Effect Of Lignosulfonate On Setting Time Of Portland Cement And Study On Its Action Mechanism

Lignosulfonate water reducer, recovered from spent sulfite pulping liquor, has theadvantages of low-price, good air-entraining and retarding performances. From the point ofresource, environment and security, it is a kind of eco-friendly product with the potential ofwidespread use. However, there are some problems limiting its application, such as lowwater-reducing efficiency, low-efficiency in improving concrete strength, and poorcompatibility between lignosulfonate and cement which can sometimes lead to quick set orexcessive retarding of concrete. By chemical modification, lignosulfonate water-reducershows a higher water-reducing efficiency, and a better workability of concrete, but theproblem of quick set （abnormal setting） induced by the poor compatibility betweenlignosulfonate and cement containing anhydrite remains unsolved, and in this case,construction becomes very difficult and properties of concrete decline sharply, which addscertain constraints to the application.In order to improve the compatibility between lignosulfonate and cement, and extend theapplication of lignosulfonate water-reducer, the reason and action mechanism of the quick setinduced by the use of lignosulfonate should be first revealed. The setting of cement is muchrelated to its hydration, so the research of action mechanism of quick set should be carried outfrom two aspects of setting time and hydration process.Firstly, the effects of ultrafiltration fraction of calcium lignosulfonate （CL）, the form ofcalcium sulfate, the alkali content and the C₃A content on setting time of Portland cementhave been systematically investigated in this paper. The result shows that the major factorsthat lead to the quick set are the low molecular weight fraction of CL, the anhydrite and C₃Acontent in cement. When the retarder is anhydrite and the C₃A content is relatively high inPortland cement, the addition of0.4%CL fraction with a low molecular weight can sharplyshorten initial set from57min to5min and final set from103min to9min.By UV, HPLC and AAS, the contents of lignin, the contents of reducing sugar and itsmonosaccharide components and the compositions of inorganic salt of the CL with differentmolecular weight fractions have been analyzed. The result shows that by ultrafiltration andfractionation, the CL fraction with a high molecular weight （CL-A） is purified, and the CLfraction with a low molecular weight （CL-C） contains most of reducing sugar and inorganicsalt. The main monosaccharide components of reducing sugar in CL are glucose, mannose,xylose, arabinose and galactose. The effects of the reductive monosaccharides and inorganic salt on setting time of Portland cement have been further studied. The result shows that theabove-mentioned five monosaccharides in the CL are the principal components of inducingthe quick set of CEM-II （the cement containing anhydrite as a retarder and relatively highC₃A content）.In order to thoroughly reveal the reason and action mechanism of quick set of the cementcontaining anhydrite induced by the use of lignosulfonate, regarding liquid phase as theresearch target, by IC, AAS and pH meter, the effects of CL and its different molecular weightfractions on the ion concentration of liquid phase in the C₃A-CaSO₄system have been studied;regarding solid phase as the research target, by XRD, TG-DSC and SEM, the effects of theCL and its different molecular weight fractions on the phase compositions and morphology ofhydration products at different time in the C₃A-CaSO₄system have been studied. Based onthe analysis of the different effects of the CL on setting times of the cement containinggypsum and another cement containing anhydrite, the following conclusions have beendrawn:When the CL is added to the C₃A-gypsum system, SO₄^2-concentration, Ca²⁺concentration and CaSO₄saturation ratio in the liquid phase all increase to some extent.At the first30min of hydration in the C₃A-gypsum system, compared with the blanksample, in the sample containing the CL, ettringite crystals of needle and rod increase in theamount, are about1²Î¼m in length, overlap each other closely; in the sample containing theCL-A fraction with a high molecular weight, the amount of ettringite is close to that of theblank sample, but ettringite crystals of needle and rod are smaller in size, are about0.5¹.5μm in length, overlap each other more closely. In the sample containing the CL or the CL-A,ettringite crystals of needle and rod are more or smaller, which can form covering layeraround the C₃A grains more tightly, and delay the hydration of the C₃A. So, the CL-A hasrelatively strong setting retarding effect on the P·II42.5R, in the high dosage of CL （≥1.0%）,the P·II42.5R cannot coagulate for a long time, and the phenomenon of excessive retardinghappens.Unlike the CL-A, the CL-C fraction with a low molecular weight has little effect on thesetting time of P·II42.5R. The result of hydration research shows that at the first30min ofhydration in the C₃A-gypsum system, in the sample containing the CL-C, the amount and sizeof ettringite crystals have little difference from that of the blank sample. As a result, the CL-Ccannot delay the hydration of the C₃A obviously, and has little effect on the setting time.In the C₃A-anhydrite system, compared with the blank sample, in the sample containingthe CL or the CL-C, Ca²⁺concentration in the liquid phase increases, but due to the rapid formation of ettringite crystal depletes large amount of sulfate, SO₄^2-concentration andCaSO₄saturation ratio in the liquid phase all decrease to some extent, especially during thefirst1min of hydration, SO₄^2-concentration and CaSO₄saturation ratio are close to zero.The CL-A fraction with a high molecular weight and the low dosage of CL （≤0.3%） canprolong the setting time of CEM-II, but within a relatively small range. The result ofhydration research shows that at the first30min of hydration in the C₃A-anhydrite system,compared with the blank sample, in the sample containing the CL-A or the CL, the amount ofettringite crystals substantially increase, the crystal sizes are bigger than that of the blanksample （the lengths are about2⁴μm in the presence of CL, and about2μm in the presenceof CL-A）, and the ettringite crystals do not overlap each other closely. The analysis suggeststhat the ettringite crystals are large in the amount, but big in size, which can form coveringlayer around the C₃A grains, but the covering layer is not tight, so the setting retarding effecton the C₃A and the CEM-II is weak.When the CL is added in the dosage of1.0%or the CL-C fraction with a low molecularweight is added in the dosage of0.4%, the setting time of CEM-II is about10min, and thephenomenon of excessive retarding happens. The result of hydration research shows that atthe first30min of hydration in the C₃A-anhydrite system, compared with the blank sample, inthe sample containing the CL-C, ettringite crystals substantially increase in the amount, andare mostly present in the form of long rods, and are much bigger than that of the blank sample（about6⁸μm in length）, and there are pore spaces between the crystals, so it is hard to formtight covering layer around the C₃A grains to hinder the diffusion of the liquid phase, and theC₃A continues to react with the anhydrite rapidly, leading to the quick set of CEM-II.By XRD, TG-DSC and SEM, the effects of the CL-C fraction with a low molecularweight and the above-mentioned five monosaccharides on the phase compositions andmorphology of hydration products of CEM-II at different time have been studied too in thispaper. The result shows that in the case of the quick set of CEM-II induced by the CL-Cfraction with a low molecular weight and arabinose, relatively fine ettringite crystals of needleand rod do not get a well growth or overlap each other. It is further demonstrated that due toettringite crystals do not form complete covering layer around the cement grains to delay thesubsequent hydration of cement and C₃A, the CEM-II reaches to setting quickly.