Author Archive

Studying Anticancer Agents with XFM

It turns out that rhenium-based compounds have been showing some promising anticancer activity as they’re stable, allow for real time imaging, structurally diverse, and have low off-site toxicity. The most commonly studied complexes are based around a Re(I) tricarbonyl core with the other three binding sites occupied by ligands of varying complexity. Researchers in the US and Australia developed a tricarbonyl Re isonitrile polypyridyl complex fac-[Re(CO)3(dmphen)(para-tolyl-isonitrile)]+, where dmphen = 2,9-dimethyl-1,10-phenanthro-line, called TRIP for short. TRIP showed promising cytotoxicity and can be imaged using confocal fluorescence microscopy, taking advantage of the emissive metal to ligand charge transfer (MLCT) state. The persistence of the emission indicates that the ligands remain bound to the Re even within cells. The complex’s cytotoxicity stems from its inducement of cells to accumulate misfolded proteins, resulting in apoptosis from the unfolded protein response (UPR). UPR induced cell death is relatively uncommon and led the researchers to find a method to characterize the speciation of TRIP in vitro. They used synchrotron X-ray fluorescence microscopy (XFM) to probe the cellular uptake and distribution of TRIP and an iodo-derivative I-TRIP by looking at elemental signals.

Figure 1. Chemical structures of TRIP and I-TRIP

I-TRIP is particularly well-suited to this type of study, as the iodine provides an additional spectroscopic handle on the isonitrile ligand absent in TRIP. Of course, the researchers had to confirm that I-TRIP possessed similar cytotoxicity and working mechanism to TRIP. Various assays and biological studies showed evidence of comparable cytotoxicity and mechanism, demonstrating that altering the substitution of the isonitrile ligand doesn’t significantly impact the bioactivity of the complex. With that settled, the experiments could move to the synchrotron to probe elemental distributions.

Figure 2. XFM elemental distribution maps of HeLa cervical cancer cells treated with either DMSO (control), TRIP, or I-TRIP.

Cells treated with both TRIP and I-TRIP show a clear Re signal, confirming that they can enter and persist in cells. Critically, the colocalization of the Re and I maps for I-TRIP samples indicate that the isonitrile ligand remains bound as a part of the Re complex inside the cells. This strongly suggests that the Re complex is intact while it induces cell death, adding to the developing mechanistic understanding of their activity. This work shows the utility of XRM as a technique to study the distribution of organometallic complexes in living cells. Additionally, the tunability and stable bioactivity of the Re complexes shows that they’re amenable to study by a wide range of techniques that will allow for further mechanistic probing.

To find out more, please read:

X-Ray fluorescence microscopy reveals that rhenium(I) tricarbonyl isonitrile complexes remain intact in vitro

Chilaluck C. Konkankit, James Lovett, Hugh H. Harris and Justin J. Wilson

Chem. Commun., 2020, 56, 6515-6518

About the blogger:

Dr. Beth Mundy is a recent PhD in chemistry from the Cossairt lab at the University of Washington in Seattle, Washington. Her research focused on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Teachable micromotors

Micromotors, for the uninitiated aka me, are a specific type of colloidal structure that harvests energy from their environment and turns it into motion. In order for them to be truly effective for possible applications though, they must be able to communicate and coordinate with one another. If you want to make microbots, they can’t just move about all willy-nilly with no regard for each other. Recently, scientists have created a mixed system where one structure sheds silver ions that cause other micromotors to accelerate, an approach that mimics natural systems like bee colonies.

Researchers in China developed a system where photochemically powered micromotors can spontaneously “teach” catalytic micromotors to oscillate without any external influence. The teachers are Janus-type microparticles composed of either polymethylmethacrylate (PMMA) or silicon dioxide particles half coated with silver. Under irradiation with KCl and H2O2, the silver can interconvert between Ag(0) and Ag(1), causing the oscillatory motion of the entire particle. In contrast, the two non-oscillatory micromotors, either polystyrene spheres half coated with platinum (PS-PT) or gold-rhodium microrods (Au-Rh), catalytically decompose H2O2 to move autonomously in standard Brownian motion. When the two types of micromotors are mixed under UV light, the motion of the non-oscillatory materials changes from random to clearly oscillating (Figure 1). The intensity of the motion change depends on the proximity of the learner to the teacher, with learners closer to the teacher displaying more intense oscillations.

