On Strong Inorganic Bonds and Cation Cages

Strong inorganic bonds were mentioned in an earlier post. The term as used is descriptive and informal. I use the term to describe a bond formed in an irreversible reaction, one that runs to completion and where the bonds formed are permanent except under extreme conditions. If you have thoughts or suggestions on how to best describe or characterize such bonds, I would very much like to hear them.

The thing about strong inorganic bonds is that they are useful. To form coatings, you must start at some point where you only have a surface to be coated and then, through a series of generally irreversible processes, you must bring the elements of your coating together and form a lasting deposit on said surface. During processing, strong inorganic bonds are useful because they control the progression and direction of the processes. When you develop coatings, you develop more than the final composition of mater ultimately formed; you must also develop the processes for applying the coating. When the coating process is complete, there is generally some requirement for permanence or persistence of the coating. For the finished coating, strong inorganic bonds are useful because they impart durability and chemical resistance.

The purpose of this blog is to discuss coatings and coating development, not to give lessons in materials science unnecessarily. So, I’m going to try to cut to the chase and rather than go into the theory behind the nature of strong ionic bonds, I'm just going to jump ahead to the answer: strong inorganic bonding is due to a combination of ion size, cation oxidation number and the cation cage.

Let’s consider the example given in the earlier post. The example was that of sodium chloride, a highly ionic inorganic compound that readily dissolves in water, and aluminum oxide, a highly stable inorganic compound that is insoluble in water. In the following table, there are some quantities given. I’m not going to go into the theory behind these quantities or explain why I selected them except to say that these quantities illustrate some important differences between the two compounds and provide clues as to why the bonding in these two materials is so very different.

Figure 1. Comparison of selected properties of aluminum oxide and sodium chloride to provide clues for and to motivate an understanding of the origin of the differences in the bonding in these materials.

Figure 1 lists ionic radii for aluminum oxide and sodium chloride. The values include the anionic radii, r_a, for oxygen and chlorine and the cationic radii, r_c, for aluminum and sodium. The oxidation numbers, z_c, for aluminum and sodium are also given. Following that, there are various derived values including the ionic radius ratio (cation-to-anion ratio) and the Dietzel field strength for both compounds. Also, for reference, the melting point and single-bond strengths are given.

One of the first things that figure 1 brings home is that the sodium chloride ions are much larger than the aluminum oxide ions. Next, the oxidation numbers for the aluminum oxide ions are larger than for the sodium chloride ions. Therein lies much of the story: small ions with large oxidation numbers promote strong inorganic bonds. Those are clues, but they don't really explain anything as they suggest a difference in degree and not a difference in kind.

When we looked only at electronegativity earlier, sodium chloride and aluminum oxide seemed very similar: they both should be about 63% ionic in character but the behavior of the sodium chloride and aluminum oxide is such that it suggests that the bonding in these materials is of two different kinds. How can we account for the differences? A short answer would be that we can ascribe the difference to the shielding and tightness of the respective cation cages.

A cation cage is a structure found in most ionic and inorganic materials. The cage is not made of cations. Rather, it is made of anions. Anions are usually significantly larger than cations with the result that in most ionic and inorganic compounds, the anions are tightly packed and the cations take up positions in the small gaps between the anions. Think of pool balls tightly packed into a rack. The anions are like the pool balls: tightly packed and touching each of their neighbors at a single point. Like the pool balls in the rack, even though they are tightly packed, there are still small gaps or spaces: interstitial sites where a smaller object could be fit in. It is in these interstitial sites where the cation cages are formed. Not every interstitial site will act as a cation ion cage. Generally only a portion will be filled and the filled sites are generally distributed in a network of filled sites so as to preserve local electroneutrality. Each cation cage traps a single cation in the center of the cage. The anions that surround the cation shield it from the repulsive forces of nearby cations. The tighter the cage and the more effective the shielding, the stronger and more stable the bonding. We listed two properties in the table, melting point and single-bond strength, where the effects of the cation cage can be detected but the effect of the cation cage can be seen in many other properties as well such as hardness and density.

