The Identification and Behavior of Hydrated Salt Crystals

(1)

B A C H E L O R T H E S I S O F R O Z E L I N E W I J N H O R S T

U N D E R T H E S U P E R V I S I O N O F

M W . D R . N . F . S H A H I D Z A D E H A N D D H R . P R O F . D R . D . B O N N

J U L Y 2 0 1 8

THE IDENTIFICATION AND

BEHAVIOR OF HYDRATED

(2)

The Identification and

Behavior of Hydrated Salt

Crystals

Rozeline Wijnhorst(10726217)

University of Amsterdam

Faculty of Science

Van der Waals-Zeeman Instituut

Soft Matter Group

Supervisor:

dr. N.F. Shahidzadeh

Second Examiner:

prof. dr. D. Bonn

Report Bachelor Project Physics and Astronomy , size 15 ECTS , conducted

between 01 - 04 - 2018 and 07 - 07 - 2018

(3)

Abstract

Sodium sulfate shows non-facetted crystals during deliquescence at a relative humidity of 96%. These crystals show deformable behavior. This report explains the identification of these facetted crystals using Raman spectroscopy. From the resulting spectra we can identify the non-facetted crystals as mirabilite (the highest hydrate of sodium sulfate). This means that we see them as crystals and not as droplets of a solution with a high saturation. Furthermore the report studies the deformable behavior of the non-facetted crystals. We compared the depth and width of a closing hole inside the crystal wall with the dissolution rate. With these results we can conclude that the deformation is not caused by the dissolution of the crystal alone. The closing of the crystal wall seems to go faster if a smaller particle moves out of the crystal. The self healing process of the closing holes, is a finite time singularity. Further research needs to be done on the forces that act on the crystal surface.

(4)

Populair Wetenschappelijke

Samenvatting

Zoutkristallen zijn korrels en als je goed kijkt, zie je dat ze facetten hebben. Met het blote oog valt er misschien niet heel veel op aan zoutkristallen. Als we zouten echter beter bestuderen onder de microscoop, dan kunnen we zien dat zouten verschillende verschijningsvormen hebben. Deze verschijningsvormen noemen we fasen en zijn afhankelijk van onder andere de luchtvochtigheid, de temperatuur en de verzadigdheid van de oplossing waaruit ze zijn ontstaan. Het zout natrium-sulfaat heeft bijvoorbeeld twee droge fasen en twee gehydrateerde fasen. Een gehydrateerd kristal, heeft een bijzondere molecuul structuur. Een zoutmolecuul is, als het gehydrateerd is, omringd door meerdere water moleculen. Natriumsulfaat heeft twee van deze fasen. Eentje met zeven water moleculen: het heptahydraat, en eentje met tien water moleculen: het decahydraat.

Tijdens het onderzoeken van natriumsulfaat bij een hoge luchtvochtigheid, hebben wij iets bijzonders waargenomen. De natriumsulfaat kristallen hebben dan geen facetten zoals verwacht, maar hebben ronde hoeken en lijken vervormbare kristalwanden te hebben. Tijdens dit project heb ik deze kristallen gedentificeerd met behulp van Raman spectroscopie. Raman Spectroscopie zegt iets over hoe moleculen zijn opgebouwd. Molecuul bindingen zijn niet zo star als ze op schematische tekeningen lijken. Je kan deze verbindingen zien als veertjes die kunnen buigen en vibreren. Als in de spectroscoop met een monochromatische laser op de moleculen geschenen wordt, veranderd de golflengte van het laserlicht omdat de molecuul bindingen trillingen in het elektromagnetisch veld veroorzaken. Het weerkaatste licht, met de veranderde golflengte, wordt opgevangen door een detector. Door de verandering in golflengte te bepalen, kan iets gezegd worden over de molecuul bindingen. Je kunt je voorstellen dat gehydrateerde zoutkristallen, vanwege de gebonden water moleculen, andere veranderingen in de golflengte van het laserlicht geven, dan de droge kristallen. Met behulp van deze methode kun je de verschillende kristal fasen van elkaar onderscheiden. Door gebruik te maken van Raman spectroscopie, hebben wij gevonden dat de kristallen met de ronde hoeken en vervormbare kristalwanden, gedentificeerd kunnen worden als het decahydraat van natriumsulfaat. Dit is bijzonder omdat de kristallen er anders uitzien en zich anders gedragen dan we zouden verwachten van decahydraat kristallen.

