Adsorption thermal energy storage works by exploiting the fact that water molecules bind strongly to the enormous internal surface area of microporous solids such as zeolites. When humid air flows over a dry zeolite bed, water vapour is taken up and the heat released during binding can be delivered to a process or a building. Charging the store reverses the process: heating the bed drives the water out, and the energy is effectively locked in the desiccated crystal until needed. Because the energy is stored chemically rather than as sensible heat, storage densities of around 180 kilowatt hours per cubic metre are theoretically achievable, several times what a hot water tank can manage. Open adsorption systems have already found commercial niches, notably in an energy-efficient dishwasher developed by Bosch Siemens Hausgeräte together with ZAE Bayern, but researchers in the field concede that adsorption heat storage remains in the early stages of commercialisation.

The central problem the German team set out to solve is one of matching. An adsorbent that performs superbly at one combination of temperature and humidity may be mediocre at another, and the conditions are dictated entirely by the application. A mobile storage unit harvesting industrial waste heat operates under very different temperatures than a sorption system built into a tumble dryer. The team therefore proposed a method in four steps: first compare candidate materials under application conditions using characteristic curves, then measure dynamic breakthrough curves in a realistic fixed bed, third validate a computer simulation of that bed against the experiments, and finally test the hydrothermal stability of the shortlisted materials by cycling them hundreds of times under the actual operating conditions. The novelty, the authors argue, lies in folding hydrothermal stability testing, which previous studies often omitted, into the design workflow itself.

The first step hinged on measuring adsorption equilibria with unusual precision. The team used a simultaneous thermal analysis instrument operated in thermogravimetric analysis mode, coupled to a humidity generator and a high-accuracy dew point mirror calibrated over a range from minus 50 to plus 95 degrees Celsius with an uncertainty of no more than plus or minus 0.1 degrees. Roughly 300 milligrams of each zeolite were exposed to carefully controlled combinations of temperature and dew point, generating the equilibrium data needed to construct so-called characteristic curves. These curves, rooted in adsorption potential theory going back to Dubinin, plot the volume of water adsorbed against the adsorption potential, a measure of the free energy change as vapour binds to the pore walls. Once a characteristic curve is known, the volumetric water uptake for any pair of adsorption and desorption conditions can be read off directly, allowing materials to be compared before a single prototype is built.

The measurements proved reliable. For the 13X material the characteristic curve was valid across adsorption potentials from 105 to 3,495 kilojoules per kilogram, and for the NaY material from 106 to 3,524 kilojoules per kilogram, with combined measurement uncertainties of only a few kilojoules per kilogram and about 0.001 cubic centimetres per gram in adsorbed volume. Crucially, adsorption and desorption measurements coincided almost perfectly, indicating that true equilibrium was reached and that neither material shows meaningful hysteresis. The curves also agreed closely with published data for the same and similar zeolites, confirming that the thermal analysis route can replace bulkier methods that require nearly 50 grams of sample. When the team then read the curves against two application scenarios, a mobile sorption storage and an industrial tumble dryer, the NaY zeolite came out slightly ahead in both cases, delivering a volumetric water turnover of 0.270 cubic centimetres per gram for the mobile scenario compared with 0.240 for the 13X, and 0.266 against 0.235 for the dryer scenario.

Static equilibrium, however, says nothing about how fast a real storage bed charges and discharges. For that, the researchers built an open-air test rig in which a blower pushes air through a countercurrent humidification column and an electric heater capable of reaching 350 degrees Celsius, before the flow enters a fixed bed of zeolite pellets. The column, 148 millimetres in diameter and filled to a height of 300 millimetres, is double-walled with evacuated perlite filling to form vacuum super insulation, minimising heat losses and temperature gradients across the bed. Dew point mirrors and Pt100 sensors continuously recorded the air conditions immediately before and after the bed. The resulting breakthrough curves, which track how temperature and humidity at the outlet evolve over time, revealed clear differences between the materials. During desorption, the NaY bed showed sharper, more pronounced transitions, which are generally preferred in applications, whereas during adsorption the 13X bed changed state more abruptly and cleanly.

These experimental curves then served as the benchmark for a one-dimensional fixed bed simulation built on four coupled partial differential equations describing the mass and energy balances of gas and sorbent. The model treats the pellets as isothermal, accounts only for mass transfer resistance in the gas boundary layer around the grains, and was run in Matlab with spatial steps of 0.25 to 0.5 centimetres and time steps of one to two minutes. When adjusted for heat losses, the simulation reproduced temperature levels and phase durations of the measured breakthrough curves very well, and even captured a striking dew point peak in the first hour of NaY adsorption runs, confirming that this behaviour was genuine start-up dynamics rather than a measurement artefact. Transitions in the simulation were somewhat sharper than in the experiments, a discrepancy attributed to airflow maldistribution and thermal capacitances not covered by the model. With validation complete, the tool can now be used to scale column geometry and operating parameters to real systems without repeating costly experiments.

The final and, in this case, decisive step was the cycling test. A purpose-built rig subjected both zeolites to 140 full cycles under conditions derived from the tumble dryer integration: desorption at 300 degrees Celsius with a dew point of 60 degrees, and adsorption at 42.5 degrees Celsius with a dew point of 30 degrees. Reference cycles at milder conditions were run periodically, and the samples were weighed to track water uptake. The 13X zeolite began with an adsorbed water mass of 0.281 grams per gram of dry material, but lost 4.6 percent within just ten cycles and finished at 36.5 percent below its starting capacity after 140 cycles, consistent with earlier reports that zeolite 13X loses roughly half its capacity when exposed to 300-degree humid conditions. The NaY zeolite, starting slightly higher at 0.292 grams per gram, lost only 1.1 percent after ten cycles and just 3.3 percent after 140, with the rate of decline slowing as cycling continued.

For the envisioned application the conclusion is unambiguous: 13XBFK should not be used under those conditions, while NaYBFK shows the hydrothermal resilience needed to sustain storage capacity over a service life. The team is careful to note that stability verdicts cannot be transferred between applications, since lower or higher temperatures and dew points change the ageing chemistry, and that economics and adsorption enthalpies, which feed into the coefficient of performance, must also enter the final decision. The authors acknowledge that their methodology is time-consuming and demands up to three separate test setups, a drawback they consider nonetheless essential for responsible sorption system design. The work was part of the AMThES project supported by the German Federal Ministry for Economic Affairs and Climate Action.

The broader significance of the study goes beyond one dryer or one storage vessel. It demonstrates a template for screening adsorbents in a way that mirrors reality: equilibrium data first, dynamics second, validated simulation third, and brutal long-term stability testing last. As industrial waste heat recovery and electrified heating push the demand for compact, high-density thermal storage, the difference between a material that fades by a few percent and one that collapses by a third could determine whether adsorption technology finally moves from the laboratory bench to the factory floor. For zeolite NaY, at least under these punishing conditions, the future looks considerably drier, and that is precisely why it lasts.

