Chemical composition of mortars made from the magnesia amidoimidophosphates system

Abstract

It is well known that hydrated magnesium ammonium phosphate compounds (MgNH4PO4.nH20) are readily formed as white crystalline precipitates when aqueous solutions of ammonium phosphates are mixed with a solution of magnesium salt [1]. The aged precipitates of the MgO-H3_x(NHe)xPO4H20 systems are hard and strong. Elucidation of the reaction mechanism of the MgO-H3_x(NH4)xPO4-H20 system [2] suggested that it would be possible to find alternative Mg 2+ ion-accepting liquids such as monoammonium and diammonium phosphate [3-6], ammonium pyrophosphate [7] and ammonium polyphosphates [8,9]. Amidoimidophosphates are attractive systems to study, since they are used as fertilizers and are known to be stronger complexing agents than polyphosphates [10]. Moreover, amidoimidophosphates are slowly converted in basic media first into amidoimidophosphates, then into ammonium phosphates [11]. Therefore, during the setting mechanism the cement would release ammonia and heat slowly, contrary to cements of the MgO-H3_x(NHe)~PO4-H20 system. The initial study reported here is concerned with establishing conditions under which amidoimidophosphates retain their structural integrity during the cement paste hardening process. We then determine whether the properties of amidoimidophosphates cements differ from those of orthophosphate cements. Magnesium oxide supplied by Martin Marietta (grade 10CR) was used. It has on average a median particle size of 10/xm and a surface area of <1 m 2 g-1. Water was deionized before use. Polyimidophosphates were obtained by heat treatment for i h at 110 °C in phosphorodiamidic acid atmosphere [12]. All magnesia-polyimidophosphate cements were prepared using the same procedure. Water (10 wt %) and polyimidophosphates (30 wt %) were mixed and stirred for 10 min. Magnesium oxide powder (60 wt %) was then added and stirred for 5 min to obtain a homogeneous and pourable paste. The paste was rapidly transferred into a 25 mm x 25 mm × 25 mm cubic vial. The paste was allowed to cure for 7 days at room temperature and humidity in the uncapped vial. X-ray diffractograms were recorded with nickelfiltered C u K , radiation using a Philips PWl010 diffractometer. Infrared spectra were recorded on a Perkin-Elmer M1710 spectrometer using the KBr disc method. 31p liquid nuclear magnetic resonance