De ronde natriumsulfaat kristallen, lijken vervormbare kristalwanden te hebben. In de ronde kristallen zitten zwarte vlekken gevangen. Als deze vlekken oplossen of uit het kristal naar buiten bewegen, wordt een gat in de kristalwand gevormd. Dit gat lijkt zichzelf weer te helen doordat de kristalwand netjes terug buigt naar zijn oude positie. Om aan te tonen dat deze buiging niet veroorzaakt wordt door het oplossen van het kristal, hebben wij de breedte en diepte van de gaten over de tijd, vergeleken met de oplossingsgraad van het kristal. Uit deze metingen blijkt dat de vervormbaarheid niet alleen door het oplossen van het kristal wordt veroorzaakt. Verder gaat het terugbuigen sneller, als een kleiner gevangen kristal zich naar buiten beweegt. Verder hebben we gezien dat de functie waarmee de kromming van de gaten naar nul gaat, een finite time singularity is. Een finite time singularity is een moment in de tijd waar een situatie radicaal veranderd, of waar de ene structuur ineens overgaat in een andere structuur. Dit verband duidt er op dat er krachten met elkaar in competitie zouden moeten zijn. We weten nog niet welke krachten dat zijn. Er moet dus meer onderzoek gedaan worden naar de krachten die op het oppervlakte van deze ronde, vervormbare natriumsulfaat kristallen werken.

(5)

Chapter 1 Introduction

Hydrates are molecules entrapped in a structure of water molecules [12]. Gas hydrates for example, are gas molecules entrapped in water molecules. They occur in seas all over the earth. Due to the pressure of the seawater, gas hydrates appear as ice[9]. They can be the energy source of the future if we are able to get the gas out the ice ourselves [7]Unfortunately there is another possibility. Gas hydrates may be the downfall of the earth. If the ice around the gas molecules melts due to climate change , tmore greenhouse gasses will be released [9]. Right now, gas hydrates are causing a lot of problems by blocking pipe lines used in the petroleum industry[6]. In this way hydrates can be either helpful or harmful. This is one of the reasons why it is of great importance, to fully understand the behavior and emergence of hydrates, by doing fundamental research on them.

The emergence of hydrates depends on di↵erent parameters including temperature, the concen-tration of the bulk solution in which the hydrate is formed and relative humidity [16]. Although there is knowledge about the chemical structure of di↵erent hydrated phases[12, 11], their oc-currence and stability on mars [2] and their influence on lime mortars [8, 13], there is no recent fundamental research on the characteristics and behavior of hydrated crystals during deliques-cence.

This thesis focuses on the phase transitions of anhydrous sodium sulphate salt during deli-quescence. We have seen the formation of a large amount of non-facetted crystals during the deliquescence process at high relative humidity. These crystals seem to have a deformable sur-face. The research is focused on identifying the crystal phase with Raman spectroscopy. Another objective was to study the odd, deformable behavior of these sodium sulphate crystals at high relative humidity.

The report is build up as follows: first we will share our observations at high relative humidity, together with some theory about the phase transitions of sodium sulfate. The subsequent chapters will discuss the identification using Raman Spectroscopy and the study on the deformable behavior of non-facetted sodium sulfate crystals which are formed prior to complete deliquescence. The report ends with the conclusions and an outlook on further research.

(7)

Chapter 2 Theory and observations

Figure 2.1: Crystal structure of mirabilite, black dots are sulfate atoms, white dots are water molecules/citeruben

In this chapter, the theory behind the phase diagrams of sulfate will be discussed in relation to our observations. First the dehydration process of sodium sulfate is described. Afterwards the deliquescence process of sodium sulphate will be explained. At the end we will discuss di↵erent hypotheses on the formation of the decahydrate of sodium sulphate during deliquescence.

2.1 Dehydration of sodium sulphate

Figure 2.2: Mirabilite crystal on macro scale

Sodium sulfate salt has di↵erent polymorphs ,with di↵erent solubilities and equilibrium relative humidities(Figure2.5a). There are two anhyd-rous forms: the stable thenardite(phase V) and the unstable phase III. Next to the anhydrous forms, sodium sulfate has two hydrated forms: the stable decahydrate mirabilite and the unstable heptahydrate which is formed during increasing or decreasing temperature [16].

