A US perspective on closing the carbon cycle to defossilize difficult-to-electrify segments of our economy.

Electrification to reduce or eliminate greenhouse gas emissions is essential to mitigate climate change. However, a substantial portion of our manufacturing and transportation infrastructure will be difficult to electrify and/or will continue to use carbon as a key component, including areas in aviation, heavy-duty and marine transportation, and the chemical industry. In this Roadmap, we explore how multidisciplinary approaches will enable us to close the carbon cycle and create a circular economy by defossilizing these difficult-to-electrify areas and those that will continue to need carbon. We discuss two approaches for this: developing carbon alternatives and improving our ability to reuse carbon, enabled by separations. Furthermore, we posit that co-design and use-driven fundamental science are essential to reach aggressive greenhouse gas reduction targets.

Physics-Based Machine Learning Models Predict Carbon Dioxide Solubility in Chemically Reactive Deep Eutectic Solvents.

Carbon dioxide (CO2) is a detrimental greenhouse gas and is the main contributor to global warming. In addressing this environmental challenge, a promising approach emerges through the utilization of deep eutectic solvents (DESs) as an ecofriendly and sustainable medium for effective CO2 capture. Chemically reactive DESs, which form chemical bonds with the CO2, are superior to nonreactive, physically based DESs for CO2 absorption. However, there are no accurate computational models that provide accurate predictions of the CO2 solubility in chemically reactive DESs. Here, we develop machine learning (ML) models to predict the solubility of CO2 in chemically reactive DESs. As training data, we collected 214 data points for the CO2 solubility in 149 different chemically reactive DESs at different temperatures, pressures, and DES molar ratios from published work. The physics-driven input features for the ML models include σ-profile descriptors that quantify the relative probability of a molecular surface segment having a certain screening charge density and were calculated with the first-principle quantum chemical method COSMO-RS. We show here that, although COSMO-RS does not explicitly calculate chemical reaction profiles, the COSMO-RS-derived σ-profile features can be used to predict bond formation. Of the models trained, an artificial neural network (ANN) provides the most accurate CO2 solubility prediction with an average absolute relative deviation of 2.94% on the testing sets. Overall, this work provides ML models that can predict CO2 solubility precisely and thus accelerate the design and application of chemically reactive DESs.

Tuning Surface, Phase, and Magnetization of Superparamagnetic Magnetite by Ionic Liquids: Single-Step Microwave-Assisted Synthesis.

Achieving colloidal and chemical stability in ferrofluids by surface modification requires multiple steps, including purification, ex situ modification steps, and operation at high temperatures. In this study, a single-step microwave-assisted methodology is developed for iron oxide nanoparticle (IONP) synthesis utilizing a series of imidazolium-based ionic liquids (ILs) with chloride, bis(trifluoromethylsulfonyl)imide, and pyrrolide anions as the reaction media, thus eliminating the use of volatile organics while enabling rapid synthesis at 80 °C as well as in situ surface functionalization. The characterized surface functionality, hydrodynamic particle size, magnetization, and colloidal stability of the IONPs demonstrate a strong dependence on the IL structure, ion coordination strength, reactivity, and hydrophilicity. The IONPs present primarily a magnetite (Fe3O4) phase with superparamagnetism with the highest saturation magnetization at 81 and 73 emu/g at 10 and 300 K, respectively. Depending on the IL coating, magnetization and exchange anisotropy decrease by 20 and 2-3 emu/g (at 35 wt % IL), respectively, but still represent the highest magnetization achieved for coated IONPs by a coprecipitation method. Further, the surface-functionalized superparamagnetic magnetite nanoparticles show good dispersibility and colloidal stability in water and dimethyl sulfoxide at 0.1 mg/mL concentration over the examined 3 month period. This study reports on the intermolecular and chemical interactions between the particle surface and the ILs under synthetic conditions as they relate to the magnetic and thermal properties of the resulting IONPs that are well suited for a variety of applications, including separation and catalysis.

Accurate Machine Learning for Predicting the Viscosities of Deep Eutectic Solvents.