Cation cages are the building blocks of inorganic coatings. Being able to control the size, shape and distribution of the cages that are formed is an essential development tool. Cation cages come in a number of different shapes depending on the packing of the anions and the cation-to-anion radius ratio. The most common cages are tetragonal, octahedral, and square antiprism. Tetragonal is the simplest.

Figure 2. An aluminum(III) ion enclosed in a cation cage. The oxygen ions are the anions which form the cage and are centered on the cage's vertices (corners) shielding the cation from the charges of other nearby cations.

In figure 2, the first view is of the cation cage shown as a wire frame with an aluminum(III) ion disposed in the center of the cage. The next two views show how the oxygen ions are disposed at the vertices (corners) of the cage and how they act together to form the cage and to shield the trapped aluminum(III) ion. I find it helpful to think of the anions as pool balls in a rack except, instead of being packed in a two-dimensional arrangement, it is packed in various three dimensional arrangements that might comprise two or more layers. Then I like to remove a few pool balls at a time and look for the different interstitial spaces that different packing arrangements can provide.

The aluminum(III) ion, when surrounded by oxygen ions, does not normally reside in a tetragonal cage. It prefers an octahedral cage because the cation-to-anion radius ratio (0.536) is too large for a tetragonal cage. However, adding cations with high oxidation numbers, z_c > 4 such as phosphorous(V), promotes packing arrangements of the oxygen anions that can cause the aluminum(III)'s cages to adopt the tetragonal form.

The aluminum(III) ion in α-Al₂O₃ is enclosed in an octahedral cation cage.

Figure 3. An aluminum(III) ion enclosed in an octahedral cage. In the second and third views, the oxygen ions are added one level at a time to help visualize how the cation is disposed within the cage. In the bottom layer, the disposition of the anions is identical to the disposition of anions for the tetragonal cation cage shown in Figure 2. The shape of the octahedral cation cage could be described as a trigonal antiprism. It usually isn't.

Figure 3 shows the aluminum(III) ion in it's most usual configuration with oxygen ions. This configuration is very useful as it is possible to substitute other cations of about the same size for a portion of the aluminum(III) ions in a coating formula. This enables selective modification of properties, such as melting point, while retaining some of the chemical stability and transport properties of the unalloyed aluminum oxide. Examples of cations that could be substituted for a portion of the aluminum(III) ions would be the iron(III) and chromium(III) ions or any trivalent ion with a radius between about  52 and 81 pm. Thus yttrium(III) would be too large and boron(III) would be too small.

Sodium chloride is unable to take on the tetragonal or octahedral cation ion cage configurations. Because its cation to anion ratio is large (0.695), it must take on the larger and more complex square anti-prism form.

Figure 4. A sodium ion enclosed in a square antiprism cage. The shape of the cage is complex and comprises eight identical vertices and has two types of faces (two square and eight triangular).

Figure 4 shows a sodium ion in a square antiprism cage. The figures 2, 3 and 4 have been drawn to a common scale to help illustrate some of the features of the different kinds of bonding and to bring out some of the nature of the strong inorganic bonds. It can be seen that the ions in sodium chloride are significantly larger than those in aluminum oxide which leads to lower Dietzel field strengths in the sodium chloride. The sodium chloride cation cage also comprises more anions leading to lower single-bond strengths.

We can conclude from these illustrations that if we want to develop inorganic coatings with good durability and chemical resistance, we should be looking for cations with high oxidation numbers, small ionic radii and a small cation-to-anion radius ratio.

Cation cages are related to the topic of Coordination Number (CN). The coordination number for an anion is the number of cations that it touches. For oxygen, this is generally two. For a cation, the coordination number is the number of anions that it touches. For the cation cages that we looked at in this post, the coordination numbers would be 4, 6 and 8 for tetragonal, octahedral, and the square antiprism, respectively.

