More information soon.

In this lecture I will provide an overview of the growing use of machine learning methods in materials science, including for the energy sector. Machine learning has been mostly used to process the huge amounts of data generated with modern equipment in materials characterization and in applications such as in sensing and biosensing. Examples will also be given of integrated approaches to employ machine learning with data of varied natures, including images, text, and scientific data. These approaches are relevant not only to develop fully (or semi-) automated systems but also to achieve efficient knowledge discovery tools.

The main thrust of our research team at the Dense Energy Carriers (DEC) Division is the conversion of CO2 to alcohols, hydrocarbons, syngas or other value-added chemicals through light-assisted electrolysis of aqueous solution containing dissolved CO2. Our strategy involves the development of a solar-driven electrolyzer with photoactive electrodes, in order to achieve the CO2 reduction reaction (CO2RR) in a gas diffusion electrode (GDE) at the cathodic compartment, while the oxygen evolution reaction (OER) takes place at the anodic compartment. Using innovative methodologies, we synthesized different “families” of n-type and p-type semiconductors based on abundant elements and we investigated their properties using several techniques, including a novel “System for Photoelectrochemical Measurements” that we assembled at the Institute of Chemistry – UNICAMP. Ilmenite (FeTiO3) and bismuth vanadate (BiVO4), n-type semiconductors, exhibited promising results as photoanodes for light-assisted OER. On the other hand, the multiphasic complex oxides based on silver-bismuth tungstates (“ABW”), or copper tungstates (“CuW”), that behave as p-type semiconductor electrodes, were evaluated as photocathodes for CO2RR. Irradiated ABW photocathodes produced syngas and formate in negative potentials as low as -0.2 V (vs. Ag/AgCl); moreover, since ABW was very photoactive towards O2 reduction, it was also applied for hydrogen peroxide production in a photoreactor for water decontamination. An air-fed GDE|ABW photocathode was associated with a TiO2 photoanode and both electrodes were externally connected to a solar cell; irradiated by a solar simulator, this solar-driven photoreactor exhibited excellent performance for removal of the antibiotic amoxicillin from water. Also, in collaboration with researchers from USA and UK, the Cu-W electrodes were investigated using experimental and theoretical approach and interesting results were obtained. The Cu-W was also photoactive for CO2RR and the issues with long-term stability can be overcome using an aminosiloxane overlayer, which act as a protective layer against photocorrosion and also as a CO2RR-to-formate selectivity promoter. Thus, our perspectives include the association of an ilmenite (or bismuth vanadate) photoanode with the CO2 fed-GDE|Cu-W/aminosiloxane photocathode, externally connected to a perovskite solar cell, in a solar-driven electrolyzer for solar fuels production through CO2 conversion.

The development of synthesis technology to design nanomaterials that fulfil all requirements for photoelectrochemical (PEC) cell application has been an arduous task to accomplish. Chemical routes derived from polymeric precursor method stands out in scientific research and in consolidated industrial process of coating, but it has failed in production of active material for PEC. Herein, a mixture of alcohols as solvent to prepare the polymeric hematite precursor solution has succeeded in creating a well-adherent and conformational hematite layer onto the irregular transparent conductive oxide (TCO) topology. This conformational morphology improved the oxide/TCO interface revealed in the Scanning Electron Transmission Microscopy (STEM) analysis, which led to a more effective charge injection and partially mitigated the shunting recombination regularly observed in mesoporous materials (reducing the TCO exposition to the electrolyte). Lastly, the well-known difficulty to identify the role of element addition was unveiled by simultaneously running a STEM map using both Energy Dispersive and Electron Energy Loss Spectroscopies (EDS/EELS). imaging. These analyses show the Zr-rich crystal/crystal interface, while Ni preferentially enriches the crystal surface (exposed to the electrolyte). This comprehensive understanding of the solvent mixing and elemental addition enables the control and design of engineered interfaces for more cost-effective, large scale production and efficient photoelectrochemical (PEC) water splitting systems and devices.

