High-Throughput Bayesian Optimization of Cement-Salt Hydrates Composites for Seasonal Thermochemical Energy Storage

Alessio Mondello, Giulio Barletta, Luca Lavagna, Matteo Fasano, Matteo Pavese, Eliodoro Chiavazzo · 2026

Read High-Throughput Bayesian Optimization of Cement-Salt Hydrates Composites for Seasonal Thermochemical Energy Storage by Alessio Mondello, Giulio Barletta, Luca Lavagna, Matteo Fasano, Matteo Pavese, Eliodoro Chiavazzo (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-19494.txt.

Introduction

Mondello and team use high-throughput Bayesian optimization to design cement-salt composites for seasonal heat storage. Salt hydrate storage offers much but couples makeup, synthesis feasibility, performance and cost in tricky ways. Searching over salt type, concentration and cement ratios, the campaign found Pareto-optimal mixes based on calcium chloride, zinc nitrate and lithium chloride. The lithium chloride mix reached about 458 kilojoules per kilogram, and new mixes beat earlier cement materials five times, though trailing silica-gel rivals.

Catalog record

CREATORAlessio Mondello, Giulio Barletta, Luca Lavagna, Matteo Fasano, Matteo Pavese, Eliodoro Chiavazzo
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE85276
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.19494

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

High-Throughput Bayesian Optimization of Cement-Salt Hydrates Composites for Seasonal Thermochemical Energy Storage Alessio Mondello, Giulio Barletta, Luca Lavagna, Matteo Fasano, Matteo Pavese, Eliodoro Chiavazzo Abstract Thermochemical energy storage (TCES) based on salt hydrates is a promising route for seasonal heat storage; however, the design of practical sorbent materials remains challenging due to a non-trivial coupling between composition, synthesis feasibility, performance, and cost. Here, focusing on salt-into-matrix cement-based composites, we demonstrate that a high-throughput experimental framework based on Bayesian optimization (BO) can be used to orchestrate the optimization process of composite materials for low-temperature TCES. The explored design space is defined by salt type, salt concentration, water-to-cement ratio, and additive-to-cement ratio, while two competing objectives are pursued in parallel, namely the specific energy and the specific energy cost. The BO-guided campaign identified Pareto-optimal composites based on CaCl 2 , Zn(NO 3 ) 2 , and LiCl, highlighting the promise of cement-salt combinations that have been only marginally explored, or not previously reported, in cement-based TCES systems. The best-performing formulation (LiCl-based), achieved an average specific energy of about 458 kJ kg − 1 458\text{\,}\mathrm{kJ}\text{\,}{\mathrm{kg}}^{-1} , whereas CaCl 2 - and Zn(NO 3 ) 2 -based composites showed lower but still competitive specific energy values combined with more favorable specific energy cost. Overall, the optimized formulations improved the specific energy of previously developed cement-based materials by up to a factor of five, although it remains below that of state-of-the-art composites based on silica gel and expanded vermiculite. Nonetheless, the present materials, notably CaCl 2 - and Zn(NO 3 ) 2 -based composites, offer an attractive cost-to-performance balance, highlighting BO as an effective strategy for accelerated TCES materials discovery. 1 Introduction Thermal energy storage (TES) is widely regarded as a key enabling technology for the large-scale integration of intermittent renewable energy sources towards decarbonization of the energy sector both in the industrial and civil sectors skevi_reviewing_2025 , spietz_thermochemical_2025 , whitepaperIWG52024 . A major challenge lies in the temporal mismatch between energy availability and demand. In particular, solar thermal energy exhibits both a daily peak during daylight hours and a seasonal peak in summer, whereas heat demand is typically higher in the evening and during winter. While short- and medium-term thermal energy storage technologies are characterized by a higher technological maturity morciano2025trending , neri2020numerical , a strong research effort is still needed to develop affordable long-term TES systems capable of storing thermal energy on a seasonal timescale skevi_reviewing_2025 , spietz_thermochemical_2025 . More specifically, three main families are typically identified according to the main mechanism by which heat is stored and released. Sensible heat storage (SHS) technologies rely on temperature variations of a storage mean; latent heat storage (LHS) technologies exploit phase transitions; and thermochemical energy storage (TCES) technologies store heat through reversible chemical or physical interactions skevi_reviewing_2025 , spietz_thermochemical_2025 . SHS and LHS are well-established for short- and medium-term applications; however, they suffer from thermal losses and relatively low energy densities, which limit their suitability for long-term storage clark_experimental_2021 , spietz_thermochemical_2025 . TCES systems, although not yet fully mature, are an attractive option for long-term applications due to the intrinsic advantages of their storage mechanism skevi_reviewing_2025 , spietz_thermochemical_2025 . TCES exhibit higher volumetric energy densities than SHS and LHS, enabling a reduction in the required plant