C&EN’s Year in Chemistry 2018

Credit: C&EN/Shutterstock

Intelligent robots may yet take our jobs; they definitely grabbed a lot of C&EN’s headlines this year. Researchers continued to explore applications of machine learning, a type of artificial intelligence used to make predictions or decisions through algorithms that can learn from large data sets. The technology powers self-driving cars and image-recognition software.

Scientists demonstrated numerous ways in which machine learning can help explore chemical space. For example, Heather Kulik of the Massachusetts Institute of Technology and colleagues identified inorganic molecules called spin-crossover complexes that could be useful as sensors or electronic switches (J. Phys. Chem. Lett. 2018, DOI: 10.1021/acs.jpclett.8b00170). Apurva Mehta of the SLAC National Accelerator Laboratory, along with collaborators, used machine learning to identify new alloys that are metallic glasses (Sci. Adv. 2018, DOI: 10.1126/sciadv.aaq1566). And the chemical company Symrise teamed up with IBM to search for new fragrances using machine learning.

Related: Is machine learning overhyped?

Meanwhile, Thomas F. Miller and coworkers at the California Institute of Technology showed how machine learning can benefit chemical modeling, demonstrating how it can predict electronic properties of molecules with high accuracy and low computational cost (J. Chem. Theory Comput. 2018, DOI: 10.1021/acs.jctc.8b00636). In related work, Adrian Roitberg of the University of Florida showed off a machine-learning-based tool that calculates molecular forces and energies and offers high performance at a low computational cost.

Synthetic chemists also experimented with machine learning. Abigail G. Doyle of Princeton University worked with colleagues there and at Merck & Co. to optimize the yield of an amination reaction by asking their algorithm to vary the reagents used (Science 2018, DOI: 10.1126/science.aar5169). Bartosz Grzybowski of Ulsan National Institute of Science and Technology and the Polish Academy of Sciences put the synthesis-planning Chematica software, which he developed and MilliporeSigma acquired, to the test. Human chemists found that the computer program charted routes to products that were at least as good as those humans have developed (Chem 2018, DOI: 10.1016/j.chempr.2018.02.002). And Alán Aspuru-Guzik of the University of Toronto is one of several chemists who applied machine learning in software capable of independently running experiments and then using the results to improve the procedures.

Do these advances mean that machine learning is living up to its hype? Chemists have mixed feelings about the field. Most agree it’s a useful tool despite overheated enthusiasm from some corners.

Related: New directions for machine learning

Credit: AP

On Jan. 1, 2018, the Chinese government severely restricted imports of plastic for recycling.

Plastic waste has been a long-standing problem, and this year the world continued to grapple with the growing mountains of discarded plastic products. Governments sought ways to curb plastic use and boost recycling efforts, companies invested in technology to recycle certain problematic plastics, and chemists created novel sustainable polymers.

China kicked off 2018 by restricting plastic waste imports with more than 0.5% contaminants, such as food residue or metals. This low threshold effectively meant that many recyclers in the US and elsewhere, which previously shipped their recyclable plastics to China for the labor-intensive sorting process, would have to find alternative solutions.

Related: Plastic waste threatens coral reefs

Also in January, the European Commission introduced its first Europe-wide plastics recycling plan, which aims to make all plastic packaging used in the region recyclable by 2030. The plan also includes the goal of recycling 55% of all that packaging material by 2030; currently, Europeans recycle 30% of their plastic waste. In May, the commission proposed a new law that bans or restricts the 10 single-use plastic products most often found on Europe’s beaches and in its seas, including plastic cutlery and cotton swab sticks.

A thermoset polymer with boron-oxygen bonds has the ability to be reshaped and recycled, thanks to this chemical transformation.

One particularly problematic polymer—polystyrene, which is common in food packaging and packing materials—currently isn’t recycled on a large scale because of cost and food contamination. This year, large companies teamed up with start-ups, hoping to make polystyrene recycling practical and cost effective. Ineos Styrolution and Agilyx are looking at using pyrolysis to regenerate the styrene monomer, while Total and Polystyvert are exploring the process of dissolving polystyrene in a solvent and recrystallizing the polymer to rid it of contaminants.

