Desalination of porous building materials by electrokinetics: an NMR study
- 1.3k Downloads
This article presents the first non destructive measurements of salt ions transport through fired-clay brick during electrokinetic desalination using nuclear magnetic resonance technique. The effect of the strength of an applied electric field on the migration of salt ions is examined by varying the electrical potential gradients from 0.75–2 V cm−1 across the specimens. The measurements show that for electrokinetic to exceed ion transport by diffusion a minimum level of applied voltage is necessary. Below this threshold salt transport by diffusion is dominant over electromigration. The effect of advection on the salt transport is studied by introducing a hydraulic gradient across the specimen. The results show that advection is a major transport process in the materials studied. To assess the relative magnitude of the various active transport processes during electrokinetic desalination, a scale analysis on the basis of dimensionless numbers is presented. The value of these numbers determines which transport mechanism will dominate the desalination process in a given sample length and time scale.
KeywordsFired-clay brick Diffusion Advection Electromigration Nuclear magnetic resonance
The deterioration of porous building materials and structures by the crystallization of water soluble salts is a well known phenomenon. The threats posed by salts to porous materials can be minimized either by controlling the environment or by removing the salts from the zone of deterioration. In the case of porous building materials salt extraction is typically accomplished by dry removal of the salt efflorescence and/or aqueous methods—a common example of which being the use of poultices [1, 2, 3] to reduce salt content of the affected object. The extraction efficiency of a poultice is largely limited by the permeability and pore size distribution of both the poultice and the substrate [4, 5]. Since the pore size distribution of a poultice can vary with its moisture content mainly due to its shrinkage during drying , it is difficult to tailor the pore size distribution of a poultice precisely to those required for optimum salt extraction efficiency. Also, the effective depth of salt extraction achieved using poultices in the case of building materials is often low, e.g. for fired-clay brick it is of the order of 20 mm . Diurnal fluctuations in temperature and relative humidity are additional factors that can significantly affect the performance of a poultice . It has been proposed that the limitations posed by poultices can be overcome by using electrokinetics as an alternative desalination method for brick masonry .
The electrokinetic method aims to remove ionic species from the zone of deterioration by an externally applied electric field. Although the electrokinetic technique has already been used to decontaminate soil  and concrete , its application in the desalination and dehumidification of masonry bricks and walls is a recent topic [6, 9, 10, 11, 12]. During the electrokinetic desalination of masonry bricks, electromigration of salt ions is considered the dominant transport mechanism and the contribution of both diffusion and advection transport processes is usually neglected. In this study the effects of both diffusion and advection in addition to electromigration on the transport of salt ions through fired-clay bricks under an applied electric field are also considered.
Until now, the effect of electrokinetic desalination has been investigated using destructive methods. These methods are limited by their inability to measure variations in moisture and salt content in porous media during the salt extraction process. Moreover, the spatial resolution is restricted (being generally in the order of 10 mm). Consequently, for this study non destructive measurements were performed during the electrokinetic desalination of building materials by using Nuclear magnetic resonance (NMR). It has been shown that NMR is a powerful technique for measuring the combined transport of moisture and Na ions in the building materials .
The objective of this work is to measure the contribution of diffusion, advection and electromigration to the overall salt transport under the influence of concentration, hydraulic and potential gradients respectively.
2 Transport processes during electrokinetic desalination
During electrokinetic desalination an applied electric field is used to mobilize salt ions from the depth of the porous material towards its surface. The removal of salt ions from the material’s surface is then typically accomplished using aqueous poulticing methods. For this purpose during electrokinetic desalination poultices in the form of wet sorbent materials (e.g. sponges, clays etc.) are applied to the surface of the porous material together with the electrodes. However, the migration of salt ions towards the surface of the porous material and the applied poultice establish a concentration gradient within the material, and thereby can induce ion diffusion. Moreover, salt can also be transported by hydraulic gradients, caused by any difference in water level across the specimen. Finally, evaporative drying of the initially wet sorbent materials can also result in salt transport due to capillary advection.
A brief description of each of the transport processes (diffusion, advection, electromigration), in saturated porous media, relevant to this study is given below.
2.4 Total mass flux
In order to study the variation in moisture and salt concentration within the brick specimen under the effect of diffusion, advection and electromigration two types of sample holders were designed to be used in the NMR set-up. The detail of both the electrokinetic and NMR set-ups is given below.
