Project results
Year 2011
Objective: Preparation and characterisation of porous media with a controlled amount of magnetic impurities
At this stage was investigated the possibility of producing porous ceramics containing different types of magnetic impurities and at different concentrations (0-10%). The porous ceramics were prepared using both the replica technique and the powder compression technique. Different magnetic impurities were tested but iron oxide (III) was finally chosen for further NMR studies. This choice is motivated by the moderate internal gradients produced inside the porous samples (<15T/m at 20MHz proton resonance frequency) and the fact that it is found in natural or fabricated porous media as soils, rocks, concrete. The magnetic properties of the produced samples were investigated using magnetic susceptibility measurements, electron spin resonance measurements and NMR relaxometry measurements. The pore size of the produced samples was determined by using the scanning electron microscopy and optical microscopy.
The results were compared with those obtained from NMR relaxation measurements indicating a good agreement. A special attention was given to a new NMR technique which monitors the magnetization decay due to diffusion in internal fields (DDIF). The pore size distribution of porous media can be then determined in a completely noninvasive manner using the DDIF technique and without a previous calibration. The DDIF technique was used for the first time in studying such samples and its limits were tested. To obtain the porosity of the produced samples gravimetric measurements were done. They were then compared with NMR measurements. It was shown that the pore size of the produced porous ceramics is not influenced by the amount of magnetic impurities and is the same for all produced samples.
The results were compared with those obtained from NMR relaxation measurements indicating a good agreement. A special attention was given to a new NMR technique which monitors the magnetization decay due to diffusion in internal fields (DDIF). The pore size distribution of porous media can be then determined in a completely noninvasive manner using the DDIF technique and without a previous calibration. The DDIF technique was used for the first time in studying such samples and its limits were tested. To obtain the porosity of the produced samples gravimetric measurements were done. They were then compared with NMR measurements. It was shown that the pore size of the produced porous ceramics is not influenced by the amount of magnetic impurities and is the same for all produced samples.
Year 2012
Objective: The influence of the internal gradients on diffusion and relaxation measurements
At this stage the dynamics of water (polar) and cyclohexane (nonpolar) molecules within the pore space of porous ceramics containing controlled and increased amount of iron oxide as magnetic impurities was investigated. The porous ceramics were fabricated using the conventional method of preparation from powders which are first dry pressed and then subject to thermal treatment. Six samples (S0-S10) with increasing concentration of iron oxide (III) were prepared by adding 0, 2, 4, 6, 8 or 10g of iron oxide to 100g of mixed powder. In order to extract the pore size distribution of the produced samples they were examined by scanning electron microscopy and the DDIF technique. A new NMR technique which relies on the attenuation of the echo train in the well-known CPMG experiment due to diffusion in internal gradients is also proposed. This technique takes into account the diffusion effects under restricted conditions on transverse relaxation measurements (center figure right). The magnetic characterization of the produced samples was done using a vibrating sample magnetometer indicating a linear dependence of the susceptibility constant with the iron oxide content. The solvents examined in our study were chosen in order to monitor the influence of the polarity on the surface relaxation process. The experimental relaxation dispersion curves (center left figure) could be compared with a two phase exchange model taking into account relaxation by interaction with paramagnetic centers on the surface of porous media. This comparison allowed us the conclusion that iron oxide clusters inside the porous matrix do not contribute to the relaxation in the frequency range of our experiments. It also allowed the determination of the transverse diffusional correlation time and of the diffusion coefficient at the liquid/solid interface. Moreover, a similar behavior in relaxation dispersion curves of water and cyclohexane filled samples could be observed, independently of the fact that water molecules are polar and cyclohexane molecules are nonpolar. On the other hand a longer correlation time was extracted in the case of water molecules as compared with the cyclohexane ones. The results of the study on fabricated porous ceramics have allowed a better interpretation of the experimental data on cement-based materials with naturally contained magnetic impurities. Thus it was shown that it is possible to study the pore size evolution during cement hydration (bottom figure left) or the influence of some external factors as temperature and carbon dioxide. Moreover information about the reaction products inside cement paste could be extracted on this basis and correlated with complementar measurements (bottom figure right). As a general conclusion we may say that our studies on porous ceramics with controlled amount of magnetic impurities contribute to the interpretation of the experimental data obtained on natural or fabricated porous media with unknown magnetic properties. The results of our studies to date were disseminated by publications in ISI journals and communicated at international conferences (see here). They were also partially included in 2 PhD Theses (S. Muncaci and M. Simina) finalised during 2012. Another outcome of the stage 2012 of the project is the acquisition of a Fast Field Cycling NMR relaxometer (top figure) unique in Romania.
Year 2013
Objective: Study of the molecule-surface interaction effects on macroscopic dynamics (translational diffusion) and location of polar and non-polar molecules under confinement
Objective: Study of the molecule-surface interaction effects on macroscopic dynamics (translational diffusion) and location of polar and non-polar molecules under confinement
As has been stated above the translational motion of the confined molecules in porous media is strongly influenced by the interaction with the surface. Until now the NMR studies of molecular dynamics in porous media mainly refer to saturated samples where the surface contribution is significantly reduced and the extracted information is incomplete. In the present stage of our project we have extended the investigations to partially saturated samples, in which the surface contribution is much more significant. In our investigations comparative studies on two types of samples were done: samples without magnetic impurities and samples with magnetic impurities.