Subject of Research: Application-oriented characterisation and simulation of zeolite materials for adsorption thermal energy storage

Article Title: Application oriented material characterisation and simulation for adsorption thermal energy storage

Article References: Schmit, H., Pöllinger, S., Schubert, T., Lävemann, E., & Hiebler, S. (2025). Application oriented material characterisation and simulation for adsorption thermal energy storage. Frontiers in Thermal Engineering, 5, Article 1513507. https://doi.org/10.3389/fther.2025.1513507

Image Credits: AI Generated

DOI: 10.3389/fther.2025.1513507

Keywords: adsorption thermal energy storage, zeolite, characteristic curve, breakthrough curves, fixed bed simulation, hydrothermal stability, thermogravimetric analysis, water adsorption, thermal energy density, tumble dryer, mobile sorption storage, cycling stability

A research team at the Institute of Physics of the Benemérita Universidad Autónoma de Puebla in Mexico, led by Azury Nava Guzmán together with Enrique Sánchez Mora, Anabel Romero López, and Mary Carmen Amador Meléndez, has now systematically mapped how a remarkably humble molecule can tame these unruly particles. Their chosen stabilizer is glycine, the simplest of the amino acids that build proteins. Writing in the Journal of Nanoparticle Research, the team shows that the amount of glycine present during synthesis acts as a dial that tunes crystallite size, particle diameter, surface charge, colloidal behavior, and even the thermal response of zero-valent iron nanoparticles, all without disturbing the metallic iron core itself. The work, published as open access on 13 October 2026, was funded by Mexico’s Secretariat of Science, Humanities, Technology and Innovation.

The synthesis itself is a classic of the field. The researchers dissolved ferric chloride in water under a flowing nitrogen atmosphere at 20 degrees Celsius, then injected sodium borohydride, a strong reducing agent, at a controlled rate. The reaction reduces iron ions to metallic iron while releasing hydrogen gas, and it is exothermic enough that the mixture warms to about 42 degrees Celsius at a pH of roughly 9. The resulting black powder was washed an extraordinary 35 times with acetone, using magnetic separation and sonication between washes, to strip away boron byproducts and residual salts. To functionalize the particles, the team simply added glycine to the iron salt solution before reduction, at three doses: 8, 16, and 32 millimoles, producing samples labeled NPs-FeG1, NPs-FeG2, and NPs-FeG3.

X-ray diffraction delivered the first surprise. Every sample, functionalized or not, showed the sharp reflections of body-centered cubic metallic iron, with characteristic peaks near 45.1, 65.3, and 82.8 degrees in two-theta. Glycine left the crystal structure untouched. But the width of the main (110) peak told a subtler story: it broadened steadily as glycine concentration rose, from a full width at half maximum of 0.94 degrees for bare nanoparticles to 1.62 degrees for the highest glycine dose. Applying Scherrer’s equation, the team calculated that crystallite size shrank progressively from 9.55 nanometers in the unfunctionalized particles to 7.81, 6.65, and finally 5.54 nanometers across the three functionalized samples. Glycine, in other words, behaves like a surfactant at the growing crystal surface, capping the domains of metallic iron before they can enlarge.

Scanning electron microscopy then revealed an apparent paradox. While the crystallites got smaller, the particles got bigger. The average diameter of unfunctionalized nanoparticles measured 58.95 plus or minus 12.80 nanometers, based on 1000 particles analyzed per sample, but the glycine-treated particles grew to 75.51, 83.36, and 83.79 nanometers for the three doses, with the two higher doses essentially indistinguishable. The resolution of the paradox lies in scale: each particle seen under the microscope is a polycrystalline aggregate containing many of the smaller crystalline domains measured by diffraction. Glycine-mediated surface interactions appear to promote the assembly of these larger particles during synthesis, even as they suppress growth within each crystalline domain. All samples, meanwhile, assembled themselves into striking chain-like structures, a hallmark of magnetic dipole-dipole interactions pulling quasi-spherical iron particles nose-to-tail.

The chains are not purely magnetic accidents. The researchers propose a layered interaction model in which surface chemistry plays a decisive role. Metallic iron exposed to water and oxygen during synthesis develops a thin skin of hydroxides and oxyhydroxides such as goethite, lepidocrocite, and ferrihydrite. Hydroxyl groups on this skin can hydrogen-bond with one another on bare particles. When glycine is present, its zwitterionic form, carrying both a positively charged ammonium group and a negatively charged carboxylate, can anchor through its amino end to the hydroxylated surface, leaving the carboxylate groups exposed to the surrounding water. Carboxylates on neighboring particles can then pair up, stitching dimers and chains together. Increasing the glycine dose raises the density of these polar groups on the surface, favoring exactly the hydrogen-bonded agglomerates the microscope revealed.

Infrared spectroscopy backed this mechanism with molecular-level detail. The functionalized particles showed weak O-H stretching bands between 3727 and 3696 inverse centimeters, indicating that glycine modifies the environment of surface hydroxyls. Bands between 2670 and 2678 inverse centimeters, tied to the ammonium group of zwitterionic glycine interacting with surface hydroxyls, appeared only in the functionalized samples. The carboxylate-asymmetric-stretch and ammonium-deformation bands near 1615 to 1565 inverse centimeters shifted progressively to lower frequencies with increasing glycine dose, and a mixed CH2 bending and NH2 torsion band near 810 inverse centimeters moved as well, both signatures of a changing vibrational environment for adsorbed glycine. Bands between 664 and 602 inverse centimeters confirmed Fe-O vibrations from the surface oxide and oxyhydroxide species, which XRD suggested must be present only in small or poorly crystalline amounts, since no bulk oxide phases were detected.

Perhaps the most practically consequential result came from zeta potential measurements, a proxy for colloidal stability. Bare nanoparticles carried a zeta potential of minus 22.95 millivolts, indicating only limited stability in water. The lowest glycine dose pushed the value to minus 32.83 millivolts, the most negative of all samples, reflecting exposed carboxylate groups and stronger electrostatic repulsion between particles. But the story reversed at higher doses: the 16 and 32 millimole samples dropped to just minus 10.07 and minus 9.68 millivolts. The team attributes this to charge shielding by dense glycine coverage, which lets hydrogen bonding and dipole-dipole forces between amino and carboxyl groups outweigh electrostatic repulsion, promoting the agglomeration seen by electron microscopy. The lesson is that more stabilizer is not always better; for zero-valent iron, the sweet spot for colloidal stability sits at the lowest glycine concentration tested.

Thermal analysis added a final dimension. Thermogravimetric curves showed multistage behavior for all samples: water desorption below 150 degrees Celsius, dehydroxylation of oxyhydroxides and gradual oxidation of metallic iron between 200 and 600 degrees, and a pronounced mass gain above 600 degrees as iron takes up oxygen. The functionalized samples added a distinct mass loss between 150 and 350 degrees Celsius corresponding to glycine decomposition, and the oxidation-related mass changes were less pronounced than in bare particles, suggesting glycine partially delays oxidation. The team also tracked the Curie transition near 770 degrees Celsius, where the particles lose their ferromagnetism and the apparent mass recorded by the balance jumps, an instrumental artifact the authors carefully disentangled from real oxidation. Raman spectra of the heat-treated residues revealed hematite and beta-FeOOH in all samples, with the functionalized residues additionally showing lepidocrocite signatures, confirming that the organic layer reshapes the thermal degradation pathway of the particles.

What makes this study notable is its systematic dose-response design applied to metallic iron rather than the iron oxides that dominate the glycine literature. Most prior work on glycine and iron nanomaterials focused on magnetite and related oxides, whose chemistry cannot simply be extrapolated to a reactive metallic core. By demonstrating that glycine concentration independently tunes crystallite growth, particle aggregation, surface charge, and thermal stability while preserving the body-centered cubic iron phase, the Mexican team has provided a practical recipe book for engineers designing zero-valent iron materials. The implications reach into groundwater remediation, where better-dispersed particles could intercept contaminants more efficiently; into catalysis, where surface chemistry governs electron transfer; and toward emerging biomedical and biosensing applications, where amino-acid coatings offer a biocompatible interface. The next step, the authors note, will be surface-sensitive techniques such as X-ray photoelectron or Mössbauer spectroscopy to pin down the exact oxidation states at the glycine-iron interface.

