WSSP Session 1 — Inorganic (Supra) Molecular Chemistry

This post is a follow-up to my previous blog post. This post will share about Dr. Anthony Cozzolino’s research at Texas Tech University on the self-assembly of nature-inspired reversed bilayer membranes using the supramolecular interactions of antimony and bismuth.

To start, Dr. Cozzolino gives a quick primer on basic chemistry. A lot of the topics covered were materials that I had learned in previous chemistry and organic chemistry courses!

Here we started with discussing how atoms are held together. When electrons are interacting and there are opposite charges between two things, we have covalent bonds between atoms. An example of this is two hydrogen atoms coming together and overlapping while the electrons end up in between the two nucleic and are binding the two nuclei together. This is written with a single line that’s representing the two electrons between the hydrogen atoms. The electrons are equally shared and right in the middle. The cost of breaking this bond is 436 kJ/mol

Similar to covalent bonds, we have polar covalent bonds between two different atoms with different electronegativities. In my organic chemistry course, we had defined polar covalent bonds as having an electronegativity difference of 0.4–1.69 between the two atoms, as written in the textbook Organic Chemistry, 8th Edition, by Paula Bruice. When a polar covalent bond is formed, electrons are shared, but not equally: the more electronegative atom pulls more electrons towards themselves. When electrons are pulled towards one atom, the nucleus is adjusted and dipole moments are established with partial positive and partial negatives.

Ionic bonds are when electrons are no longer shared but instead are transferred. Bruice defines this bond as having an electronegativity difference greater than 1.7 between the two atoms. An example is lithium, a metal, reacting with fluorine. Lithium donates an electron to fluorine, so lithium becomes positively charged and fluorine is negatively charged. The strength of the ionic bond can be determined through electrostatic attraction following Coulomb’s law: the force of attraction is proportional to the charge on the two species divided by the distance squared.

Next, we discussed various relevant trends present in a periodic table.

One key trend is size: in general, it decreases from left to right and from bottom to top. Another key trend is the metallic character: the elements become more metallic as they get away from helium (He). The ability to lose an electron, also known as ionization energy, is another important trend. Things are easy to ionize, or able to easily remove electrons, at the bottom left-hand corner and harder to ionize at the top right corner. On the other hand, gaining an electron is more favorable at the top, between flooring and chlorine. These elements have high electron affinity and can easily gain electrons. Electronegativity is the ability of an element to attract shared electrons towards itself. There are high electronegativities in the top right-hand corner and they lower as one travels farther away from fluorine. In all, these trends help us figure out what elements can be paired together to form covalent bonds.

Now, we discussed ways to measure atoms and molecules. Their size, particularly, is relevant here.

If we have an isolated molecule or atom and it encounters another atom or molecule and they don’t form an attractive attraction (meaning they don’t stick to each other), the closest distance they can approach each other is the van der Waals radius. The units to measure the size of the atoms and molecules are angstroms, and the molecules of particular interest are hydrogen, oxygen, and antinomy (Sb).

For the atomic surface, computationally, we can think of where the electrons are. The electrons are what make up the size of the atom. Electrons are less likely found further away from the positively charged nucleus. We can draw isosurfaces and find the 0.002 au where 99.8% of electron density resides.

The covalent radius is when atoms share electrons. The distance between the nucleus and the halfway point between the points. About exactly halfway exactly is when two atoms are the same. The covalent radius of hydrogen is smaller than the van der Waals radius for hydrangea. This is small for other elements as well.

The ionic radius is the measured radius when two electrons stick together and are held together by pure electrostatic attraction. Assuming they’re hard spheres, we can get the radius. H+ is the hydrogen that loses electrons to form a hydrogen cation. H- is when the hydrogen gains an electron. In an H- molecule, there are two electrons for each proton, so the radius is huge compared to the van der Waals radius of the neutral hydrogen atom and the covalent radius. The size of the radius can change depending on how many electrons the molecule has.

There are various depictions of molecules in chemists. Take butane, a 4-chain alkane, or benzene, a ring with 3 double bonds. In organic chemistry, particularly, there are various representations that chemists can use to display these hydrocarbons that have bonds coalescing on carbons and hydrogens only. Dr. Cozzolino’s research lab is particularly interested in the properties of the surface of the molecule.

So, from there, we were introduced to supramolecular chemistry, the study of the weak interactions between molecules.

Supramolecular chemistry deals with building complex systems using small molecules or bonds. It is a powerful way of making complex structures. For example, the relatively weak interactions bring together almost every component in the human body. Although these are relatively weak interactions, they are very precise and reproducible in such a way that it is error-free.