Related: Recycling needs a revamp

Scientists also explored ways to solve the plastics problem with polymer innovations. Researchers led by Zhibin Guan at the University of California, Irvine, created a thermoset plastic that can be reshaped and recycled, which are features that most thermoset plastics, such as melamine and epoxy resin, lack. The thermoset that Guan’s group developed features dynamic boron-oxygen networks. The polymer can be converted to its starting monomer, a diboronic acid, via heating in boiling water (J. Am. Chem. Soc. 2018, DOI: 10.1021/jacs.8b03257).

Recyclability was also a priority for Eugene Y.-X. Chen, of Colorado State University. Chen’s group developed a recyclable polymer with mechanical strength and thermostability that are comparable to popular plastics’ properties. The new polymer, made from γ-butyrolactone fused to a trans cyclohexane group, was thermally stable and crystalline, features that could make it competitive with commonly used plastics. By applying heat or a chemical treatment, the chemists could quantitatively recover its monomer in a pure state, ready for repolymerization (Science 2018, DOI: 10.1126/science.aar5498).

Credit: CREDIT

This strained lactone can be polymerized with a catalyst. That polymer returns to the original monomer in the presence of zinc and heat.

“Solving the increasingly worsening worldwide plastics pollution problem takes a whole-society approach that will require the effort and cooperation of all relevant stakeholders, from plastics inventors to producers, retailers, consumers, and recyclers,” Chen says. “As chemistry led to the creation of the plastics modern life depends on, it will undoubtedly contribute key solutions to the current plastics problem.”

Related: Fighting ocean plastics at the source

Credit: ACS Catal.

Hydrogenating an inactive, oxygen-covered palladium nanoparticle leads to α-palladium hydride, an active catalyst for hydrogen peroxide production. Further hydrogenation forms β-palladium hydride, which produces water.

Solid catalysts, often in nanoparticle form, drive the overwhelming majority of industry’s major chemical processes. Yet many molecular subtleties of these reactions remain unknown, challenging scientists’ ability to improve these catalytic transformations.

Related: Single Atoms Mediate Reaction

In 2018, researchers designed new nanocatalysts and uncovered mechanistic details about the way they mediate reactions. The findings may help reduce the energy use and cost of some large-scale industrial processes.

At Karlsruhe Institute of Technology, for example, a group led by Dmitry E. Doronkin devised an X-ray spectroscopy–based method that tracked structural changes in palladium nanocatalysts as they converted hydrogen and oxygen to hydrogen peroxide.

Manufacturers produce nearly 5 million metric tons of H₂O₂ annually, mainly from anthraquinone via an energy-intensive, multistep process. Making the commodity—a commercial bleach, disinfectant, and oxidizer—directly from hydrogen and oxygen gases could save energy. But because of thermodynamics, the direct reaction of the gases forms water instead of H₂O₂.

Related: Making Hydrogen Peroxide Directly

By spying on catalyst changes that occurred on the fly inside a reactor, the team found that if it held the H₂-to-O₂ ratio between 0.5 and 2.0, hydrogen converted metallic palladium particles to α-palladium hydride, an active catalyst that makes hydrogen peroxide. At higher ratios, hydrogen converted the α-hydride to the β form, the catalytic culprit that makes water (ACS Catal. 2018, DOI: 10.1021/acscatal.7b03514).

Doronkin says the team has now scaled up the process, boosting the H₂O₂ yield by a factor of 100. He adds that “a large industrial company” has expressed interest in the work, but he’s keeping the details under wraps for now.

Credit: Franklin Tao/Univ. of Kansas

Tucked inside the pores of a zeolite, a single-atom rhodium catalyst couples methane, carbon monoxide, and oxygen to make acetic acid at low temperatures (Rh = green; Al = gold; O = red; C = blue).

Another example of catalyst sleuthing this year came from Franklin “Feng” Tao and coworkers at the University of Kansas. That team synthesized an alumino­silicate zeolite known as ZSM-5 and deposited isolated rhodium cations inside its micropores. The team showed that its new single-atom catalyst combined methane, carbon monoxide, and oxygen, producing acetic acid at around 150 °C.

Global production of acetic acid, a reagent for making monomers and other compounds, sits around 14 million metric tons. Commercial routes to the organic acid often involve methanol, which is made via industrial processes that run at temperatures as high as 1,000 °C. A lower-temperature route to acetic acid could lead to significant energy savings.

