Deterioration Mechanisms: Difference between revisions

From Saltwiki
Jump to navigation Jump to search
(106 intermediate revisions by 3 users not shown)
Line 1: Line 1:
<bibimport />
Autoren: [[Benutzer:AStahlbuhk|Amelie Stahlbuhk]]
 
Authors: [[user:MSteiger|Michael Steiger]], [[user:Hschwarz|Hans-Jürgen Schwarz]]  
<br><br>
<br><br>
back to [[SaltWiki:Community_portal]]  
zurück zu [[Salzwiki:Portal]]  


== Abstract  ==
== Abstract  ==


The models are presented that can lead to salt-induced damage.
The different salt crystallization processes that can lead to deterioration of porous inorganic materials are discussed.  
 
 
== Introduction  ==


== Introduction==
Much of our cultural heritage is constituted  by porous inorganic materials such as stone, brick, mortars and renders. Salts will enter these porous materials, e.g., from rising damp, and their crystallization in the pores will lead to the materials' deterioration; a process that depends on the relative humidity (RH) and temperature. When the RH decreases, water from the salt solution will be released into the atmosphere until the deliquescence relative humidity (DRH) of the salt in question is reached, at this point a saturated solution will result; while below it the salt will crystallize. If the RH increases, exceeding the DRH, the salt will absorb water vapor from the air and tend to form a saturated solution and, if the RH continues to increase, it will be further diluted. This shows that changes in RH around the DRH of the salt will lead to alternating crystallization and deliquescence cycles. The cycling around the DRH is responsible for the deterioration induced and that can be attributed to crystallization pressure; repeated cycling inducing more damage. One of the critical factors in developing an effective crystallization pressure is the formation of a supersaturated solution  <bib id="Steiger:2005"/>. Crystallization from a supersaturated solution will only occur at a lower RH or temperature than those required for forming saturated solutions. Supersaturated solutions form through water vapor evaporation, if for example, the RH drops, or by lower temperatures.  Dissolution of a metastable phase can result in a supersaturated solution, which will be discussed a following section <bib id="Steiger.etal:2008"/>. Whether the presence of salts will result in a deterioration process depends on the nature of the salt, or salt-mixture, and its properties, but mainly by the environmental conditions, including the amount of water vapor available. These mechanisms, based on theoretical models, are only applicable for porous inorganic materials.


Salts can cause damage in porous materials such as sandstone, brick and paint layers but also by reaction on suitable surfaces. Damaging processes always occur in connection with water, i.e. vice versa, without water - in whatever form - no damage processes will occur.
==Crystallization within a pore==


The damage mechanisms are described in connection with inorganic non-metallic surfaces, or in inorganic non-metallic porous systems. Damage on metallic materials are usually based on different principles and are dealt with in a separate chapter.
The concentration of a salt solution within a pore will depend on environmental factors such as relative humidity (RH) and temperature (T). When the RH drops, the concentration will increase as water vapor will be released from the solution; if the amount evaporated is such that the solubility of the salt in question is attained, then crystallization of the salt will occur. This RH is known as the deliquescence RH (DRH). Capillary transportation of the solution to the material's surface, where the main evaporation occurs, will increase the evaporation rate. As drying continues, the evaporation front will move into the interior of the porous material, as the evaporated moisture will move faster than the capillary movement to the evaporation front <bib id="Hall.etal:1984"/>. Therefore, salts transported by the solution can crystallize both on the material's surface, i.e., efflorescence, as well as in the area below the surface as subflorescence. Where crystallization will actually occur depends on the drying conditions, the composition of the salt solution in question, its concentration and the material's properties <bib id="Espinosa-Marzal.etal:2010"/>.


The damage mechanisms are based on models of the crystallization process and the changes in the physical and chemical conditions in the pore space. It is expected that in some cases more than one damage process is active.
==Historical Background==


== Overview to the topics of decay mechanisms  ==
Currently, our understanding of the models to explain the deterioration of porous materials and crystallization pressure from salts resulted from controversial discussions since the 20th century. The growth of a crystal in the presence of obstacles was referred to as linear crystallization pressure (e.g., <bib id="Becker.etal:1916"/>, <bib id="Taber:1916"/>, <bib id="Correns.etal:1939"/>, <bib id="Duttlinger.etal:1993"/>). A growing crystal, when it crystallizes from a supersaturated solution, can generate a maximum pressure when growing in one direction and reaching an obstacle; a film solution is present between the crystal surface and the obstacle to allow its ongoing growth <bib id="Taber:1916"/>. Important studies have been carried out by  <bib id="Correns:1926"/>, who classified the crystallization pressure as a function of the volume increase of the crystal (hydrostatic crystallization pressure), hydration pressure as well as those referred to as linear crystallization pressure.


In the literature (e.g. <bib id="Correns:1926" />, <bib id="Winkler:1975" />, <bib id="Duttlinger.etal:1993" />) again and again the following types of damage mechanisms are mentioned that are directly associated with salts or be enhanced by salts, which will be discussed briefly below:
However, others (<bib id="Taber:1916"/>), pointed out the influence of the inter-facial energy that should reach a given value so that the crystallization could take place. He pointed out that a growing crystal is not attached to the surface, there being a solution film between it and the surface, to allow for an ion exchange to  take place, and that can be attributed to the inter-facial energy between crystal and pore-wall (<bib id="Correns:1926"/>, <bib id="Bruhns.etal:1913"/>, <bib id="Weyl:1959"/>.


<br clear=all>  
Some experiments, where both a loaded and an unloaded alum crystal in a saturated solution were grown, showed that only the unloaded crystal grew thus requiring a re-examination of the linear crystallization pressure, and that furthermore the pressure developed could be correlated better with a volume increase rather than the pressure of the crystal. This could also be applied to the hydration pressure, since an anhydrous or a low hydrated phase would have a smaller volume than the fully hydrated one (<bib id="Bruhns.etal:1913"/>). The hydro-static pressure is given by the increased volume resulting from the crystallization of the salt, since the volume of salt and saturated solution is greater than that of the supersaturated solution. Relevant contributions regarding the hydration pressure can be found in <bib id="Mortensen:1933"/>, who realized that the hydration pressure of an anhydrate crystal could be calculated as follows;


=== [[Linear Crystallisation Pressure]] ===
Δ<i>p</i><sub>hydr</sub>=(Δ<i>nRT</i>/Δ<i>V</i><sub>m</sub>)∙ln(RH/RH<sub>eq</sub>) (Eq.1)


The linear growth pressure is the pressure to which a crystal can grow or vice versa the maximum pressure that a growing crystal can exert.
where Δ<i>n</i> is the difference in water molecules per mole of salt of the lower hydrate <i>n</i><sub>1</sub> with the higher hydrate <i>n</i><sub>2</sub>; Δ<i>V</i><sub>m</sub> is the difference between the molar volume of both phases; RH the relative humidity at which the hydration reaction occurs, and RH<sub>eq</sub> the equilibrium moisture at the temperature <i>T</i> corresponding to the hydration-dehydration equilibrium. The hydration pressure corresponds to the maximum developed pressure that a growing crystal at higher hydration can exert on the pore wall, since at a higher pressure, dehydration would result <bib id="Mortensen:1933"/>, <bib id="Steiger.etal:2014"/>.


The linear growth pressure is particularly dependent on the degree of supersaturation of the salt solution. The calculations in most of the literature go back to Correns and Steinborn <bib id="Correns.etal:1939" /> .  
Both Correns and Steinborn (<bib id="Correns.etal:1939"/>) also studied the "crystallization pressure". For the case of the "linear growth pressure" they gave a formula for the chemical potentials for which the degree of supersaturation of the solution defined the intensity of the developed pressure, as follows:


<br clear=all>
<i>p</i>=(<i>RT</i>/<i>V</i><sub>m</sub>)∙ln<i>S</i>=(<i>RT</i>/<i>V</i><sub>m</sub>)∙ln(<i>c</i>/<i>c</i><sub>0</sub>)  (Eq. 2)


=== [[Porenabhängiger Crystallisation Pressure]]  ===
where ∆<i>p</i> is the crystallization pressure; <i>R</i> the gas constant; <i>T</i> the temperature; <i>V</i><sub>m</sub> the molecular volume of the salt; <i>S</i> the supersaturation; <i>c</i> the concentration of the supersaturated solution, and <i>c<sub>0</sub></i> the concentration of the saturated solution.