The high energy-consuming that increases every year and the elevated environmental pollution caused by an exhaustive use of fossil fuels has boosted the search for renewable, sustainable, and clean energy sources. In this context, the use of solar energy has gained a lot of attention mainly because of its high energy capacity. Among the ways of using solar radiation such as heat, photovoltaic conversion, and energy chemical storage, the last is a promising and less developed approach. Currently, the photoelectrochemical cell (PEC) is the main technology for converting solar into chemicals through the production of high energy density molecules. In this scenario, one of the main applications of PEC is the production of hydrogen from the photoelectrochemical water electrolysis. Hydrogen is considered an attractive energy carrier because of its high energy density (122 kJ mol-1) with an energy/mass ratio about 2.5 times higher than fossil fuels. Additionally, hydrogen as a fuel generates nothing beyond water as product, and because of that is considered an environmentally friendly fuel. The semiconductors play an important role in PEC devices, they are responsible for absorbing the incident solar photons and converting them into electrons and holes. In a p-type semiconductor, the electrons photogenerated migrate to the semiconductor/electrolyte interface becoming available for reduction reactions. Thus, p-type semiconductors work as photocathodes in the PEC. On the other side, n-type semiconductors work as photoanodes acting in the water oxidation. Our efforts are drawn to fabricate sustainable, inexpensive, and highly efficient photoanodes and photocathodes. Our trajectory to fabricate good candidates as photoanodes begins with the development of WO3 and BiVO4 by low-cost approach, using spray-coating as method of deposition, and continues until the investigation of the effect of metal phosphate cocatalyst on the photocurrent for the oxygen production. Our photocathodes are based on chalcogenide compounds. Electrodeposited Sb2Se3 and Sb2S3 and their association with Pt and MoSx cocatalysts provide high photocurrent for hydrogen generation. By engineering defects by alloying, using other cocatalyst beyond Pt, and investigating different electron transport layers and substrates, we could develop CuGaSe2 and Cu2ZnSnSe4 – based photocathodes with high photocurrents and high efficiencies of solar energy to hydrogen. It is not the end yet, and still there is a long way to go, to improve the photocathodes and photoanodes we have been developing and making it possible for clean and renewable energy.

Pseudocapacitors combine electrostatic ion adsorption onto high surface area materials of electric double layer capacitors with ultra-fast faradaic reactions typical from batteries. In this talk, we are going to highlight our advances on combining the best of the both devices and a new class of operando characterization analyses.

Ionic liquids play an important role considering the world energy storage scenario as it could be used as additive to improve the efficiency of many devices. This talk will be focused on C16MIMBr-, an Ionic Liquid Crystal, which is able to form different microstructures according to the solvents medium. As a consequence, a long life cycle battery could be achieved.

Li-air batteries`s technology face plenty of operation challenges since the construction of the cathode electrode up to the device conditioning and test. Knowledge in this field is fundamental to achieve noteworthy advances. In this presentation it will be addressed the study of operation parameters pointing out our research group breakthroughs.

In situ and operando experiments using conventional and synchrotron radiation sources: a journey from electrochemistry to perovskite solar cells
Recent publications and environmental disasters confirm the critical moment that humanity is facing due to the global warming caused by the non- environmentally friendly human activity. To boost the transition from a fossil fuels-based economy to a more sustainable one, the field of chemistry, through the development and application of new materials for the storage and utilization of renewable energy, plays a key role.In the Dense Energy Carriers (DEC) division at CINE, we work in the design of new materials to capture and use solar light. To measure the physicochemical properties and structure of the materials, the reaction selectivity, catalysts activity and stability we use several in situ techniques, ranging from conventional (like FTIR and Raman) to most sophisticated, asX-ray synchrotron-based techniques.In this presentation, we will summarize important results using in situexperiments at different synchrotron radiation facilities around the globe to probe halide perovskite formation, stability, and composition. We will also present our most recent development for studies involving single nanoparticle experiments at the Carnauba beamline at SIRIUS, Brazil.In situ experiments using conventional techniques like UV-vis, FTIR, Raman, etc. have proven to be extremely valuable to answer several important questions to our projects.We will show examples of FTIR in situ experiments applied to understand photoelectrochemical CO2 reduction, photochemical water splitting [8] andstability of the perovskite quantum dots.