Figure 1. Change in movement of non-oscillating micromotors when exposed to oscillating “teacher” structures.

In fact, the Au-Rh rods will demonstrate more intense oscillations than the PMMA-Ag particles. The researchers propose a mechanism where the PMMA-Ag particles release silver ions as they oscillate, which then deposit onto the Au-Rh rods. The silver increases the catalytic activity of the rods and then, given the operating conditions, undergoes the same redox process that causes oscillation in the PMMA-Ag system.

Figure 2. Proposed mechanism of silver release and adsorption onto Au-Rh rods.

This hypothesized mechanism is supported by the development of oscillatory behavior by the Au-Rh rods in under reaction conditions where the PMMA-Ag particles are replaced by silver ions in solution. The silver on the rod surface isn’t merely adsorbed – it forms into small silver nanoparticles which can be seen via electron microscopy, making a new trimetallic structure. These nanoparticles change the trajectories of the rods, causing them to move in circles. While this system isn’t perfect, the student structures have imperfect memories and cannot teach one another, it provides a strategy for working with groups of micromotors to move towards coordinated motion and further applications.

To find out more, please read:

Non-oscillatory micromotors ‘‘learn’’ to oscillate on-the-fly from oscillating Ag micromotors

Chao Zhou, Qizhang Wang, Xianglong Lv and Wei Wang

Chem. Commun., 2020, Advance Article

About the blogger:

Dr. Beth Mundy is a recent PhD in chemistry from the Cossairt lab at the University of Washington in Seattle, Washington. Her research focused on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Aluminum-Based Liquid Coordination Complexes

Before I get into the research, I just have to say that I think ionic liquids (ILs) are just cool. An ionic compound that’s a liquid? Mind blown. Not only are they interesting, but those unique properties make them attractive for industrial use. One class of ionic liquids generating research focus is halometallate ionic liquids (HILs), generated by reacting a metal halide and an organic halide salt. In particular, chloro-alluminate ILs have been some of the most promising for applications in industrial acid catalysis and are composed of anionic aluminate species. However, recent work has found mixtures of aluminum chloride and N/O/S donors produce a liquid Lewis acidic compounds, referred to as liquid coordination complexes (LCCs). These LCCs are potential replacements for HILs, as they’re typically easier and cheaper to prepare. Further studies have found ionic species by 27Al NMR, drawing parallels between LCCs and HILs. By combining AlCl3 and polar organic solvents, researchers in the US screened for novel LCC or HIL reactivity in catalysis.

This straightforward approach allowed researchers to easily tune the ratio of AlCl3 and solvent to find mixtures with desired properties. They chose the nitrogen donor 1-methylimidazole, N-Mim, and an oxygen donor N-methyl-2-pyrrolidone (O-NMP) as their solvents for testing given their wide availability and relevance to organic chemical reactivity. The selected solvent and AlCl3 were mixed at room temperature and found in all cases to form heterogenous mixtures. When heated to 100 oC mixtures with molar fractions of AlCl3 between 0.33 and 0.6 formed viscous liquids, many of which became solids at room temperature. These compounds were then crystalized for x-ray crystallographic analysis, where their structures confirmed the association of the aluminum with the nitrogen (Figure 1) or oxygen.

Figure 1. 50% probability ellipsoid plot of AlCl3(N-Mim), showing coordination between the aluminum and nitrogen

In the N-Mim system, 27Al NMR showed the formation of a single aluminum-containing species at AlCl3 molar fractions of 0.5 and below, with exchange occurring when higher concentrations of N-Mim are present. The O-NMP system proved more challenging to characterize crystallographically, potentially due to the formation of larger oligomeric complexes in the solid phase and increased disorder in the O-NMP ligand. However, 27Al NMR proved insightful and showed the presence of multiple aluminum-containing species, including several different stoichiometries of aluminum-solvent adducts.

Figure 2. Aluminum NMR spectra of aluminum/O-NMP complexes showing speciation over a range of different stoichiometries.