Crystals in the large pores grow at the expense of those in the small pores. The crystals in the large pores grow at the lower chemical potential first until they fill up the pore space  and then even further until its chemical potential is increased by increasing the pressure in the large pore at the level of the small pore. Consequently, at a pore distribution with two distinct maxima damage caused by crystallization of salts can be expected <bib id="Snethlage:1984" />.
Since their various experiments using different crystal surfaces always produced lower values than the theoretical ones, they attributed this to the different inter-facial energies. In subsequent equations, the crystallization pressure was also given as a function of the supersaturation, which allowed different ways of expressing them (e.g., <bib id="Goranson:1940"/>, <bib id="Buil:1983"/>).
<br clear=all>


=== [[Hydrostatic Crystallisation Pressure]]  ===
Everett (<bib id="Everett:1961"/>) considered the crystallization of ice and suggested a theory through which the crystallization pressure was the consequence of stability differences between the larger crystals. These differences result from the disparities in curvature at the interface between the solid and liquid phases, which lead to different chemical potential for crystals with different sizes. However, he did not consider the supersaturation, his equation:


For some salts <bib id="Correns.etal:1939" />, the volume of 'salt + solution' is greater than that of the saturated solution. If this salt crystallizes in the pores and seals the pores, an additional salt crystallization produces a hydrostatic pressure on the pore walls.
<i>p</i>=<i>γ</i><sub>cl</sub>∙(d<i>A</i>/d<i>V</i>)  (Eq. 3)
<br clear=all>


=== [[Hydration Pressure]] ===
describes the pressure differences between two crystals of different size; where <i>γ</i><sub>cl</sub> is the inter-facial energy between the solid and the liquid phase, <i>A</i> is the surface, and <i>V</i> is the volume. For spherical crystals this can be rewritten  


Zahlreiche bauschädliche Salze liegen abhängig von Temperatur und relativer Luftfeuchte in verschiedenen Hydratstufen vor, d.h. es existieren von ihnen Salzphasen mit zum Teil deutlich unterschiedlichem Wassergehalt. Die Bildung der neuen Phase ist sowohl nach kompletter Auflösung des alten Kristalls, als auch durch Umbau des Kristalls im festen Zustand vorstellbar. Der Einbau von Wasser in das Kristallgitter ist mit einer Volumenzunahme verbunden und übt somit einen Druck auf eine Begrenzung aus.
∆<i>p</i>=2<i>γ</i><sub>cl</sub>∙[(1/<i>r</i><sub>1</sub>)-(1/<i>r</i><sub>2</sub>)]  (Eq. 4)
<br clear=all>  


<!--
The pressure that develops through crystal growth in the larger pores with the radius <i>r</i><sub>2</sub> is appropriate, since crystal growth in smaller pores is not thermodynamically favored. The crystal developing in a larger pore will grow as long as the chemical potential of the crystal corresponds to that of a crystal in the smaller pore <i>r</i><sub>1</sub> (<bib id="Everett:1961"/>, <bib id="Steiger:2005"/>).
=== [[Osmotic Pressure]]  ===


Durch Verdunstungs- und Durchfeuchtungsvorgänge entstehen in einem porösen und salzbelasteten System Bereiche mit unterschiedlich konzentrierten Lösungen. Stehen solche Lösungen direkt miteinander in Verbindung, versucht das Lösungsmittel durch Diffusion von den Bereichen mit niedriger Konzentration in die Bereiche mit höherer Konzentration einen Konzentrationsausgleich herbeizuführen. Dadurch kann ein "osmotischer Druck" erzeugt werden.  
The two approaches of Correns and Everett were considered contradictory deterioration mechanisms for many years, since one considered supersaturation and the other the crystal curvature as responsible for the crystallization pressure developed. Both approaches were followed in parallel, and that of Correns was criticized by its very high supersaturation. In the case of the Everett theory, because data for the reliable pore sizes were available, it was reconsidered, which was not the case for the supersaturation <bib id="Steiger:2005"/>.


<br clear=all>
==Crystallization Pressure==
-->
=== [[Chemomechanical Effects]]  ===


Die Eigenschaften eines Stoffes sind nicht nur von seiner Zusammensetzung abhängig, sondern auch von seinen Umgebungsbedingungen. Temperatur, Druck und auch Stoffe mit denen er in Kontakt steht und kommt, beeinflussen seine Eigenschaften. Durch Salzlösungen können infolge der Veränderungen der elektrischen Eigenschaften diese ein signifikant anderes Verhalten zeigen als bei Abwesenheit von Salzen. Insbesondere das [[Zeta-Potential]], also das die elektrische Doppelschicht kennzeichnende elektrische Potential, spielt in diesem Zusammenhang eine Rolle. Für viele Materialien besteht eine enge Wechselbeziehung zwischen Zeta-Potential und verschiedenen mechanischen Eigenschaften derart, dass z.B. Härte, Bruchzähigkeit oder Festigkeit maximal werden, wenn das Zeta-Potential verschwindet. Über diesen Mechanismus können Salze das Verhalten von Materialien beeinflussen. Dabei können in der Doppelschicht bei Nichtmetallen so hohe Feldstärken entstehen, dass die Beweglichkeit von Strukturfehlern (Versetzungen, Risse) beeinflusst wird. Dadurch werden auch die technischen Eigenschaften verändert, ohne dass spezifische Reaktionen mitwirken müssen.  
Deterioration of a porous material from salt crystallization can only occur when the crystal(s) grows in a supersaturated solution against an obstacle, such as a pore wall (e.g., <bib id="Correns.etal:1939"/>). As the crystal grows further, a mechanical load, i.e., strain, is applied to the porous material. Another requirement for crystal growth against a wall is the presence of a solution film between the crystal and the wall <bib id="Weyl:1959"/> (a thickness of about 1 nm <bib id="Scherer:1999"/>) through which the ion transport can occur. The reason for the presence of this solution film are the repulsive forces between the crystal and the pore wall, otherwise there would be no further growth <bib id="Steiger:2005"/>. In tests where a crystal was clamped between two glass-plates to evaluate the developed force, the presence of the solution film between crystal and glass-plates could be visually observed and its relevance confirmed <bib id="Desarnaud.etal:2016"/>. The repulsive pressure between the two components defines the maximal crystallization pressure, since if this value is increased a direct contact between crystal and wall would result thus ending the crystal growth. Since the repulsive pressure has a slight dependence on the wetting properties and the surface charge, this implies that different salts and porous materials will develop diverse crystallization pressures <bib id="Desarnaud.etal:2016"/>.


<br clear=all>
A crystal growing in a pore is exposed to anisotropic pressure. Those faces/surfaces of the crystal which grow towards the pore-wall, are subjected to a higher pressure than the free surfaces in contact with the solution and subjected to its hydrostatic pressure. The difference between these two pressures is the crystallization pressure. Since crystal solubility is dependent on the applied pressure and increases with it, the free, unloaded surfaces/faces and those that are loaded, i.e., under pressure, will have different solubilities. This means, that comparing the concentration of the solution at the loaded face with that found by the free surface, the latter will be supersaturated <bib id="Steiger:2005"/>. The equation for the crystallization pressure can be given as:   
 
∆<i>p</i>=(<i>RT</i>/<i>V</i><sub>m</sub>)∙ln(<i>a</i>/<i>a</i><sub>0</sub>) (Eq. 5).
 
where ∆<i>p</i> is the crystallization pressure, i.e., the difference between the pressure on a loaded face <i>p</i><sub>c</sub> and the hydrostatic pressure of the free faces <i>p</i><sub>l</sub>; <i>a</i> is the activity of the supersaturated solution, and <i>a</i><sub>0</sub> that of the saturated solution. By using activities the non-ideal behaviour of the salt solutions is taken into account <bib id="Steiger:2005"/>.
 
This equation shows a significant similarity to that suggested by Correns and Steinborn (<bib id="Correns.etal:1939"/>), however, we are considering the non-ideal behaviour of salt solutions by using the activity instead of concentrations. Furthermore, Correns neglected the anisotropic pressure to which the crystal is subjected.
 
