Introduction　Relying　on　their　advantages　of　high　water　reducing　rate,　low　dosage,　and　strong　molecular　structure　design　ability,　polycarboxylate　superplasticizers　(PCEs)　have　become　the　most　widely　used　admixture　in　the　field　of　concrete　admixtures.　Adsorption　of　PCEs　onto　the　surfaces　of　cement　particles　is　the　prerequisite　for　the　dispersing　performance　of　PCE　molecules.　Therefore,　the　influences　of　PCE　molecules　with　different　structures　on　the　dispersion　properties　of　cement　slurry　can　be　understood　by　revealing　the　adsorption　process　of　PCE　and　the　conformation　characteristics　of　PCE　after　adsorption.　It　should　be　noted　that　both　the　adsorption　process　and　the　conformation　of　the　adsorbed　PCE　involve　interfacial　interactions　between　the　organic　and　inorganic　phases,　which　is　the　dominant　factor　determining　the　overall　properties　of　the　material.　Recently,　molecular　dynamics　(MD)　simulations　have　been　applied　to　the　PCEs　to　elucidate　its　working　mechanism.　However,　the　molecular　size　of　these　model　PCEs　was　much　smaller　than　those　typically　used　for　industrial　applications　in　order　to　reduce　computation　cost.　Furthermore,　only　pure　water　was　used,　or　only　Na+,　Ca2+　and　Cl−　were　additionally　added　to　mimic　a　cement　pore　solution.　The　above　methods　make　the　results　unable　to　accurately　reflect　the　adsorption　mechanism　of　PCE.　Therefore,　to　better　understand　and　effectively　control　the　performance　of　PCE,　more　in-depth　studies　at　the　atomic　and　molecular　scales　are　needed.　Our　research　team　innovatively　used　MD　simulations　in　which　the　molecular　size　of　the　model　PCE　is　consistent　with　that　of　PCE　commonly　used　for　industrial　applications　and　simulates　cement　pore　solutions.　In　addition,　the　adsorption　mechanism　of　sulfonation　modification　of　PCE　was　studied　by　combining　the　structure,　dynamics　and　stability　of　interfacial　connections.　Methods　The　MD　simulations　used　the　Gromacs-4.6.7　software　package　and　a　conventional　oplsaa　force　field　with　RESP　charges.　The　UFF　force　field　was　used　for　the　substrate.　The　equilibrium　MD　simulation　was　conducted　under　the　NPT　ensemble　for　a　total　run　time　of　20　ns　at　a　relaxed　liquid　configuration　(25　℃).　The　solution　environment　simulations　were　performed　using　3000　water　molecules.　A　relaxation　system　was　selected　and　the　energy　was　minimized　by　the　steepest　descent　method　with　a　termination　gradient　of　100　kJ/(mol∙nm)　before　the　relaxation.　The　system　temperature　maintained　constant　through　a　Nosé–Hoover　thermostat,　and　periodic　boundary　conditions　were　applied　to　all　three　dimensions.　Long-range　electrostatics　within　a　relative　tolerance　of　1×10−6　is　calculated　by　the　particle　mesh　Ewald　method　and　a　cut-off　distance　of　1　nm　was　used　as　the　Ewald　interaction　and　van　der　Waals　interactions.　The　bond　lengths　of　hydrogen　atoms　were　constrained　by　the　LINCS　algorithm.　A　leap-frog　algorithm　was　used　with　a　time　step　of　1　fs.　Results　and　discussion　Under　cement　solution　conditions,　the　side　chains　of　PAA-g-PALS　molecules　were　not　entangled　with　the　backbone　(present　a　network-like　conformation),　while　a　small　part　of　the　side　chains　of　PAA-g-HPEG　molecules　will　curl　onto　the　backbone　(most　of　the　side　chains　curled　into　a　mushroom　shape).　In　addition,　the　rotation　radius　of　PAA-g-PALS　(7.8　nm)　was　significantly　larger　than　PAA-g-HPEG　(7.1　nm)　due　to　the　large　amount　of　—SO3–　with　strong　electrostatic　repulsion.　Part　of　the　backbone　of　PAA-g-HPEG　was　adsorbed　to　the　particles　(the　side　chains　curl　and　wrap　the　backbone),　while　PAA-g-PALS　was　adsorbed　to　the　substrate　surface　by　the　＂flat＂　mode　of　the　individual　side　chains　(the　backbone　and　other　side　chains　extend　into　the　solution).　Compared　with　PAA-g-HPEG,　the　binding　point　of　sulfation　modified　PCE　and　the　substrate　is　relatively　scattered　and　increases　the　binding　surface　with　the　substrate,　which　is　consistent　with　the　calculation　of　binding　force　PAA-g-PALS　(174.3　kJ/mol)＞　PAA-g-HPEG　(134.5　kJ　/　mol).　The　vertical　heights　of　the　individual　PAA-g-HPEG　and　PAA-g-PALS　molecules　on　the　substrate　surface　were　6.8　nm　and　19.7　nm,　respectively,　and　the　coverage　rate　was　26.4%　and　21.2%,　respectively.　When　considering　the　polymer　cross-dimensions　parallel　and　perpendicular　to　the　particle　surface,　the　penetration　volume　formed　by　PAA-g-PALS　on　the　substrate　surface　increased　by　132.64%　relative　to　PAA-g-HPEG.　Conclusions　The　main　conclusions　of　this　paper　are　summarized　as　following.　PAA-g-HPEG　and　PAA-g-PALS　tend　to　be　comb　and　network　conformations　in　cement　slurry　environment,　respectively.　Sulfonated　modification　dispersed　the　distribution　of　binding　points　between　PCE　and　the　substrate,　improved　the　binding　force　between　PCE　and　the　substrate,　extended　the　polymer　far　away　from　the　substrate,　and　significantly　increase　the　permeability　volume　(by　about　132.64%).　The　present　results　not　only　provide　new　insights　into　the　action　mechanism　of　PCE　at　the　atomic　level,　but　also　provide　a　theoretical　basis　for　promoting　the　development　of　high　performance　cement-based　materials.　©　2024　Chinese　Ceramic　Society.　All　rights　reserved.