By using spectroscopy and other methods, the Kansas team determined that the catalyst takes the form of anchored RhO₅ species and charted its role in making acetic acid in an 18-step reaction mechanism (Nat. Commun. 2018, DOI: 10.1038/s41467-018-03235-7).

Tao notes that in a follow-up study, his group found that the single-atom rhodium catalyst can couple ethane with H₂O₂ to make acetic acid at temperatures as low as 25 °C, another finding that could potentially benefit an industrial catalytic process.

Related: Single-atom catalysts gained a toehold

Credit: Nature

Genome editing (left) prevented damage in a mouse cochlea with mutations in Tmc1 (right). Scale bar = 50 μm.

The year 2018 was a roaring one for hearing loss research. A study published at the end of 2017 reverberated throughout the field. A team led by David R. Liu of Harvard University and Zheng-Yi Chen of Harvard Medical School and Massachusetts Eye and Ear used CRISPR-Cas9 genome editing to prevent mutation-related ear damage and protect hearing in mice (Nature 2017, DOI:10.1038/nature25164).

The edited gene in that study—Tmc1—drew more attention this year when a different team confirmed that the protein it encodes plays a key role in converting sound vibrations to electrical signals (Neuron 2018, DOI:10.1016/j.neuron.2018.07.033), a detail scientists had been trying to nail down for almost two decades.

Related: Scientists confirm key protein that helps convert sound vibration to electrical signals

AZD5438 is a compound with potent protective abilities against cisplatin-induced hearing loss in mice.

The protein, TMC1, helps form the pore of an ion channel on inner-ear sensory cells. The channel’s pore opens in response to vibrations caused by sound waves. Once the channel opens, ions flow into the sensory cell, generating electric currents that trigger a series of signaling events that convert the sound waves into something the brain can process.

Chemistry played a key role in this discovery. Led by David Corey of Harvard Medical School and Jeffrey Holt of Boston Children’s Hospital, the scientists systematically modified one amino acid residue after another in the TMC1 protein in mice. They wanted to know whether adding a bulky charged group to each amino acid would block the ion flow and prevent the sound-induced electric currents. Because most of these chemical modifications inhibited the currents, the researchers were able to conclude that TMC1 forms the ion channel’s pore.

This discovery is great news for the 400 million people around the world who experience hearing loss, says Karen Avraham of Tel Aviv University. Corey notes that gene therapies that replace the mutant gene or correct it could be helpful for treating naturally occurring TMC1 mutations in humans. The genome-editing research performed by Liu and Chen in mice is an early-stage demonstration of that approach.

In related work, Jian Zuo and coworkers at St. Jude Children’s Research Hospital showed the potential of using small molecules, such as AZD5438, to treat hearing loss in mice (J. Exp. Med. 2018, DOI:10.1084/jem.20172246). The research focused on hearing loss induced by the cancer drug cisplatin, but, as Alan Cheng of Stanford University points out, the targets the group identified with its molecules may hold promise for treating other forms of hearing damage.

Related: Small molecules shield mice against hearing loss

Credit: Oxford PV

This tandem device (15.6 cm × 15.6 cm) features a silicon solar-cell bottom layer and a perovskite cell overcoat to maximize photovoltaic efficiency.

For more than 20 years, scientists and engineers have been driven by the tantalizing promise of converting sunlight to electricity with high efficiency using inexpensive materials. In 2018, a number of those energy harvesters set new performance records, milestones that pushed photovoltaic devices closer to meeting global energy needs. Perovskites have grabbed more solar-cell headlines in recent years than any other category of emerging photovoltaics because of their rapid improvement. These types of light-sensitive materials, such as (CH₃NH₃)PbI₃, have the perovskite crystal structure and ABX₃ stoichiometry.

Other types of low-cost photovoltaics, such as ones based on polymers or quantum dots, have improved slowly in the efficiency with which they convert sunlight to electricity. They’ve reached efficiency values in the low teens.

