organoaluminium compound in a solvent, yes, it’s labelled: diethylaluminium chloride. Then titanium tetrachloride is drawn up with a pipette and is added. Only now does it turn yellow. After a short time, this precipitate forms. Gradually it turns cloudy. Now the entire mixture must be transferred to the other reaction vessel. That is done with the help of a siphon. Now it’s beginning. Here you see the reaction, here the stirrer, which you saw before, and how the catalyst is introduced with a siphon, which enters below as a cloud. At the top it becomes less and less. The whole thing took about ten minutes. Now the stirrer is switched on and the ethylene flow is started, right … left …, what you see: for the time being the flow rate is the same on both sides; it leaves as quickly as it enters. That’s the upper part. Now it’s quite rapid, and now you see a vigorous flow of ethylene on the right. On the left it’s much weaker, although it is being fed in at a furious rate, the ethylene is being completely absorbed. There’s even a slight vacuum. The temperature rises. You can see, it’s just under 50 degrees … 60 degrees. And now you can see already how the precipitate is forming. That’s … the polymerisation reaction, running now for around 20 minutes. You can already see how the whole thing is getting thicker. Nothing is coming out here anymore, here’s a weak vacuum. Now you see the material everywhere in the upper part of the reactor. The ethylene stream is measured again with a rotameter. It’s becoming increasingly viscous. If you don’t have a very powerful stirrer, it remains stationery at the end. This is now one hour after the start. That’s how the whole thing looks now. Now … the decomposition starts, now alcohol comes in and air. You’ll soon see, now it turns snow-white. And then you see the filtering process and this film …The material is still hot. You can see the vapours rising from it. And this way you can … the liquid is sucked out from below and the whole thing is pressed. Afterward, it can be removed from the filter. The result is a heap of white power, which is not shown in the film. Finally, the film shows a few objects made from the material. Here you have the tube again and a few …, the first plastic films, which were not very successful, which we made ourselves. Thank you. Thank you very much for the applause. However, I’m not quite finished. Please allow me just a few more minutes. I have tried to trace the entire causal chain leading from my beginnings 40 years ago to an important new finding for the industry. This development was always driven by experimentation and observation. The whole secret was actually nothing more than us – I speak in the plural because, of course, many diligent people were involved – exploring a still uncharted but doubtlessly interesting field of organic chemistry with eyes open and minds alert. For decades we did not think in our wildest dreams that our investigations into organometallic compounds would have such ramifications. Given the events in Mühlheim over the past decade, some of my colleagues here If so, you do us an injustice. Because it was always driven by our endeavour to expand our scientific knowledge. Of course it was gratifying that the results of our work allowed us a great deal of independence and freedom in our scientific research. But otherwise I regard the things many outsiders believe to be particularly important and decisive more as by-products of our actual work and, I must admit, often as very annoying by-products. Annoying insofar as everything our research produced necessarily arose from my personal need to investigate many outlying fields. However, that limits the options for continuing to follow my own recipe for success, namely to wander without prejudice through the fields of organic chemistry with eyes wide open, searching in all directions. I’m now in a position at my institute to see that research along these lines is being continued by outstanding young colleagues with great success, and eventually we all find ourselves in the situation where we have to recognize that our main purpose is to pave the way for the next generation. Another aim of my lecture was to show you that it’s possible to start in organic chemistry, like Mr. Otto Hahn, and to be successful while sticking to the field of organic chemistry. But I have to qualify that a little, because our entire work also had an inorganic element in that it was the organometallic compounds, meaning the products of a kind of symbiosis between the two main disciplines of chemistry, which ultimately led to success. I have always believed that you should not draw sharp lines between individual subdisciplines, and I have therefore always felt that I’m first and foremost a chemist, and not, for example, a macromolecular chemist or an organic chemist or a plastics expert. And it is my wish that, despite the ongoing process of specialisation, similar basic attitudes will continue to be possible in the future.

Karl Ziegler (1964)

From Triphenylmethyl to Polyethylene - Less Well-Known Facts About the Developments Leading to the Invention Made in Muelheim (German presentation)

Karl Ziegler (1964)

From Triphenylmethyl to Polyethylene - Less Well-Known Facts About the Developments Leading to the Invention Made in Muelheim (German presentation)

Comment

The present lecture is the first of two Karl Ziegler ever gave in Lindau. It is remarkable for several aspects, particularly for its honesty. Ziegler gives a very clear account of the complete scientific development that led to the invention of the famous Ziegler-Natta catalyst (1963 Nobel Prize in Chemistry to Karl Ziegler and Giulio Natta). In doing so, he does not leave out or sugarcoat any detour taken. He also makes it unmistakably clear that he never intended to develop a catalyst and initially did not even have a remote interest in technical applications of his organic chemistry research. Looking back from the point of view of his Nobel Prize worthy results, the experiments he describes thus paint the picture of a rather windy road to a hardly anticipated success.

However, Ziegler also points out that as soon as the far-reaching importance of his results became apparent, he followed up on them with quite some curiosity, focus and vigour. Towards the end of his talk, he describes his approach as an unbiased, impartial and attentive hike through the world of chemistry, combining aspects of the organic, inorganic and technical manifestations of the field. He thereby raises a strong point in favour of the benefits of fundamental, untargeted, interdisciplinary research.

This point is emphasized further when one considers the impact of Ziegler’s work on science and society. Today, just as in 1964, the year of the talk, Ziegler-Natta catalysts are used worldwide for the large scale industrial manufacture of polyethylene (from ethylene) and polypropylene (from propylene), two of the most common and widely employed plastics. The total global annual production of these two materials is well in excess of 100 million tonnes, generating an annual market of around 200 billion EUR. In fact, it is highly unlikely that you do not have a Ziegler-Natta based plastic in your reach while you are reading this. Water bottles, plastic bags, stationery, but also car parts and laboratory equipment are just a few examples.

From a chemical perspective, the importance of transition metal to carbon bonds (e.g. carbon to titanium bonds in the case of the Ziegler-Natta catalyst) in chemical catalysis surged in the second half of the 20th century. The result are several important technical processes such as the Monsanto and CATIVA processes for the large-scale production of acetic acid from methanol as well as a plenty of Nobel Prizes in Chemistry, the most recent ones being the 2001 Prize to William S. Knowles, Ryoji Noyori and K. Barry Sharpless, the 2005 Prize to Yves Chauvin, Robert H. Grubbs and Richard R. Schrock as well as the 2010 Prize to Richard F. Heck, Ei-ichi Negishi and Akira Suzuki. The way towards these impressive developments was paved by Ziegler’s early success with transition metal catalysis, a success, which according to Ziegler himself, essentially depended on the academic freedom he enjoyed during his career.

David Siegel

Cite


Specify width: px

Share

COPYRIGHT

Cite


Specify width: px

Share

COPYRIGHT


Related Content