Towards a Sustainable Approach for Renewable Energy as a basis for a Multicolor Economy
Nowadays with the environmental concern and the need to spread development all around the word new emphasis has to be given for different ways to produce energy and chemicals besides conventional sources as oil and coal. In fact, there is a linear relationship between available energy and human welfare, which includes education, health and basic living infrastructure, among other things. This reflects in the human developing index. Large amount of the world population (something as 4,5 billions of people) has an average of 2-3 Kwatt per-capita, which means 10 to 50 times less than developed countries. If a world development is to happen and the CO 2 emissions are related to energy generation based on fossil fuels, renewable energies must play a decisive role. Taking these into consideration renewable energy will play a capital role and the concept of energy matrix based in many sources is potential possibility. However, sustainability has to account for providing suitable solutions. Bearing all this in mind, this work proposes a set of criteria that may be a tool to discriminate and classify possible solutions and how they may be put together to provide energy and chemical in a sustainable fashion. In doing so the concept of green economy is expanded to multicolor economy and this will the focus of this presentation. As part the energy matrix potentially may come from intermittent sources as photovoltaic and Wind energies, advanced energy storage devices are required to match the user demand with generation and this will be addressed taking into account the different applications and scenarios, always considering the sustainability as an issue.

We have witnessed an unprecedented revolution in the domestic supply of shale hydrocarbon reserves, which has the potential to lower the cost of our nation’s electricity, power, chemicals, and fuels for the next century. We will discuss how the Engineering Research Center for Innovative and Strategic Transformation of Alkane Resources (CISTAR) is developing the science and technology to use shale gas resources responsibly to generate chemicals and fuels and provide a transition to a sustainable energy future, while reducing greenhouse gas emissions and promoting environmentally responsible development of this opportune national resource.

Semiconductor nanostructures with tunable photoresponse can capture the visible and near IR photons quite effectively. By controlling the size and composition one can tune the photoresponse of semiconductor nanostructures in the entire visible region. The key advantage of semiconductor nanostructures lies in designing thin film solar cells with low temperature processing. These advantages significantly decrease the energy payback time since less energy is consumed (and hence a lower carbon footprint) during their manufacture. The early studies which focused on the synthesis of semiconductor quantum dots and elucidation of charge carrier dynamics have led to the evolution of high efficiency metal halide perovskite solar cells. In the present talk light induced charge carrier generation and transport across interfaces which are important in the operation of solar cells will be discussed. Recent developments in utilizing semiconductor quantum dots in light energy conversion will be discussed.

Porous ceria: 20 mol% gadolinia (20GDC – Gadolinia-Doped Ceria) ceramic matrices were prepared by thermal removal of sacrificial alkali carbonates, tape casting and freeze-drying techniques. Composite membranes were prepared with those porous 20GDC samples by infiltration of molten eutectic alkali carbonates (EAC). The pore structure of the membranes was evaluated by analysis of scanning electron microscopy (FEG-SEM) images. The deconvolution of the electrochemical impedance spectroscopy (EIS) diagrams measured in the solid-to-molten EAC temperature range (300 oC to 700 oC) allowed for evaluating the optimum operation temperature for carbon dioxide permeation. FEG-SEM images, EIS diagrams, Arrhenius plots for 20GDC/(Na,K)2CO3 will be presented.

Controlling the synthesis parameters during radiation-induced grafting (RIG) is a key feature to design anion-exchange membranes (AEMs) with enhanced properties that lead to high performances and stability. The present study provides a systematic investigation of the effect of RIG methods on physicochemical properties of LDPE-based AEMs with similar degree of grafting (DoG) and ion exchange capacity (IEC). Samples grafted by two different methods have the same molecular structure, but distinct physicochemical properties due to markedly differences in the degree of crosslinking. Grafting step directly influences the internal structure and morphology of AEMs. Detailed characterization showed that RIG method determines the mechanical properties, water transport, and the distribution of ionic groups in AEMs. These characteristics have direct impact on fuel cell performance, anion-conductivity and AEM stability.