I think that that is enough talking about bonding for now. It's time to put some of what we have discussed to use and to see how it plays out in the real coating world. What we have discussed is going to be very useful and you can bet that the cation cages will play a significant role. Our upcoming discussion will cover many fields of application and, while what we have discussed so far will be useful, there are many big surprises to come. I am looking forward to it.

The next post will be about how to form coatings and will discuss the many ways that aluminum oxide coatings can be formed as they are very widely used in architecture, manufacturing and the power generating industries.

On the Coating of Aluminum, Memory and the Aluminum(III) Ion

The subject of aluminum and coating of aluminum came up in a recent post. The subject is an interesting one with many surprises that illustrate processes used to form coatings and the types of coatings that can be formed. Before we get started, there are a couple of things about aluminum that should be noted.

Perhaps the most important thing to know about aluminum is it has a long memory. A very long memory. When we say that a material has memory, we are referring to a tendency of materials to retain the effects of their processing history. The effects of early processing steps tend to fade with additional processing but some effects are lasting and can even change the outcome of later processes. We sometimes use the word memory to describe that tendency. Aluminum remembers everything that ever happened to it from the time is extracted from the parent bauxite ore (using the Bayer process) and reduced (using the Hall-Héroult process) and then through all the handling, shaping, forming, heat treating, tempering, fabrication and use. Most materials possess some degree of memory but aluminum’s memory is exceptional: it is eidetic.

The second thing to know about aluminum is its ability to form strong and stable bonds with oxygen. This ability is so important to the coating developer that it is worth considering why it is so and, along the way, to introduce some important concepts. The first concept we should cover is electronegativity. Electronegativity is the attraction that an atom has for electrons. All atoms are electronegative to some degree. Aluminum atoms are not as strongly electronegative as oxygen atoms with the result that when aluminum and oxygen react to form a compound, the oxygen swipes all three of the aluminum atom’s valence electrons (valence electrons are the electrons in an atom’s outer electron shell) producing the aluminum(III) ion (which is a cation because it has a positive charge. If the charge were negative, it would be an anion).

The aluminum(III) ion does not behave like ordinary “highly ionic” ions such as the sodium ion. Ions are highly ionic when the bonding between the cation and the anion is principally electrostatic due to the ions’ opposite electric charges. The sodium ion is a cation and it is what makes sea water taste salty. The salt in sea water is primarily sodium chloride which is table salt. When you put a spoonful of table salt in a glass of water and stir it, it readily dissolves because the water is a polar liquid and can easily separate the sodium and chlorine ions to form a salt solution.

The aluminum(III) ion is not highly ionic when the anion is oxygen, and it is not soluble in water even though the difference in electronegativity between aluminum and oxygen is about the same as the difference in electronegativity between sodium and chlorine. Unlike the ions of sodium and chlorine, the ions of aluminum and oxygen form a strong and stable inorganic bond which is what makes the aluminum(III) ion so useful to the coating developer. The strong inorganic bond keeps the coating from washing away in the rain and imparts durability and chemical resistance.

Several concepts have been touched on in this post: memory or the lasting imprint of processing on a material, electronegativity, the types of ions and the broad spectrum of inorganic bonding. In future posts, I hope to show how these are very useful tools in the coating developer’s toolbox. In the next post, I plan to discuss the nature of strong inorganic bonds and show why understanding them is so valuable.

Reflecting on the Bloodhound SSC Project

Rod Macleod brought up the subject of the Bloodhound SSC project. The subject was new to me, but there is an excellent entry in Wikipedia about it that explains what the project is about (building a supersonic car) and speaks to why it is being done, how, and what is hoped to be gained.

