Plasma-Treated Cellulose Fibers Make Cement Composites Tougher, Study Finds

Cellulose fibers have long been touted as a green alternative to the synthetic reinforcements used in construction materials, but they carry a stubborn weakness: they love water. When embedded in cement, these plant-derived fibers soak up moisture, swell, and gradually become mineralized by the alkaline pore solution that permeates the matrix. Over time, the fibers embrittle, lose their flexibility, and the fiber–matrix interface degrades, eroding the very toughness that makes cellulose attractive in the first place. A new study published in Discover Industrial Chemistry and Materials offers a strikingly elegant fix, and its central finding is counterintuitive enough to reshape how researchers think about waterproofing natural fibers: the most water-repellent treatment is not the one that produces the strongest composite.

The research, led by Laís Salviano de Oliveira Schiavi of the Universidade Federal de Lavras in Brazil, together with colleagues at UNESP, UFSCar, and UFLA, used plasma-enhanced chemical vapor deposition, or PECVD, to coat bleached eucalyptus cellulose fibers with thin organosilicon films derived from hexamethyldisiloxane, commonly abbreviated HMDSO. The technique is inherently solvent-free, a significant environmental advantage over conventional wet-chemical treatments such as silanization or polymeric impregnation, which often require long processing times and reagents that complicate industrial adoption. In the plasma reactor, HMDSO vapor, oxygen, and argon were introduced into a low-pressure chamber where a 13.56 MHz radiofrequency field ignited a non-thermal discharge, fragmenting the precursor molecules and depositing a nanometer-scale coating onto roughly 180 grams of dispersed cellulose pulp per run.

What makes HMDSO particularly versatile as a precursor is that the chemistry of the resulting film can swing dramatically depending on the gas mixture. With abundant oxygen, the plasma fragments oxidize into silica-like structures rich in polar Si–O bonds. With less oxygen and more precursor, methyl-containing organosilicon groups survive, producing hydrophobic, wax-like surfaces. The Brazilian team exploited this tunability by running four treatment conditions with nominal HMDSO concentrations ranging from 2.3 to 70.0 percent, labeled ORG-T02, ORG-T30, ORG-T50, and ORG-T70, while holding deposition time at 600 seconds and applied power at 100 watts so that only the precursor concentration varied. Reference films deposited on glass coverslips and polished aluminum plates confirmed the chemical progression: the lowest concentration yielded an oxidized, silica-like coating, while the highest produced a clearly hydrophobic surface with a water contact angle of approximately 109 degrees.

Crucially, the plasma treatment left the fibers themselves intact. Scanning electron microscopy showed that the characteristic fibrous morphology survived all four treatments without visible degradation, and X-ray diffraction confirmed that the crystalline structure of cellulose was preserved, with crystallinity indices hovering between roughly 78 and 80 percent across treated samples compared with 81.5 percent for untreated pulp. Energy-dispersive X-ray spectroscopy detected silicon on every treated fiber surface and none on the untreated reference, confirming that HMDSO-derived species had been deposited. In other words, the modification was strictly a surface phenomenon, exactly what a reinforcing fiber needs if its bulk mechanical contribution is to be retained during composite processing and service.

The moisture results, however, revealed that wettability and water uptake do not map onto each other in any simple linear way. After 192 hours in a nearly saturated atmosphere, the most hydrophobic treatment, ORG-T70, reduced moisture absorption by only 9.26 percent relative to untreated cellulose. The intermediate condition, ORG-T30, performed best, cutting moisture uptake by 12.20 percent, while ORG-T50 and ORG-T02 actually increased absorption by 36.91 and 43.91 percent respectively. The authors attribute this apparent paradox to the heterogeneous nature of plasma deposition on fibrous substrates: contact angle measurements probe only the local region touched by a droplet, whereas moisture absorption integrates the response of the entire fiber assembly, including coating continuity, coverage, and packing effects that a single droplet cannot capture.

To probe the molecular underpinnings of the fiber–coating interaction, the team turned to computational chemistry. Using the Quantum Theory of Atoms in Molecules, or QTAIM, framework with B3LYP calculations, they modeled a cellulose fragment approaching an HMDSO molecule at varying distances. The simulations showed no chemical reaction between the two species; instead, the most favorable configuration occurred at an intermolecular separation of about 1.67 angstroms and was dominated by hydrogen bonding between the cellulose hydroxyl group and the oxygen atom of HMDSO. The atomic properties of the interacting hydrogen satisfied the Koch and Popelier criteria for hydrogen bond formation. The authors are careful to note that this simplified model does not reproduce the full plasma environment, but it lends theoretical plausibility to an adsorption mechanism that would explain how the coating anchors to cellulose without disrupting its structure.