3.1 Electrokinetic set-up
In order to study the transport of Na ions through fired-clay bricks three types of experiments were performed. First, the influence of concentration gradient, i.e. diffusion, on the transport of Na ions was studied by applying sponges, soaked with de-mineralized water, to the uncovered faces of the brick. Secondly, the transport of Na ions under the effect of a hydraulic gradient (i.e. advection) was observed by introducing a constant height difference in the water level of 2 cm across the sample. The combined effect of both the hydraulic and potential gradients i.e. advection–electromigration on the salt transport was also studied by exposing the specimen to both the potential and hydraulic gradients simultaneously. Thirdly, the effect of the intensity of applied electric field on the migration of Na ions was studied by using platinum mesh electrodes applied together with the wet sponges across the specimen.
In both diffusion and electromigration experiments the sponges were replaced after measuring each profile (i.e. after 30 min) to avoid any back diffusion of salt ions from the sponges to the brick and to suppress any variation in pH caused by the oxidation and reduction reactions at the anode and cathode respectively. Platinum electrodes were used to avoid the introduction of secondary corrosion products. During the electromigration experiments the current I (A) across the brick was continuously monitored and it was observed that it remained approximately constant during the entire experiment with in ~10%.
3.2 NMR set-up
This NMR set up uses conventional electromagnets that produce magnetic field of 0.96 T. In order to measure quasi-simultaneously both the hydrogen and Na profiles during the desalination treatment, a specially designed RF circuit was incorporated . A constant magnetic field gradient of 0.1 T m−1 was applied using Anderson coils, giving a one-dimensional spatial resolution of 1 and 4 mm for the hydrogen and Na respectively. First the moisture content in the small region of the sample near the centre of the RF coil is measured. Next the frequency is changed from H to Na and the Na content in that region is measured. After these two measurements the sample is moved in the horizontal direction and the moisture and Na contents are measured. This procedure is repeated until complete moisture and Na profile is measured.
4 Results and discussion
4.1 Diffusion experiment
By using the above given initial and boundary conditions in Eq. 17 the diffusion coefficient of Na ions in the brick of the order of 0.80 × 10−9 m2 s−1 is obtained by curve fitting and is in good agreement with the reported values .
4.2 Advection experiment
By taking this value for the speed of the Na front as the volumetric flux, i.e. v h, and using it in Eq. 4 the intrinsic permeability of the fired-clay brick used in this experiment, of the order of 6 × 10−13 m2, was obtained and it is comparable to the permeability value of 10−12 m2 as reported by van der Heijden et al. .
4.3 Electromigration experiments
From the slope of linear fit (Fig. 7b) the effective mobility (μ eff) of the Na ions obtained is of the order of 4.43 × 10−9 m2 s−1 V−1. However, this value is lower than the mobility derived from the diffusion experiment according to Eq. 9; which equals to 32 × 10−9 m2 s−1. This difference in mobilities might be attributed to the fact the actual potential gradients across the sample are smaller than the applied potential gradients due to polarization of electrodes.
4.4 Diffusion versus advection and electromigration
It can be seen that within the time scale of the experiments and the length of the specimen used, up to applied field strength of 1.20 V cm−1 diffusion is dominant over electromigration. At applied field strengths of 1.58 and 2 V cm−1 electromigration dominates the transport process. However, in these experiments advection was shown to be dominant over both diffusion and electromigration.
The results presented in Fig. 8 shows that, in comparison with diffusion and electromigration, salts can be efficiently removed by hydraulic advection from relatively high permeable materials like fired-clay bricks. However, the in situ removal of salts by hydraulic advection from building structures has severe limitations. Even in the case of very simple building structures made up of relatively high permeable materials i.e. fired-clay bricks it is difficult to expose the entire building to a hydraulic gradient.
In order to overcome these difficulties the injection-poultice technique has been developed for in situ use in the brickwork  where salt transport is accomplished by advection. In this method water is injected through deep holes bored into the joints of wall and the salt contaminated water is sucked back from the same side of the wall to the sponges either by capillary pressure due to drying or by pumping water using the pumps at the adjacent outer layer of the wall.
However, in the case of low permeable building materials e.g. mortar and concrete the salt transport by hydraulic advection is practically impossible due to extremely high hydraulic gradients requirement. In this type of materials electromigration is amongst the possible ways to accomplish salt extraction.
4.5 Scale analysis of transport processes
The results presented in the previous subsections show that all the transport processes i.e. diffusion, advection and electromigration can have significant effect on the transport of salt ions through fired-clay bricks. The dominance of any particular transport process can be determined by undertaking a scale analysis. The scale analysis regarding competition between diffusion–advection and diffusion-migration is made on the basis of Peclet number, Pe, defined as the ratio between the rates of transport by convection to the molecular diffusion . In this study the Pe has been used to determine the competition between the processes that lead to directional mass transport (i.e. advection and electromigration) to the random mass transport i.e. diffusion. In the case of advection-electromigration, where both the processes lead to a directional mass transport, the scale analysis is made by introducing a new dimensionless number, ξ. In this analysis the characteristic length is the length of the specimen (i.e. 6 cm) that was used during the experiments.