Subject of Research: Effect of glycine concentration on the physicochemical properties of glycine-functionalized zero-valent iron nanoparticles

Article Title: Glycine-functionalized zero-valent iron nanoparticles: effect of glycine concentration on physicochemical properties

Article References: Guzmán, A. N., Mora, E. S., López, A. R., & Meléndez, M. C. A. (2026). Glycine-functionalized zero-valent iron nanoparticles: effect of glycine concentration on physicochemical properties. Journal of Nanoparticle Research, 28(10), Article 264. https://doi.org/10.1007/s11051-026-06787-8

Image Credits: AI Generated

DOI: 10.1007/s11051-026-06787-8

Keywords: zero-valent iron, nanoparticles, glycine, surface functionalization, X-ray diffraction, zeta potential, colloidal stability, FTIR spectroscopy, thermogravimetric analysis, Raman spectroscopy, environmental remediation, nanoparticle synthesis

The research team, led by Borui Lu, Yue Wang, Peiyao Yu, Mingzhe An, Ziruo Yu, and Yu Zhang, investigated L-(+)-tartaric acid at dosages of 0, 0.10, 0.15, and 0.20 percent of cement mass, holding the water-to-cement ratio fixed at 0.40 and using a fixed polycarboxylate superplasticizer dosage of 0.02 percent to keep the admixture background constant. They paired two parallel series: aggregate-free cement pastes for setting time, thermogravimetric analysis, X-ray diffraction, and electron microscopy, and one-to-three cement-to-sand mortars for fluidity, mercury intrusion porosimetry, and mechanical testing. This paired design allowed the same chemical perturbation to be followed from paste-scale hydration chemistry through to mortar-scale pore structure and strength. Specimens were cured at 20 degrees Celsius and above 95 percent relative humidity and examined at 7, 28, and 56 days, giving a cross-age picture that most earlier tartaric acid studies, which focused on early hydration alone, did not attempt.

The fresh-state results show a textbook dosage response. Both initial and final setting times increased continuously as the tartaric acid dose rose, with the delay becoming steep at the higher dosages. Mortar fluidity followed the same pattern: the flow-table spread grew from 214 millimeters for the control mix to 242 millimeters at 0.20 percent, an increase of about 13.1 percent under the fixed superplasticizer dosage. Because cement setting occurs when hydration products connect neighboring particles into a load-bearing network, the measured delay points directly to a longer period before that network reaches percolation. The authors interpret the combined setting and flow data as progressively stronger early-age retardation, consistent with a molecular picture in which tartrate ions coordinate calcium in solution and adsorb on clinker and hydrate surfaces, modifying dissolution and the conditions required for calcium silicate hydrate nucleation.

Quantitative X-ray diffraction with Rietveld refinement, using a 50 percent corundum internal standard and a common constrained five-phase protocol across all twelve datasets, provided an independent marker of the same trend. At 7 days, residual C3S, the reactive alite phase that drives early strength, rose systematically from 13.45 percent in the control paste to 17.33, 19.53, and 20.35 percent at the 0.10, 0.15, and 0.20 percent dosages respectively. Residual beta-C2S showed a much weaker dosage trend, indicating that the retardation acts primarily on alite. By later ages, residual C3S declined in all mixtures, demonstrating that silicate consumption continued after the strongly retarded early stage rather than being permanently suppressed. This distinction between delayed and diminished hydration is central to the study’s argument that retardation can be a benefit rather than a penalty.

Portlandite, the crystalline calcium hydroxide that cement hydration produces in abundance, served as a second chemical record. Thermogravimetric analysis, quantifying portlandite from the dehydroxylation mass loss between 400 and 550 degrees Celsius, showed contents falling monotonically at 7 days from 18.75 percent in the control to 18.01, 16.26, and 14.43 percent with increasing dosage. Yet by 56 days the picture had inverted: the 0.10 and 0.15 percent pastes reached 22.58 and 23.12 percent portlandite, exceeding the control’s 21.12 percent. The quantitative XRD data mirrored this early suppression and later recovery, and the agreement between the two independent techniques at 7 days provides a cross-method consistency check. The divergence in absolute values at later ages reflects the different analytical signals involved, mass loss versus diffraction intensity, and the authors interpret each dataset within its own measurement basis.

Scanning electron microscopy with energy-dispersive X-ray spectroscopy added local chemical evidence for the microstructural story. A representative elongated plate-like crystal in the 0.15 percent paste at 7 days gave a calcium-rich composition with a calcium-to-silicon atomic ratio of about 3.55 and no detectable aluminum or sulfur, a signature compatible with portlandite once the silicon contribution from adjacent silicate hydrate within the interaction volume is considered. Across all dosages and ages, similar plate-like crystals appeared with dosage- and age-dependent changes in size, abundance, and spatial arrangement. This matters because portlandite habit is known to be sensitive to the chemical environment of growth: organic molecules can alter crystal number, size, and growth pattern, and dissolved ions can shift the relative stability of specific crystal faces. The tartaric acid environment thus appears to leave a legible fingerprint on the crystal architecture of the hardened paste.

The pore network of the corresponding mortars evolved in step with the phase history. Mercury intrusion porosimetry, interpreted in terms of pore-entry or threshold dimensions rather than true pore sizes, showed that for the 0.15 percent mortar, cumulative intrusion at the smallest displayed entry diameter of about 11 nanometers fell from 0.0395 milliliters per gram at 7 days to 0.0314 milliliters per gram at 56 days, a deficit relative to the control that widened from 9.6 percent to 22.2 percent. The principal differential peaks stayed within a narrow 32 to 40 nanometer band throughout curing, meaning the characteristic entry scale was stable while the accessible volume changed most. By 56 days the 0.15 percent mixture had the lowest cumulative intrusion of all four mortars, linking the sustained post-retardation hydration documented in the pastes to a progressively denser, less accessible pore network in the mortars.

Strength development crowned the sequence. Both flexural and compressive strengths increased as the dosage rose from zero to 0.15 percent and then declined at 0.20 percent, making the 0.15 percent mix the best performer at every age. Relative to the control, its flexural strength was higher by 20.0, 14.6, and 9.6 percent at 7, 28, and 56 days, while compressive strength exceeded the control by 33.7, 16.0, and 11.7 percent at the same ages. The advantage was largest at 7 days and narrowed as the control caught up in hydration, consistent with the established understanding that strength in hydrated cement is governed jointly by porosity, pore structure, and degree of hydration. Notably, the 0.20 percent mixture, despite showing the lowest early cumulative intrusion at 7 and 28 days, did not match the 0.15 percent strength, suggesting that excessive retardation leaves the hydration degree too far behind for pore refinement alone to compensate.

The study also raises an intriguing stereochemical question. Previous work on brushite bioceramics and calcium sulfate cements found that L-(+)-tartaric acid and other L-enantiomeric alpha-hydroxycarboxylic acids regulated crystal growth and enhanced mechanical performance more strongly than their D counterparts, identifying molecular handedness as a possible control variable in calcium-mineral crystallization. The present dataset characterizes only the L stereoisomer in ordinary Portland cement, so matched comparisons of L-, D-, meso-, and racemic tartaric acid systems remain an open experimental frontier. The authors likewise note that the portlandite observations connect plausibly to carbonation behavior, since both the amount and spatial form of calcium hydroxide influence how carbonation progresses, motivating future measurements of carbonation depth and dimensional change.