With modified fibers in hand, the researchers extruded fiber-cement composites containing 60 percent high-early-strength Portland cement, 33 percent agricultural limestone, 5 percent cellulose fibers, and small amounts of a methylcellulose thickener and a carboxylic polyether rheology additive. Extrusion, which compacts the material and aligns fibers along the flow direction, was chosen for its ability to produce dense specimens at low water-to-cement ratios. Before any mechanical testing, the team verified that the treated fibers would not sabotage cement curing: using the Hofstrand inhibition index, every formulation scored far below the 10 percent threshold that signals problematic hydration inhibition, with values between roughly 0.18 and 0.40 percent. The plasma coatings, in short, did not interfere with the chemistry that turns cement paste into solid matrix.

The decisive test came after accelerated aging. Specimens cured for 28 days were subjected to repeated cycles of 8 hours of UVA exposure at 365 nanometers followed by 4 hours of moisture conditioning, a protocol adapted from ASTM G154-23, and then retested after 10 and 20 cycles. The results overturned the hydrophobicity hierarchy. After 20 cycles, composites reinforced with fibers treated under the ORG-T02, ORG-T30, and ORG-T50 conditions showed modulus of rupture values 22.54, 38.01, and 6.62 percent higher, respectively, than the composite made with untreated fibers, while the most hydrophobic formulation, F70-ORG, fell 9.19 percent below the untreated reference. Modulus of elasticity followed the same pattern, with F30-ORG 37.72 percent above the untreated composite and F70-ORG 9.75 percent below, and toughness told a similar story, with F02-ORG and F30-ORG exceeding the aged untreated reference by 57.14 and 42.85 percent.

The fate of the highly hydrophobic F70-ORG formulation is particularly instructive. At 28 days, before any aging, it posted the best moisture-related numbers of the batch, with apparent porosity reduced by 20.67 percent and water absorption by 24.43 percent relative to the untreated fiber composite. Yet those gains evaporated under cyclic conditioning, and by 20 cycles the formulation was mechanically underperforming the untreated control. Maximum hydrophobicity, the study concludes, does not correspond to the best composite response. The authors caution that repeated moisture exposure allows continued hydration of residual cement phases, which may densify the matrix and complicate the interpretation, and that no direct interfacial adhesion measurements were performed, so the differences after aging cannot be attributed exclusively to fiber modification. All composites nonetheless remained within the water absorption limits of the Brazilian standard NBR 7581-2 for corrugated fiber-cement products.

The broader significance of the work lies in its reframing of a design assumption. Rather than chasing ever-lower water contact angles, the results suggest that intermediate surface conditions, exemplified by ORG-T30, strike a more favorable balance between moisture protection and whatever interfacial interaction the fibers retain with the cementitious matrix. Because PECVD is solvent-free, fast, and tunable through gas composition alone, it offers the construction industry a scalable route to upgrading a renewable reinforcement that has been hampered by durability concerns for decades. The authors emphasize that their 20-cycle protocol is a comparative tool rather than a predictor of field service life, and that longer exposures plus direct interface characterization will be needed to confirm how persistent the benefits are. Even so, the message is clear and likely to resonate across materials science: when it comes to protecting natural fibers in cement, moderation beats extremes.

Subject of Research: HMDSO plasma modification of cellulose fibers for improving the water resistance and mechanical durability of extruded fiber-cement composites

Article Title: HMDSO plasma treatment of cellulose fibers influences water resistance and mechanical performance of extruded fiber cement composites

Article References: HMDSO plasma treatment of cellulose fibers influences water resistance and mechanical performance of extruded fiber cement composites. (n.d.). https://doi.org/10.1007/s44508-026-00025-8

Image Credits: AI Generated

DOI: 10.1007/s44508-026-00025-8

Keywords: cellulose fibers, HMDSO, plasma treatment, PECVD, fiber-cement composites, hydrophobicity, moisture absorption, accelerated aging, modulus of rupture, cement hydration, organosilicon coatings, construction materials