Development and characterisation of injectable calcium phosphate cements for use in vertebroplasty

The global objective of this thesis was to understand how the starting components of brushite cements influence the cement properties relevant for its use in vertebroplasty. Therefore, this work focussed on the following cement properties : mechanical strength, X-ray opacity and heat release upon setting. To carry out the global objective, a test protocol was first developed to characterise the mechanical properties of calcium phosphate cements. The Mohr's circles representation allowed to understand that for a material like our cement, ultimate stress measured with the Brazilian test can only underestimate tensile strength because the compressive/tensile strength ratio is lower than 8. Nevertheless, the Brazilian test remains a test that is easy to use with brittle materials and requires simple sample shape. Therefore, it was considered as an appropriate test for our cements, and was used as a way to measure the indirect tensile strength of the cements tested in the run of this work. Refining two of the starting powders, MCPM and CSH, allowed to increase significantly both compressive and diametral tensile strengths. Milling of these powders provided a five-fold increase in axial compression, and a three-fold increase in diametral compression. The mechanical strengths of cements made with milled MCPM and CSH were the highest measured in this thesis work : 4.1 MPa diametral tensile strength and 22.1 MPa compressive strength. The mechanical properties of the cement were found to be strongly dependent on both CSH granulometry and chemical purity. A computer model was developed to calculate the linear X-ray attenuation coefficient of any material of known composition. It was then validated by experimental measurements. The standard linear X-ray attenuation coefficient a cement should exhibit to be used in vertebroplasty was found to amount 2.47 cm-1 for the conditions simulated. It was calculated from an acrylic cement composition considered a standard in vertebroplasty. Calculation of the attenuation coefficient of a brushite cement revealed that this material is not radiopaque enough for safe use in vertebroplasty. Therefore, we added iodinated molecules to a brushite cement formulation so as to produce the required X-ray opacification. We found that incorporation of Iopentol, a contrast media commonly used in X-ray imaging applications, does not impair the mechanical properties of the cement and does not alter the setting time of the reaction either. Finally, a custom-made calorimeter was developed for measurement of kinetic and thermometric parameters of cements. One single measurement allowed calculation of the maximum reaction rate of the setting reactions, the working and setting times of the cement pastes, the degree of conversion of the cements, and the temperature increase induced by the setting reaction. The degree of hydration of the calcium sulfate was found to have a significant influence on the kinetics of the setting reaction. The reaction is accelerated by calcium sulfate dihydrate because it provides nuclei for heterogeneous precipitation of brushite. In contrast, calcium sulfate hemihydrate was found to slow down the reaction by releasing sulfate ions in the mixing liquid. As the reaction is faster when CSD is used, heat production occurs faster and accumulates in the system. As a consequence, the temperature elevation during setting is more important.

Insights on chemical and physical chloride binding in blended cement pastes

Fabien Jacques Michel Georget, Julien Nicolas Gonthier, Karen Scrivener, William Wilson

This study investigates chloride binding in blended cement pastes exposed to 0.5 M NaCl solutions (with and without pH adjustment) using X-ray diffraction and energy-dispersive X-ray spectroscopy image analysis (edxia). The aim is to better understand the effects of the binder type, the water-to-binder ratio and the pH on the chemical binding in AFm phases and the physical binding on C-A-S-H. Results show that the binding cannot be predicted from AFm and C-A-S-H contents alone because competing ions in the system affect both the Friedel's salt solid solution chemistry and the C-A-S-H binding capacity. Notably, the high content of aluminous hydrates in LC3 systems leads to a high chemical binding even if Friedel's salt solid solutions have relatively low chloride contents (particularly at a higher pH). On the contrary, the CEMIII/A paste showed low binding because of relatively high sulfate and magnesium contents which compete for incorporation/adsorption in aluminous hydrates (AFm, ettringite and hydrotalcite).