For now, the practical takeaway is a carefully bounded one. At a water-to-cement ratio of 0.40, 0.15 percent L-(+)-tartaric acid best balanced early hydration control with later hardened-state development, combining pronounced setting delay and improved flow with continued silicate consumption, the lowest late-age accessible pore volume, and the highest measured strengths. Because tartaric acid’s retarding action is strongly concentration-dependent, and because dosage expressed per cement mass translates into different effective concentrations in the mixing water at different water-to-cement ratios, the optimum window is expected to shift with mixture design. The authors caution that verification across other ratios and cement chemistries is needed before broader dosage recommendations can be made. Even so, the work demonstrates that a modest, wine-cellar-simple molecule, dosed with precision, can orchestrate the entire arc of cement hydration from first stiffening to final strength.

Subject of Research: Effects of L-(+)-tartaric acid dosage on hydration, microstructure, and strength development of Portland cementitious materials

Article Title: Hydration regulation, microstructural evolution and strength development of cementitious materials modified by L-(+)-tartaric acid

Article References: Lu, B., Wang, Y., Yu, P., An, M., Yu, Z., & Zhang, Y. (2026). Hydration regulation, microstructural evolution and strength development of cementitious materials modified by L-(+)-tartaric acid. Case Studies in Construction Materials, 25, Article e06609. https://doi.org/10.1016/j.cscm.2026.e06609

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06609

Keywords: cement hydration, tartaric acid, retarding admixture, portlandite, Rietveld QXRD, thermogravimetric analysis, mercury intrusion porosimetry, compressive strength, microstructure, C3S, pore structure, construction materials

The principle behind thermal analysis is elegantly simple. Every food has a chemical signature, and that signature dictates how it behaves under heat. DSC measures heat flow into or out of a sample relative to a reference as temperature changes, capturing melting points, crystallization temperatures, glass transitions, and protein denaturation events. TGA, by contrast, continuously records changes in sample mass under a controlled atmosphere, revealing moisture content, decomposition temperatures, and degradation pathways. Because adulteration, substitution, or raw material variation alters these thermal events—shifting melting temperatures, changing enthalpy values, or reshaping mass-loss curves—each food generates a characteristic thermal fingerprint that fraud cannot easily disguise.

What makes these techniques so attractive compared with conventional methods is practicality. Chromatographic techniques such as gas chromatography-mass spectrometry and liquid chromatography-mass spectrometry remain the gold standard for definitive molecular identification, but they demand extensive sample preparation, expensive instrumentation, solvents, and long analysis times. DSC and TGA require minimal or no sample preparation, tiny sample quantities, no solvents, and only minutes of analysis time. This aligns squarely with the principles of green analytical chemistry, and it makes the techniques well suited to high-throughput routine screening where hundreds of samples must be triaged quickly.

Edible oils illustrate the power of the approach most vividly. Premium oils such as extra virgin olive oil are among the most frequently adulterated commodities in the world, and their market value depends on botanical origin, geography, and production method. Because each oil’s triacylglycerol composition is unique, each produces a distinctive crystallization profile that adulteration measurably distorts. Recent studies reviewed by the authors show DSC combined with chemometrics detecting sunflower, corn, and soybean adulteration in extra virgin olive oil with quantitative prediction errors as low as a few percent. In flaxseed oil work, linear discriminant analysis achieved 99.5 percent classification accuracy for rapeseed adulteration, while artificial neural networks delivered quantitative predictions with a correlation coefficient of 0.996. Data fusion of DSC with GC-MS in walnut oil studies pushed support vector machine classification to 100 percent accuracy.

Honey, another fraud magnet, has proven equally amenable to thermal fingerprinting. The glass transition temperature of honey is highly sensitive to sugar composition and water content, so adding cheap syrups such as high-fructose corn syrup, rice syrup, or inverted sugar systematically lowers the glass transition and alters heat capacity changes. One research team used DSC with partial least squares regression to successfully predict adulterant concentrations in sunflower honey contaminated with five different syrups at levels of 5 to 20 percent. More strikingly, recent work has coupled DSC with artificial intelligence: convolutional neural networks boosted by synthetic minority over-sampling improved classification accuracy of adulterated stingless bee and Tualang honeys from a baseline of 24 to 67 percent up to 60 to 91 percent, while a separate study using graph-based semi-supervised learning achieved accurate classification of corn syrup adulteration with only a limited number of labeled samples.

Beyond oils and honey, the review documents thermal analysis succeeding across a remarkable range of matrices. In dairy, DSC coupled with machine learning models including gradient boosting machines and multilayer perceptrons achieved 100 percent classification accuracy in detecting formaldehyde, whey, urea, and starch adulteration in raw bovine milk. Thermogravimetry combined with chemometrics detected starch in commercial yogurt above the legal 1 percent threshold with over 83 percent accuracy. DSC has resolved species-specific thermal fingerprints in beef, pork, rabbit, and chicken for solvent-free halal authentication, and TGA with derivative thermogravimetry quantified corn, husk, and straw adulteration in ground roasted Arabica coffee with a detection limit of just 0.36 percent by mass.

The techniques also serve quality control well beyond fraud detection. DSC-derived oxidation induction times and onset temperatures correlate strongly with conventional peroxide and anisidine values, providing rapid indicators of oil deterioration during storage. Glass transition measurements have been shown to predict chemical stability: in condensed biopolymer matrices, storing lipids below the glass transition slowed oxidation by restricting molecular diffusion, suggesting DSC can forecast shelf life without months of storage trials. DSC further characterizes cocoa butter polymorphism—critical to the gloss and snap of chocolate—starch gelatinization, and protein denaturation, with recent nano-DSC work resolving the thermal transitions of six individual whey proteins by manipulating pH and ion chelation conditions.

The authors are candid about the limitations. Thermal profiles are influenced by cultivar, geography, climate, processing history, oxidation, and experimental conditions such as heating rate, sample mass, and thermal history, all of which can complicate comparisons between laboratories. Overlapping thermal events in complex matrices can obscure individual transitions, and sensitivity to low-level adulteration is limited when adulterants behave thermally like the authentic product. The review also raises a pointed methodological question the field has largely ignored: what is the minimum detectable thermal shift below which DSC- and TGA-based claims become unreliable? Without systematic detection thresholds for each food matrix, cross-study comparisons remain largely qualitative. High classification accuracies reported after machine learning integration, the authors caution, should not be mistaken for universal detection capability, particularly when models rest on small datasets without independent external validation.

The future, however, looks increasingly intelligent. The review outlines a roadmap in which deep learning extracts discriminative features directly from raw thermal curves, explainable AI techniques such as SHAP reveal which temperature ranges drive classification decisions, and transfer learning compensates for small thermal databases. Federated learning could allow laboratories and regulators to train shared models without sensitive data ever leaving its site of origin, while digital-twin frameworks could couple thermal sensors to real-time, in-line quality monitoring under Industry 4.0 infrastructure. Portable miniaturized DSC devices may eventually bring authentication to the point of collection, and hyphenated systems combining thermal analysis with FTIR, Raman, or mass spectrometry promise multimodal fingerprints capturing variations no single technique can detect.

The review’s bottom line is measured but optimistic: DSC and TGA will not replace chromatography or spectroscopy as confirmatory reference methods, but positioned as rapid, solvent-free screening tools within a complementary analytical strategy, they offer a practical route to catching fraud before it reaches consumers. Realizing that vision will require standardized protocols, large and diverse reference databases built on FAIR data principles, interlaboratory validation, and eventual regulatory acceptance. If those pieces fall into place, the humble thermogram—once confined to polymer labs—may become one of the food industry’s most powerful anti-fraud weapons.