A pressure develops on the pore structure when the unloaded crystal face is in contact with the supersaturated solution, since the supersaturation will be decreased by the crystal growth of the unloaded crystal face (as long as they are available). This reduces the supersaturation and the pressure on the loaded face cannot be sustained. Since the original solution was saturated with regards to the loaded face, it will fall below saturation which will result in the dissolution of the loaded face thus reducing the pressure to maintain equilibrium. Both these effects lead to short-term high crystallization pressures that can only operate under supersaturated conditions. Therefore, the pressure build-up through crystallization does not occur under equilibrium conditions; it depends on a kinetic and dynamic process that is subjected to diffusion and crystal growth rate of unloaded crystal faces (<bib id="Steiger:2009"/>, <bib id="Scherer:2004"/>).
 
==Clarification==
 
When a salt dissolves, the following equation applies:
 
M<sub><i>ν</i>M</sub>X<sub><i>ν</i>X</sub>∙<i>ν</i><sub>0</sub>H<sub>2</sub>O⇌<i>ν</i><sub>M</sub>M<sup><i>z</i>M+</sup>+<i>ν</i><sub>X</sub>X<sup><i>z</i>X-</sup>+<i>ν</i><sub>0</sub> H<sub>2</sub>O    (Eq. 6)
 
where M are the cations; X the anions, <i>ν</i> the number of ions M and X; <i>z</i> the charge of the corresponding ions and <i>ν</i><sub>0</sub> the number of water molecules.
 
The activity of the dissolved salt is <i>a</i>=<i>a</i><sub>M</sub><sup><i>ν</i><sub>M</sub></sup>∙<i>a</i><sub>X</sub><sup><i>ν</i><sub>X</sub></sup>∙<i>a</i><sub>w</sub><sup><i>ν</i><sub>0</sub></sup>, the ion activity product, where <i>a</i><sub>w</sub> is the water activity.
 
Expressing the activity as the activity coefficient and the molality as <i>a</i><sub>i</sub>=<i>γ</i><sub>i</sub>∙(<i>m</i><sub>i</sub>/m<sup>0</sup>), where <i>a</i><sub>i</sub> corresponds to the ion activity; <i>γ</i><sub>i</sub> the activity coefficient of the ions; <i>m</i><sub>i</sub> the molality of the ions in the solution, and where <i>m</i><sup>0</sup>=1 mol/kg, then the crystallization pressure of a crystal in contact with the solution can be calculated, as long as the activity coefficients and the water activity are known. 
 
It follows then that a more detailed equation for Δ<i>p</i> can be found. In case of a single salt, it is possible to introduce an average activity coefficient 
<i>γ</i><sup>±</sup>=(<i>γ</i><sub>M</sub><sup><i>ν</i><sub>M</sub></sup><i>γ</i><sub>X</sub><sup><i>ν</i><sub>X</sub></sup>)<sup>(1/<i>ν</i>)</sup> but taking into consideration the simplification <i>m</i><sub>i</sub>=<i>ν</i><sub>i</sub>∙<i>m</i>. This then corresponds to the activity <i>a</i>=(<i>ν</i><sub>M</sub><sup><i>ν</i><sub>M</sub></sup><i>ν</i><sub>X</sub><sup><i>ν</i><sub>X</sub></sup>)∙(<i>mγ</i><sub>±</sub>)<sup>ν</sup>∙<i>a</i><sub>w</sub><sup><i>ν</i><sub>0</sub></sup>
where <i>ν</i> corresponds to the number of ions that result from the salt dissolution. From this the following equation results:
 
Δ<i>p</i>=<i>νRT</i>/<i>V</i><sub>m</sub>∙[ln(<i>m</i>/<i>m</i><sub>0</sub>)+ln(<i>γ</i><sub>±</sub>/<i>γ</i><sub>±,0</sub>)+(<i>ν</i><sub>0</sub>/<i>ν</i>)∙ln(<i>a</i><sub>w</sub>/<i>a</i><sub>w,0</sub>)]    (Eq. 7)
 
Next, the influencing factors and the individual terms of the equation will be considered in more detail. The first term in brackets is the supersaturation expressed as concentration. The second and third term in brackets consider the non-ideal behaviour of the concentrated salt solution(s). If one considers an ideal behaviour, the second term is not required, since the average activity coefficient would be 1, however, the third term will remain. The water activity decreases with increasing molality and is always smaller in salt solutions than in pure water. However, since <i>ν</i><sub>0</sub>, i.e., the number of water molecules is relevant, it will not contribute in the case of anhydrous salts. Next to <i>ν</i><sub>0</sub> in the equation, is the composition of the salt with various components (such as ions and the mentioned water molecules). If the crystallization water is discarded, and with it the term that includes the water activity, the calculated crystallization pressure will not be too large (<bib id="Steiger:2005b"/>). However, if the influence of the ions is also discarded the calculated crystallization pressure will be significantly influenced, since <i>ν</i> also appears in the term before the brackets. Hence, the number of ions is directly proportional to the crystallization pressure, and if this is not taken into account, the calculation will be considerably lower, about a factor of 2 or more (<bib id="Steiger:2005b"/>, <bib id="Flatt.etal:2007"/>, <bib id="Coussy:2006"/>). If the non-ideal behaviour is neglected, the influence on the crystallization pressure will depend on the salt and can be greater or smaller. With increasing supersaturation, the activity coefficient increases and with it the pressure, and vice-versa. Considering an hydrated salt, the term with the water activity will decrease the calculated pressure, so that a pressure increase from the activity coefficients will be compensated (<bib id="Steiger:2005b"/>). On the other hand, the molar volume of the salt is inversely proportional to the crystallization pressure. This means, that salts with a low molar volume at equivalent supersaturation will develop higher pressures than those with larger molar volumes. To be considered is that some salts cannot reach high supersaturations <bib id="Steiger:2005"/>.
 
These considerations make it clear that Correns and Steinborn (<bib id="Correns.etal:1939"/>)in their study did not consider the non-ideal behaviour of the salt, but more relevantly, they neglected the composition of the salt, since the influence of the ion number was not considered and therefore the calculated crystallization pressures were too low (<bib id="Steiger:2005b"/>, <bib id="Flatt.etal:2007"/>, <bib id="Coussy:2006"/>).
 
==Crystallization in small pores==
 
In the case of small pores the effect of the crystal size on their solubility has to be considered. The smaller the crystal size, the higher the solubility, so that crystals in small pores require higher concentrations for their growth. The influence of the size of spherical crystals on their solubility can be calculated by the following equation (Ostwald-Freundlich equation):
 
ln<i>a</i><sub>0</sub>/<i>a</i><sub>∞</sub>=(2<i>γ</i><sub>cl</sub><i>V</i><sub>m</sub>)/<i>rRT</i> (Eq. 8)
 
where <i>a</i><sub>0</sub> and <i>a</i><sub>∞</sub> are the thermodynamic solubility products of the smaller crystal with regards to an infinitely large crystal; 
<i>γ</i><sub>cl</sub> is the interfacial energy between the crystal and the solution, and <i>r</i> the radius of the crystal <bib id="Steiger:2005b"/>.
 
In the case of a round pore having a small cylindrical entrance, a model for two dissimilar sized pores is required. If the solution is saturated with regards to the smaller crystal at the pore entry <i>r</i><sub>1</sub>, then it will be supersaturated with regards to the larger crystal in the spherical pore <i>r</i><sub>2</sub>. Therefore the larger crystal will continue its growth and apply pressure to the pore wall, until equilibrium is reached. Since the required concentration at the free surface (the small crystal) is greater than for the large crystal, then growth can only proceed under pressure against the pore-wall. Applying the crystallization equation for pressure using the saturation activity of the larger crystal (instead of the activity of the saturated solution), with that of the smaller crystal at the pore entry (instead of the solution activity), and the last equation for solubility dependence, the following equation for calculating the crystallization pressure is obtained: 
   
   
=== [[Dynamic synergetic Effects]]  ===
∆<i>p</i>=2<i>γ</i><sub>cl</sub>∙[(1/<i>r</i><sub>1</sub>)-(1/<i>r</i><sub>2</sub>)] (Eq. 4)
 
The equation corresponds to that of <bib id="Everett:1961"/>; it is a special case for a given geometry based on a pressure build-up following supersaturation. Thus, the equations of Correns and Everett are interconvertible, i.e., exchangeable, and describe the same deterioration mechanism for specific geometries, and therefore are not contradictory <bib id="Steiger:2005"/>. In the case of the described geometry, the supersaturation will increase with respect to a decreasing pore entry. For other geometries, other equations for the crystallization pressure can be developed (<bib id="Steiger:2005"/>, <bib id="Scherer:1999"/>, <bib id="Flatt:2002"/>, <bib id="Steiger:2006"/>.
 