PERGAMON-ELSEVIER SCIENCE LTD2022

Propriétés physico-chimiques et ostéogéniques d'un biociment hydraulique à base de phosphates de calcium

Biocements made from β-TCP - H3PO4 - H2O and β-TCP - MCPM - H2O mixtures were studied in order to obtain a better control of their setting time and mechanical strength (β-TCP = Ca3(PO4)2; MCPM = Ca(H2PO4)2.H2O). The effects of factors like the purity of the β-TCP powder, its particle size distribution, the carnet composition or the presence of additives were investigated. More fundamental studies were also done on the reactions controlling the setting time, i.e. β-TP dissolution and DCPD (CaHPO4.2H2O) precipitation. The influence of additives on the kinetics of these controlling reactions were studied in order to establish how they influenced the setting time. To complete these experiments, cements made from β-TCP - MCPM - CSH - H2O mixtures were implanted in rabbit tibias (CSH = CaSO4.l/2H2O). The aging behavior in vivo and in vitro were compared. Results show that many factors affect the physico-chemical properties of the cements, particularly setting time and tensile strength. However, these factors have only a very limited number of pathways in which they can act. For example, setting time can only change when either β-TCP dissolution rate, DCPD germination rate or interparticular free volume are modified. Tensile strength only depends on DCPD weight fraction (binder for β-TCP powder), porosity and microstructure. Sulfate, pyrophosphate and citrate ions delay the cement's setting time in the following order: sulfate < citrate < pyrophosphate. These three ions inhibit DCPD crystal growth in the same order. Therefore, these results suggest that all the ions which inhibit DCPD crystal growth are potential setting time delaying additives. The concentration range in which sulfate ions act on setting time is limited between 0 and 0.1 M; beyond this concentration sulfate ions precipitate as CSD (CaSO4.2H2O). As DCPD and CSD structures are nearly identical, the presence of CSD crystals speeds up the setting reaction by acting as seeds for DCPD crystals. This phenomenon provokes a decrease in the setting time and a refinement of the microstructure with a consequent increase in the tensile strength. Accurate control of the cement composition is very important, i.e. volume and concentration of the phosphoric acid solution. Results show that an excess of phosphoric acid provokes the recrystallization of DCPD into DCP, a process which strongly decreases tensile strength. A modification of the microstructure is also observed when initial β-TCP specific surface area is changed. An increase in the surface area gives an increase in the tensile strength and a decrease in the setting time. In our experiments, a fivefold increase of specific surface area decreases the setting time by a factor of three and doubles the tensile strength. These results show that β-TCP particle size distribution has a very strong effect on the physico-chemical properties of our cements. The comparison between in vitro and in vivo tests prove that in vivo results cannot be anticipated by in vitro experiments. However, in vivo results are extremely positive: our cement is biocompatible, bioresorbable and osteoconductive. These results show that after one month, our cements are closely bonded to living bone, and after four months, our cements are nearly completely resorbed and replaced by new bone (except for dense and large β-TCP particles).

EPFL1993

Autogenous Shrinkage and Hydration Kinetics of SH-UHPFRC under Moderate to Low Temperature Curing Conditions