Related: Perovskite progress pushes tandem solar cells closer to market

In June, Oxford PV unveiled a silicon-perovskite tandem cell that was certified by an independent agency to have a conversion efficiency of 25.2%. That device pairs a silicon cell bottom layer with a perovskite cell overcoat to broaden the range of wavelengths it can absorb. The combo cell edged out common commercial silicon cells, which achieve about 22% efficiency, and bested the roughly 23% efficiency of the record-holding all-perovskite cell.

Just weeks later, Oxford PV announced that it beat its own record with another silicon-perovskite tandem cell—this one boasting 27.3% efficiency. “We are continuously improving our tandem solar cells,” says Chris Case, Oxford PV’s chief technology officer. Hinting at another record breaker, he adds, “We expect to have more exciting performance to report shortly.”

Researchers used these two compounds as electron-accepting materials to make a tandem organic solar cell with record-setting efficiency.

Perovskite cells weren’t the only ones that made headlines this year. Researchers at Nankai University debuted an organic tandem cell with a conversion efficiency of 17.3%, topping the previous record for organic cells of 14% (Science 2018, DOI: 10.1126/science.aat2612).

Organic photovoltaics are made by pairing two organic materials, a sunlight-absorbing electron donor and an electron acceptor. Led by Yongsheng Chen, the Nankai team used computations to find two well-matched sets of these materials, especially ones that exploit the sun’s plentiful near-infrared light. The study guided the researchers to a new electron acceptor, known as O6T-4F, and another acceptor known as F-M, which they had developed earlier.

This year’s solar-cell advances are impressive and encouraging, says Notre Dame University’s Prashant V. Kamat, who has been active in photovoltaics since the 1970s. Yet for perovskites and organic cells alike, he notes, “stability issues still haunt practical devices.” Additional improvements in long-term durability will help translate these scientific advances into widely used commercial technology, he says.

Related: Organic photovoltaic window device generates electricity and blocks heat

Credit: Nat. Clim. Change

This world map depicts the risk of various kinds of malnutrition caused by elevated CO₂ levels.

As intense hurricanes pounded the Atlantic and Gulf coasts and record-breaking wildfires raged in California this year, scientists rang loud alarms about global warming. Soon, they warned, it will be too late to prevent some of the worst effects of climate change, including wildfires, flooding, sea level rise, famine, and destruction of coral reefs.

In October, the Intergovernmental Panel on Climate Change (IPCC) released a report that suggested that to prevent these calamities and limit warming to 1.5 °C above preindustrial levels—the figure chosen in the Paris Agreement—drastic changes are needed. By 2030, net carbon dioxide emissions must fall by 45% relative to 2010 levels, and net emissions must taper to zero by 2050. Yet countries, including the US, are already failing to meet the reduction pledges made in the Paris Agreement only two years ago. In November, the US Global Change Research Program released the second volume of its Fourth National Climate Assessment, which outlines dire climate change consequences for the US, including a 10% economic contraction by century’s end and decreased agricultural production. The Trump administration criticized the report, which was mandated by Congress.

Related: Elevated atmospheric CO₂ may leave parts of the world more vulnerable to malnutrition

Scientists this year also reported data suggesting that elevated atmospheric levels of CO₂ will have more complex effects than simply increasing the planet’s global temperature: they will change the Earth’s chemistry in fundamental ways.

Researchers at the University of Southern California found that warming oceans will spur marine bacteria to fix more nitrogen from the atmosphere, converting dinitrogen into molecules that organisms can feed on. Such a change in ocean chemistry could bring new life to areas of the ocean that are nutrient-poor dead zones today, with unknown ecological effects (Nat. Clim. Change 2018, DOI: 10.1038/s41558-018-0216-8).

Researchers have known that on land, high levels of CO₂ are bad news for agriculture, causing a drop in the nutrient levels of plants such as wheat and rice. The mechanism behind this phenomenon is unknown. In August, researchers at Harvard University predicted that, given this nutrient loss and current emission trajectories, carbon emissions will cause more cases of malnutrition. By 2050, an additional 175 million people may be zinc deficient, and an additional 122 million may be protein deficient, the authors project (Nat. Clim. Change 2018, DOI: 10.1038/s41558-018-0253-3).