Also for smaller pores, the pressure will only develop when the pore solution is supersaturated with respect to free surfaces, i.e., an unloaded surface. For smaller pores this is the consequence of the different solubility presented by the curving of surfaces. The higher the curvature, the higher will the solubility be and therefore the supersaturation with respect to the less curved faces where crystal growth occurs. In this case, the supersaturation and the crystallization pressure will be in equilibrium, so that they can operate for longer periods. Nonetheless, it is critical to take into account that the influence of crystal size on its solubility will only be relevant for radius <0.1 µm (<bib id="Steiger:2005"/>) and to attain higher pressures a pore diameter of <10-50 nm (<bib id="Steiger:2009"/>) is required, and this occurrence happens only for some construction materials. For larger pores the crystallization pressure is not in equilibrium (see above). In this case, the slow evaporation will result in highly concentrated solutions when they are not in contact with all crystal faces. If there is no sufficient contact with a free surface, the supersaturation can work for longer time, and the pressures maintained (<bib id="Steiger:2005b"/>). High supersaturation can be achieved through fast evaporation. If the diffusion of ions in the solution is interrupted, this may result in growth at loaded faces, when free faces are no longer in contact with the solution (<bib id="Flatt:2002"/>).
 
==Special case of hydration pressure==


Beim Austrocknen von Salzlösungen bleiben i.a. poröse Filme zurück, die Wasser oder Lösung transportieren können. Diese porösen Strukturen sind infolge ihrer hohen spezifischen Oberfläche extrem hygroskopisch und schon Änderungen der Luftfeuchtigkeit, in jedem Fall aber direkte Wasserzufuhr, ändern auf der Oberfläche der Porenwandung aufsitzende Salzstrukturen in Größe und Form. Diese Deformationen führen zu Schubspannungen an der Substratoberfläche, die zeitlich veränderlich sind. Salzinduzierte Schäden sind nach dieser Vorstellung <bib id="Puehringer:1983" /> das Resultat von vielen aufeinander folgenden Belastungszyklen mit verhältnismäßig kleiner Intensität.  
Some salts can crystallize with different degrees of hydration, for example, magnesium sulfate (MgSO<sub>4</sub>∙<i>n</i>H<sub>2</sub>O), where n ranges from 1,4-7. The different hydration number results in that the higher hydrated phases have a larger molecular volume, which were held responsible for exerting pressure on the pore structure for a long time(e.g.,<bib id="Correns:1926"/>). Two different type of hydration reactions can be established.


<br clear=all>  
One corresponds to the hydration via water vapor, which is dependent on the air moisture in the environment. If the RH value is below the deliquescence moisture of the lower hydrate, or the anhydrous phase, then the hydration occurs as a solid phase reaction. In this case, the maximum pressure that the hydrating crystal can induce on the pore wall can be calculated with the equation for hydration pressure given by Mortensen <bib id="Mortensen:1933"/> (see Historical Background section). During crystallization, the supersaturation of the solution is the driving force for the exerted pressure, while during hydration via water vapor (RH<DRH of the lower hydrate) it corresponds to the supersaturation of the water vapor in the environment in relation to the equilibrium water vapor pressure of the hydrated phase at the given temperature. In this case, if the equation of Mortensen is used, it has to be considered that there is a maximal RH above which the equation is no longer valid. If the moisture in the air is above the deliquescence moisture of the higher hydrated phase, then no hydration pressure can develop <bib id="Steiger.etal:2014"/>.  
=== [[Hygric Effects]] ===


Viele poröse Baustoffe erfahren in Abhängigkeit vom Feuchtegehalt ein Dehnen und Schwinden, die wenn sie in steigem Wechsel auftreten, zu starken Schäden führen können. In welchem Ausmaß diese Prozesse erfolgen, hängt von der Zusammensetzung, der Porenstruktur und dem Feuchteangebot ab. Sind diese Stoffe zusätzlich mit hygroskopischen Salzen belastet, so stellt sich eine neue, im Vergleich zum salzfreien Stoff, höhere Gleichgewichtsfeuchte ein, die von der Salzart und -konzentration abhängt (siehe auch <bib id="Vogt.etal:1993" />).
If the RH during moisture absorption is above the DHR of the lower hydrated phase, then hydration will continue via other mechanisms. The lower hydrate forms a saturated solution, out of which a higher hydrate will crystallize out (dissolution and recrystallization process). Also in this case the volume increase is not the reason for the pressure increase on the porous structure, but the crystallization of the higher hydrate phase from a supersaturated solution. Thus, this mechanism is a special case of the crystallization pressure as a result of crystallization from supersaturated solutions. And this mechanism is also responsible for the pressure developed through moisture absorption of liquid water in a porous system where an anhydrous or lower hydrated phase is present <bib id="Steiger:2003b"/>.
<br clear=all> 


=== [[Thermic Effects]] ===
For instance, the system [[sodium sulfate]]-water is a good example. If a porous material is moistened with a solution of [[thenardite]] at room temperature, then a saturated solution with regards to [[thenardite]] will form. Looking at the phase diagram of the system, it is evident that the solution will be saturated with regards to [[mirabilite]], and that the supersaturation increases with lowering temperature. When [[mirabilite]] crystallizes out of a highly supersaturated solution then, if sufficient pores are filled, high pressures can develop in the porous structure. This can be observed during wetting, as well as during wetting of the anhydrous or lower hydrated phases at RH up to or above the DRH, where the hydrated phase can form a supersaturated solution. In general terms, this occurs through the dissolution of a metastable phase to allow the formation of a supersaturaed solution with regards to a more stable phase, so that during its crystallization high pressures can develop <bib id="Steiger.etal:2008"/>.


Da feuchte Baustoffe deutlich bessere Wärmeleiter sind als trockene Baustoffe, kann eine erhöhte Materialfeuchte zu einer Veränderung der Temperaturunterschiede in einem Baustoff und damit zu einer Veränderung des thermischen Dehnens und Schwindes führen. Da ein Temperaturausgleich schneller erfolgt, werden die thermischen Spannungen in der Regel abnehmen.
==Consideration of crystallization pressure within building deterioration==


=== [[New  findings according to decay mechanisms]] ===
Salt crystallization will induce damage and deterioration in porous inorganic materials. Salts, as well as their solution(s), are present in the buildings environment, e.g., rising damp, so that there is a direct interchange between them. Changes in RH will result in cycling between crystallization and dissolution, especially when they occur around the DRH. Rain, as well as condensation induced by temperature changes will also result in the introduction of water or hydration of the salts already present in the material. Some salts will change phases with only temperature variation, depending on their stability within the temperature range, for example, [[mirabilite]]. However, in buildings the concentration of salts seldom reaches those that were used in laboratory studies to understand the damage mechanism.


In den letzten Jahren wurde durch intensive Forschungsrarbeiten diese bisherigen Modelle der Schadensmechanismen durch Salze überprüft und mussten z.T. modifiziert werden.
Furthermore, for the case of crystallization pressure the points discussed in the Clarification section are valid. Small pores, that are necessary for an equilibrium of the crystallization pressure, are not frequently found in most construction materials, but can be found in cement. Therefore, crystallization pressure in real buildings is not found in equilibrium, but rather as a dynamic process (<bib id="Steiger:2009"/>, <bib id="Scherer:2004"/>. The high supersaturation conditions occur only in short events and are reduced by the crystallization on free surfaces.