When the two systems were side-by-side compared for Lewis acidity and catalytic activity for the alkylization of benzene, the clear winner was the O-NMP system. The O-NMP-AlCl3 complex with an aluminum molar fraction of 0.6 was both the most Lewis acidic, determined by an acetonitrile IR probe, and the most catalytically active. It gave full conversion with a selectivity of almost 80%, while the N-Mim complex with the same mole fractions produced only a 32% conversion with no significant increase in selectivity. Complexes with less aluminum showed no signs of catalytic activity and were less Lewis acidic. The high activity of the AlCl3/O-NMP system can be explained by its possession of both a highly Lewis acidic cation [AlCl2(O-NMP)2]+ and a highly Lewis acidic anion [Al2Cl7], whose presence was identified in the NMR experiments. This work demonstrated a straightforward method to synthesize LCC-based catalysts with high activity, while providing some general guidance on the suitability of O-donor ligands for further study.

To find out more, please read:

Are ionic liquids and liquid coordination complexes really different? – Synthesis, characterization, and catalytic activity of AlCl3/base catalysts

Rajkumar Kore, Steven P. Kelley, Anand D. Sawant, Manish Kumar Mishra and Robin D. Rogers

Chem. Commun., 2020, Advance Article

About the blogger:

Dr. Beth Mundy is a recent PhD in chemistry from the Cossairt lab at the University of Washington in Seattle, Washington. Her research focused on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Finding early warnings in chemical systems

Our world runs on systems of ever-increasing complexity, be they natural or human created. Their behavior can generally be modeled and operates within a normal range… until it hits a tipping point at a border between two behavioral regimes. Once that threshold is reached the system acts in dramatically different, and often destructive, ways. However, there are often warning signs right around the tipping point that can serve to alert careful observers that the danger zone is near. By finding ways to identify and monitor this behavior either contingency plans can be put in place or the problem corrected before it becomes catastrophic. While much of this has been studied for large systems like the stock market or ecosystems, researchers recently applied this to chemical systems.

The utility of identifying early warning signals isn’t limited to preventing massive changes; it can also help bound regions of different functionality in chemical systems. To explore this, researchers used a trypsin oscillator system they previously developed that has two modes of flow (Figure 1), either sustained oscillations in trypsin concentration or dampened oscillations that eventually lead to a steady concentration of trypsin. By changing flow rate, temperature, or reagent concentration the researchers could tune the behavior of the system to test either active or passive monitoring schemes to find early warning signs. Since they knew what conditions produced each mode, they could operate the system right at the boundary and watch its behavior.

Figure 1. Schematic of trypsin system with examples of oscillation patterns.

In the “active” experiments the researchers intervened in the system’s normal functioning and watched how its response, the “critical slow down phenomena,” changed as the conditions were brought closer to the boundary. In “passive” experiments, the researchers simply brought the system close to the boundary and monitored the shape of the oscillation waves focusing on their full-width half maxima (FWHM). In the series of active sensing experiments, the researchers waited to see how long it took the system to re-establish sustained oscillations after a change in temperature. They expected to see this recovery time increase as the system was brought closer to the edge, aka the “critical slow down phenomena.” In their experiment they elevated the temperature of the system to 49.0 oC and observed how long it took to recover to either 24.3 oC (well within the sustained regime) or 20.9 oC (near the behavioral boundary). They saw the expected dramatic increase in time to return to sustained oscillation, with the system recovering in 10.7 hrs at 24.3 oC, but taking three times as long, 32.8 hrs, when the final temperature was 20.9 oC (Figure 2). The experimental results were further validated by theoretical modeling, where experimentally challenging positions at the very edge of the boundary could be explored and showed the same general behavior.

Figure 2. Top: active monitoring experiments at two different temperatures showing the differences in time needed to recover sustained oscillations. Bottom: comparison of modeled and experimental recovery times.

The theoretical modeling was further used to create a map of passive monitoring space, looking at the variations in FWHM. They saw that as the system moves away from the center of the oscillation regime the FWHM increases as the peaks broaden. This becomes more pronounced as the conditions approach the behavioral boundary. This provides a baseline knowledge with multiple types of measurements to compare the stability of the simple system over a range of conditions. Additionally, this system and others like it can serve as the base for modeling more complex systems of which they are a part.