Take the uniqueness of snowflakes: these are really remarkable structures. Unique snowflakes result from the ability of water molecules to stick together to other molecules in a way that is not rote such that each snowflake is unique. Yet, nature has its own way to make this unpredictable behavior very predictable by making sure there are multiple points of attachment for molecular design.

Crystals are a repeating motif in 3D chemistry. We can use x-rays to probe and get a picture of how molecules are organized in 3-dimensional space. Distances help us determine the interactions.

Hydrogen bonding occurs when you have a hydrogen bonded to a very electronegative atom. It’s represented as H-X where X is N, O, F, as these elements pull electrons away from the hydrogen to make it partially positively charged. On another molecule, like water, some of the electrons are existing as lone pairs, and if they sit on an element that has a partial negative charge, there can be an attraction between the opposite partial charges that bring them together. The evidence for the weak interactions occurring and holding the molecule together are well defined in the image above. They are highly organized in a way that maximizes the number of hydrogen bonds that occur.

Now, we ask the relevant question of whether this really matters? Why is hydrogen bonding so important?

Pictured above is a graph of what the earth would look like without hydrogen bonding. We know that water boils at 100 degrees celsius. When diatomic hydrogen is paired with different group 16 elements, Te, Se, S, and O, there is a trend that the boiling point goes down (indicated by the red line). If we follow that trend, we would expect that water would boil close to -100 Celsius, but that is not the case. This is due to hydrogen bonding and its unique properties. If there were no hydrogen bonding, ice wouldn’t float and flakes wouldn’t freeze. H2O (or water) is an anomaly in terms of boiling point, which is an important and unique property. It does not follow the trends as we get smaller in size, in terms of the group 16 elements.

A key idea of supramolecular chemistry is self-assembly: the process by which two molecules, different or the same, recognize each other and spontaneously come together to form a supramolecular. Nature created this since if everything didn’t fold and come together similarly each time, the body couldn’t function. Each action or mechanism of the human body arises from coding molecules and the ability to recognize each other very specifically. Nature has had time to perfect this and is a master at doing it.

Below is an image of Sulfate (So42-) and double-stranded RNA. There are very specific sites where the hydrogens can bond to sulfate and recognize it, as depicted by the orange regions.

Double standard DNA has a way in which the base pairs recognize each other, self-assemble, and zip up the DNA in a very specific way. The hydrogen bonds between two molecules don’t do anything by themselves, but they do recognize each other very physically and self-assembly.

Dr. Cozzolino does not use hydrogen and its bonding but instead uses pnictogen bonding through which he is designing and studying a new set of tools of supramolecular chemistry.

Pnictogens are elements in the same group as nitrogen. The design and study of this new tool of using pnictogens and figuring out how to be useful and complement what we can already do with hydrogen bonding is a new area of supramolecular chemistry. We start by discussing the weak interactions of the p-block of the periodic table, where the pnictogens reside and compare them to hydrogen bonding. The distance between the bond in interactions between p-block elements is an intermediate distance and is based on the strength.

Chalcogens are the elements under nitrogen — such as oxygen, sulfur, selenium, tellurium, and polonium. The heaviest of these elements are engaging in chalcogen bonding.

Underlying principles of this are the same, but the shape of the cells that contain these are slightly different. We observe a bent shape in chalcogens, and they can form two polar bonds. Opposite each polar bond is the site where we can form the chalcogen bonds where one of these is a weak interaction. They occur 180 degrees from the primary bond but 90 degrees from the other bond.

The pnictogens that Dr. Cozzolino focuses on nitrogen (N), phosphorus (P), arsenic (As), antimony (Sb), and bismuth (Bi). The heaviest ones (the last three) can form the primary bonds. The number of bonds that can form progress: iodine forms 1, tellurium forms 2, and thus antimony should be able to form 3. As these molecules occur 180 degrees from the primary bond but 90 degrees from the other bond, there is an implication that the molecules are capable of site recognition. The elements have a lone pair that can interact with a chalcogen, and vice versa. Thus, we get this double interaction and self-relativity can be observed for chalcogens and pnictogens.

Looking at an electrostatic map (fun note- I initially learned about these in my organic chemistry class!), we know that the colors depict electron densities. Computationally and manually, we can locate electron densities. Defining an isosurface as representing the surface of the molecule, the electron density can be probed conceptually by taking a point charge and driving it around the molecule at every point, and seeing if it is attracted or repelled. When we find the regions on the surface of the molecule that are attractive or republish, that indicates something is charged. This is what the colors on the electrostatic map show. The blue region is a region representing the Carbon-Iodine bond and is attractive to a negative point charge of something that has electrons (lone pairs). In the center, we can see the characteristic electrostatic potential Vmax which is the value that leads to predictable supramolecular interactions.

