Introduction Conventional concrete is a porous heterogeneous quasi-brittle material, which has some problems such as poor toughness, low flexural tensile strength and poor durability, and cannot meet the needs of modern construction for high strength, high crack resistance and high durability of concrete materials. The existing methods used to toughen concrete are fiber reinforcement and matrix toughening. Among them, fiber is an extrinsic toughening. The toughness of cement matrix can be improved with polymers and nanomaterials. In addition, compared with mixing polymers directly, the modified method of in-situ polymerization of organic monomers can improve the flexural strength of cement pastes. However, the effect and mechanism of this method on toughening concrete are still unclear. Therefore, the effect of in-situ polymerization of acrylamide (AM) monomer and wollastonite (CS) whisker on the mechanical properties, hydration heat release, phase composition and microstructure of concrete were investigated to provide an idea for toughening concrete matrix.Methods Raw materials were P·Ⅱ 52.5 Portland cement, quartz sand with the fineness modulus of 2.6-2.9 and basalt with the particle sizes of 5-16 mm. The composite modified materials used were all adscititious, in which an organic monomer was acrylamide (AM), an initiator was ammonium persulphate (APS), and a crosslinking agent was N,N'-methylenebisacrylamide (MBA). The inorganic modified material was wollastonite (CS) whisker, and its micromorphology was dense and short rod fibrous. The water-cement ratios (W/C) of concrete and cement paste were 0.47 and 0.40, respectively. The dosages of CS were 1% (in mass fraction), 2% and 4% by the mass of cement. The dosages of AM were 3% and 5% by the mass of cement, the dosages of APS were 3% and 5% by the mass of AM, and the dosage of MBA was 0.1% by the mass of AM.The compressive strength, flexural stress-strain curve and fracture energy of concrete specimens (100 mm×100 mm×100 mm and 100 mm×100 mm×400 mm) were tested after curing under standard curing condition (20 ℃, 95% RH) for 3, 7 d and 28 d. The hydration exothermic curve of the cement paste in the early 72 h was recorded at 20 ℃ by a model TAM-Air micro-calorimeter (TA Instruments Co., USA). After the cement paste was cured for 3, 7 d and 28 d, the phase composition was qualitatively and quantitatively analyzed by a model D8 ADVANCE X-ray diffractometer ( Bruker Co., Germany). The hydration degree of cement was determined by a model TG-209F3 thermal gravimetric analyzer (Netzsch Co., Germany), and the functional groups were characterized by a model Nicolet IS10 Fouri infrared spectrometer (Thermo Scientific Co., USA). The microstructure was observed by a model Quanta 3D FEG scanning electron microscope (Fei Co., USA).Results and discussion The effect of single-doped CS on the flexural strength of concrete is not significant, but it can improve the compressive strength to a certain extent at 3 d and 7 d in the early hydration stage. Adding AM has a negative effect on the bending resistance of concrete at each stage. The addition of APS and MBA can make AM crosslinked and polymerized in cement hydrated matrix, which can significantly improve the peak flexural stress and fracture energy. However, the film forming characteristics of AM polymerization lead to a poor compatibility with cement matrix, directly affecting the hydration process of cement and reducing the compressive strength. Under the synergistic effect of CS and AM polymerization induced by APS, the maximum flexural stress and fracture energy of concrete at 28 d are increased by 31.5% and 91.4%, respectively, compared with control group. Note that the compressive strength of concrete at 28 d remains unchanged.Hydration heat results show that the main exothermic processes of AM monomer polymerization and cement hydration are not synchronized. AM polymerization can immediately release a lot of heat and delay the hydration reaction of cement, while the addition of CS has little effect on the hydration process of cement. Neither CS nor AM polymerization can change the types of cement hydration products, but CS can participate in cement hydration reaction and increase calcium hydroxide (CH) content. Under the action of initiator, AM monomer polymerization can promote ettringite (AFt) formation and reduce CH content. The addition of CS can fill the void in the hydrated paste, and promote the hydration reaction of cement, thus reducing the porosity of concrete. However, AM monomer polymerization can form an organic plasticizing zone, strengthen the bonding force between hydration products such as CH, and interact with needle-shaped AFt, effectively forming an organic-inorganic interpenetration network structure.Conclusions AM in-situ polymerization and CS could improve the toughness of concrete. The 28-d flexural stress peak and fracture energy were increased by 31.5% and 91.4%, respectively, and the compressive strength remained unchanged. CS whiskers could promote cement hydration, exert filling effect and reduce the porosity. AM polymerization could build a plasticized zone in cement pastes and weaken stress concentration in concrete. The synergistic effect enhanced the bonding force between the hydration products, formed an organic-inorganic interpenetrating network structure, optimized the structure of concrete across scales, and provided an effective technology for toughening concrete matrix.