To find out more, please read:

Early warning signals in chemical reaction networks

Chem. Commun., 2020, Advance Article

Oliver R. Maguire, Albert S. Y. Wong, Jan Harm Westerdiep and Wilhelm T. S. Huck

About the blogger:

Dr. Beth Mundy recently received her PhD in chemistry from the Cossairt lab at the University of Washington in Seattle, Washington. Her research focused on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Ketone Reduction with Magnesium

Making catalysts from non-precious metals has been a challenge embraced by chemists over the past few decades. In particular, the ability of alkaline earth metal catalysts to be both highly stable and highly reactive has made them attractive for study. Magnesium complexes can catalyze hydroelementation across a carbon-oxygen double bond with a wide scope of substrates. However, it isn’t enough to just perform this reaction. For meaningful reactivity the catalyst would be enantioselective and prior to this report only one example of a stereoselective magnesium-based system existed in the literature. A persistent barrier to developing this reactivity has been the propensity of alkaline earth metal hydrides (a likely step in the catalytic cycle) to form contact-ion pairs with available anions. Based on this framework, researchers in Germany developed a catalyst motif involving borohydride-alkaline earth metal adduct formation to enhance stereoselectivity.

They drew on their precious work utilizing manganese complexes that demonstrated enantioselectivity, applying it to the magnesium system. They could straightforwardly create the magnesium alkyl complexes by mixing ligand and magnesium alkyl precursors. The complexes were tested for hydroelemetation using acetone as a model substrate. The rate and selectivity of the reaction was significantly impacted by the choice of reducing agent, with non-boron reductants producing low yields and low enantioselectivity. Altering the ligand backbone didn’t improve selectivity, but purifying an assembled precatalyst and slowly raising the temperature from -40 oC to RT increased enantiomeric excess to 96%. Even more exciting, using this general reaction format high enantioselectivity was seen for a wide range of aryl alkyl ketones screened with no adverse reactivity seen towards ester moieties (Figure 1). This system also catalyzes reactions with α-substituted, cyclic alkyl aryl, and dialkyl ketones, but with lowered enantioselectivity. However, the catalyst is tolerant of a range of functional groups and remains selective for carbonyls implying broad possible utility.

Figure 1. Substrates tested for hydroelementation using a magnesium catalyst with conversion and stereoselectivity numbers.

To more fully understand the system, the researchers undertook experiments to find possible intermediates in the catalytic cycle. The reactivity at room temperature required consistent low temperature work to prevent further progress of the intermediates along the catalytic cycle. In the absence of a boron-derived reducing agent, the catalyst forms a highly labile, and expected, alkoxide intermediate that is not observable under catalytic conditions. When the precatalyst is reacted with excess borane it forms a borohydride intermediate that was crystallographically observed.

Figure 2. Solid state molecular crystal structure of borohydride intermediate.

When this complex was reacted with a fluorinated ketone, variable temperature boron and fluorine NMR was used to identify transient species. Upon addition of the ketone immediate shifts in the boron signal to two species, one of which remains when the temperature increases to -60 oC. The unstable species can be attributed to a ketone borohydride complex, but the dynamic nature of the system makes crystallizing these intermediates very challenging. DFT calculations suggest a low-energy transition state that favors the (S)-product and accounts for the catalyst’s consistent stereoselectivity.

To find out more, please read:

Borohydride intermediates pave the way for magnesium-catalysed enantioselective ketone reduction

Vladislav Vasilenko, Clemens K. Blasius, Hubert Wadepohl and Lutz H. Gade

Chem. Commun., 2020, 56, 1203-1206

About the blogger:

Beth Mundy is a PhD candidate in chemistry in the Cossairt lab at the University of Washington in Seattle, Washington. Her research focuses on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Color switching with iridium complexes

Piezochromic luminescent materials (PCLMs) exhibit emissive behavior in response to mechanical stimuli and have captured extensive scientific attention, and not just because they’re awesome. Most display reversible bicolor switching, but a lack of universal design principles has hampered development of PCLMs with multiple colors. Utilizing the polymorphs of organic dyes, which generally have different luminescence profiles, is a promising approach towards creating multicolor switches. Iridium (III) complexes can be separately both phosphorescent and polymorphic, but no examples exhibiting both properties had been experimentally shown prior to this work.

