Introduction Phosphates are one of commonly used inorganic retarding agents in Portland cement, including monophosphates (i.e., sodium phosphate) and polyphosphate compounds (i.e., sodium pyrophosphate, sodium tripolyphosphate, and sodium hexametaphosphate). Different phosphate salts exhibit distinct retarding mechanisms. Among these, sodium phosphate is extensively investigated as an alkaline metal monophosphate-type retarder. However, there exist varying interpretations of the retarding action of sodium phosphate in different studies, potentially impacting its practical application, and necessitating a clear elucidation of its retarding mechanism. Ammonium dihydrogen phosphate as an acid phosphate can also serve as a retarder and exhibits a more enduring retarding effect rather than the commonly used sodium phosphate. Compared to PO43-, the volatility of NH4+ or its potential to mitigate the influence of alkaline metal ions on the cement hydration and the larger ionization constant of H2PO4- can have a quicker reaction with alkaline cement particles, affecting the hydration process. Some work indicate that ammonium dihydrogen phosphate has a favorable retarding effect on cementitious slurries, but its specific retarding mechanism remains less explored. It is thus necessary for the analysis of their impact on the kinetics of Portland cement hydration to conduct a comparative study between ammonium dihydrogen phosphate and sodium phosphate.Methods Cement used was conformed to the GB 8076—2008 standard with a strength grade of 42.5. Ammonium dihydrogen phosphate and sodium phosphate were of analytical reagent grade (AR). The dosage of ammonium dihydrogen phosphate and sodium phosphate was 0.5% (in mass fraction), with a fixed water-to-cement ratio of 0.4. Cement samples for microstructural tests were prepared in 60 mL plastic bottles. The cement paste was prepared via dissolving ammonium dihydrogen phosphate and sodium phosphate, and mixed with cement. The well-mixed cement paste was placed in the plastic bottles and cured under natural conditions at (20±2) ℃ until the specified age.Cement setting time and paste flowability were determined according to GB/T 1346—2011 and GB/T 8077—2000 standards, respectively. The hydration exothermic curve of the cement paste in the early 72 h at 20 ℃ was recorded by a model TAM isothermal calorimeter. The zeta potential within the first 10 minutes of cement paste hydration was measured by a model DT-1200 electroacoustic spectrometer. The element concentration of the pore solution was measured by a model PS-6 inductively coupled plasma optical emission spectrometer (Baird Co., USA). The thermogravimetric analysis was conducted by a model STA449C synchronous thermal analyzer (NETZSCH Co., Germany). The X-ray diffraction (XRD) patterns of the samples were obtained by a model Advance D8 fully automated powder diffractometer (Bruker Co., Ltd., Switzerland). The surface morphology and elemental identification of the samples were observed by a model Nova NanoSEM230 scanning electron microscope.Results and discussion The incorporation of sodium phosphate and ammonium dihydrogen phosphate results in a corresponding extension of both initial and final setting time of the cement paste. Ammonium dihydrogen phosphate with an equal mass exhibits longer initial and final setting time, compared to sodium phosphate. Compared to the pure cement group, the addition of sodium phosphate and ammonium dihydrogen phosphate improves the flowability of the cement paste and delays the time-dependent loss of flowability, showing a more pronounced retarding effect.Hydration heat, thermogravimetric and XRD results indicate that sodium phosphate and ammonium dihydrogen phosphate both can inhibit the dissolution rates of C3S and C3A, thus retarding the formation of hydration products. This leads to a reduction in the rate of heat evolution during Portland cement hydration, an extension of the induction period of hydration, a decrease in the total heat evolved during the first three days of hydration, and an overall inhibition of Portland cement hydration. However, an inhibitory effect of ammonium dihydrogen phosphate is dominant.Based on the analysis of zeta potential and SEM results, after the addition of sodium phosphate and ammonium dihydrogen phosphate in cement, a coating layer is formed on the surface of the cement particles. Sodium phosphate precipitates as calcium phosphate on the surface of Portland cement particles, while ammonium dihydrogen phosphate precipitates as hydroxyapatite. Ammonium dihydrogen phosphate exhibits a greater retarding effect rather than sodium phosphate due to its higher adsorption capacity of hydroxyapatite for inorganic ions, compared to calcium phosphate.The early pore solution analysis indicates that ammonium dihydrogen phosphate can be adsorbed more rapidly on Portland cement particles, leading to the rapid precipitation of solid-phase products. In addition, no nitrogen (N) element appears when analyzing the early pore solution of the cement paste with ammonium dihydrogen phosphate, indicating that NH4+ rapidly escape into air as ammonia when adding to the Portland cement paste. This escape of NH4+ effectively prevents the influence of alkali metal ions in sodium phosphate on the early hydration of Portland cement.Conclusions Ammonium dihydrogen phosphate and sodium phosphate both were capable of prolonging the setting time of Portland cement and retarding the time-dependent loss of paste flowability with ammonium dihydrogen phosphate exhibiting a more pronounced retarding effect. These agents could inhibit the dissolution rates of C3S and C3A, thereby reducing the rate of heat evolution during Portland cement hydration, extending the induction period of hydration, lowering the total heat evolved during the first three days of hydration, and overall suppressing Portland cement hydration. However, the inhibitory effect of ammonium dihydrogen phosphate was dominant. Compared to sodium phosphate, ammonium dihydrogen phosphate could be more rapidly adsorbed on Portland cement particles, leading to the rapid precipitation of solid-phase products, and thereby inhibiting cement hydration. Sodium phosphate precipitated as calcium phosphate on the surface of Portland cement particles, suppressing cement hydration, while ammonium dihydrogen phosphate precipitated as hydroxyapatite on the surface of Portland cement particles, similarly inhibiting cement hydration.