== Literature  ==
== Literature  ==


<bibprint />  
<biblist/>  


[[Category:Schadensmechanismen]] [[Category:MSteiger]] [[Category:HSchwarz]] [[Category:R-MSteiger]] [[Category:Review]][[Category:InProgress]]
[[Category:Schadensmechanismen]] [[Category:Stahlbuhk,Amelie]][[Category:R-MSteiger]]

Revision as of 18:33, 26 September 2019

Autoren: Amelie Stahlbuhk

zurück zu Salzwiki:Portal

Abstract[edit]

The different salt crystallization processes that can lead to deterioration of porous inorganic materials are discussed.


Introduction[edit]

Much of our cultural heritage is constituted by porous inorganic materials such as stone, brick, mortars and renders. Salts will enter these porous materials, e.g., from rising damp, and their crystallization in the pores will lead to the materials' deterioration; a process that depends on the relative humidity (RH) and temperature. When the RH decreases, water from the salt solution will be released into the atmosphere until the deliquescence relative humidity (DRH) of the salt in question is reached, at this point a saturated solution will result; while below it the salt will crystallize. If the RH increases, exceeding the DRH, the salt will absorb water vapor from the air and tend to form a saturated solution and, if the RH continues to increase, it will be further diluted. This shows that changes in RH around the DRH of the salt will lead to alternating crystallization and deliquescence cycles. The cycling around the DRH is responsible for the deterioration induced and that can be attributed to crystallization pressure; repeated cycling inducing more damage. One of the critical factors in developing an effective crystallization pressure is the formation of a supersaturated solution [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
. Crystallization from a supersaturated solution will only occur at a lower RH or temperature than those required for forming saturated solutions. Supersaturated solutions form through water vapor evaporation, if for example, the RH drops, or by lower temperatures. Dissolution of a metastable phase can result in a supersaturated solution, which will be discussed a following section [Steiger.etal:2008]Title: Crystallization of sodium sulfate phases in porous materials: The phase diagram Na2SO4–H2O and the generation of stress
Author: Steiger, Michael; Asmussen, Sönke
Link to Google Scholar
. Whether the presence of salts will result in a deterioration process depends on the nature of the salt, or salt-mixture, and its properties, but mainly by the environmental conditions, including the amount of water vapor available. These mechanisms, based on theoretical models, are only applicable for porous inorganic materials.

Crystallization within a pore[edit]

The concentration of a salt solution within a pore will depend on environmental factors such as relative humidity (RH) and temperature (T). When the RH drops, the concentration will increase as water vapor will be released from the solution; if the amount evaporated is such that the solubility of the salt in question is attained, then crystallization of the salt will occur. This RH is known as the deliquescence RH (DRH). Capillary transportation of the solution to the material's surface, where the main evaporation occurs, will increase the evaporation rate. As drying continues, the evaporation front will move into the interior of the porous material, as the evaporated moisture will move faster than the capillary movement to the evaporation front [Hall.etal:1984]The entry doesn't exist yet.. Therefore, salts transported by the solution can crystallize both on the material's surface, i.e., efflorescence, as well as in the area below the surface as subflorescence. Where crystallization will actually occur depends on the drying conditions, the composition of the salt solution in question, its concentration and the material's properties [Espinosa-Marzal.etal:2010]Title: Advances in Understanding Damage by Salt Crystallization
Author: Espinosa-Marzal, Rosa M.; Scherer, George W.
Link to Google Scholar
.

Historical Background[edit]

Currently, our understanding of the models to explain the deterioration of porous materials and crystallization pressure from salts resulted from controversial discussions since the 20th century. The growth of a crystal in the presence of obstacles was referred to as linear crystallization pressure (e.g., [Becker.etal:1916]Title: Notes on the Linear Force of Growing Crystals
Author: Becker, G.F.; Day, A.L.
Link to Google Scholar
, [Taber:1916]Title: The Growth of Crystals under External Pressure
Author: Taber, Stephen
Link to Google Scholar
, [Correns.etal:1939]Title: Experimente zur Messung und Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.; Steinborn, W.
Link to Google Scholar
, [Duttlinger.etal:1993]Title: Salzkristallisation und Salzschadensmechanismen
Author: Duttlinger, Werner; Knöfel, Dietbert
Link to Google Scholar
). A growing crystal, when it crystallizes from a supersaturated solution, can generate a maximum pressure when growing in one direction and reaching an obstacle; a film solution is present between the crystal surface and the obstacle to allow its ongoing growth [Taber:1916]Title: The Growth of Crystals under External Pressure
Author: Taber, Stephen
Link to Google Scholar
. Important studies have been carried out by [Correns:1926]Title: Über die Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.
Link to Google Scholar
, who classified the crystallization pressure as a function of the volume increase of the crystal (hydrostatic crystallization pressure), hydration pressure as well as those referred to as linear crystallization pressure.

However, others ([Taber:1916]Title: The Growth of Crystals under External Pressure
Author: Taber, Stephen
Link to Google Scholar
), pointed out the influence of the inter-facial energy that should reach a given value so that the crystallization could take place. He pointed out that a growing crystal is not attached to the surface, there being a solution film between it and the surface, to allow for an ion exchange to take place, and that can be attributed to the inter-facial energy between crystal and pore-wall ([Correns:1926]Title: Über die Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.
Link to Google Scholar
, [Bruhns.etal:1913]Title: Über die sogenannte "Kristallisationskraft"
Author: Bruhns, W.; Mecklenburg, W.
Link to Google Scholar
, [Weyl:1959]Title: Pressure Solution and the Force of Crystallisation - A Phenomenological Theory
Author: Weyl, Peter K.
Link to Google Scholar
.

Some experiments, where both a loaded and an unloaded alum crystal in a saturated solution were grown, showed that only the unloaded crystal grew thus requiring a re-examination of the linear crystallization pressure, and that furthermore the pressure developed could be correlated better with a volume increase rather than the pressure of the crystal. This could also be applied to the hydration pressure, since an anhydrous or a low hydrated phase would have a smaller volume than the fully hydrated one ([Bruhns.etal:1913]Title: Über die sogenannte "Kristallisationskraft"
Author: Bruhns, W.; Mecklenburg, W.
Link to Google Scholar
). The hydro-static pressure is given by the increased volume resulting from the crystallization of the salt, since the volume of salt and saturated solution is greater than that of the supersaturated solution. Relevant contributions regarding the hydration pressure can be found in [Mortensen:1933]Title: Die 'Salzprengung' und ihre Bedeutung für die regionalklimatische Gliederung der Wüsten
Author: Mortensen, Hans
Link to Google Scholar
, who realized that the hydration pressure of an anhydrate crystal could be calculated as follows;

Δphydr=(ΔnRTVm)∙ln(RH/RHeq) (Eq.1)

where Δn is the difference in water molecules per mole of salt of the lower hydrate n1 with the higher hydrate n2; ΔVm is the difference between the molar volume of both phases; RH the relative humidity at which the hydration reaction occurs, and RHeq the equilibrium moisture at the temperature T corresponding to the hydration-dehydration equilibrium. The hydration pressure corresponds to the maximum developed pressure that a growing crystal at higher hydration can exert on the pore wall, since at a higher pressure, dehydration would result [Mortensen:1933]Title: Die 'Salzprengung' und ihre Bedeutung für die regionalklimatische Gliederung der Wüsten
Author: Mortensen, Hans
Link to Google Scholar
, [Steiger.etal:2014]Title: Weathering and Deterioration
Author: Steiger, Michael; Charola A. Elena; Sterflinger, Katja
Link to Google Scholar
.

Both Correns and Steinborn ([Correns.etal:1939]Title: Experimente zur Messung und Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.; Steinborn, W.
Link to Google Scholar
) also studied the "crystallization pressure". For the case of the "linear growth pressure" they gave a formula for the chemical potentials for which the degree of supersaturation of the solution defined the intensity of the developed pressure, as follows:

p=(RT/Vm)∙lnS=(RT/Vm)∙ln(c/c0) (Eq. 2)

where ∆p is the crystallization pressure; R the gas constant; T the temperature; Vm the molecular volume of the salt; S the supersaturation; c the concentration of the supersaturated solution, and c0 the concentration of the saturated solution.