An example of a dehydration process, is the dehydration of the decahydrate. The decahydrate has ten water molecules for one mo-lecule of sodium sulfate. The sulfate atoms are surrounded by eight oxygen atoms that form an octahedral. The two other water molecules are interstitial(Figure 2.1).[12] When this crystalline structure grows a mirabilite crystal is formed. Mirabilite crystals are transparant and 2 The Identification and Behavior of Hydrated Salt Crystals

(8)

CHAPTER 2. THEORY AND OBSERVATIONS

have clear facets (Figure2.2).When mirabilite crystals are dehydrated, either by increasing the temperature or by decreasing the relative hu-midity, the mirabilite turns into thenardite(Figure 2.3a)[16]. The occurring thenardite is white

500 nm

Figure 2.3: crystal (a) showing thenardite after dehydration of mirabilite, (b) thenardite powder, (c) porous structure of thenardite on nano scale, made with high resolution scanning microscopy and has the same form as the mirabilite crystal that it used to be.

200 nm

Figure 2.4: Equilibrium morphology of thenardite, made with high resolution scanning microscopy The morphology of these crystals on nanoscale, is the equilibrium morphology of thenardite(Figure

2.5a) [4]. However, a more investigation on the microscopic structure after dehydration shows a porous structure (Figure2.3c). When this crystal is crushed, a white powder remains (Figure2.3b). The phase diagram of sodium sulfate (Figure2.5a), confirms these observations. The equilibrium relative humidity of mirabilite at room temperature is 96 %. During dehydration the relative humidity is decreased. During dehydration mirabilite merges into thenardite. It is possible that during the dehydration, phase III is formed in between the mirabilite and thenardite phases(Figure

2.5a)[16].

2.2 Deliquescence of sodium sulphate

When the anhydrous powder is hydrated by increasing the relative humidity to 96 %, we would expect the same process in reverse a group of crystals is formed. The crystals in the middle of the group have facets, as we expect from mirabilite crystals.However the crystals on the side, next to the solution have smooth surfaces(Figure 2.6). These crystals seem to be deformable. When an entrapped black spot (probably undissolved thenardite, mirabilite or air bubbles) inside the crystal dissolves, a hole is left in the crystal wall. The hole seems to close by itself, like the exocytosis in biological cells. (Figure2.7) [10].These non-facetted crystals are not observed during the dehydration of mirabilite or the mirabilite precipitation directly from a salt solution. The formation of the non-facetted crystals seems to be a one-way process. An important detail is, that the non-facetted crystals only emerge if the deliquescence process starts with thenardite powder which is made of dehydrated mirabilite. They will not emerge if the deliquescence starts with thenardite that is been formed in a drying droplet.

If we have a look at the t/RH phase diagram (Figure 2.5a)[16] we would expect these non-facetted crystals to be mirabilite, as it appears at 96 % relative humidity. The equilibrium solubility

(9)

Phase diagram of Na2So4

RH

T (°C)

Phase III

Anhydrous

(phase V)

Hepta-hydrate

Mix of phase V

and

decahydrate

Decahydrate

Phase diagram of Na

₂

SO

₄

[3]

Phase diagram of Sodium Sulfate

Figure 2.5: (a) RH/T Phase diagram of sodium sulfate (b) Solubility/T Phase diagram of Sodium Sulfate [16]

(10)

of thenardite lies at 3,68 m/molkg-1[16]. This means that during deliquescence of the thenardite powder, a salt solution at 3,5 molal is formed. As the equilibrium solubility of mirabilite is lower, the mirabilite crystals can then precipitate and lower the concentration to 1.4 molal (Figure

2.5b)[16].

Facets

Smooth surface

Figure 2.6: Sodium sulfate, RH = 96%, facets on the inside and non-facetted corners on the outside

T = 0 s _T = 4 s T = 9 s

T = 25 s T = 54 s

Figure 2.7: Surface deformation of non-facetted sodium sulfate crystal at RH = 96%

2.3 Thermodynamics during deliquescence

The common hypothesis on the formation of mirabilite during deliquescence, states that at first the anhydrous form dissolves. The hydrate is then formed in the highly concentrated solution

(11)

[17]. If this is the case we would expect one big mirabilite crystal to emerge. On the contrary, several crystals are formed during deliquescence. This would be less thermodynamically favorable because of the bigger total surface created in this way [14]. This process has not been studied any further for this thesis. Further research need to be done, before conclusions could be done, about whether the mirabilite crystals are formed out of the solution or is it related to the thenardite nano crystals expansion.

(12)

Chapter 3 Identification with Raman

Spectroscopy

In this chapter the identification of the non-facetted crystals using Raman spectroscopy will be described. First we will explain the obtaining and use of reference spectra. Afterwards we will explain the methods and results of Raman spectroscopy during the deliquescence of the thenardite powder. The chapter will end with the conclusions of the identification.