Figure 1. Structure of the piezochromic iridium complex.

Researchers in China and the UK synthesized a novel cationic Ir(III) complex with a ligand rationally designed for PCL behavior (Figure 1). The bulky ligand should disrupt close crystal packing and leave the material prone to collapse to an amorphous solid when under external pressure. The ligands also consist of heterocycles compounds, which are more likely to have polymorphs, possibly due to the range of non-covalent interactions between different atoms in the solid phase. These cyclic ligands are also quite rigid with limited intramolecular rotational motion, which in studies of aggregates showed that decreasing intramolecular vibrations and rotations can increase luminescence intensity. With all those design considerations put in place, the researchers generated three powder samples of the iridium complex with different emission profiles (blue, yellow, and green). Reproducibly switching the emission color is simple, done by either mechanical grinding or applying solvent.

Figure 2. Photographs and schematic of tricolor switching iridium complex.

The initial yellow complex showed low emission when solvated by acetonitrile, but altering the solvent profile to be 90% water induces aggregation and a significant increase in emission. Incorporating the complex into a rigid PMMA matrix similarly enhances its emission, which supports the hypothesis that the rigidity and corresponding diminishment of intramolecular rotations and vibrations serve to enhance the emission. All three solid polymorphs can be obtained directly via recrystallization from different solvents or by cycling through grinding and/or adding small amounts of solvent. Despite the clearly different optical properties, the samples all had identical 1H NMR spectra. The researchers turned to powder X-ray diffraction spectroscopy (PXRD) to gain additional information about the spatial arrangement of the atoms within the molecule and the overall crystallinity of the solids. They saw that the green and blue forms of the complex form distinct crystalline phases, whereas the yellow complex is completely amorphous. The blue crystals show the molecules in antiparallel coupling, while the molecules in the green crystal pack in a herringbone arrangement. However, both pack quite loosely and are thus easily collapsed to an amorphous form when under pressure. Molecular modeling used the obtained molecular confirmations to model the energy levels of the HOMO and LUMO. The models showed that the increased planarity of the phenyltriazole moieties and increase twist of the pyridine-based ligand cause the higher HOMO-LUMO gap, shifting the emission to higher energies.

Interestingly, the impact of solvent interactions with the molecules, particularly during crystallization, cannot be discounted. The researchers observed multiple hydrogen-bonding interactions with the iridium complex and either toluene or ethanol. Given the relative ease of switching, as a proof-of-concept the researchers patterned a tree and sky scene starting from the yellow form of the complex and adding ethanol or toluene with a capillary (Figure 3). The scene could be erased and repatterned with no obvious molecular performance degradation.

Figure 3. Photographs of the tree patterned from the three forms of the iridium complex.

To find out more, please read:

Reversible tricolour luminescence switching based on a piezochromic iridium(III) complex

Tianzhi Yang, Yue Wang, Xingman Liu, Guangfu Li, Weilong Che, Dongxia Zhu, Zhongmin Su and Martin R. Bryce

Chem. Commun., 2019, 55, 14582-14585.

About the blogger:

Beth Mundy is a PhD candidate in chemistry in the Cossairt lab at the University of Washington in Seattle, Washington. Her research focuses on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Tuning Zeolite Catalysis with Organic Molecules

Zeolites, a class of porous alumina-silicate materials, are industrially critical adsorbents and catalysts. Their highly robust nature and wide range of structural types (over 200!) make them suited to a range of applications. In particular, the general zeolite topology and pore size are selected to match and stabilize the intermediates of a chemical reaction. However, the tunability of zeolites is limited when compared to molecular catalysts, making them more like a solvent than, say, an enzyme. An active field of research is bridging the gap between the robust, scalable zeolites and highly controllable homogenous catalysts. Recent work identified organic residues maintained with the zeolite pores as key in the transformation of methanol to hydrocarbons. Previous fundamental studies demonstrated that a wide range of carbonyl and carbonyl derivative compounds promote the dehydration of methanol to dimethyl ether (DME).