4.5.1 Diffusion versus advection
For diffusion to be in equilibrium with advection i.e. for Pe = 1, v h should be equal to 1.32 × 10−8 m s−1, which corresponds to height difference of less than 0.13 mm. In order for diffusion to dominate over advection, i.e. for Pe ≪ 1, the difference in water level across the brick should be much less than 0.1 mm.
At first such a significant effect on salt transport due to a relatively low pressure difference of 200 Pa caused by a 2 cm height difference across the fired-clay bricks seems to be unrealistic. However, depending on the permeability of materials even in the presence of extremely small pressure differences, as it is shown in the calculation for Peclet number, advection can dominate the diffusion transport. Poupeleer et al.  while performing diffusion experiments noted that the density differences in liquid in response to the variation in salt concentration in different compartments of a diffusion cell produced enough hydrostatic pressure differences across a ceramic brick to make advection a dominant transport mechanism.
4.5.2 Diffusion versus electromigration
4.5.3 Advection versus electromigration
The corresponding Na front positions when the hydraulic and potential gradients are applied in opposite directions are represented by open triangles as shown in Fig. 9b. This figure also includes the front position of Na when hydraulic and potential gradients are applied in the same direction (open circles) and when only hydraulic gradient is applied (solid squares). In the case of only advection same results are again used that have already been presented in Fig. 5b. For both the advection-electromigration experiments the data represented by solid squares represent the position of Na front when no electric field was applied across the specimen. Although the removal of Cl ions is also important as they are responsible for the corrosion of e.g. reinforcements in concrete but the present study is limited to Na ions as in our NMR setup the sensitivity for Cl is too low.
It is possible to measure the transport of Na ions during electrokinetic desalination non-destructively using NMR. In the case of electromigration the salt (Na ions) removal rate is proportional to the applied electric field across the non-reactive porous materials. For electromigration to be dominant over diffusion a minimum level of electric field is necessary. Below this threshold electric field, diffusion is dominant over electromigration. Salt (Na ions) transport by advection is quite significant in the case of fired-clay bricks. The transport of Na ions can be enhanced if hydraulic and potential gradients are applied in the same direction, but if applied in opposing directions the rate of salt (Na ions) removal can be reduced or completely halted.
The results presented in this article and the conclusions drawn are based on relatively short term measurements (i.e. up to three and half hours). Moreover, the transport of salt ions only for one type of building materials i.e. fired-clay brick has been studied. In order to get better understanding about the role played by diffusion, advection and electromigration on the ionic transport the long term measurements on different types of porous building materials will be part of the future study.
Thanks are due to Hans Dalderop and Jef Noijen for providing technical assistance, and also to Gijs van der Heijden for much useful discussion. We are grateful to the Higher Education Commission, Pakistan and Nuffic, the Netherlands for their financial support and administrative assistance during this project.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
- 1.Auras M (2008) Salt weathering on buildings and stone sculptures. In: Proceedings from the international conference, 22–24 Oct 2008, Technical University of Denmark, The National Museum Copenhagen, CopenhagenGoogle Scholar
- 2.Sawdy A, Heritage A, Pel L (2008) Salt weathering on buildings and stone sculptures. In: Proceedings from the international conference, 22–24 Oct 2008, Technical University of Denmark, The National Museum Copenhagen, CopenhagenGoogle Scholar
- 3.Verges-Belmin V, Siedel H (2005) Desalination of masonries and monumental sculptures by poulticing: a review. Restor Build Monum 11(6):1–18Google Scholar
- 5.Sawdy A, Lubelli B, Voronina V, Pel L (2010) Optimizing the extraction of soluble salts from porous materials by poultices. Stud Conserv 55(1):26–40Google Scholar
- 8.Bertolini L, Yu SW, Page CL (1996) Effects of electrochemical chloride extraction on chemical and mechanical properties of hydrated cement paste. Adv Cem Res 8(31):93–100Google Scholar
- 14.Atkins P, de Paula J (2002) Atikin’s physical chemistry. Oxford University Press, LondonGoogle Scholar
- 21.Abragam A (1961) The principles of nuclear magnetism. Oxford, ClarendonGoogle Scholar
- 24.Matano C (1932) On the relation between the diffusion-coefficients and concentrations of solid metals (the nickel-copper system). Jpn J Phys 8:109–113Google Scholar
- 26.Poupeleer AS, Carmeliet J, Roels S, Van Gemert D (2003) Validation of the salt diffusion coefficient in porous materials. Int J Restor Build Monum 9(6):663–682Google Scholar