The increased availability of natural gas due to the advancements in exploration and extraction technologies has boosted the interest in the direct conversion of methane to C2 molecules via oxidative coupling of methane (OCM) reactions. The design of catalysts by tailoring size, shape, and oxygen vacancies has a significant impact on their performance. Within this context, in this work, the fine-tuning of structural defects of ceria rod-like and cube-like shaped nanoparticles were performed via La3+ doping (LaxCe1-xO2-1/2) in molar ratios of x=5-90 mol %. The morphology control was achieved by varying the hydrothermal synthesis temperature from 110 °C to 180 °C. The characterization of LaxCe1-xO2-1/2 catalysts was performed by scanning electron microscopy, high-resolution electron microscopy, x-ray diffraction, and Raman spectroscopy. In the LaxCe1-xO2-1/2 nanorods, the fluorite crystalline structure is essentially maintained even for x = 0.6. The Raman spectra analysis indicates that the addition of La leads and increases the formation of different amounts of extrinsic and intrinsic oxygen. In addition, the catalytic performance of the new nanomaterials for the OCM reaction was studied at 750 °C. The reagents CH4 and O2 and the products H2, CO, CO2, C2H4, and C2H6, were measured by an in-line mass spectrometer. The catalytic tests for OCM reaction show that the LaxCe1-xO2-1/2 (x=50%) nanorods catalyst have 35% of methane conversion, 68% of selectivity, and 23% of yield for the formation of C2 products. The concentration of reactive oxygen species in the LaxCe1-xO2-1/2 (x=50%) nanorods was crucial to improving the conversion rates and selectivity.

In recent years, numerous machine learning applications in materials science research have flourished. Despite this widespread use, machine learning techniques are not straightforward and often require complex engineering to pre-process data and fit models properly. Different methods can help in this process, such as the bias-variance. This method decomposes the error in terms of bias, a component related to the models capability, and variance related to sampling. Results using a widely used benchmark dataset (QM9) and machine learning algorithm (Kernel Ridge Regression), and state-of-the-art representations show that they tend to fail in adequatelypredicting the atomization energy of a particular subset of molecules with planar structure. We emphasize the necessity of research teams with different expertise to conduct multidisciplinary research of machine learning and materials science.

With the advent of graphene in 2004, the scientific community has been motivated to explore new two-dimensional (2D) materials due their novel properties. Transition metal dichalcogenides (TMDs), due their optical bandgap in visible region and high absorption coefficient, has been largely studied in the scenario of photovoltaics (PV).[1] In these 2D systems, due the quantum confinement, excitonic effects are fundamental for the correct description of the optical bandgap. More importantly is the possibility to tune this property modifying the surround dielectric environment, making easier the enhancement of solar harvesting efficiency. However the large computational cost of DFT+BSE formalism can make this study very impeding, as a solution we developed an alternative methodology based in the combination of DFT with Wannier based Tight-Binding methods. We developed our own TB code to solve the exciton problem with, reducing significantly the computational cost, but also keeping a good accuracy. Our solution also allows to effectively control the dielectric environment by the choice of proper Coulomb potential. In this work we put our method to proof, studying 72 2H TMDs monolayers with the objective to understand the significance of excitonic effects in the efficiency our type II TMD heterojunctions solar cells. Through the investigation of structural, electronic, optical and excitonic properties of these monolayers, we could understand how excitons can change the heterojunction type and tunes power conversion efficiency (PCE). Only a small group of 14 monolayers showed the necessary properties for PV applications, being MoTe2/MoSe2, MoSe2/WS2 and MoTe2/WS2 heterojunctions the most promising compounds with an efficiency around 20%, which is attractive for commercial use.[2] Futhermore the present methodology has been applied for perovskites based materials.[3].