Researchers at BP used methyl mono- and di-carboxylate esters to dehydrate methanol to DME at low temperatures. The mild reaction conditions allowed for high selectivity for DME while eliminating convoluting side reactions. They added either methyl formate or methyl n-hexanoate to a series of zeolite with pores ranging from narrow to wide. At a 5 mol% concentration relative to methanol they saw significant increases in DME production, particularly for the medium and wide pores. Systematic testing of carboxylate chain length found that increasing chain length increased turnovers occurred until methyl n-hexanoate, after which no further benefits were observed as the n-methyl hexanoate had already saturated the catalyst (Figure 1). All proved highly selective for converting methanol to DME with no observed hydrocarbon formation.

Figure 1. Production of DME on a medium-pore zeolite with methyl carboxylate esters of varying chain lengths.

The experimental results were coupled with theoretical work modeling the energetics of the adsorption of the ester onto the zeolite. The calculations showed an increase in adsorption energy with increased chain length, attributed to van der Waals interactions.

Figure 2. Transition state predicted by molecular modeling with methanol attacking the organic promoter adsorbed on the zeolite catalyst.

They also gave even higher energies to molecules with two carboxylate esters, like dimethyl adipate. In fact, the strongly binding molecules produced increased catalysis at loadings as low as 0.001% with respect to methanol. The promoters can be easily switched by changing the input, demonstrating the reversibility of binding at the active site. Additional molecular modeling was used to study possible transition states to develop a catalytic cycle. A proposed transition state involves a direct reaction between the methanol and the organic promotor, however specific evidence has yet to be seen. Additional work examining the role of the water present as a co-adsorbate and its impacts on transition states has yet to be done. Overall, the use of various organic molecules as promotors for the dehydration of methanol to DME on various zeolite catalysts was explored. This represents exciting fundamental study of industrially-relevant chemistry with significant room for future work.

To find out more, please read:

Getting zeolite catalysts to play your tune: methyl carboxylate esters as switchable promoters for methanol dehydration to DME

Benjamin J. Dennis-Smither, Zhiqiang Yang, Corneliu Buda, Xuebin Liu, Neil Sainty, Xingzhi Tan and Glenn J. Sunley

Chem. Commun., 2019, 55, 13804-13807.

About the blogger:

Beth Mundy is a PhD candidate in chemistry in the Cossairt lab at the University of Washington in Seattle, Washington. Her research focuses on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Designing Syntheses with Machine Learning

I don’t know if you’ve looked at the structure of pharmaceuticals recently, but most novel drugs are rather complicated. Identifying promising new targets is just the start for synthetic chemists; they then need to figure out how to use a series of reactions to take simple (and commercially available) molecules and transform them into a new drug. They also must predict all possible side reactions and products given a set of reaction conditions, particularly when a range of functional groups are involved. Historic approaches involved manual curation of reaction rules, limited by personal experience and the state of the accessed chemical literature. Newer approaches seek to create templates directly from data but are defined by available data sets and cannot reliably extrapolate. The emergence of machine learning offers the opportunity to move beyond traditional templating and atom mapping of reactants to products. It also offers to take full advantage of novel technologies and address problems with dataset bias and ineffective modeling systems.

In a collaboration between academics in the UK and industrial scientists in the US, researchers used Molecular Transformer, an attention-based machine translation model, to perform both reaction prediction and retrosynthesis analysis after training on a publicly available dataset. Instead of atom mapping, which moves atoms from the reactants to the products, Molecular Transformer (MT) relies on SMILES text strings, which represent structures in a line format. A unique aspect of this work is the validation and training performed using proprietary data of drug targets from Pfizer. They used three datasets: the first a literature standard from the US Patent and Trade Office (USPTO), the second from internal medicinal chemistry projects in Pfizer, and the final a diverse range of 50,000 reactions from US patents (USPTO-R). Building on previous research from the authors, they trained the MT on both the Pfizer data and the initial USPTO data sets. They found that the Pfizer data provided the most accurate product predictions and that the MT could also return a confidence rating to determine the probability the prediction is correct.