Since their various experiments using different crystal surfaces always produced lower values than the theoretical ones, they attributed this to the different inter-facial energies. In subsequent equations, the crystallization pressure was also given as a function of the supersaturation, which allowed different ways of expressing them (e.g., [Goranson:1940]The entry doesn't exist yet., [Buil:1983]Title: Thermodynamics and Experimental Study of the Crystallization Pressure of Water Soluble Salts
Author: Buil, Michel
Link to Google Scholar
).

Everett ([Everett:1961]Title: The thermodynamics of frost damage to porous solids
Author: Everett, D.H.
Link to Google Scholar
) considered the crystallization of ice and suggested a theory through which the crystallization pressure was the consequence of stability differences between the larger crystals. These differences result from the disparities in curvature at the interface between the solid and liquid phases, which lead to different chemical potential for crystals with different sizes. However, he did not consider the supersaturation, his equation:

p=γcl∙(dA/dV) (Eq. 3)

describes the pressure differences between two crystals of different size; where γcl is the inter-facial energy between the solid and the liquid phase, A is the surface, and V is the volume. For spherical crystals this can be rewritten

p=2γcl∙[(1/r1)-(1/r2)] (Eq. 4)

The pressure that develops through crystal growth in the larger pores with the radius r2 is appropriate, since crystal growth in smaller pores is not thermodynamically favored. The crystal developing in a larger pore will grow as long as the chemical potential of the crystal corresponds to that of a crystal in the smaller pore r1 ([Everett:1961]Title: The thermodynamics of frost damage to porous solids
Author: Everett, D.H.
Link to Google Scholar
, [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
).

The two approaches of Correns and Everett were considered contradictory deterioration mechanisms for many years, since one considered supersaturation and the other the crystal curvature as responsible for the crystallization pressure developed. Both approaches were followed in parallel, and that of Correns was criticized by its very high supersaturation. In the case of the Everett theory, because data for the reliable pore sizes were available, it was reconsidered, which was not the case for the supersaturation [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
.

Crystallization Pressure[edit]

Deterioration of a porous material from salt crystallization can only occur when the crystal(s) grows in a supersaturated solution against an obstacle, such as a pore wall (e.g., [Correns.etal:1939]Title: Experimente zur Messung und Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.; Steinborn, W.
Link to Google Scholar
). As the crystal grows further, a mechanical load, i.e., strain, is applied to the porous material. Another requirement for crystal growth against a wall is the presence of a solution film between the crystal and the wall [Weyl:1959]Title: Pressure Solution and the Force of Crystallisation - A Phenomenological Theory
Author: Weyl, Peter K.
Link to Google Scholar
(a thickness of about 1 nm [Scherer:1999]Title: Crystallization in pores
Author: Scherer, George W.
Link to Google Scholar
) through which the ion transport can occur. The reason for the presence of this solution film are the repulsive forces between the crystal and the pore wall, otherwise there would be no further growth [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
. In tests where a crystal was clamped between two glass-plates to evaluate the developed force, the presence of the solution film between crystal and glass-plates could be visually observed and its relevance confirmed [Desarnaud.etal:2016]The entry doesn't exist yet.. The repulsive pressure between the two components defines the maximal crystallization pressure, since if this value is increased a direct contact between crystal and wall would result thus ending the crystal growth. Since the repulsive pressure has a slight dependence on the wetting properties and the surface charge, this implies that different salts and porous materials will develop diverse crystallization pressures [Desarnaud.etal:2016]The entry doesn't exist yet..

A crystal growing in a pore is exposed to anisotropic pressure. Those faces/surfaces of the crystal which grow towards the pore-wall, are subjected to a higher pressure than the free surfaces in contact with the solution and subjected to its hydrostatic pressure. The difference between these two pressures is the crystallization pressure. Since crystal solubility is dependent on the applied pressure and increases with it, the free, unloaded surfaces/faces and those that are loaded, i.e., under pressure, will have different solubilities. This means, that comparing the concentration of the solution at the loaded face with that found by the free surface, the latter will be supersaturated [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
. The equation for the crystallization pressure can be given as:

p=(RT/Vm)∙ln(a/a0) (Eq. 5).

where ∆p is the crystallization pressure, i.e., the difference between the pressure on a loaded face pc and the hydrostatic pressure of the free faces pl; a is the activity of the supersaturated solution, and a0 that of the saturated solution. By using activities the non-ideal behaviour of the salt solutions is taken into account [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
.

This equation shows a significant similarity to that suggested by Correns and Steinborn ([Correns.etal:1939]Title: Experimente zur Messung und Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.; Steinborn, W.
Link to Google Scholar
), however, we are considering the non-ideal behaviour of salt solutions by using the activity instead of concentrations. Furthermore, Correns neglected the anisotropic pressure to which the crystal is subjected.

A pressure develops on the pore structure when the unloaded crystal face is in contact with the supersaturated solution, since the supersaturation will be decreased by the crystal growth of the unloaded crystal face (as long as they are available). This reduces the supersaturation and the pressure on the loaded face cannot be sustained. Since the original solution was saturated with regards to the loaded face, it will fall below saturation which will result in the dissolution of the loaded face thus reducing the pressure to maintain equilibrium. Both these effects lead to short-term high crystallization pressures that can only operate under supersaturated conditions. Therefore, the pressure build-up through crystallization does not occur under equilibrium conditions; it depends on a kinetic and dynamic process that is subjected to diffusion and crystal growth rate of unloaded crystal faces ([Steiger:2009]Title: Mechanismus der Schädigung durch Salzkristallisation
Author: Steiger, Michael
Link to Google Scholar
, [Scherer:2004]Title: Stress from crystallization of salt
Author: Scherer, George W.
Link to Google Scholar
).

Clarification[edit]

When a salt dissolves, the following equation applies:

MνMXνXν0H2O⇌νMMzM++νXXzX-+ν0 H2O (Eq. 6)

where M are the cations; X the anions, ν the number of ions M and X; z the charge of the corresponding ions and ν0 the number of water molecules.

The activity of the dissolved salt is a=aMνMaXνXawν0, the ion activity product, where aw is the water activity.

Expressing the activity as the activity coefficient and the molality as ai=γi∙(mi/m0), where ai corresponds to the ion activity; γi the activity coefficient of the ions; mi the molality of the ions in the solution, and where m0=1 mol/kg, then the crystallization pressure of a crystal in contact with the solution can be calculated, as long as the activity coefficients and the water activity are known.

It follows then that a more detailed equation for Δp can be found. In case of a single salt, it is possible to introduce an average activity coefficient γ±=(γMνMγXνX)(1/ν) but taking into consideration the simplification mi=νim. This then corresponds to the activity a=(νMνMνXνX)∙(±)νawν0 where ν corresponds to the number of ions that result from the salt dissolution. From this the following equation results:

Δp=νRT/Vm∙[ln(m/m0)+ln(γ±/γ±,0)+(ν0/ν)∙ln(aw/aw,0)] (Eq. 7)

Next, the influencing factors and the individual terms of the equation will be considered in more detail. The first term in brackets is the supersaturation expressed as concentration. The second and third term in brackets consider the non-ideal behaviour of the concentrated salt solution(s). If one considers an ideal behaviour, the second term is not required, since the average activity coefficient would be 1, however, the third term will remain. The water activity decreases with increasing molality and is always smaller in salt solutions than in pure water. However, since ν0, i.e., the number of water molecules is relevant, it will not contribute in the case of anhydrous salts. Next to ν0 in the equation, is the composition of the salt with various components (such as ions and the mentioned water molecules). If the crystallization water is discarded, and with it the term that includes the water activity, the calculated crystallization pressure will not be too large ([Steiger:2005b]Title: Crystal growth in porous materials: II. The influence of crystal size
Author: Steiger, Michael
Link to Google Scholar
). However, if the influence of the ions is also discarded the calculated crystallization pressure will be significantly influenced, since ν also appears in the term before the brackets. Hence, the number of ions is directly proportional to the crystallization pressure, and if this is not taken into account, the calculation will be considerably lower, about a factor of 2 or more ([Steiger:2005b]Title: Crystal growth in porous materials: II. The influence of crystal size
Author: Steiger, Michael
Link to Google Scholar
, [Flatt.etal:2007]Title: A commented translation of the paper by C.W. Correns and W. Steinborn on crystallization pressure
Author: Flatt, Robert J.; Steiger, Michael; Scherer, George W.
Link to Google Scholar
, [Coussy:2006]The entry doesn't exist yet.). If the non-ideal behaviour is neglected, the influence on the crystallization pressure will depend on the salt and can be greater or smaller. With increasing supersaturation, the activity coefficient increases and with it the pressure, and vice-versa. Considering an hydrated salt, the term with the water activity will decrease the calculated pressure, so that a pressure increase from the activity coefficients will be compensated ([Steiger:2005b]Title: Crystal growth in porous materials: II. The influence of crystal size
Author: Steiger, Michael
Link to Google Scholar
). On the other hand, the molar volume of the salt is inversely proportional to the crystallization pressure. This means, that salts with a low molar volume at equivalent supersaturation will develop higher pressures than those with larger molar volumes. To be considered is that some salts cannot reach high supersaturations [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
.