footprint spietz_thermochemical_2025 . Moreover, long-term storage is feasible because energy is stored in the form of chemical potential: if the physical separation between adsorbate and adsorbent is properly managed, the stored energy remains preserved and is released only when the two active components are brought into contact skevi_reviewing_2025 , clark_experimental_2021 . For low-temperature TCES, water vapor is the most commonly used adsorbate due to its safety, availability, and favorable operating conditions zhang_research_2025 . The selection of the most suitable sorbent material remains an open and highly active research topic within the scientific community spietz_thermochemical_2025 , liu_systematic_2025 . The optimal choice depends strongly on the type of TCES system (open or closed) and on the operating conditions, such as available charging temperature and discharging humidity yang_salt_2023 , liu_systematic_2025 . Hydrated salts are among the most extensively investigated sorbent materials because of their high theoretical energy density spietz_thermochemical_2025 , yang_salt_2023 , zhang_research_2025 . These salts store and release heat through successive dehydration and hydration cycles; however, their practical implementation is hindered by well-known drawbacks, including deliquescence, swelling, mechanical instability, and limited cyclability clark_experimental_2021 , yang_salt_2023 , liu_systematic_2025 . To overcome these limitations, the incorporation of hydrated salts into a stable porous host matrix has been widely explored in the literature, and the resulting materials are commonly referred to as “composite salt in porous matrix” (CSPM) systems gordeeva_composites_2012 . The porous matrix provides mechanical support, promotes salt dispersion, mitigates agglomeration, and could improve heat and mass transfer within the material zhang_research_2025 , liu_systematic_2025 . In addition to conventional porous matrices such as silica gel, zeolites, vermiculite, and other mineral supports zhang_research_2025 , liu_systematic_2025 , cement-based materials have recently been investigated as innovative and low-cost host matrices for hydrated salts clark_experimental_2021 , salustro_thermal_2024 , mohapatra_salt_2023 , skevi_reviewing_2025 . The use of cement as host matrix offers several potential advantages, including robust structural integrity and the presence of a multiscale porous network capable of accommodating and dispersing salt clark_experimental_2021 , salustro_thermal_2024 , lavagna_insight_2022 . Despite this potential, the formulation of cement–salt composites for TCES remains a highly non-trivial design problem. The performance of the final material depends on multiple coupled variables, including the salt chemistry, salt loading, binder composition, and additive content, which jointly affect thermochemical performance, processability, stability, and cost of the materials. As a result, the associated search space is combinatorial and strongly non-linear. A full experimental exploration of this space would require thousands of synthesis–characterization cycles, each involving sample preparation, curing, granulation, and sorption analysis over timescales of weeks. This makes conventional trial-and-error strategies, one-factor-at-a-time studies, or even dense grid-based screening impractical for the efficient identification of optimal formulations. In recent years, data-driven optimization strategies have emerged as powerful tools for accelerating materials discovery and reducing the experimental burden associated with expensive searches ling_high-dimensional_2017 , lookman_active_2019 , gomez-bombarelli_automatic_2018 , janet_accelerating_2018 , jin_machine-learning-assisted_2022 . Among these approaches, Bayesian optimization (BO) has proven particularly effective for navigating large experimental spaces under tight resource constraints. By combining probabilistic surrogate models with acquisition functions that balance exploration and exploitation, BO enables the efficient identification of optimal materials or processing conditions using relatively small datasets frazier_tutorial_2018 , brochu_tutorial_2010 , shahriari_taking_2016 . This methodology has been successfully applied to a wide range of materials problems, enabling sample-efficient discovery of high-performance materials in chemistry, catalysis, and energy storage systems guo_bayesian_2023 , yang_bayesian_2022 , khatamsaz_bayesian_2023 , langner_beyond_2020 , agarwal_discovery_2021 , zhang_exploring_2023 , bonke_multi-variable_2024 , zhang_optimizing_2023 , doan_quantum_2020 , niu_accelerated_2025 , yik_accelerating_2025 , li_computational_2025 . A few examples of using sequential learning algorithms for the optimization of thermal energy storage materials have been reported in purely numerical studies trezza_minimal_2022 . However, its application to TCES materials remains largely unexplored, especially for cement-based salt hydrate composites, where synthesis/stability constraints and multi-objective trade-offs play a central role. In this work, building on the preliminary studies by Lavagna et al. lavagna_cementitious_2020 , Clark et al. clark_experimental_2021 , and Salustro et al. salustro_thermal_2024 , we develop a high-throughput, data-driven experimental framework based on BO

Read the complete text