Sodium-ion batteries (SIBs) are promising candidates for large scale energy storage technologies, mainly due to their relatively low costs and chemical similarity with lithium-ion batteries (LIBs).1 In this context, ionic liquids (ILs) are an interesting electrolyte class due to their high chemical and thermal stability and relative conductivity, which combined can lead to high-performance electrolytes. We have performed a set of Density Functional Theory (DFT) calculations and classical molecular dynamics (MD) simulations for ILs based electrolytes, focusing on the understanding between theionic structure and electrolyte performance. Using DFT calculations we obtained the electrochemical windows (ECW) for 25 ILs containing imidazolium and ammonium cations paired with several sulfonyl-imide-based anions. We observed that small changes in the ionic structure, like the replacement of the imidazolium ring by an alkyl group, can lead to a significant increase in the ECW values. Besides, our calculations showed that the ammonium anions can reach electrochemical Windows (ECWs) that exceed 5:0 eV, which is a good range for high-performance batteries. From the atomistic MD simulations, we observed that the addition of the Na+ ion led to the slowing down of the ions transport, which is related to the formation of large [NaxAniony]x – y aggregates. A detailed evaluation of the aggregate structure showed that one Na+ interacts with more than one anion at the same time, forming a "bridge" interaction between them, that increases aggregate size upon the addition of more sodium-ions.3 Besides, the use of small and asymmetrical ions are an interesting approach to increase the IL ionic conductivities. Using coarse-grained models and the structure factor (S(q)) analysis, we observed that the aggregates are dispersed in the whole liquid structure, and the large aggregates lead to long range interactions between the ions. At the same time, the large aggregates change the Na+ diffusion mechanism, at high mole fractions the sodium-ions are moving across different solvation shells in the hopping diffusion mechanism. While at low mole fractions, e.g. small aggregates, the diffusion is mainly based on the vehicular mechanism. This change increases the sodium-ion apparent transport number and the electrolyte performance at higher mole fractions. So far, we have shown that the joining of classical molecular dynamics simulations, e.g., atomistic and coarse-grained models, and DFT calculations are a power-full tool for the screening of ionic liquids electrolytes5.

Overview of CINE activities on methane conversion using electrochemistry
Modern society is rooted in fossil oil-derived materials that are ubiquitous in our daily life. To achieve the stringent targets of reduction in CO2 emissions in the next decades, the decarbonization of the (petrochemical) industry is imperative. This is a challenging development yet to be fully tackled and that has received much less attention in the research community as compared to the research on low carbon power generation technologies that are nowadays already available and cost-effective. The solid background of our research group on fuel cell technologies, including pioneer work on the use of ethanol as fuel, allowed us to extend the application of such devices as (electro)chemical reactors for value-added molecules using methane and other strategic feedstocks. This overview of the M2P program will cover the main results of the six projects dedicated to methane conversion reactions using fuel cell-like electrochemical reactors. As a new research topic being developed in our group, we have put a great effort in building up a productive scientific environment with modern facilities that enable us to tackle essentially any electrochemical reaction under different conditions from low to high temperature. Such infrastructure along with a vibrant research team have translated in important results which include, new materials such as high-performance anionic polymeric membranes, ion-conducting composite separation membranes, shape-controlled thermally stable nanoceramics for high-temperature catalysis, and in operando detection of methanol in methane-fed fuel cells and photocatalytic reactors.

CMSC Achievements: Computational Investigation of Perovskites Based Materials
Recently, the international energy agency indicated that photovoltaics (PV) will increase its contribution from nearly 1% in the world energy matrix to about 30% to reach net-zero CO2 emissions. Several technologies have contributed to the present status of photovoltaics (e.g., silicon, CaTe, etc), however, higher efficiency and low-cost materials are required to reach those expectations. The CMSC program focus on two classes of materials, namely, (a) short-term applications based on perovskites materials and (b) long-term applications based on two-dimensional van der Waals heterostructures. In this talk, we will focus our discussion on the most important scientific achievements obtained by the CMSC program for perovskites-based materials, which includes application of density functional theory calculations to understand the role of cations substitutions in the classical perovskite structures, as well as the investigation of double perovskites formation, zero-dimensional perovskites, the role of octahedron distortions in the electronic and excitonic properties. Finally, we will discuss also the role of binders in the stability of two-dimensional perovskites.

Education and Dissemination of Knowledge: Achievements, Challenges, and Future Plans
CINE has the mission to produce knowledge at the research frontier, and in parallel to transfer technlogy to the business sector, in the field of new energies. Thus, internal and external education and dissemination of knowledge plays a critical role to the CINE success. At this talk, we will summarize the most important achievements so far, challenges, the and role of digital integration for our future.

We will talk about the Hytron’s history, a company founded as a spin-off from Hydrogen Laboratory at UNICAMP, and the importance of hydrogen as an energy vector for Brazil.