Figure 1. Sample syntheses predicted by Molecular Transformer for various bioactive molecules of interest.

While synthesis predictions can easily be checked, it’s harder to confirm accuracy with retrosynthesis since there is not a single correct answer. The researchers used the broad USPTO-R to train MT, which consistently outperformed both a benchmark template-based program and another literature machine learning method also trained on USPTO-R. When tested on the Pfizer dataset, the MT performed best with 31.5% accuracy despite the datasets coming from different regions of chemical space (which increased to 91% when MT was trained on Pfizer data). Figure 1 shows several predicted routes for the synthesis of bioactive molecules as predicted by MT, which generally agree with established syntheses. These data suggest the highly generalizable nature of MT as a tool for developing novel pharmaceutically interesting molecules.

To find out more, please read:

Molecular Transformer unifies reaction prediction and retrosynthesis across pharma chemical space

Alpha A. Lee, Qingyi Yang, Vishnu Sresht, Peter Bolgar, Xinjun Hou, Jacquelyn L. Klug-McLeod and Christopher R. Butler

Chem. Commun., 2019, 55, 12152-12155.

About the blogger:

Beth Mundy is a PhD candidate in chemistry in the Cossairt lab at the University of Washington in Seattle, Washington. Her research focuses on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

MOF-Derived Solid-State Lithium-Oxygen Batteries

Just in case you weren’t aware, it turns out that lithium-based batteries are kind of a big deal. While the Nobel-winning batteries have already revolutionized consumer electronics, further development requires batteries with even higher energy densities. Enter: lithium-oxygen batteries (LOBs) with theoretical energy densities of 3500 W h/kg. LOBs come in non-aqueous, aqueous, hybrid, and solid-state varieties based on their electrolytes. Given the previous safety issues for lithium-based batteries with liquid electrolytes (remember the exploding phones?), solid-state electrolytes have attracted substantial research attention. Specifically, Li1+xAlxGe2x(PO4)3, or LAGP, shows promise given its high Li+ transport number and electrochemical stability over a wide window. These solid-state electrolytes need to be combined with new catalytically active high surface area cathode materials that will not react with the lithium and degrade, a persistent issue with MOFs.

Figure 1. Schematic of an assembled all solid-state lithium-oxygen battery.

Researchers in China and Japan have combined LAGP electrolyte with NiCo2O4 (NCO) nanoflakes as the catalytically active cathode material. They then assembled full solid-state batteries, the structure of which is shown in Figure 1, for electrochemical and stability testing. The LAGP was prepared using previously established methods and found to exhibit the expected high stability and lithium mobility. To prepare the nanoflakes, the researchers annealed cobalt-based MOFs on a sacrificial carbon substrate then dipped them in a Ni(NO3)2 solution for nickel doping and annealed once more. This leaves the final nanostructured metal oxide, with the elemental composition confirmed by TEM elemental mapping. As a conveniently freestanding electrode material, the nanoflakes were then loaded in as the cathode.

Once assembled, the researchers tested the full all solid-state LOBs for stability and performance. They demonstrated high discharge capacity and electron transfer efficiency with charge and discharge potentials well within the electrochemical window of the LAGP electrolyte. These are attributable to the high lithium ion mobility and the porous bimetallic nature of the cathode. To confirm that the incorporation of nickel impacted the overall device performance, the pure cobalt nanoflakes were used as the cathode.

Figure 2. Cycling performance of cobalt (left) and cobalt-nickel cathodes (right) at a current density of 100 mA/g.

As seen in Figure 2, the cobalt-only batteries exhibit significant capacity loss in only 35 cycles whereas the NCO cathodes showed no degradation after 90 cycles. While cycling the NCO electrodes, the reversible formation of Li2O2, a common discharge product, occurred in the open pores of the cathode. These pores allow the 500 nm Li2O2 particles to form and dissolve without disrupting the structure of the cathode and give a more stable battery. This research brings completely solid-state lithium-oxygen batteries one step closer to reality.

To find out more, please read:

All solid-state lithium–oxygen batteries with MOF-derived nickel cobaltate nanoflake arrays as high-performance oxygen cathodes

Hao Gong, Hairong Xue, Xueyi Lu, Bin Gao, Tao Wang, Jianping He and Renzhi Ma

Chem. Commun., 2019, 55, 10689-10692.