In the first set of samples the diffusion of water and cyclohexane molecules inside partially saturated porous media without magnetic impurities (VitraPor#4 and VitraPor#5) was measured as a function of the filling factor. It is observed an increase of the diffusion coefficient in the case of cyclohexane filling VitraPor#5 (1 um pores) by reducing the filling factor (top figure left). This increase is attributes to the vapor phase contribution to the diffusion process. On the other hand, in the case of Vitrapor#4 (10 um pores) no enhance due to vapor phase contribution was observed. These results are in agreement with the Monte-Carlo simulations using both the plug and the surface model distribution of the liquid (top figure right) inside pores. The results on VitraPor#4 sample showed us that the contribution of the vapor phase can be also neglected in the case of porous ceramics with the pores in the same range. This is an important result that allows a simpler interpretation of the relaxation data.
The second set of samples under investigation consists of porous ceramics with controlled amount of magnetic impurities (produced in the previous stage of the project). In this case the diffusion measurements with the conventional pulse field gradient techniques are unreliable due to the fact that the internal gradients may exceed by orders of magnitude the external gradients of the NMR instrument. That is why in our investigations we use the Fast Field Cycling NMR relaxometry as a tool in extracting the diffusion coefficient at the surface of porous media. Thus, it was possible to monitor the evolution of the diffusion coefficient at the surface of porous media for different filling factors. The relaxation dispersion curves (middle figures) show a completely different behavior in the case of water and cyclohexane. This behavior has been observed also from transverse relaxation measurements and can be associated with the polar character of molecules which determines different distributions of the two fluids inside porous ceramics. A theoretical model that assumes as the dominant relaxation mechanism the protons interaction with the paramagnetic centers located on the sample surface was developed. The model considers the case of partially saturated porous media and takes into account the difference in molecular distribution on the surface. The model was applied also to fit the relaxation dispersion curves acquired in the case of cement paste (natural magnetic impurities) during the hydration process. The results have shown that the diffusion coefficient of water molecules at the interface of cement grains is reduced by an order of magnitude as compared with its bulk value. However, the diffusion coefficient at the surface of cement grains does not change during the early stage (<3h) of hydration.The results of our studies to date were disseminated by publications in ISI journals and communicated at international conferences (see here). They were also partially included in 1 PhD Thesis (A. Pop) which will be finalised during 2014.
In the first set of samples the diffusion of water and cyclohexane molecules inside partially saturated porous media without magnetic impurities (VitraPor#4 and VitraPor#5) was measured as a function of the filling factor. It is observed an increase of the diffusion coefficient in the case of cyclohexane filling VitraPor#5 (1 um pores) by reducing the filling factor (top figure left). This increase is attributes to the vapor phase contribution to the diffusion process. On the other hand, in the case of Vitrapor#4 (10 um pores) no enhance due to vapor phase contribution was observed. These results are in agreement with the Monte-Carlo simulations using both the plug and the surface model distribution of the liquid (top figure right) inside pores. The results on VitraPor#4 sample showed us that the contribution of the vapor phase can be also neglected in the case of porous ceramics with the pores in the same range. This is an important result that allows a simpler interpretation of the relaxation data.
The second set of samples under investigation consists of porous ceramics with controlled amount of magnetic impurities (produced in the previous stage of the project). In this case the diffusion measurements with the conventional pulse field gradient techniques are unreliable due to the fact that the internal gradients may exceed by orders of magnitude the external gradients of the NMR instrument. That is why in our investigations we use the Fast Field Cycling NMR relaxometry as a tool in extracting the diffusion coefficient at the surface of porous media. Thus, it was possible to monitor the evolution of the diffusion coefficient at the surface of porous media for different filling factors. The relaxation dispersion curves (middle figures) show a completely different behavior in the case of water and cyclohexane. This behavior has been observed also from transverse relaxation measurements and can be associated with the polar character of molecules which determines different distributions of the two fluids inside porous ceramics. A theoretical model that assumes as the dominant relaxation mechanism the protons interaction with the paramagnetic centers located on the sample surface was developed. The model considers the case of partially saturated porous media and takes into account the difference in molecular distribution on the surface. The model was applied also to fit the relaxation dispersion curves acquired in the case of cement paste (natural magnetic impurities) during the hydration process. The results have shown that the diffusion coefficient of water molecules at the interface of cement grains is reduced by an order of magnitude as compared with its bulk value. However, the diffusion coefficient at the surface of cement grains does not change during the early stage (<3h) of hydration.The results of our studies to date were disseminated by publications in ISI journals and communicated at international conferences (see here). They were also partially included in 1 PhD Thesis (A. Pop) which will be finalised during 2014.
