We report the first nuclear magnetic resonance ͑NMR͒ two-dimensional correlation T 1 -T 2 and T 2 -T 2 measurements of hydrating cement pastes. A small but distinct cross peak in the two-dimensional relaxation spectrum provides the first direct evidence of chemical exchange of water between gel and capillary pores occurring over the first 14 days of hydration. A correlation of features along the line T 1 =4T 2 provides strong supportive evidence for the surface diffusion model of 1 H nuclear spin relaxation in cements and for a multimodal discrete pore size distribution. Differences in detail of the results are reported for white cement paste and white cement paste with added silica fume. Both the method and the theory presented can be applied more widely to other high surface area materials with other reactive surface areas.

Nuclear Magnetic Resonance (NMR) cryoporometry is a technique for non-destructively determining pore size distributions in porous media through the observation of the depressed melting point of a confined liquid. It is suitable for measuring pore diameters in the range 2 nm -1 µm, depending on the absorbate. Whilst NMR cryoporometry is a pertabative measurement, the results are independent of spin interactions at the pore surface and so can offer direct measurements of pore volume as a function of pore diameter. Pore size distributions obtained with NMR cryoporometry have been shown to compare favourably with those from other methods such as gas adsorption, DSC thermoporosimetry, and SANS. The applications of NMR cryoporometry include studies of silica gels, bones, cements, rocks and many other porous materials. It is also possible to adapt the basic experiment to provide structural resolution in spatially dependent pore size distributions, or behavioural information about the confined liquid.

No abstract

Nuclear magnetic resonance (NMR) relaxation times are shown to provide a unique probe of adsorbate–adsorbent interactions in liquid-saturated porous materials. A short theoretical analysis is presented, which shows that the ratio of the longitudinal to transverse relaxation times (T1/T2) is related to an adsorbate–adsorbent interaction energy, and we introduce a quantitative metric esurf (based on the relaxation time ratio) characterising the strength of this surface interaction. We then consider the interaction of water with a range of oxide surfaces (TiO2 anatase, TiO2 rutile, γ-Al2O3, SiO2, θ-Al2O3 and ZrO2) and show that esurf correlates with the strongest adsorption sites present, as determined by temperature programmed desorption (TPD). Thus we demonstrate that NMR relaxation measurements have a direct physical interpretation in terms of the characterisation of activation energy of desorption from the surface. Further, for a series of chemically similar solid materials, in this case a range of oxide materials, for which at least two calibration values are obtainable by TPD, the esurf parameter yields a direct estimate of the maximum activation energy of desorption from the surface. The results suggest that T1/T2 measurements may become a useful addition to the methods available to characterise liquid-phase adsorption in porous materials. The particular motivation for this work is to characterise adsorbate–surface interactions in liquid-phase catalysis.

No abstract

Summary Increasing flooding-solution viscosity with polymers provides a favorable mobility ratio compared with brine flooding and hence improves volumetric sweep efficiency. Flooding with a polymer solution exhibiting elastic properties has been reported to increase displacement efficiency, resulting in a sustained doubling of the recovery enhancement compared with the use of conventional viscous-polymer flooding (Wang et al. 2011). Flooding with viscoelastic-polymer solutions is claimed also to increase recovery more than expected from changes in capillary number alone (Wang et al. 2010). This increase in displacement efficiency by viscoelastic polymers is reported to occur because of changes in the steady-state-flow profile and enhancements in oil stripping and thread formation. However, within the industry there are doubts that a genuine effect is observed, or that improvements in displacement efficiency occur with field-applicable flow regimes (Vermolen et al. 2014). In this study, we demonstrate that flooding with viscoelastic-polymer solutions can indeed increase recovery more than expected from changes in capillary number. We show a mechanism of fluctuations in flow at low Reynolds number by which viscoelastic-polymer solutions provide improvements in displacement efficiency. The mechanism, known as elastic turbulence, is an effect previously unrecognized in this context. We demonstrate that the effect may be obtained at field-relevant flow rates. Furthermore, this underlying mechanism explains both the enhanced capillary-desaturation curves and the observation of apparent flow thickening (Delshad et al. 2008; Seright et al. 2011) for these viscoelastic solutions in porous media. The work contrasts experiments on flow and recovery by use of viscous and viscoelastic-polymer solutions. The circumstances under which viscoelasticity is beneficial are demonstrated. The findings are applicable to the design of formulations for enhanced oil recovery (EOR) by polymer flooding. A combination of coreflooding, micromodel flow, and rheometric studies is presented. The results include single-phase and multiphase floods in sandstone cores. Polymer solutions are viscoelastic [partially hydrolyzed polyacrylamide (HPAM)] or viscous (xanthan). The effects of molecular weight, flow rate, and concentration of the HPAMs are described. The data lead us to suggest a mechanism that may be used to explain the observations of improved displacement efficiency and why the improvement is not seen for all viscoelastic-polymer floods.

The pulsed-field gradient (PFG)-NMR technique has been applied to study molecular diffusion of organic liquids within mesoporous materials used in heterogeneous catalysis, in order to assess the effect of chemical functionalities on the effective self-diffusivity of the probe molecule within the pore space. True tortuosity values of the porous matrix can be calculated from the ratio of the unrestricted free self-diffusivity to the self-diffusivity within the pore space only when the small liquid-phase probe molecules do not have any chemical functionality that interacts within the solid phase (e.g., alkanes). The use of molecules with reactive chemical functionalities gives values heavily dependent on the physical and chemical interactions within the porous medium; hence, these values cannot be defined as tortuosity. Polyols showed an interesting behavior of enhanced rate of self-diffusion within the confined pore space, and this is attributed to the ability of the porous medium to disrupt the extensive intermolecular hydrogen bonding network of polyols.

It is known that internal magnetic field gradients in porous materials, caused by susceptibility differences at the solid-fluid interfaces, alter the observed effective Nuclear Magnetic Resonance transverse relaxation times T2,eff. The internal gradients scale with the strength of the static background magnetic field B0. Here, we acquire data at various magnitudes of B0 to observe the influence of internal gradients on T2-T2 exchange measurements; the theory discussed and observations made are applicable to any T2-T2 analysis of heterogeneous materials. At high magnetic field strengths, it is possible to observe diffusive exchange between regions of local internal gradient extrema within individual pores. Therefore, the observed exchange pathways are not associated with pore-to-pore exchange. Understanding the significance of internal gradients in transverse relaxation measurements is critical to interpreting these results. We present the example of water in porous sandstone rock and offer a guideline to determine whether an observed T2,eff relaxation time distribution reflects the pore size distribution for a given susceptibility contrast (magnetic field strength) and spin echo separation. More generally, we confirm that for porous materials T1 provides a better indication of the pore size distribution than T2,eff at high magnetic field strengths (B0>1 T), and demonstrate the data analysis necessary to validate pore size interpretations of T2,eff measurements.