These considerations make it clear that Correns and Steinborn ([Correns.etal:1939]Title: Experimente zur Messung und Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.; Steinborn, W.
Link to Google Scholar
)in their study did not consider the non-ideal behaviour of the salt, but more relevantly, they neglected the composition of the salt, since the influence of the ion number was not considered and therefore the calculated crystallization pressures were too low ([Steiger:2005b]Title: Crystal growth in porous materials: II. The influence of crystal size
Author: Steiger, Michael
Link to Google Scholar
, [Flatt.etal:2007]Title: A commented translation of the paper by C.W. Correns and W. Steinborn on crystallization pressure
Author: Flatt, Robert J.; Steiger, Michael; Scherer, George W.
Link to Google Scholar
, [Coussy:2006]The entry doesn't exist yet.).

Crystallization in small pores[edit]

In the case of small pores the effect of the crystal size on their solubility has to be considered. The smaller the crystal size, the higher the solubility, so that crystals in small pores require higher concentrations for their growth. The influence of the size of spherical crystals on their solubility can be calculated by the following equation (Ostwald-Freundlich equation):

lna0/a=(2γclVm)/rRT (Eq. 8)

where a0 and a are the thermodynamic solubility products of the smaller crystal with regards to an infinitely large crystal; γcl is the interfacial energy between the crystal and the solution, and r the radius of the crystal [Steiger:2005b]Title: Crystal growth in porous materials: II. The influence of crystal size
Author: Steiger, Michael
Link to Google Scholar
.

In the case of a round pore having a small cylindrical entrance, a model for two dissimilar sized pores is required. If the solution is saturated with regards to the smaller crystal at the pore entry r1, then it will be supersaturated with regards to the larger crystal in the spherical pore r2. Therefore the larger crystal will continue its growth and apply pressure to the pore wall, until equilibrium is reached. Since the required concentration at the free surface (the small crystal) is greater than for the large crystal, then growth can only proceed under pressure against the pore-wall. Applying the crystallization equation for pressure using the saturation activity of the larger crystal (instead of the activity of the saturated solution), with that of the smaller crystal at the pore entry (instead of the solution activity), and the last equation for solubility dependence, the following equation for calculating the crystallization pressure is obtained:

p=2γcl∙[(1/r1)-(1/r2)] (Eq. 4)

The equation corresponds to that of [Everett:1961]Title: The thermodynamics of frost damage to porous solids
Author: Everett, D.H.
Link to Google Scholar
; it is a special case for a given geometry based on a pressure build-up following supersaturation. Thus, the equations of Correns and Everett are interconvertible, i.e., exchangeable, and describe the same deterioration mechanism for specific geometries, and therefore are not contradictory [Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
. In the case of the described geometry, the supersaturation will increase with respect to a decreasing pore entry. For other geometries, other equations for the crystallization pressure can be developed ([Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
, [Scherer:1999]Title: Crystallization in pores
Author: Scherer, George W.
Link to Google Scholar
, [Flatt:2002]Title: Salt damage in porous materials: how high supersaturations are generated
Author: Flatt, Robert J.
Link to Google Scholar
, [Steiger:2006]Title: Crystal growth in porous materials: Influence of supersaturation and crystal size
Author: Steiger, Michael
Link to Google Scholar
.

Also for smaller pores, the pressure will only develop when the pore solution is supersaturated with respect to free surfaces, i.e., an unloaded surface. For smaller pores this is the consequence of the different solubility presented by the curving of surfaces. The higher the curvature, the higher will the solubility be and therefore the supersaturation with respect to the less curved faces where crystal growth occurs. In this case, the supersaturation and the crystallization pressure will be in equilibrium, so that they can operate for longer periods. Nonetheless, it is critical to take into account that the influence of crystal size on its solubility will only be relevant for radius <0.1 µm ([Steiger:2005]Title: Crystal growth in porous materials: I. The crystallization pressure of large crystals
Author: Steiger, Michael
Link to Google Scholar
) and to attain higher pressures a pore diameter of <10-50 nm ([Steiger:2009]Title: Mechanismus der Schädigung durch Salzkristallisation
Author: Steiger, Michael
Link to Google Scholar
) is required, and this occurrence happens only for some construction materials. For larger pores the crystallization pressure is not in equilibrium (see above). In this case, the slow evaporation will result in highly concentrated solutions when they are not in contact with all crystal faces. If there is no sufficient contact with a free surface, the supersaturation can work for longer time, and the pressures maintained ([Steiger:2005b]Title: Crystal growth in porous materials: II. The influence of crystal size
Author: Steiger, Michael
Link to Google Scholar
). High supersaturation can be achieved through fast evaporation. If the diffusion of ions in the solution is interrupted, this may result in growth at loaded faces, when free faces are no longer in contact with the solution ([Flatt:2002]Title: Salt damage in porous materials: how high supersaturations are generated
Author: Flatt, Robert J.
Link to Google Scholar
).

Special case of hydration pressure[edit]

Some salts can crystallize with different degrees of hydration, for example, magnesium sulfate (MgSO4nH2O), where n ranges from 1,4-7. The different hydration number results in that the higher hydrated phases have a larger molecular volume, which were held responsible for exerting pressure on the pore structure for a long time(e.g.,[Correns:1926]Title: Über die Erklärung der sogenannten Kristallisationskraft
Author: Correns, Carl W.
Link to Google Scholar
). Two different type of hydration reactions can be established.

One corresponds to the hydration via water vapor, which is dependent on the air moisture in the environment. If the RH value is below the deliquescence moisture of the lower hydrate, or the anhydrous phase, then the hydration occurs as a solid phase reaction. In this case, the maximum pressure that the hydrating crystal can induce on the pore wall can be calculated with the equation for hydration pressure given by Mortensen [Mortensen:1933]Title: Die 'Salzprengung' und ihre Bedeutung für die regionalklimatische Gliederung der Wüsten
Author: Mortensen, Hans
Link to Google Scholar
(see Historical Background section). During crystallization, the supersaturation of the solution is the driving force for the exerted pressure, while during hydration via water vapor (RH<DRH of the lower hydrate) it corresponds to the supersaturation of the water vapor in the environment in relation to the equilibrium water vapor pressure of the hydrated phase at the given temperature. In this case, if the equation of Mortensen is used, it has to be considered that there is a maximal RH above which the equation is no longer valid. If the moisture in the air is above the deliquescence moisture of the higher hydrated phase, then no hydration pressure can develop [Steiger.etal:2014]Title: Weathering and Deterioration
Author: Steiger, Michael; Charola A. Elena; Sterflinger, Katja
Link to Google Scholar
.

If the RH during moisture absorption is above the DHR of the lower hydrated phase, then hydration will continue via other mechanisms. The lower hydrate forms a saturated solution, out of which a higher hydrate will crystallize out (dissolution and recrystallization process). Also in this case the volume increase is not the reason for the pressure increase on the porous structure, but the crystallization of the higher hydrate phase from a supersaturated solution. Thus, this mechanism is a special case of the crystallization pressure as a result of crystallization from supersaturated solutions. And this mechanism is also responsible for the pressure developed through moisture absorption of liquid water in a porous system where an anhydrous or lower hydrated phase is present [Steiger:2003b]Title: Salts and Crusts
Author: Steiger, Michael
Link to Google Scholar
.