Subject of Research: Applications of differential scanning calorimetry and thermogravimetric analysis in food quality control and authenticity verification

Article Title: Recent applications of DSC and TGA in food quality control and authenticity verification: Current progress and future directions

Article References: Pourmoradian, A., Mohammadi, A., & Barzegar, M. (2026). Recent applications of DSC and TGA in food quality control and authenticity verification: Current progress and future directions. Food Chemistry: X, 39, Article 104587. https://doi.org/10.1016/j.fochx.2026.104587

Image Credits: AI Generated

DOI: 10.1016/j.fochx.2026.104587

Keywords: food fraud, food authentication, differential scanning calorimetry, thermogravimetric analysis, thermal analysis, adulteration detection, olive oil, honey, chemometrics, machine learning, oxidative stability, glass transition

Guizhou’s reputation for fine tea is rooted in its geography. The province sits at high altitude with a temperate climate, abundant rainfall, and fertile soils, conditions that favor the slow growth and rich chemical development of tea trees. Yet traditional quality assessment, which depends on sensory panels or isolated chemical indicators, has long been criticized as subjective, slow, and incomplete. Quality is shaped by a tangled web of factors, including cultivar, cultivation and processing techniques, geographical location, climate, and harvest time, making it difficult to capture with any one-dimensional test. The research team, publishing in Food Science & Nutrition, argued that a comprehensive, multiparameter evaluation using objective chemometric methods has become a pressing need in modern tea research.

The five teas examined were Qingyu tiny kuding tea, Guiding yunwu tribute tea, Suiyang mountain silver flower tea, Shiqian moss tea, and Pu’an black tea, all purchased from farmers’ markets in Guiyang and botanically authenticated. Each sample was dried at 85 degrees Celsius for twelve hours, ground, sieved, and tested in triplicate, with relative standard deviations held below two percent. The researchers then deployed an impressive analytical arsenal: a NETZSCH thermogravimetric analyzer to track how each tea decomposes under heat, an A300 amino acid analyzer to quantify twenty amino acids, and an ICP-OES spectrometer to measure seventeen trace elements, alongside determinations of combustion heat, fat, ash, and crude fiber content.

The thermogravimetric results were strikingly varied. Decomposition began at temperatures ranging from 39.7 degrees Celsius in Pu’an black tea to 60.3 degrees Celsius in Guiding yunwu tribute tea, and final residual masses ranged from 18.39 percent for the tribute tea to 44.35 percent for the black tea, indicating substantial differences in moisture binding, organic matrix composition, and inorganic residue. Every tea showed two major weight-loss stages, with the fastest decomposition occurring between roughly 316 and 337 degrees Celsius, and prominent exothermic peaks near 110 to 117 degrees Celsius with peak areas between 201.90 and 282.00 joules per gram. Using gray pattern recognition, the team also quantified combustion stability, finding that Suiyang mountain silver flower tea was the most thermally stable, followed by Guiding yunwu tribute tea, Shiqian moss tea, Qingyu tiny kuding tea, and finally Pu’an black tea.

The elemental analysis delivered both reassurance and surprises. Arsenic was detected in only one sample, Suiyang mountain silver flower tea, at a very low concentration of 0.0369 micrograms per gram, while mercury and scandium were absent from all five teas. Lead levels ranged from 0.89 to 2.51 micrograms per gram, with the lowest found in Pu’an black tea. On the nutritional side, manganese concentrations were remarkable, spanning from 86.2 micrograms per gram in Qingyu tiny kuding tea to nearly 700 micrograms per gram in Shiqian moss tea, while zinc peaked at 848.23 micrograms per gram in Suiyang mountain silver flower tea. Iron, magnesium, copper, and barium all varied systematically across the teas, creating elemental fingerprints that the authors attribute to differences in soil geochemistry, root uptake capacity, and agricultural practices across Guizhou’s diverse producing regions.

Amino acid profiling proved equally revealing. All five teas contained seven essential amino acids, namely threonine, valine, methionine, isoleucine, leucine, phenylalanine, and lysine, though tryptophan and methionine sulfoxide were not detected. The ratio of essential to total amino acids ranged from 9.78 to 26.20 percent, and theanine, the amino acid most responsible for tea’s savory sweetness, showed enormous variation, from just 114.45 micrograms per gram in Qingyu tiny kuding tea to a striking 18,603.86 micrograms per gram in Guiding yunwu tribute tea. Serine reached 13,527 micrograms per gram in Shiqian moss tea, while arginine dominated Qingyu tiny kuding tea at 7,492.73 micrograms per gram. The authors suggest these differences arise from complex interactions between cultivar-specific nitrogen metabolism and environmental factors such as altitude, temperature, rainfall, and light intensity.

To turn this mountain of data into a coherent ranking, the researchers applied entropy factor analysis, a technique that fuses information theory with factor analysis to distill dozens of correlated variables into a handful of independent factors. Thirty-eight variables were compressed into four entropy factors that together explained 100 percent of the variance. The first factor, weighted most heavily, captured minerals and amino acids including magnesium, manganese, serine, and phenylalanine; the second reflected combustion heat, ash, fat, and theanine; the third covered crude fiber, aluminum, and aspartic acid; and the fourth grouped combustion stability with barium, iron, and sodium. Combining the four factors with weights proportional to their eigenvalues produced a composite score for each tea, and the final ranking placed Qingyu tiny kuding tea first with a score of 0.9953, followed by Shiqian moss tea, Suiyang mountain silver flower tea, Pu’an black tea, and Guiding yunwu tribute tea at negative 0.5711.

The team then subjected the data to two independent validation approaches. Entropy factor cluster analysis grouped the five teas into three distinct clusters: Suiyang mountain silver flower tea, Shiqian moss tea, and Pu’an black tea clustered together, possibly reflecting similar fermentation processes and regional cultivation conditions, while Guiding yunwu tribute tea and Qingyu tiny kuding tea each formed their own groups, consistent with their distinctive oxidative profiles and traditional processing. Meanwhile, a supervised orthogonal projections to latent structures discriminant analysis model achieved exceptional discrimination, with R-squared values of 0.887 for the predictors and 0.998 for the responses, and a predictive Q-squared of 0.997. Permutation testing and leave-one-out cross-validation confirmed the model was not overfitting, and twenty-one variables, including proline, arginine, alanine, copper, manganese, and theanine, emerged as key differentiators with importance scores above one.

The implications extend well beyond academic curiosity. A validated, multiparameter fingerprinting system could transform tea authentication, helping regulators detect mislabeled or adulterated products, assisting producers in quality control and market positioning, and guiding consumers toward teas whose nutritional and safety profiles have been rigorously verified. Because trace element patterns reflect the mineral signature of local soils, the same framework could support geographical origin tracing, a growing concern in premium food markets worldwide. The authors note that the approach also provides a mechanistic understanding of why teas differ, linking observable quality grades to cultivar traits, soil chemistry, and processing choices.

The researchers are candid about the limitations of their work. Only five teas from a single province were examined, and the framework’s robustness will need confirmation across broader geographic regions, additional cultivars, and more diverse processing types. Future studies, they say, will expand the sample set and add parameters such as catechins, caffeine, polyphenols, aroma compounds, and sensory evaluation. Still, the study marks a meaningful step toward replacing the taster’s subjective verdict with a reproducible, multidimensional measurement, one that treats a cup of tea not as a matter of opinion but as a chemical fingerprint waiting to be read.