In pnictogen bonding, antimony and bismuth particularly engage in weak supramolecular interaction. Conducting this bonding in a predictable way is the next relevant area of interest as we will have some sort of idea of the outcome. As shown elbow, we will use the nomenclature of “secondary bonding interactions’ ‘ to term the bonding of interest.

While there are many structures observed with pnictogen bonds, no one was really using them on purpose. Dr. Cozzolino wondered why and presumed it was due to no good rules or strategies present. Thus, this is what he proposed in the early part of his research program: design molecules on purpose that perform these interactions in a predictable way.

Regions that want electrons contain pnictogens. The three arrows on the trigonal pyramid-shaped pnictogen molecule represent the sites for pnictogen bonding. Those regions or sites interact with molecules that have electrons they are willing to share (termed nucleophilic). The research focuses on 3 primary bonds such that there’s space for 3 pnictogen bonds (form 3 pnictogen interactions per pnictogen). If the molecule has one pair, it has the ability to self-recognize and self-assemble with a second molecule of itself and to do this over and over forming 3 pnictogen bonds.

The design that Dr. Cozzolino’s lab came up with was designed such that there was enough space for the bonds. The elements besides the pnictogens were oxygen or nitrogen and now even sulfur. But these electronegative elements caused the primary bonds to be polar. Thus, a cage was developed that restricted the bond angles to be 90 degrees. If they weren’t in a cage, they would flatten a bit at about 97 degrees which wouldn’t leave enough room for the 3 new pnictogens to form pnictogen bonds. So, thus, the proposed structure allows for the formation of the actual bonds.

Once the basic structural unit is established, the natural question that comes next is whether we can combine these basic units to form larger structures. And yes- this is possible! A dimer is a combination of two interactions (essentially a double interaction). These molecules can propagate one another. A tetramer is a triple interaction. We can keep building bigger models.

It was found that long-tails can promote anti-parallel arrangement through DFT calculations. Thus, reverse bilayers and bonds should be able to be formed. This was determined with antimony initially through computational models.

To determine which pnictogen was best, we can use any compound that has similar characteristics as hydrogen (they should stick and have equally similar bonds). Some of these chemicals are phosphorus, arsenic, antimony, and bismuth. These can be used as they are already used in life, are not toxic, and are already available. Arsenic, though, should be avoided as it might be a bit of a problem. Antimony is a bit of an unknown. Bismuth though, Pepto Bismol is already relevant, so this compound may be useful as well. The fundamental study consisted of figuring out which molecule to use. Students aligned/created these molecules and computationally explored them.

Once the molecule of interest was identified, the next step was making cages. As depicted in the figure above, the blue regions on the phosphorus representative at the element are not really attractive towards a negative charge. But, pentafluoride-iodo-benzene is. The value of the iodine is 125 Kj/mol which is vastly different from 16 Kj/mol. The phosphorus case does not form pnictogen bonds but repels anything that comes in. In arsenic, cages do form pnictogen bonds, but only form one per pnictogen instead of the three that was the ultimate goal. There was electrostatic potential there, it was slightly positive light blue but still smaller than molecules that do form supramolecular bonds reliably. In an antimony cage, there were three pnictogen bonds per antimony. The distances were quite short and the bond length was 2 Angstroms and van der Waals radius was around 3 angstroms. This is somewhere in the middle as it was short as opposed to the van der Waals radius but longer than a single bond. This structure was expanded upon and it was determined that this was the right choice. The bismuth structure was very tight and it wasn’t easily characterized as there was not a lot of experiential evidence.

The head of the structure is what engages in pnictogen bonding and the tail points in and leaves a little bit of space. The tail can be made a little bit longer. The molecules can sublime to the gas state and break every pnictogen bond. If you could measure the heat of sublimation — energy cost to break the bonds — we could get a good experimental guess about the strength of the pnictogen bonds. This is the sort of data that you get from differential standing calorimetry. The change in temperature measures how much heat is flown in for the constant decrease in temperature. As shown in the graph below, the area under the curve is the sublimation enthalpy of both systems is 100 KJ/mol. Dividing by the three number of interactions per molecule, we can estimate the strength of each interaction as 33 Kj/mol.