While thinking about the Bloodhound project, a lot of interesting memories came to mind. In the course of developing coatings, I have had the pleasure of meeting and working with many top-notch scientists and engineers. As a group, they sometimes get the reputation for being dull and rather boring; they’re just not the ones that you expect to be the life of the party. I think that this reputation is unfortunate and undeserved. But, I also think that this reputation is understandable when you consider how focused these people are. Their attention is tightly focused on their work. They don’t just plot-up their data and call it a day. They study their data, memorize it, agonize over it and reflect on it. Obviously, this is not something that everyone does or can do or would choose to do if they could. It takes a certain passion. To an outsider, one who did not understand what they were going through, or one who was not similarly afflicted, their demeanor could come off as distant, detached and disinterested.

Actually, I have found that, as a generality, these people are very approachable, especially after regular office hours when the hubbub of the day’s activities have subsided and the quiet of the evening has set in. Then, in that setting, you can have some very interesting conversations and, if you are so inclined, you can learn some very interesting things. In the space of 40 minutes or maybe a few hours, you can learn more than you could have learned in your whole career. And, it’s not things that you could ever learn from a text book; it’s much better than that. They tell you about their frustrations and the challenges that they are up against. If they have a complaint, their complaint is usually directed toward the basic cussedness of nature. Nature has a way of staying a few steps ahead of us and is always more complex than we had hoped. Text books don’t reflect this and the popular view of science is that science provides all the answers. It doesn’t. The best science can usually do is to provide some of the questions, a few at a time.
That’s one of the things that make the Bloodhound SSC project so interesting. What the scientists and engineers are doing here is that they are putting something to a test. What they are putting to the test is not the Bloodhound craft itself or even their knowledge and understanding of the underlying science and technology. What they are putting to the test is their curiosity and their ability to learn.

So, what does all of this have to do with being a coating developer? Quite a lot, actually. Rod limited his question to the outer surfaces of the Bloodhound craft. On something that complex, the outer surfaces, well – please forgive me, they hardly scratch the surface of the potential for coating applications. Wherever you have a surface, an interface between two media, you have a potential application for coatings, especially if one of the media is abrasive, erosive, corrosive, or chemically or energetically aggressive. The Bloodhound craft is loaded with these, but I think that, for the most part, that these surfaces are well characterized and provided for. Those surfaces really don’t worry me. There is the matter of the bespoke hybrid rocket engine. That’s not something I know much about. A number of my co-workers worked on the NASA shuttle engines right up until their recent retirement. They will know much more than I do.

The thing that captures my attention about this project is the wheels. The wheels are to be made of aluminum. Now, aluminum is inherently a soft, ductile, malleable material with a very low specific gravity and a fairly high specific strength. When we say that a material is soft, we are referring to how it responds to pressing a hard indenter into it. It is a standard test. The way that this test works is that you place a hard indenter on the surface of the material that you would like to test. Then, you apply some pressure. What happens is that the hard indenter sinks into the softer material that you are testing. When you remove that pressure, there is some rebound. The surface of the softer material tends to return back to its original shape. That sounds all well and good, but if you carefully examine the test point under a microscope you will see something very interesting. Provided that you applied enough pressure while making your test, you will find a mark on the surface of the softer material and you can use the size of that mark to determine some rather fundamental properties of the material you are testing. This test is actually a rather common quality control test, and it is used in many industries. (What happens, you ask, if the hardness of the indenter and the material you are testing are equal or very close. Ah, well, you would have to ask. But we don’t have time for that right now.)

The thing is, aluminum is inherently very soft. If you push something into it, it tends to leave a mark. As the Bloodhound craft travels over the 12 mile test track, it is going to encounter a lot of very hard materials, and they will leave a mark. Actually, I don’t think trying to coat the wheels is a good idea. No coating would adhere or last at those speeds. No. The thing that I think should be done is to coat the test track to hold down the dust and to ensure that the hardness of the material that the wheels roll over is less than the hardness of the aluminum wheels.