Scientists say there are steps we can take to avoid the worst of these climate change disasters. The IPCC report suggests meeting its stringent targets by implementing more carbon-capture technology and deploying nuclear and renewable energy sources. Others are calling for a tax on carbon emissions. This year’s memorial Nobel Prize in economics honored William D. Nordhaus of Yale University and Paul M. Romer of New York University for their work on carbon taxes and how technological innovation can spur economic growth.

Related: Last chance to curb greenhouse gas emissions and climate change to limit catastrophic effects, international panel says

Credit: Scimat/Science Source

A number of gut microbe species like Bifidobacterium adolescentis are affected by even a single course of broad-spectrum antibiotics.

This year scientists continued to unravel the complicated ties between human health and the thousands of microbial strains living on and inside our bodies. These microbes have evolved with humans for millennia and have been considered benign, if not helpful, to us. But as researchers at the University of Sheffield discovered, that’s not always the case. The team found that a noninfectious strain of skin bacteria called Micrococcus luteus promotes development of pathogenic Staphylococcus aureus infections in mice (Nat. Microbiol. 2018, DOI: 10.1038/s41564-018-0198-3).

What’s most interesting, says Trinity College Dublin’s Rachel M. McLoughlin, who wasn’t involved in the study, is that even dead M. luteus cells promoted infection, which means that dying, antibiotic-susceptible bacteria may actually help antibiotic-resistant bacteria survive.

Related: Two studies analyze the aftermath of antibiotics

M. luteus’s betrayal aside, many bacteria strains are still beneficial, especially in the gut. Several studies showed that disrupting this microbial community could have negative effects on our health.

One major disturbance comes from our use of antibiotics. A small study in humans reported that after a single course of treatment, several bacterial species in the gut did not return even after about six months (Nat. Microbiol.2018, DOI: 10.1038/s41564-018-0257-9). Another study, this one in mice, revealed that the antibiotic-triggered loss of some gut microbes messed with immune cells called macrophages and that this disruption could cause inflammatory conditions (Sci. Transl.Med. 2018, DOI: 10.1126/scitranslmed.aao4755). In both studies, the same strain of bacteria, called Bifidobacteria, went missing after the antibiotic treatment.

Scientists also learned how moving to a new country can affect gut microbes. In one study, members of two communities who emigrated from Southeast Asia to the US experienced a decrease in gut microbe diversity, losing many of their native strains and picking up other US-associated strains, though not as many as they lost. This change in microbiome diversity may predispose the immigrants to obesity and other metabolic disorders (Cell 2018, DOI: 10.1016/j.cell.2018.10.029). The team proposed that diet affected the composition of the immigrants’ microbiomes but suspect other causes also play a role.

Dan Knights, a microbiologist at the University of Minnesota who led the study, says the researchers are starting to dissect those possible contributing factors. They’re also studying whether the microbiome change can directly contribute to certain disorders.

“We’re just at the opening stages of understanding exactly which microbes matter in different diseases, but it’s very clear that having the wrong overall community of microbes can contribute to a wide range of human diseases,” Knights says.

Related: Immigrating to the U.S. leads to loss of microbiome diversity

Credit: ACS Cent. Sci.

Using electron crystallography to analyze microcrystals, chemists in California were able to determine the molecular structures of various organic compounds, seen in this micrograph (grid holes are 2 μm across).

Using electrons rather than X-rays to determine the structures of biomolecules seemingly achieved its zenith in 2017, when researchers won the Nobel Prize in Chemistry for innovations in cryo-electron microscopy. But electrons can also be used to determine the structures of small molecules, and in 2018 chemists garnered widespread attention for demonstrating how practical this technique, known as electron crystallography, can be.

Related: Cryo-electron microscopy innovators win 2017 Nobel Prize in Chemistry

Two teams working independently showed they could analyze tiny crystals of organic compounds by cooling the crystals and zapping them with an attenuated electron beam. Electrons interact more strongly with the molecules in crystals than X-rays do, so scientists need less material to get enough information to determine a structure.