Dr. Cozzolino aims to develop and use the new tool to develop self-assembly of reversed bilayer vesicles with pnictogen bonds. He believed that you could make bilayers (but ones that were not actual bilayers). First, we make the tail longer by using two carbons to fill the space entirely. The molecules were forced to adopt the different supramolecular conformation which is exactly what happened. The three carbons in the tail could no longer form the helical structure and the molecules arrange themselves in an up-down fashion to give a 2D inverted bilayer when compared to what nature does. The natural approach to bilayers (ones that can be found in the cell membrane) can be viewed in water (but it’s not restricted to that). There are trials in the middle that are nonpolar and heads on the outside that is polar. The head and tails stay separate from each other.

The inverse molecule was actually constructed by a Welch scholar! The head was capable of recognizing other heads through pnictogen bonds and the tails started to form crystals that were analyzable. This formed the exact same inverted bilayer stricture which was the first example of an inverse bilayer as compared to normal biological systems.

The next order of business was determining what would happen when these were put into a solution (as this was no longer working as a natural system). Dry acetone was used for sonicating and then the structure and size of the particles were analyzed using dynamic light scattering (which is the scattering of laser light off particles). There was a homogeneous mixture and the prequel did not dissolve completely. Scanning electron microscopy was used and the exterior surface and its topology were measured.

The dynamic light scattering helped form the bilayer vesicles from dynamic light scattering. With a fair degree of confidence, we would imagine a spherical structure that consists of the biliary with head on the side and tail pointing out. We can now imagine what it actually looks like in the homogenous solution. The downside of this system is that it falls apart in water as molecules are unstable in water. Thus, to overcome this, we can switch to sulfur for stability. When the oxygen is swapped for sulfur, there is a tradeoff between stability and bond polarity.

Antimony thiolate cages were developed for the bilayers. They were held together by nitrogen bonds. The bonds were created and a narrow size distribution and bigger vesicles were developed compared to average, previous ones.

It was found that water doesn’t cause the molecules to decompose. Fluorescent dye was used to visualize the molecules and to get rid of the dye, post-synthetic encapsulation was used. This entails shining light of a specific wavelength and measuring the wavelength.

The antimony and bismuth were swapped out to examine the role it potentially plays in the vesicles. There is molecular-level control over size as they form at about 200 nm in size. This is something that you can’t really control with phospholipids so it was an interesting find.

The starting material is moisture sensitive. Taking the starting material, the methyl groups are put in the presence of something that can support it in the solvent. The solid precipitates out and it can be purified in various ways.

It was found that the stronger the pnictogen bonds, the smaller the size. It’s much narrower for bismuth: smaller but also more uniform in size. Switching out the pnictogen has some important implications on size and structure.

The vesicles that are made are very large. Under control conditions, when we try to make the vesicles, it was observed that there are small vesicles on average of uniform size with longer tails. The total length has been found to influence the vesicle size. But under control conditions where we try to make the vesicles — we observe that we get small vesicles on average of uniform in size with longer tails.

When molecules move around, they can shift the membrane (even flip flop). It’s pretty slow but there are enzymes that assist with this. Vesicles can fuse or fall apart. This dynamic behavior is of interest, and there is a new system that is being worked upon.

The small molecule is gathered and a cage is formed. The long tail still forms the inverse bilayers in the crystal structures, and insulation occurs using a fluorescent dye. After dialysis, we can see that the vesicles of display dynamic behavior.

Vesicles are cast on a dye slide and as shown in the image above, in the circle, they appear to merge. The dynamic behaviors of their moves are something that needs to be studied more carefully.

The vesicles spontaneously align with each other and merge into one long strand. This needs to be reproduced and studied further, but it does appear that not only can we create vesicles, but that they have some dynamic behavior that at least is loosely mocking what nature and natural cells can do.

Here’s a quick TL DR summary of pnictogen bonding.

Throughout this talk, we strived to understand some of the fundamental principles of pnictogen bonding including the facts that pnictogen bonds can direct the self-assembly of reversed bilayer vesicles, molecular parameters can influence size, the use of sulfur leads to tables molecules and vesicles, and that pnictogen bonds expand the toolset available for supramolecular chemistry.

Some of the current open questions are determining how stable these systems are physically, how dynamic these systems are, whether diffusion occurs across the membrane, what different functions can be incorporated into the system, whether mixed systems are stable, whether these systems can be made amenable to physiological conditions, and what different artificial cell-like functions can be introduced.

In all, this session by Dr. Cozzolino was very informative, intriguing, and demonstrated how applicable chemistry can be!