Subject of Research: Multidimensional chemometric quality assessment of five Guizhou teas using thermogravimetric, amino acid, and trace element analyses

Article Title: Multidimensional Quality Assessment of Guizhou Teas Integrating Thermogravimetric, Amino Acid, and Trace Element Analyses

Article References: Zhou, L., & Huang, C. (2026). Multidimensional Quality Assessment of Guizhou Teas Integrating Thermogravimetric, Amino Acid, and Trace Element Analyses. Food Science & Nutrition, 14(10), Article e72431. https://doi.org/10.1002/fsn3.72431

Image Credits: AI Generated

DOI: 10.1002/fsn3.72431

Keywords: tea quality, Guizhou, thermogravimetric analysis, amino acids, trace elements, chemometrics, entropy factor analysis, OPLS-DA, theanine, food authentication, heavy metals, combustion stability

The technology, described in Environmental Science and Pollution Research, is called the Advanced Thermal Processor, or ATP, and its defining feature is that it operates continuously rather than in batches. Contaminated cuttings are fed through a sealed hopper and screw conveyor into a slightly inclined rotating drum, where they are heated simultaneously by two different mechanisms. Microwave energy penetrates the material volumetrically, exciting polar molecules such as water throughout the bulk of the waste rather than merely at its surface. Infrared heaters arranged along the drum provide uniform surface heating, filling in the thermal gaps that microwaves alone can leave. The researchers argue that this hybrid approach solves the two problems that have dogged microwave treatment in earlier studies: uneven heating and poor scalability to industrial throughput.

To test the system, the team collected six samples of water-based drill cuttings contaminated with sour crude oil from real Middle Eastern offshore drilling operations, then processed them in two batches through a pilot-scale prototype under inert, low-oxygen conditions maintained by nitrogen purging. The treatment was deliberately gentle: solids left the drum at discharge temperatures of just 60 to 100 degrees Celsius, after residence times of roughly 15 to 30 minutes. Despite these mild conditions, the results were striking. Oil removal efficiencies ranged from 95.0 to 97.8 percent across all six samples, and in every case the residual oil-on-cuttings fell below the 1 percent regulatory threshold. Total Petroleum Hydrocarbons and Total Oil and Grease, the two standard measures of hydrocarbon contamination, were each reduced by more than 96 percent.

The low operating temperature is the headline finding. Conventional thermal desorption and incineration systems typically run at 300 to 500 degrees Celsius or higher, demanding large energy inputs and risking secondary emissions. Thermogravimetric analysis of the treated solids showed why the hybrid process can be so much cooler. In one sample, a major mass loss of about 21.17 percent occurred at around 106 degrees Celsius, corresponding to the release of water and light volatile hydrocarbons. Only minor additional losses appeared at higher temperatures between 275 and 413 degrees, indicating that what remained was mostly thermally stable inorganic rock. In other words, most of the contamination that regulators care about volatilises below 110 degrees when heat is delivered volumetrically, without overheating or degrading the solid matrix.

The physics behind this selectivity is well understood. Water is a highly polar molecule and absorbs microwave radiation far more efficiently than hydrocarbons, so microwave energy preferentially drives off moisture, opening a porous structure in the cuttings that enhances heat and mass transfer. Previous studies have shown that microwave-treated residues are more porous than those dried by conventional electric heating, which relies on thermal conductivity and tends to leave higher residual oil. Microwaves can also promote pyrolysis of petroleum hydrocarbons, making them particularly suited to oil-based contamination. Earlier work demonstrated that microwave radiation could reduce n-paraffin content in cuttings contaminated with non-aqueous fluids below discharge limits, and that the recovered organic phase retains its chemical composition well enough to be reused in fresh drilling fluid.

The Brunel study also mapped the limits of the process. A clear inverse trend emerged between initial water content and oil removal efficiency: the sample with the highest moisture, 24 percent by weight, achieved the lowest removal at roughly 95 percent, while drier samples approached 97.8 percent. The explanation is energetic competition. Because water soaks up microwave energy so readily, wetter feed demands more of the available power for evaporation, leaving less to volatilise oil. Water may also form a transient barrier around oil droplets, impeding heat transfer and delaying volatilisation. A similar, weaker trend appeared with total fluid content: samples above about 23 percent combined oil and water showed marginally lower efficiency, possibly because rapid heating creates insulating vapour layers around droplets, a phenomenon analogous to the Leidenfrost effect. The practical lesson is that feed pre-conditioning, particularly dewatering with a decanting centrifuge, is essential to squeeze maximum performance from the system.

The prototype was designed as a complete process train rather than a laboratory curiosity. After centrifuge dewatering and homogenisation, cuttings pass through an airlock hopper into the rotating drum, where nitrogen purging and a slight positive pressure keep oxygen out, a critical safety measure in the ATEX Zone 2 environment of a drilling platform. Volatilised oil and water vapour are continuously extracted, scrubbed of entrained particles, and condensed; an oil-water separator then recovers oil suitable for reuse and produces water with residual oil content between 0 and 15 parts per million. Non-condensable gases are recirculated as carrier gas, cutting emissions further. The treated solids emerge dry and powder-like through a sealed screw conveyor and rotary airlock, with a water mist suppressing dust. Oxygen levels are monitored continuously, with automatic shutdown protocols and emergency nitrogen purging standing by.

The regulatory context gives the work its urgency. Since 2001, OSPAR has banned offshore discharge of cuttings contaminated with oil-based or synthetic-based fluids unless residual oil is below 1 percent by dry weight, and in 2012 it added stringent sediment monitoring and hydrocarbon reporting requirements to support benthic ecosystem recovery. Other frameworks, including the United States EPA Effluent Guidelines, the Barcelona Convention and the London Protocol, regulate discharge but none are as strict. The stakes extend beyond ecology: returned drilling fluids can carry heavy metals such as nickel, mercury, cadmium, chromium, copper and lead, elements associated with chronic diseases including cancer. Life cycle assessments of conventional management options, from land spraying to solidification and reinjection, show that solidification carries the highest environmental burden because of cement and lime consumption, while reinjection, though comparatively benign, simply relocates contaminants to deeper formations.

What distinguishes the ATP from earlier microwave remediation efforts is continuity. Pilot-scale microwave studies dating back more than a decade proved the chemistry but stalled at the batch scale, unable to match the relentless output of a drilling operation, particularly during top-hole sections where cuttings production peaks. By integrating microwave and infrared heating into a continuously fed rotating drum with automated vapour recovery, the Brunel team has produced a system whose throughput can in principle be scaled with drum size and feed rate rather than being capped by batch cycle times. The ability to remove oil, water and hydrocarbons in a single processing step, rather than the multiple stages conventional methods often require, further improves the operational economics of on-site treatment.

The authors are careful about what remains to be done. The experimental programme analysed here was conducted some years before publication, and the available records did not preserve the individual analytical protocol numbers behind the oil, TPH and TOG determinations, a transparency note that underscores the need for full-scale validation. The team itself flags that future work should focus on full-scale demonstration and detailed energy and economic assessments before the technology can be considered proven at industrial scale. Even so, the combination of results, above 95 percent oil removal on real field samples, compliance with the strictest discharge thresholds, and treatment temperatures below 110 degrees, makes a credible case that hybrid microwave-infrared desorption could turn one of offshore drilling’s most stubborn waste problems into a manageable, and potentially resource-recovering, side stream of production.