Deep eutectic solvents (DESs) are emerging as environmentally friendly designer solvents for mass transport and heat transfer processes in industrial applications; however, the lack of accurate tools to predict and thus control their viscosities under both a range of environmental factors and formulations hinders their general application. While DESs may serve as designer solvents, with nearly unlimited combinations, this unfortunately makes it experimentally infeasible to comprehensively measure the viscosities of all DESs of potential industrial interest. To assist in the design of DESs, we have developed several new machine learning (ML) models that accurately and rapidly predict the viscosities of a diverse group of DESs at different temperatures and molar ratios using, to date, one of the most comprehensive data sets containing the properties of over 670 DESs over a wide range of temperatures (278.15-385.25 K). Three ML models, including support vector regression (SVR), feed forward neural networks (FFNNs), and categorical boosting (CatBoost), were developed to predict DES viscosity as a function of temperature and molar ratio and contrasted with multilinear and two-factor polynomial regression baselines. Quantum chemistry-based, COSMO-RS-derived sigma profile (σ-profile) features were used as inputs for the ML models. The CatBoost model is excellent at externally predicting DES viscosity, as indicated by high R2 (0.99) and low root-mean-square-error (RMSE) and average absolute relative deviations (AARD) (5.22%) values for the testing data sets, and 98% of the data points lie within the 15% of AARD deviations. Furthermore, SHapley additive explanation (SHAP) analysis was employed to interpret the ML results and rationalize the viscosity predictions. The result is an ML approach that accurately predicts viscosity and will aid in accelerating the design of appropriate DESs for industrial applications.

Vibrations and Phase Stability in Mixed Valence Antimony Oxide.

α-Sb2O4 (cervantite) and ß-Sb2O4 (clinocervantite) are mixed valence compounds with equal proportions of SbIII and SbV as represented in the formula SbIIISbVO4. Their structure and properties can be difficult to calculate owing to the SbIII lone-pair electrons. Here, we present a study of the lattice dynamics and vibrational properties using a combination of inelastic neutron scattering, Mössbauer spectroscopy, nuclear inelastic scattering, and density functional theory (DFT) calculations. DFT calculations that account for lone-pair electrons match the experimental densities of phonon states. Mössbauer spectroscopy reveals the ß phase to be significantly harder than the α phase. Calculations with O vacancies reveal the possibility for nonstoichiometric proportions of SbIII and SbV in both phases. An open question is what drives the stability of the α phase over the ß phase, as the latter shows pronounced kinetic stability and lower symmetry despite being in the high-temperature phase. Since the vibrational entropy difference is small, it is unlikely to stabilize the α phase. Our results suggest that the α phase is more stable only because the material is not fully stoichiometric.

Predictive understanding of the surface tension and velocity of sound in ionic liquids using machine learning.

Knowledge of the physical properties of ionic liquids (ILs), such as the surface tension and speed of sound, is important for both industrial and research applications. Unfortunately, technical challenges and costs limit exhaustive experimental screening efforts of ILs for these critical properties. Previous work has demonstrated that the use of quantum-mechanics-based thermochemical property prediction tools, such as the conductor-like screening model for real solvents, when combined with machine learning (ML) approaches, may provide an alternative pathway to guide the rapid screening and design of ILs for desired physiochemical properties. However, the question of which machine-learning approaches are most appropriate remains. In the present study, we examine how different ML architectures, ranging from tree-based approaches to feed-forward artificial neural networks, perform in generating nonlinear multivariate quantitative structure-property relationship models for the prediction of the temperature- and pressure-dependent surface tension of and speed of sound in ILs over a wide range of surface tensions (16.9-76.2 mN/m) and speeds of sound (1009.7-1992 m/s). The ML models are further interrogated using the powerful interpretation method, shapley additive explanations. We find that several different ML models provide high accuracy, according to traditional statistical metrics. The decision tree-based approaches appear to be the most accurate and precise, with extreme gradient-boosting trees and gradient-boosting trees being the best performers. However, our results also indicate that the promise of using machine-learning to gain deep insights into the underlying physics driving structure-property relationships in ILs may still be somewhat premature.

Photochemically-Driven CO2 Release Using a Metastable-State Photoacid for Energy Efficient Direct Air Capture.

One of the grand challenges underlying current direct air capture (DAC) technologies relates to the intensive energy cost for sorbent regeneration and CO2 release, making the massive scale (GtCO2 /year) deployment required to have a positive impact on climate change economically unfeasible. This challenge underscores the critical need to develop new DAC processes with substantially reduced regeneration energies. Here, we report a photochemically-driven approach for CO2 release by exploiting the unique properties of an indazole metastable-state photoacid (mPAH). Our measurements on simulated and amino acid-based DAC systems revealed the potential of mPAH to be used for CO2 release cycles by regulating pH changes and associated isomers driven by light. Upon irradiating with moderate intensity light, a ≈55 % and ≈68 % to ≈78 % conversion of total inorganic carbon to CO2 was found for the simulated and amino acid-based DAC systems, respectively. Our results confirm the feasibility of on-demand CO2 release under ambient conditions using light instead of heat, thereby providing an energy efficient pathway for the regeneration of DAC sorbents.