Mohammadhadi Kazemi Kamyab

Strain Hardening Ultra High Performance Fiber Reinforced Concrete (SH-UHPFRC) were used successfully over the last 8 years in numerous cast in place rehabilitation applications as long lasting waterproofing layers (20 to 30 mm thickness) and, combined with rebar, to increase the structural load bearing capacity of existing bridge decks or building slabs (50 to 70 mm thickness). In composite applications on reinforced concrete substrates, SH-UHPFRC provide a deformation capability (strain hardening) larger than their free shrinkage which makes localized macro cracking at service state very unlikely. On the other hand some specific conditions of microclimatic state of the substrate (moisture and thermal conditions) and climatological (temperature and humidity) of the site such as casting at low temperatures (winter conditions) might increase the autogenous shrinkage and thus eigenstresses of the freshly cast SH-UHPFRC and hinder the development of its tensile strength or/and deformability. Combination of these specific circumstances could compromise the SH-UHPFRC protective functions due to the occurrence of localized macro cracks, making the objective of using SH-UHPFRC as a onetime intervention strategy in rehabilitation application obsolete. Advancement in hydration at early age leads to formation of partially emptied capillary spaces contributing to self-desiccation, decrease of the relative humidity and increase of the capillary depression. Since water is under depression, the solid porous material is under compression forcing the material to shrink. From the aforementioned, it can be deduced that the autogenous shrinkage is a force driven phenomenon: a force is applied on the ageing viscoelastic porous skeleton and the apparent structural response of the bulk cementitious material is the autogenous shrinkage strain. Based on the aforementioned two levels of investigation were considered in this thesis: I. Material level: ageing viscoelastic porous skeleton of hydrating cementitious material II. Structural level: autogenous shrinkage development under the action of its driving forces The objective of this thesis was to study the development of the autogenous shrinkage and its relation with the development of microstructure and mechanical properties in SH-UHPFRC cured at moderate and low temperature conditions (1 to 20°C). Focus was put on: (1) kinetics and magnitude of hydration of silica fume and cement, (2) activation energy of the processes, (3) stiffness (E modulus and Poisson ratio) development, and (4) states of water and pore development. Finally a cross link analysis of information obtained from material and structural level was carried out to better highlight the mechanisms and driving forces of autogenous shrinkage. CM22-TKK, a SH-UHPFRC developed at MCS for rehabilitation applications, on the basis of CEMTECmultiscale® fibrous mixes, was used. Silica fume is one of the major constituents of CM22-TKK (26% by mass of cement). Extent of the silica fume pozzolanic reaction significantly affects the pore size distribution of the UHPFRC matrix which then influences the development of the relative humidity, self-desiccation and finally autogenous deformation. Detailed experimental investigation performed showed that only maximum 8% by mass of cement out of 26% silica fume used in the recipe reacts in moderate and low temperature curing conditions. Finally, a model of silica-fume degree of reaction as a function of curing temperature (5 to 250 °C) was proposed on the basis of a detailed literature survey of existing works with low W/B mixes. The activation energy of the thermo-mechanical properties was evaluated for thermo (cumulative heat of hydration) and mechanical properties development (such as dynamic Elastic modulus, dynamic Poisson’s ratio and compressive strength). A single activation energy Ea of 27.4 KJ/mol could successfully model the activation of all properties. Since maturity was not a proper reference to compare the autogenous deformations at various temperatures, a more fundamental reference such as degree of hydration was used. Experimental investigation was carried out using different techniques such as isothermal calorimetry and 29Si solid NMR to evaluate the development of the degree of hydration of cement and silica fume in UHPFRC at moderate and low temperature curing conditions. The ultimate degree of hydration of cement in CM22-TKK for a closed system was 0.34. There was a close correspondence between the experimental results and the predictions of the volumetric phase distribution models of Waller (0.28) and Jensen (0.32). Finally, the De Schutter et al. 1996 model was used to describe the mechanical properties (Emodulus and Poisson ratio) as function of degree of hydration. Nondestructive and non-invasive 1H-NMR was used to evaluate volumetric phase distribution development at 20°C, especially the classification of the water types (capillary, gel and C-S-H interlayer water) which brought fully new and valuable information about the hydration kinetics and pore size development. The results were also compared with Jensen volumetric phase distribution model to check the advantages and limitation of both approaches. Finally experimental investigation was carried out to follow the development of autogenous shrinkage and eigenstresses (in case of restrained shrinkage) of CM22-TKK cured in isotherm conditions (1 to 20°C). Eigenstresses were obtained and used to highlight the effect of the viscous response of the material at the different temperatures investigated. The previously mentioned results on the kinetics of the silica fume reaction, cement hydration and evolution of states of water and development of stiffness were then put into perspective with the trends observed on the autogenous shrinkage and eigenstresses development.

EPFL2013

Autogenous Shrinkage and Hydration Kinetics of SH-UHPFRC under Moderate to Low Temperature Curing Conditions

Mohammadhadi Kazemi Kamyab

EPFL2013

Propriétés physico-chimiques et ostéogéniques d'un biociment hydraulique à base de phosphates de calcium

EPFL1993