About the blogger:

Beth Mundy is a PhD candidate in chemistry in the Cossairt lab at the University of Washington in Seattle, Washington. Her research focuses on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Mechanical Stress Turns These Dendrimers Blue

We all know what happens when materials take too much mechanical stress – they eventually break.

What if you could easily tell when something like a support was close to its maximum stress, before it undergoes a catastrophic event, just by looking at it? One option is to incorporate a mechanochromic polymer, a polymer that changes color when under sufficient mechanical stress, to provide a visual indicator that a material has reached a specific stress threshold. The polymers don’t need to be entirely composed of mechanochromically active moieties to exhibit useful properties; many studies have focused on a single active mechanophore at the center of a large polymer chain. In fact, the mechanical force is greatest at the center of a chain and is directly proportional to the length of the chains. This holds for polymers in solution but hasn’t been extensively studied in the types of bulk systems useful for applications.

Recently, researchers in Japan set out to characterize the effects of chain length and branching on mechanochromic dendrimers, polymers with monodisperse and regularly branched globular structures. Showing that dendrimers exhibit mechanochromism is already a novel result, but their well-defined nature allowed the researchers to draw correlations between structure and bulk responsiveness. They employed diarylbibenzylfuranone (DABBF) as the mechanochromic moiety since it generates arylbenzofuranone (ABF) radicals, which are blue, air-stable, and electron paramagnetic resonance spectroscopy (EPR) active, when exposed to mechanical force (Figure 1).

Figure 1. Structure of the DABBF moiety and the active ABF radicals generated by its dissociation.

These characteristics allow for straightforward qualitative and quantitative analysis. The team coupled the DABBF moiety with two series of dendrimers, with increasing generations having larger and more highly branched monomer units, to create a range of molecular weights and degrees of branching for study. The dendrimers showed a color change from white to blue (Figure 2) when ground in a ball mill, which was used to ensure the reproducibility of the force applied to all samples.

Figure 2. Photographs of the first (top) and second (bottom) mechanochromic dendrimers before and after grinding, showing the color change associated with the generation of ABF radicals.

EPR measurements confirmed the presence of the ABF radicals in the samples after milling, demonstrating that the color change is due to the cleavage of the DABBF. The integrated EPR spectra were used to quantitatively determine the percentage of DABBF moieties that dissociated. The responsiveness of the dendrimers increased exponentially with increasing generation and branching. However, the primary factor governing ABF generation was found to be molecular weight. Two dendrimers with different levels of chain entanglement, but similar molecular weights, exhibited comparable cleavage ratios.  The question then became does molecular weight increase the transfer efficiency of force to the DABBF or does the increased steric bulk make it harder for the ABF radicals to recombine? To probe the kinetics of this process, the researchers varied the grinding time and saw that within 5 minutes all the highly branched samples reached their maximum dissociation level. Additionally, monitoring the ABF recombination showed that even after 6 hours approximately 95% of the radicals remained dissociated in all 3rd and 4th generation dendrimers. These data suggest that the enhancement in responsiveness can be attributed to better force transmission to the DABBF.

This work shows mechanoresponsiveness in a range of dendrimers with varying degrees of branching and rigidity. Not only did they demonstrate novel activity, but the researchers also probed the mechanism of the enhanced activity with increasing molecular weight. This initial study opens avenues to explore polymer rigidity, surface functionality, and other dendrimer features to design new, functional materials.

To find out more, please read:

Mechanochromic dendrimers: the relationship between primary structure and mechanochromic properties in the bulk

Takuma Watabe, Kuniaki Ishizuki, Daisuke Aoki, and Hideyuki Otsuka

Chem. Commun., 2019, 55, 6831-6834.

About the blogger:

Beth Mundy is a PhD candidate in chemistry in the Cossairt lab at the University of Washington in Seattle, Washington. Her research focuses on developing new and better ways to synthesize nanomaterials for energy applications. She is often spotted knitting in seminars or with her nose in a good book. You can find her on Twitter at @BethMundySci.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)