Microwave Regeneration and Thermal and Oxidative Stability of Imidazolium Cyanopyrrolide Ionic Liquid for Direct Air Capture of Carbon Dioxide.

Understanding the oxidative and thermal degradation of CO2 sorbents is essential for assessing long-term sorbent stability in direct air capture (DAC). The potential degradation pathway of imidazolium cyanopyrrolide, an ionic liquid (IL) functionalized for superior CO2 capacity and selectivity, is evaluated under accelerated degradation conditions to elucidate the secondary reactions that can occur during repetitive absorption-desorption thermal-swing cycles. The combined analysis from various spectroscopic, chromatographic, and thermal gravimetric measurements indicated that radical and SN 2 mechanisms in degradation are encouraged by the nucleophilicity of the anion. Thickening of the liquid and gas evolution are accompanied by 50 % reduction in CO2 capacity after a 7-day exposure to O2 under 80 °C. To prevent long exposure to conventional thermal heating, microwave (MW) regeneration of the CO2 -reactive IL is used, where dielectric heating at 80 and 100 °C rapidly desorbs CO2 and regenerates the IL without any measurable degradation.

One-Dimensional Glassy Behavior of Ultraconfined Water Strings.

Water is renowned for its anomalous behaviors, which can be linked to a distributed H-bond network in bulk water. Ultraconfinement of the water molecule can remove H-bonding, leaving only molecular water. In natural cordierite crystals, water is trapped in an orthorhombic channel with an average diameter of 5.7 Å, running through the center of the unit cell parallel to the c-axis. Calorimetric measurements reveal the existence of a one-dimensional (1D) glass linked to this water. In these channels, water molecules in truncated, sparse 1D strings interact only via dipole-dipole correlations. A physical 1D glass is formed from these strings. This unusual state can be explained by a modified Ising model. This model predicts a dependence of the glass transition temperature, Tg, on the size of these domains. This is confirmed experimentally.

Dialing in Direct Air Capture of CO2 by Crystal Engineering of Bisiminoguanidines.

Direct air capture (DAC) technologies that extract carbon dioxide from the atmosphere via chemical processes have the potential to restore the atmospheric CO2 concentration to an optimal level. This study elucidates structure-property relationships in DAC by crystallization of bis(iminoguanidine) (BIG) carbonate salts. Their crystal structures are analyzed by X-ray and neutron diffraction to accurately measure key structural parameters including molecular conformations, hydrogen bonding, and π-stacking. Experimental measurements of key properties, such as aqueous solubilities and regeneration energies and temperatures, are complemented by first-principles calculations of lattice and hydration free energies, as well as free energies of reactions with CO2, and BIG regenerations. Minor structural modifications in the molecular structure of the BIGs are found to result in major changes in the crystal structures and the aqueous solubilities within the series, leading to enhanced DAC.

Quantum Cascade Laser Infrared Spectroscopy for Online Monitoring of Hydroxylamine Nitrate.

We describe a new approach for high sensitivity and real-time online measurements to monitor the kinetics in the processing of nuclear materials and other chemical reactions. Mid infrared (Mid-IR) quantum cascade laser (QCL) high-resolution spectroscopy was used for rapid and continuous sampling of nitrates in aqueous and organic reactive systems, using pattern recognition analysis and high sensitivity to detect and identify chemical species. In this standoff or off-set method, the collection of a sample for analysis is not required. To perform the analysis, a flow cell was used for in situ sampling of a liquid slipstream. A prototype was designed based on attenuated total reflection (ATR) coupled with the QCL beam to detect and identify chemical changes and be deployed in hostile environments, either radiological or chemical. The limit of detection (LOD) and the limit of quantification (LOQ) at 3σ for hydroxylamine nitrate ranged from 0.3 to 3 and from 3.5 to 10 g·L-1, respectively, for the nitrate system at three peaks with wavelengths between 3.8 and 9.8 µm.

Improved ZnS nanoparticle properties through sequential NanoFermentation.