On Developing Coatings

On and off, I have been developing coatings for over 30 years. It has been one of my favorite occupations. Coatings are very interesting from a scientific point of view (they have lots of surface area) and from a technical point of view. Coatings are used mostly to protect and preserve products, but they can also be used to impart properties to the base material of the product that the product would not otherwise have, such as color and other optical properties such as emissivity. Properties such as easy release of ice, water replency, and antifouling can also be imparted. In one project, I was assigned the task of developing a non-catalytic coating. I was asked to do this by NASA. They were working on hypersonic vehicle designs.

Hypersonic vehicles go faster than the speed of sound. Much faster, say somewhere between 5 and 25 times the speed of sound. At those speeds, the skin of the hypersonic vehicle gets hot, very hot. How hot, well, that depends on a lot of things such as the shape of the vehicle and properties of the skin of the plane such as emissivity. That's the second time I have mentioned emissivity so I should explain what it is. Emissivity is an optical property of a surface that controls how efficiently the surface radiates heat: the higher the emissivity, the more efficiently the surface radiates heat and, in an environment such as that formed in hypersonic travel, emissivity controls how hot or cool a surface becomes. Emissivity is thus an important property to control, and it can be controlled with coatings. For the environments created by hypersonic travel, though, it is not the only important surface property that needs to be controlled. Catalicity, or catalytic efficiency, is an important surface property that can have dramatic effects on the transfer of heat from the air flowing around the vehicle to the surface of the vehicle.

So why do hypersonic vehicles get hot? It's for the same reasons that meteoroids burn up on entering the earth's atmosphere. The usual explanation given for this is atmospheric friction. This explanation is not quite right. There is some pretty complex physics and chemistry going on. As a meteoroid or hypersonic vehicle travels through the atmosphere, a shock wave forms in front of it. That shock wave contains some very energetic gas, which is one of the reasons that meteors are visible streaks of light in the night sky. The gas in the shock layer has enough energy to make the gas glow and even to melt the surface of the meteoroid. For hypersonic vehicles, melting the surface would seem to be something that we want to avoid. That's not always true. The coatings on the surface of the Apollo space capsules were designed to melt and "ablate" during reentry. Ablation was used to absorb the heat and then evaporate taking the heat away and keeping the space capsule cool. For reusable hypersonic vehicles, ablation is not a great strategy. Something more permanent is needed and you need a different strategy. To find that strategy, looking at the details of what's going on is crucial. I said that some pretty complex physics and chemistry is going on in that shock layer. One thing that is happening is that the gas is dissociating and ionizing. The gas in our atmosphere primarily comprises two gases: nitrogen and oxygen. Other gases are present in very small quantities such as argon, water vapor and carbon dioxide. Nitrogen and oxygen, in the earth's atmosphere, are present as the molecules dinitrogen and dioxygen. In other words, nitrogen is present as a molecule comprising two nitrogen atoms, dinitrogen, and oxygen is present as a molecule comprising two oxygen atoms, dioxygen. This is important because, in the shock wave, the molecules of nitrogen and oxygen can dissociate into their atomic forms. At first, this is a good thing. Dissociation of these molecules into their atomic forms uses up some of the energy in the shock wave so there is less energy available for heating the meteor or hypersonic vehicle. Oxygen tends to dissociate first and then nitrogen at higher energies. So far, so good. But, here's the down side: once you have atomic oxygen and nitrogen present in the gas surrounding the meteoroid or atmospheric vehicle, they are chemically very active. That's where the properties of the surfaces becomes important. If the surface is a good catalyst, and the surface of most meteoroids are, the transfer of heat to the surface is greatly enhanced by the recombination of the atomic oxygen and nitrogen. If recombination can be avoided, you get the upside of dissociation, less sensible heat in the surrounding gas, without the downside of more efficient transfer of heat. Hence, the request by NASA to develop non-catalytic coatings for its hypersonic vehicle designs.

As you can see, developing coatings is a highly interesting field and one that takes you into many areas of science and technology that you would not initially expect. And that's what I love about developing coatings: there's so much to learn.