Subject of Research: Hybrid microwave-infrared thermal desorption treatment of oil-contaminated offshore drill cuttings

Article Title: Hybrid microwave–infrared thermal desorption for remediation of oil-contaminated offshore drill cuttings: field-sample evaluation and process performance

Article References: Babaei-Mahani, R., Fereidounpour, A., He, Y., & Scholes, P. (2026). Hybrid microwave–infrared thermal desorption for remediation of oil-contaminated offshore drill cuttings: field-sample evaluation and process performance. Environmental Science and Pollution Research. https://doi.org/10.1007/s11356-026-38229-2

Image Credits: AI Generated

DOI: 10.1007/s11356-026-38229-2

Keywords: drill cuttings, microwave heating, infrared heating, thermal desorption, offshore drilling, oil contamination, waste treatment, OSPAR Convention, hydrocarbon removal, thermogravimetric analysis, environmental remediation, Brunel University London

The research, led by Lukas Fliri of Aalto University and BOKU University together with colleagues at Aalto and VTT Technical Research Centre of Finland, focused on fibers spun with the Ioncell technology, a process that dissolves pre-hydrolyzed kraft birch pulp in a superbase-derived ionic liquid and dry-jet wet spins it into filaments. Unlike the native cellulose I allomorph found in wood and cotton, these regenerated fibers crystallize in the cellulose II form, the same allomorph that matters for carbon fiber production. Yet, as the authors point out, nearly all of the fundamental mechanistic data on cellulose pyrolysis has been gathered on cellulose I, while studies on cellulose II have mostly reported end-material properties without the underlying physics and chemistry. That gap made it difficult for the team to compare their own carbonization experiments with the literature, so they decided to build the reference dataset themselves.

The experimental design deliberately echoes a classic: the influential 1960s work by Tang and Bacon on the carbonization of Fortisan cellulose II fibers. Fliri and colleagues heated Ioncell fibers from room temperature up to 1100 degrees Celsius under helium in a simultaneous thermal analyzer coupled to a mass spectrometer, using heating rates spanning 2.5 to 40 kelvin per minute. Tiny samples of around 3.5 milligrams minimized heat and mass transfer distortions, and the evolved gases, water, carbon monoxide and carbon dioxide, were quantified through a calibration procedure based on calcium oxalate monohydrate. In parallel, batch carbonizations were carried out in a tubular furnace under nitrogen at nine target temperatures between 300 and 1100 degrees Celsius, with no thermostabilization pretreatment and no tension applied to the fibers, so that the intrinsic behavior of the material could be observed without process tricks that boost yield.

The thermal analysis revealed the familiar fingerprint of cellulose pyrolysis: a violent mass loss concentrated in a window of only 70 to 100 degrees, with onset temperatures between 288 and 322 degrees Celsius depending on heating rate. At the peak of this event, cellulose chains depolymerize and expel levoglucosan, the sugar anhydride that dominates the volatile stream, alongside furans, dehydrated sugars, water and carbon oxides. Crucially, the slower the heating, the more char survived. Char yield at 1100 degrees Celsius fell log-linearly as the heating rate rose, and the combined yields of water and carbon dioxide tracked the char yield almost perfectly. This is exactly what the two-competing-pathway model of cellulose pyrolysis predicts: long residence times below 300 degrees favor dehydration reactions that lock carbon into char precursors, while rapid heating funnels the material toward volatilization as levoglucosan instead.

The kinetic analysis added quantitative teeth to this picture. Using Friedman’s isoconversional method, the team found that the apparent activation energy of Ioncell fiber pyrolysis sits in a narrow band of roughly 190 to 220 kilojoules per mole up to about 80 percent conversion, consistent with a single dominant step: the rupture of the beta-O-4 glycosidic bond that links the glucose units of cellulose. Beyond that conversion level, the activation energy climbs sharply, signaling a shift from depolymerization to reactions within the solid carbonaceous residue. A first-order kinetic model, fitted with activation energies of 227 to 231 kilojoules per mole and pre-exponential factors matching literature values for cellulose I, described the mass loss curves well, but the authors caution that this is a phenomenological description of the main mass loss step, not proof of a simple one-step mechanism, and the parameters should not be extrapolated to isothermal treatments or temperatures above about 600 degrees.

One of the more surprising results concerns heat. The apparent heat of pyrolysis, measured by integrating the differential scanning calorimetry signal, came out between roughly 180 and 250 kilojoules per kilogram across all heating rates, in close agreement with decades of measurements on microcrystalline cellulose and filter paper. When the team ran control experiments on pure levoglucosan, the heat of its thermal evaporation matched the apparent heat of cellulose pyrolysis almost exactly. The implication is striking: the endothermic demand of cellulose pyrolysis is largely just the energy needed to evaporate levoglucosan. From a process engineering standpoint, this reaction heat is small compared with the roughly 650 kilojoules per kilogram of sensible heat required simply to bring the fibers up to their degradation temperature, a fact that matters for anyone designing industrial carbonization furnaces.

The batch carbonization experiments then traced how the fibers’ properties evolve with the final heat treatment temperature. Yields dropped steeply between 300 and 500 degrees and then plateaued between 6.8 and 9.4 percent by weight. Elemental analysis showed hydrogen and oxygen draining away as carbon accumulated, and a van Krevelen plot revealed two distinct linear regimes: below about 400 degrees the hydrogen-to-carbon and oxygen-to-carbon ratios follow a slope of roughly two hydrogens per oxygen, the signature of dehydration, while above that temperature decarboxylation and decarbonylation take over. Notably, oxygen loss largely stops above 600 degrees, and X-ray photoelectron spectroscopy confirmed that thermally stable oxygen functionalities remain embedded in the carbon even at 1000 degrees. These trapped oxygen atoms, the authors argue, act like reticulation agents, promoting five- and seven-membered rings that curve and distort the growing aromatic network and prevent the formation of a graphite-like structure, a key reason cellulose-derived carbons are non-graphitizing.

Morphology told a parallel story. The round cross-section of the Ioncell filaments survived carbonization intact, but the diameter shrank by roughly 60 percent by the time the fibers reached 600 degrees, after which further shrinkage was minimal. Fiber diameter correlated linearly with carbon content and log-linearly with carbonization yield, and the team interprets the shrinkage as the surface degradation of oriented cellulose microfibrils followed by their coalescence into denser cross-linked furanic and polyaromatic structures. Raman spectroscopy completed the picture: a polyaromatic network was already detectable at 400 degrees, and the ratio of the disorder-induced D band to the graphitic G band jumped from 0.59 to 0.93 between 600 and 700 degrees, then climbed more slowly toward 1100 degrees. The D-to-G intensity ratio correlated linearly with the hydrogen-to-carbon ratio above 600 degrees, indicating that large aromatic domains grow through dehydrogenation and radical condensation of smaller ring systems. Estimated in-plane aromatic cluster sizes grew from about 1.0 to 1.4 nanometers between 600 and 800 degrees and then stalled.

Perhaps the most consequential conclusion is what the study did not find. Older carbon fiber literature, still frequently cited, holds that the cellulose II crystal structure follows a special carbonization mechanism tied to the recondensation of four-carbon intermediates. The new data show no evidence of such a peculiarity: the pyrolysis behavior of Ioncell fibers matched the classic two-pathway models, and the high-temperature evolution agreed with general models for non-graphitizable carbons. The authors also observed an unexplained systematic mass loss above 800 degrees, but only at the slowest heating rate of 2.5 kelvin per minute, a phenomenon not previously reported that they flag for further confirmation. By identifying the temperature windows where dehydration, volatilization, dehydrogenation and aromatization each dominate, the dataset gives carbon fiber researchers a practical map for optimizing thermostabilization and carbonization schedules, and a common reference point for comparing cellulose II precursors of every kind. For a field trying to make carbon fiber cheap, renewable and strong enough for the mass market, that map may prove as valuable as the fibers themselves.