Year 2014
Objective:Study of the molecule-surface interaction effects on rotational dynamics of molecules confined inside porous media without magnetic impurities
Objective:Study of the molecule-surface interaction effects on rotational dynamics of molecules confined inside porous media without magnetic impurities
During the research stage 2014 was investigated the effect of molecule-surface interaction on the rotational dynamics of molecules confined inside model porous media without magnetic impurities. It is already known that the rotational correlation time of confined molecules is strongly influenced by the interaction with the surface and intrinsically by the surface properties. Information about the rotational correlation time can be extracted via NMR techniques using two approaches: i) from relaxation data measurements as a function of relaxation field using the FFC technique; ii) by comparing the echo attenuation due to dipolar correlations in a stimulated echo experiment with the attenuation in a spin echo experiment. All these experiments were performed in our studies both as a function of temperature (figure a) and filling factor (figure d). The samples under investigation were the porous glasses Vycor (4nm pores diameter), Vitrapor#5 (1 micrometer pores), VitraPor#4 (10 micrometer pores) and porous ceramics prepared in the previous stages of the project having the pores size in the range of 10 micrometers. The results on these samples were compared with those obtained on white cement paste having the capillary pores in the same range of dimensions. The investigated fluids were both polar and nonpolar. As representatives of polar molecules water and dimethyl sulfoxide were chosen. The cyclohexane was used as a representative of nonpolar molecules. In order to separate the bulk contribution from the surface layer contribution the samples were first filled with deuterated water then evacuated and refilled again with the investigated solvent (figure c). A difference in the behavior of polar molecules as compared with the nonpolar ones was observed both as a function of filling factor, frequency and temperature. The presence of a liquid layer below the freezing temperature was observed. It was concluded that the relaxation in the unfrozen layer is dominated by the reorientations mediated by translational displacements. Our experiments were based both on conventional NMR relaxometry techniques (Hahn echo, stimulated echo, CPMG) and the Fast Field Cycling technique that allows observation of nuclear spin at different frequencies. It was also investigated the possibility of using dipolar correlation effect on the stimulated echo as a tool to study molecular dynamics at the surface. Thus it was demonstrated the presence of oscillations in the decay of the stimulated echo (figure b) as a function of the first interpulse interval in a stimulated echo experiment. These oscillations which were not observed in the Hahn echo experiment for the same parameters may be attributed to the dipolar correlations.The results of our studies to date were disseminated by publications in ISI journals and communicated at international conferences (see here). They were also partially included in 1 PhD Thesis (A. Pop) which is about to have the public presentation.
Year 2015
Objective: Study of the molecule-surface interaction effects on rotational dynamics of molecules confined inside porous media with magnetic impurities
In the reporting period were studied the effects introduced by the surface-molecule interaction on the dynamics of rotation of molecules confined inside porous ceramics containing controlled amounts of magnetic impurities. The aim of these investigations was to identify a behavior pattern for the molecules in porous media with controlled amount of magnetic impurities that can be compared to the results of natural porous media having similar pore sizes and surfaces with similar properties. Our studies were based both on conventional relaxometry techniques as CPMG (figures a and b) and the Fast Field Cycling technique (figure c), allowing the observation of the nuclear spin at different frequencies. The main advantages of FFC technique in comparison with the classical relaxometry at a unique frequency are its sensitivity to a broader spectrum of molecular motions and that is not influenced by internal diffusion gradients. The FFC drawback compared with CPMG technique at a certain frequency is the lower signal/noise ratio and consequently the longer duration of the experiments (up to 20 hours in the case of our samples).The investigations have revealed an increased relaxation rate by increasing the temperature (figures a and c) despite of the drop in the rotational correlation time. This increase can be attributed to partial desaturation of samples at high temperatures and thus to an increased role of the surface of interchange contributing to the relaxation phenomena. The experiments performed around liquid-solid phase transition (figure b) allowed us the identification of the plastic component of cyclohexane with a very small dependence on temperature and no dependence on magnetic impurity. This observation will be very useful in the final stage of the project when cement based materials (figure d) naturally contained magnetic impurities will be studied.
Year 2016
Objective: Application of the previous knowledge to the investigation of rotational dynamics of water molecules confined inside cement paste
Objective: Application of the previous knowledge to the investigation of rotational dynamics of water molecules confined inside cement paste
In the present stage the dynamics of water molecules saturating a cement paste prepared with silica nano-particles and hydrating at two different temperatures was investigated. The investigations referred to capillary pores (a) and were performed using the Fast Field cycling technique. The results have shown a reduction in the duration of the dormancy stage of hydration both by increasing the temperature and by the presence on silica nano-particles. Furthermore a reduction in the size of capillary pores could be detected (b). Fitting of the relaxation dispersion curves (c) with a relaxation model that takes into account the dipolar interaction of water protons with the paramagnetic centers located on the surface of capillary pores, it is possible to determine the transverse correlation time at the pore surface. It is observed a reduction of the correlation time by increasing the temperature both during the dormancy stage and the early stage of acceleration. Note however that the data could not be fitted with the theoretical model for later hydration stages when the so called gel-like structure (C-S-H) inside the cement paste becomes significant and the pores are partially saturated with liquid. In that case an exchange process between inter-C-S-H pores and the capillary pores needs to be considered as a relaxation mechanism.