Sequential NanoFermentation (SNF) is a novel process which entails sparging microbially produced gas containing H2S from a primary reactor through a concentrated metal-acetate solution contained in a secondary reactor, thereby precipitating metallic sulfide nanoparticles (e.g., ZnS, CuS, or SnS). SNF holds an advantage over single reactor nanoparticle synthesis strategies, because it avoids exposing the microorganisms to high concentrations of toxic metal and sulfide ions. Also, by segregating the nanoparticle products from biological materials, SNF avoids coating nanoparticles with bioproducts that alter their desired properties. Herein, we report the properties of ZnS nanoparticles formed from SNF as compared with ones produced directly in a primary reactor (i.e., conventional NanoFermentation, or "CNF"), commercially available ZnS, and ZnS chemically synthesized by bubbling H2S gas through a Zn-acetate solution. The ZnS nanoparticles produced by SNF provided improved optical properties due to their smaller crystallite size, smaller overall particle sizes, reduced biotic surface coatings, and reduced structural defects. SNF still maintained the advantages of NanoFermentation technology over chemical synthesis including scalability, reproducibility, and lower hazardous waste burden.

Synthesis and structure of synthetically pure and deuterated amorphous (basic) calcium carbonates.

It is generally believed that H2O and OH- are the key species stabilizing and controlling amorphous calcium carbonate "polyamorph" forms, and may in turn control the ultimate crystallization products during synthesis and in natural systems. Yet, the locations and hydrogen-bonding network of these species in ACC have never been measured directly using neutron diffraction. We report a synthesis route that overcomes the existing challenges with respect to yield quantities and deuteration, both of which are critically necessary for high quality neutron studies.

CO2 Capture from Ambient Air by Crystallization with a Guanidine Sorbent.

Carbon capture and storage is an important strategy for stabilizing the increasing concentration of atmospheric CO2 and the global temperature. A possible approach toward reversing this trend and decreasing the atmospheric CO2 concentration is to remove the CO2 directly from air (direct air capture). Herein we report a simple aqueous guanidine sorbent that captures CO2 from ambient air and binds it as a crystalline carbonate salt by guanidinium hydrogen bonding. The resulting solid has very low aqueous solubility (Ksp =1.0(4)×10-8 ), which facilitates its separation from solution by filtration. The bound CO2 can be released by relatively mild heating of the crystals at 80-120 °C, which regenerates the guanidine sorbent quantitatively. Thus, this crystallization-based approach to CO2 separation from air requires minimal energy and chemical input, and offers the prospect for low-cost direct air capture technologies.

Neutron Scattering of Residual Hydrogen in 1,4-Dioxane-d8 Liquid: Understanding Measurements with Molecular Dynamics Simulations.

That incoherent scattering from protiated molecular liquids adds a constant background to the measured scattering intensity is well-known, but less appreciated is the fact that coherent scattering is also induced by the presence of hydrogen in a deuterated liquid. In fact, the scattering intensity can be very sensitive, in the small-q region, with respect to the amounts and distribution of residual H in the system. We used 1,4-dioxane liquid to demonstrate that the partial structure factors of the HD and DD atom pairs contribute significantly to intermolecular scattering and that uncertainty in the extent of deuteration account for discrepancies between simulations and measurements. Both contributions to uncertainty have similar magnitudes: scattering interference of the hydrogen-deuterium pair, and complementary interference from the deuterium-deuterium pair by virtue of chemical inhomogeneity. This situation arises in practice since deuteration of liquids is often 99% or less. A combined experimental and extensive computational study of static thermal neutron scattering of 1,4-dioxane demonstrates the foregoing. We show, through simulations, that the reason for the differences is the content of protiated dioxane (vendors quote 1%). We estimate that up to 5% (at 298 K and at 343 K) protiated molar fraction may be involved in generating the scattering differences. Finally, we find that the particular distribution of hydrogen in the protiated molecules affects the results significantly; here, we considered molecules to be either fully protiated or fully deuterated. This scenario best reconciles the computational and experimental results, and leads us to speculate that the deuteration synthesis process tends to leave a molecule either fully deuterated or fully protiated. Although we have used 1,4-dioxane as a model liquid, the effects described in this study extend to similar liquids, and similar systematic experimental/computational studies can be performed to either understand measurements or calibrate/validate molecular dynamics models.

Pyrolysis of phenethyl phenyl ether tethered in mesoporous silica. Effects of confinement and surface spacer molecules on product selectivity.