For instance, the system sodium sulfate-water is a good example. If a porous material is moistened with a solution of thenardite at room temperature, then a saturated solution with regards to thenardite will form. Looking at the phase diagram of the system, it is evident that the solution will be saturated with regards to mirabilite, and that the supersaturation increases with lowering temperature. When mirabilite crystallizes out of a highly supersaturated solution then, if sufficient pores are filled, high pressures can develop in the porous structure. This can be observed during wetting, as well as during wetting of the anhydrous or lower hydrated phases at RH up to or above the DRH, where the hydrated phase can form a supersaturated solution. In general terms, this occurs through the dissolution of a metastable phase to allow the formation of a supersaturaed solution with regards to a more stable phase, so that during its crystallization high pressures can develop [Steiger.etal:2008]Title: Crystallization of sodium sulfate phases in porous materials: The phase diagram Na2SO4–H2O and the generation of stress
Author: Steiger, Michael; Asmussen, Sönke
Link to Google Scholar
.

Consideration of crystallization pressure within building deterioration[edit]

Salt crystallization will induce damage and deterioration in porous inorganic materials. Salts, as well as their solution(s), are present in the buildings environment, e.g., rising damp, so that there is a direct interchange between them. Changes in RH will result in cycling between crystallization and dissolution, especially when they occur around the DRH. Rain, as well as condensation induced by temperature changes will also result in the introduction of water or hydration of the salts already present in the material. Some salts will change phases with only temperature variation, depending on their stability within the temperature range, for example, mirabilite. However, in buildings the concentration of salts seldom reaches those that were used in laboratory studies to understand the damage mechanism.

Furthermore, for the case of crystallization pressure the points discussed in the Clarification section are valid. Small pores, that are necessary for an equilibrium of the crystallization pressure, are not frequently found in most construction materials, but can be found in cement. Therefore, crystallization pressure in real buildings is not found in equilibrium, but rather as a dynamic process ([Steiger:2009]Title: Mechanismus der Schädigung durch Salzkristallisation
Author: Steiger, Michael
Link to Google Scholar
, [Scherer:2004]Title: Stress from crystallization of salt
Author: Scherer, George W.
Link to Google Scholar
. The high supersaturation conditions occur only in short events and are reduced by the crystallization on free surfaces.

Literature[edit]

[Becker.etal:1916]Becker, G.F.; Day, A.L. (1916): Notes on the Linear Force of Growing Crystals. In: Journal of Geology, 24 (4), 313-333, UrlLink to Google Scholar
[Bruhns.etal:1913]Bruhns, W.; Mecklenburg, W. (1913): Über die sogenannte "Kristallisationskraft". In: Sechster Jahresbericht des Niedersächsischen Geologischen Vereins zu Hannover, (), 92-115Link to Google Scholar
[Buil:1983]Buil, Michel (1983): Thermodynamics and Experimental Study of the Crystallization Pressure of Water Soluble Salts. In: F.H. Wittmann (eds.): Materials Science and Restoration, Lack und Chemie, Filderstadt, 373-377.Link to Google Scholar
[Correns.etal:1939]Correns, Carl W.; Steinborn, W. (1939): Experimente zur Messung und Erklärung der sogenannten Kristallisationskraft. In: Zeitschrift für Kristallografie, (101), 117-133Link to Google Scholar
[Correns:1926]Correns, Carl W. (1926): Über die Erklärung der sogenannten Kristallisationskraft. In:: Preuss. Akad. der Wissensch, Sitzungsband, de Gruyter, 81-88.Link to Google Scholar
[Coussy:2006]The entry doesn't exist yet.
[Desarnaud.etal:2016]The entry doesn't exist yet.
[Duttlinger.etal:1993]Duttlinger, Werner; Knöfel, Dietbert (1993): Salzkristallisation und Salzschadensmechanismen. In: Snethlage, Rolf (eds.): Jahresberichte Steinzerfall - Steinkonservierung 1991, Ernst & Sohn, (Berlin), 197-213.Link to Google Scholar
[Espinosa-Marzal.etal:2010]Espinosa-Marzal, Rosa M.; Scherer, George W. (2010): Advances in Understanding Damage by Salt Crystallization. In: Accounts of Chemical Research, 43 (6), 897-905, UrlLink to Google Scholar
[Everett:1961]Everett, D.H. (1961): The thermodynamics of frost damage to porous solids. In: Transactions of the Faraday Society, 57 (), 1541-1551Link to Google Scholar
[Flatt.etal:2007]Flatt, Robert J.; Steiger, Michael; Scherer, George W. (2007): A commented translation of the paper by C.W. Correns and W. Steinborn on crystallization pressure. In: Environmental Geology, 52 (2), 187-203, 10.1007/s00254-006-0509-5Link to Google Scholar
[Flatt:2002]Flatt, Robert J. (2002): Salt damage in porous materials: how high supersaturations are generated. In: Journal of Crystal Growth, 242 (3), 435-454Link to Google Scholar
[Goranson:1940]The entry doesn't exist yet.
[Hall.etal:1984]The entry doesn't exist yet.
[Mortensen:1933]Mortensen, Hans (1933): Die 'Salzprengung' und ihre Bedeutung für die regionalklimatische Gliederung der Wüsten. In::, Justus Perthes, 130-135.Link to Google Scholar
[Scherer:1999]Scherer, George W. (1999): Crystallization in pores. In: Cement and Concrete Research, 29 (), 1347-1358Link to Google Scholar
[Scherer:2004]Scherer, George W. (2004): Stress from crystallization of salt. In: Cement and Concrete Research, 34 (4), 1613–1624, UrlLink to Google Scholar
[Steiger.etal:2008]Steiger, Michael; Asmussen, Sönke (2008): Crystallization of sodium sulfate phases in porous materials: The phase diagram Na2SO4–H2O and the generation of stress. In: Geochimica et Cosmochimica Acta, 72 (17), 4291-4306, UrlLink to Google Scholar
[Steiger.etal:2014]Steiger, Michael; Charola A. Elena; Sterflinger, Katja (2014): Weathering and Deterioration. In: Siegesmund S.; Snethlage R. (eds.): Stone in Architecture, Springer Verlag Berlin Heidelberg, 223-316, 10.1007/978-3-642-45155-3_4.Link to Google Scholar
[Steiger:2003b]Steiger, Michael (2003): Salts and Crusts. In: Brimblecomb, Peter (eds.): Air Pollution Reviews – Vol. 2: The effect of air pollution on the Built Environment, Imperial College Press, 133-181.Link to Google Scholar
[Steiger:2005]Steiger, Michael (2005): Crystal growth in porous materials: I. The crystallization pressure of large crystals. In: journal of Crystal Growth, 282 (3), 455-469, Url, 10.1016/j.jcrysgro.2005.05.007Link to Google Scholar
[Steiger:2005b]Steiger, Michael (2005): Crystal growth in porous materials: II. The influence of crystal size. In: Journal of Crystal Growth, 282 (3), 470-481, 10.1016/j.jcrysgro.2005.05.008Link to Google Scholar
[Steiger:2006]Steiger, Michael (2006): Crystal growth in porous materials: Influence of supersaturation and crystal size. In: Fort, Rafael; Alvarez de Buergo, Monica; Gomez-Heras, Miquel; Vazquez-Calvo, Carmen (eds.): Heritage, Weathering and Conservation: Proceedings of the International Heritage, Weathering and Conservation Conference (HWC-2006), 21-24 June 2006, Madrid, Spain, Taylor & Francis, 245-251.Link to Google Scholar
[Steiger:2009]Steiger, Michael (2009): Mechanismus der Schädigung durch Salzkristallisation. In: Schwarz, Hans-Jürgen; Steiger, Michael (eds.): Salzschäden an Kulturgütern: Stand des Wissens und Forschungsdefizite, Eigenverlag, 66-80.Link to Google ScholarFulltext link
[Taber:1916]Taber, Stephen (1916): The Growth of Crystals under External Pressure. In: American Journal of Science, (41), 532-556Link to Google Scholar
[Weyl:1959]Weyl, Peter K. (1959): Pressure Solution and the Force of Crystallisation - A Phenomenological Theory. In: Journal of Geophysical Research, 64 (11), 2001-2025Link to Google Scholar