Subject of Research: Pyrolysis dynamics and physicochemical property development during the carbonization of Ioncell regenerated cellulose II fibers for carbon fiber production

Article Title: Tracking carbonization in Ioncell cellulose II fibers: pyrolysis dynamics and development of the physicochemical properties

Article References: Fliri, L., Yildirim, Z., Meinander, K., Guizani, C., & Hummel, M. (2026). Tracking carbonization in Ioncell cellulose II fibers: pyrolysis dynamics and development of the physicochemical properties. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13716-6

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13716-6

Keywords: carbon fibers, cellulose II, Ioncell, pyrolysis, carbonization, levoglucosan, char yield, Raman spectroscopy, thermogravimetric analysis, biobased materials, turbostratic carbon, kinetic modeling

The research team, led by Désiré Ndahirwa of UniLaSalle with colleagues including Hélène Lenormand, Hafida Zmamou and Nathalie Leblanc, focused on a question that has dogged the field for years. Wood ash contains reactive silica, alumina and calcium-bearing phases that can, in principle, behave like cement itself, reacting with water through hydraulic and pozzolanic pathways. In practice, however, pastes made from pure wood ash are weak and porous. Previous studies had shown that replacing moderate amounts of cement with wood ash often reduces compressive and flexural strength, and most work had examined substitution levels below 50 percent. Almost nothing was known about what happens at very high replacement levels, where wood ash dominates the mixture, and the stiffness of such materials, measured as Young’s modulus, had rarely been quantified despite its importance for structural design.

To close that gap, the researchers gathered four locally sourced wood ashes from the Normandy region of France. Two of them, designated WFA3 and WBA, came from the combustion of wood pellets and were delivered as wet sludge, requiring oven drying and crushing before use. The other two, WFA8 and WFA9, arrived as dry fine powders from a local heating plant operated by Coriance in Mont-Saint-Aignan, where boilers with power inputs of 6 and 8 megawatts burn forestry wood chips, bocage wood chips and pallet residues at temperatures between 900 and 1100 degrees Celsius. The team prepared seventeen paste formulations in total: four containing only wood ash, twelve blending 80 to 95 percent wood ash with 5 to 20 percent ordinary Portland cement, and a reference paste of pure cement, all compacted with a mini-Proctor device to boost density and cured for up to 28 days.

The chemical analysis alone revealed why wood ash is such a tricky raw material. The four ashes were dominated by calcium oxide, silica, potassium oxide and sulfur trioxide, but in wildly varying proportions. Sulfate contents reached 15.4 percent in WFA8 and 14 percent in WFA9, far above the 4 to 5 percent limit set by ASTM standards for pozzolans, and their combined pozzolanic oxides fell well below the required thresholds. In plain terms, these ashes do not qualify as conventional pozzolans, yet X-ray diffraction showed they carry crystalline phases such as portlandite, calcite, albite, alite, dolomite and sylvite that can still participate in binding reactions. The variability is a direct consequence of differing feedstocks, boiler designs and combustion temperatures, and it means each ash must be evaluated on its own terms rather than lumped into a single category.

When it came to mechanical performance, the effect of the small cement additions was unmistakable. Pure wood ash pastes managed compressive strengths of only 0.13 to 1.65 megapascals at 28 days, but raising the cement content to 20 percent lifted those values substantially. The best performer was the WFA8-based blend, which reached 5.49 megapascals, while WFA9 and WFA3 pastes achieved 4.01 and 3.87 megapascals respectively at the same dosage. The researchers attribute the gains to a richer supply of hydration products, including calcium silicate hydrate gel, portlandite and ettringite, formed as the cement’s tricalcium silicate reacts with water and progressively densifies the paste matrix. Stiffness told the same story: the modulus of elasticity, estimated from the linear portion of stress-strain curves, climbed with cement content, curing time and bulk density in three of the four ash families, with the highest values consistently recorded in mixtures containing 20 percent cement.

One ash refused to follow the script. The wood bottom ash, WBA, behaved atypically across every measurement. Its pastes lost compressive strength between 7 and 28 days at certain dosages, its modulus of elasticity peaked at 7 days and then declined, and thermogravimetric analysis found no detectable portlandite whatsoever. X-ray diffraction offered an explanation: the WBA pastes contained no alite, the calcium silicate phase that drives strength development in hydrating cement, and their dominant crystalline phases were calcite and quartz. Adding 20 percent cement pushed the estimated calcite content up from about 47 to 63.5 percent while quartz fell, a signature of carbonation of calcium silicate hydrate or the formation of complex, less efficient hydrate phases, both of which are associated with increased porosity and weaker binding.

Scanning electron microscopy added a visual dimension to the story. Under the microscope, pure cement paste appeared dense and well packed, bristling with the products of hydration, while pastes made from 100 percent wood ash showed loosely arranged particles riddled with interparticle voids, along with unreacted ash grains and dark fragments of unburnt wood. With 20 percent cement added, the microstructure tightened, porosity dropped and hydration products proliferated, with needle-like ettringite crystals, gel-like calcium silicate hydrates, plate-shaped portlandite and rhombohedral calcite all visible. The ashes from the heating plant, WFA8 and WFA9, consistently produced more compact matrices than the pellet-derived WFA3 and the bottom ash WBA, underlining how much origin and processing shape a material’s destiny.

The mineralogical detective work also turned up some genuinely unexpected chemistry. In the WFA8 blends, introducing cement promoted the formation of alunite, a potassium aluminum sulfate hydroxide phase not present in the unblended paste, alongside an array of compounds including syngenite, arcanite, serandite and harmotome. Thermogravimetric analysis at 7 and 28 days complemented the diffraction data, identifying calcium silicate hydrates and ettringite dehydrating between 50 and 200 degrees Celsius, AFm phases such as calcium monocarboaluminate and hemicarboaluminate decomposing between 200 and 300 degrees, portlandite dehydroxylating between 400 and 500 degrees, and carbonates releasing carbon dioxide from 500 to 800 degrees. The two techniques agreed closely, with the single discrepancy being ettringite in the bottom ash pastes, which thermal analysis detected but diffraction did not, likely because its concentration fell below the instrument’s detection limit.

What emerges from the study is a nuanced but practical message. Low doses of ordinary Portland cement, between 5 and 20 percent, can meaningfully upgrade pastes in which wood ash makes up as much as 95 percent of the solid content, provided the ash is fine and reasonably reactive. The improvements in stiffness, strength and microstructure are real, even if the resulting materials remain suited to low-strength applications such as lightweight binders rather than load-bearing concrete. The decisive variable, the authors conclude, is the variability of the ash itself: its origin, chemistry and mineralogy govern everything from phase development to porosity. That insight carries weight well beyond Normandy. With millions of tons of biomass ash generated annually and cement production responsible for a major share of global carbon dioxide emissions, even modest cement dosages that transform a landfill-bound waste into a functional building material represent a meaningful step toward circular, lower-carbon construction. The next challenge will be standardizing how ashes are characterized and selected, so that builders can trust what is in the bag before it ever reaches the site.

Subject of Research: The effect of low ordinary Portland cement content on the stiffness, microstructure and mineralogical composition of wood ash-based pastes

Article Title: Effect of low ordinary Portland cement content on stiffness, microstructure and mineralogical composition of wood ash pastes

Article References: Ndahirwa, D., Lenormand, H., Zmamou, H., Chenot, E., Potel, S., & Leblanc, N. (2026). Effect of low ordinary Portland cement content on stiffness, microstructure and mineralogical composition of wood ash pastes. Cleaner Engineering and Technology, 34, Article 101303. https://doi.org/10.1016/j.clet.2026.101303

Image Credits: AI Generated

DOI: 10.1016/j.clet.2026.101303

Keywords: wood ash, ordinary Portland cement, Young's modulus, compressive strength, X-ray diffraction, thermogravimetric analysis, scanning electron microscopy, calcium silicate hydrate, ettringite, porosity, biomass ash, sustainable construction