There has been expanding interest in exploring porous metal oxides as a confining environment for organic molecules resulting in altered chemical and physical properties including chemical transformations. In this paper, we examine the pyrolysis behavior of phenethyl phenyl ether (PPE) confined in mesoporous silica by covalent tethers to the pore walls as a function of tether density and the presence of cotethered surface spacer molecules of varying structure (biphenyl, naphthyl, octyl, and hexadecyl). The PPE pyrolysis product selectivity, which is determined by two competitive free-radical pathways cycling through the two aliphatic radical intermediates (PhCH·CH(2)OPh and PhCH(2)CH·OPh), is shown to be significantly different from that measured in the liquid phase as well as for PPE tethered to the exterior surface of nonporous silica nanoparticles. Tailoring the pore surface with spacer molecules further alters the selectivity such that the PPE reaction channel involving a molecular rearrangement (O-C phenyl shift in PhCH(2)CH·OPh), which accounts for 25% of the products in the liquid phase, can be virtually eliminated under pore confinement conditions. The origin of this change in selectivity is discussed in the context of steric constraints on the rearrangement path inside the pores, surface and pore confinement effects, pore surface curvature, and hydrogen bonding of PPE with residual surface silanols supplemented by nitrogen physisorption data and molecular dynamics simulations.

Nonfunctionalized polydimethyl siloxane superhydrophobic surfaces based on hydrophobic-hydrophilic interactions.

Superhydrophobic surfaces based on polydimethyl siloxane (PDMS) were fabricated using a 50:50 PDMS-poly(ethylene glycol) (PEG) blend. PDMS was mixed with PEG, and incomplete phase separation yielded a hierarchic structure. The phase-separated mixture was annealed at a temperature close to the crystallization temperature of the PEG. The PEG crystals were formed isothermally at the PDMS/PEG interface, leading to an engineered surface with PDMS spherulites. The resulting roughness of the surface was studied by scanning electron microscopy (SEM) and atomic force microscopy (AFM). The PDMS spherulites, a few micrometers in diameter observed from SEM images, were found to have an undulated (rippled) surface with nanometer-sized features. The combination of micrometer- and nanometer-sized surface features created a fractal surface and increased the water contact angle (WCA) of PDMS more than 60°, resulting in a superhydrophobic PDMS surface with WCA of >160°. The active surface layer for the superhydrophobicity was approximately 100 µm thick, illustrating that the material had bulk superhydrophobicity compared to conventional fluorocarbon or fluorinated coated rough surfaces. Theoretical analysis of the fractal surface indicates that the constructed surface has a fractal dimension of 2.5, which corresponds to the Apollonian sphere packing.

Confinement effects on product selectivity in the pyrolysis of phenethyl phenyl ether in mesoporous silica.

Pyrolysis of phenethyl phenyl ether confined in mesoporous silicas by covalent grafting results in significantly increased product selectivity compared with fluid phases.

Pyrolysis of mesoporous silica-immobilized 1,3-diphenylpropane. Impact of pore confinement and size.

Mesoporous silicas such as SBA-15 and MCM-41 are being actively investigated for potential applications in catalysis, separations, and synthesis of nanostructured materials. A new method for functionalizing these mesoporous silicas with aromatic phenols is described. The resulting novel hybrid materials possess silyl aryl ether linkages to the silica surface that are thermally stable to ca. 550 degrees C, but can be easily cleaved at room temperature with aqueous base for quantitative recovery of the organic moieties. The materials have been characterized by nitrogen physisorption, FTIR, NMR, and quantitative analysis of surface coverages. The maximum densities of 1,3-diphenylpropane (DPP) molecules that could be grafted to the surface were less than those measured on a nonporous, fumed silica (Cabosil) and were also found to decrease as a function of decreasing pore size (5.6-1.7 nm). This is a consequence of steric congestion in the pores that is magnified at the smaller pore sizes, consistent with parallel studies conducted using a conventional silylating reagent, 1,1,3,3-tetramethyldisilazane. Pyrolysis of the silica-immobilized DPP revealed that pore confinement leads to enhanced rates and altered product selectivity for this free-radical reaction compared with the nonporous silica, and the rates and selectivities also depended on pore size. The influence of confinement is discussed in terms of enhanced encounter frequencies for bimolecular reaction steps and pore surface curvature that alters the accessibility and resultant selectivity for hydrogen transfer steps.

Pore size effects in the pyrolysis of 1,3-diphenylpropane confined in mesoporous silicas.

A new method for derivatizing mesoporous silicas, SBA-15 and MCM-41, with a substituted phenol is described, and pore confinement and surface curvature are shown to impact the reaction rate and product selectivity for the pyrolysis of surface-immobilized 1,3-diphenylpropane.