The technique wasn’t new, but it had largely gone unnoticed by organic chemists. That changed after the two reports appeared. A team led by Tim Gruene, a scientist at the Paul Scherrer Institute, published a paper on the subject in mid-October (Angew. Chem., Int. Ed. 2018, DOI: 10.1002/anie.201811318). Researchers led by Jose A. Rodriguez, Hosea M. Nelson, and Tamir Gonen of the University of California, Los Angeles, along with the California Institute of Technology’s Brian M. Stoltz, reported similar work the next day on the preprint server ChemRxiv (2018, DOI: 10.26434/chemrxiv.7215332.v1), with the publication appearing in early November (ACS Cent. Sci. 2018, DOI: 10.1021/acscentsci.8b00760).

The researchers showed they could apply electron crystallography to crystals as small as 100 nm in all dimensions. Crystals studied with X-ray crystallography typically need to be at least 5 µm in all dimensions. Because electron crystallography can study such vanishingly small amounts of material, chemists can use the technique to study the myriad compounds that don’t form large, high-quality crystals. “We have been in talks with established vendors and some truly unexpected parties regarding strategies to get cryogenic electron crystallography to the broader organic chemistry community,” Nelson says. “There are some challenges that need to be overcome, for sure, but we think it will happen in the near future.”

Related: Electron Diffraction Technique Reveals Structure Of Vanishingly Small Protein Crystals

Chemists noted that the technique has limitations. While the analysis requires only a small amount of material, that material must be crystalline. And at the moment, the technique can determine only a molecule’s relative stereochemistry, not its absolute stereochemistry.

Nevertheless, the reports generated excitement. Derek Lowe, author of the popular chemistry and pharma blog In the Pipeline, hailed the reports as ushering in a new era of small-molecule organic structure determination.

Related: Electron crystallography could be a powerful tool for organic chemists

Ecigarettes have exploded in popularity among teens. According to the US Food and Drug Administration, their use by US high school students surged from 12% in 2017 to 21% in 2018, despite federal regulations barring anyone under 18 years of age from buying e-cigarettes.

In particular, e-cigarette maker Juul has dominated the market, as well as news coverage, this year. “The whole Juul thing is completely blowing up,” says Portland State University chemist James Pankow, referring to the product’s growing use and press attention. While the company says that its product is intended to help tobacco smokers quit, “Juuling” has taken off among young consumers, who are drawn to the product’s discreet design and appealing e-liquid flavors, such as mint and mango, surveys suggest. In May, Pankow and Portland State colleagues David Peyton and Anna K. Duell reported another explanation for Juul’s success: chemistry.

Related: E-cigarettes' chemistry may explain their popularity among teens

Using nuclear magnetic resonance spectroscopy, the team measured the amounts of two major forms of nicotine, protonated and free-base nicotine, in a variety of e-liquid brands and found that Juul e-liquids had the lowest amounts of free-base nicotine, which is associated with harsh throat sensations when inhaled (Chem. Res. Toxicol. 2018, DOI: 10.1021/acs.chemrestox.8b00097).

The researchers think that Juul products are popular in part because they deliver high amounts of nicotine in a way that’s easily inhalable. The group says another important aspect of the aerosolized e-liquids that scientists should study is the particles’ size. Smaller particles penetrate deeper into the respiratory system than larger ones and are more easily absorbed by the body, possibly making it easier to get a bigger nicotine hit from each puff, Pankow explains.

E-cigarette devices are also used to deliver cannabidiol (CBD), a nonpsychoactive cannabinoid that is legal with a prescription in most states to treat conditions like insomnia and pain. Hundreds of CBD e-liquids are available for purchase online and are completely unregulated, says Virginia Commonwealth University’s Justin L. Poklis.

Poklis and his team analyzed nine CBD e-liquids from one manufacturer after receiving a tip from an outside source questioning their purity. By using mass spectrometry, the group detected a synthetic cannabinoid called 5F-ADB, reported to have euphoric and other psychoactive effects, in four of the products. In one of the products, the researchers detected dextromethorphan, a compound found in cough medicine (Forensic Sci. Int. 2018, DOI: 10.1016/j.forsciint.2018.10.019).

The study doesn’t report the amount of the contaminants in the e-liquids, but Poklis says his team plans to publish those data in a follow-up study.

“Until you know how much of the material there is and compare it to its bioactivity and toxicity, it’s hard to know how significant” these findings are, Portland’s Peyton says. However, he says, if the manufacturer intentionally added these compounds, that would be alarming.

Related: Mass spectrometry measures chemical exposures in e-cigarette users’ mouths