3.1 Reference Spectra

Raman spectroscopy was used, because the hydrated crystal consists for a big part of water. As water absorbs lots of infra red radiation, Raman spectroscopy is more suitable for measurements on hydrated salt crystals [15].For the identification the Renishaw InVia Microscope is used to take the Raman spectra. We used a laser wavelength of 532 nm and 100 percent laser power. All the measurements have taken place on room temperature (around 21 C). To identify to which

Reference Fingerprint Ramanspectra

0 1000 2000 3000 4000 5000 6000 300 500 700 900 1100 1300 1500 1700 In te ns ity Wavenumber(cm-1) 989 994 1101,1132,1153 618 982 1115 622,633,649 994 1101,1132,1163 446, 459 452, 467 452, 467 452 thenardite crystal soluCon (1.4 molal) thenardite powder mirabilite 618,629 1117 622,633, 649 1153

Figure 3.1: Reference Raman Spectra: mirabilite, thenardite crystal, 1.4 molal solution and thenardite powder

(13)

CHAPTER 3. IDENTIFICATION WITH RAMAN SPECTROSCOPY

reference spectra(mirabilite, thenardite in crystal form, thenardite powder and 1.4 molal sodium sulfate solution(Figure3.1)

The reference spectra are made in the so called fingerprint area. This area shows peaks, that are specific for the di↵erent phases of sodium sulfate[15,5,1]. The di↵erent phases have peaks in the same area’s, but the position and amount of peaks, di↵er for all the phases(Figure3.2). The thenardite crystal and the thenardite powder, show the same fingerprint spectrum(Figure 3.1). This means that they have the same molecular structure [15].

Figure 3.2: Expected peak positions for the di↵erent phases of sodium sulfate [5]

If we compare the OH-stretching peaks of the thenardite that is formed in a drying solu-tion(Figure3.3), and the anhydrous powder (Figure3.4)we see no di↵erence either. Both anhyd-rous forms do not consist of water. The mirabilite and the 1.4 molal solution have OH-stretch peaks that are in line with the literature values(Figure 3.5, (Figure3.6), (Figure3.2) [5, 1]. The fact that no di↵erence is seen between the anhydrous thenardite powder and the thenardite from the drying droplet is unexpected as they behave di↵erently during deliquescence.

(14)

CHAPTER 3. IDENTIFICATION WITH RAMAN SPECTROSCOPY 0 200 400 600 800 1000 1200 1400 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Inte nsity wavenumber(cm-1) OH-stretch area, thenardite, 100%laserpower, 532nm, 20x long working distance objecFve

Figure 3.3: Extended Raman Spectrum in OH-stretch area for thenardite crystal

0 200 400 600 800 1000 1200 1400 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Inte nsity wavenumber(cm-1) OH-stretch area, Na2So4 anhydrous powder, 100%laserpower, 532nm, 20x long working distance objecIve

Figure 3.4: Extended Raman Spectrum in OH-stretch area for thenardite powder

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Inte nsity wavenumber(cm-1) OH-stretch area, mirabilite, 100%laserpower, 532nm, 20x long working distance objecFve 3471

Figure 3.5: Extended Raman Spectrum in OH-stretch area for mirabilite

0 10000 20000 30000 40000 50000 60000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 inte nsity Wavenumber(cm-1) OH-stretch area, Na2So4 solu>on(1.4molal), 100%laserpower, 532nm, 20x long working distance objec>ve 3471

(15)

3.2 Raman Spectroscopy during deliquescence

The deliquescence process took place inside a climate chamber.The climate chamber is made of a slide of aluminium, plexiglass walls and a glass slide on top. Inserting wetted tissues in the box increases the relative humidity. The anhydrous thenardite powder is located on the bottom on the aluminium slide.

The box is placed in the InVia Renishaw raman microscope. Spectra were taken every 4,2 seconds with a wavelength of 532 nm, 100 % laser power under a 20x objective with a long working distance. The spectrometer has an inaccurany of_{±2cm 1. During deliquescence, a shift}

0 5000 10000 15000 20000 300 400 500 600 700 800 900 1000 1100 1200 Inte nsity wavenumber(cm-1) Shi7 in Raman spectra during deliquescence Na2SO4 (532nm, 20x, 100% laser intensity) Symmetric

bending Symmetric stretch

Ti me( s) 0 1700 850 An;-Symmetric bending An;-Symmetric bending

Figure 3.7: Sequence of Raman Spectra during deliquescence of Sodium Sulfate starting with thenardite powder

in all the peaks is observed. First we will compare the shift of the peak in the SO4-symmetric stretch areas, as it is the most obvious indicator for each phase.

During the deliquescence the SO4-symmetric stretch peak shifts from 994 cm 1 towards 989 cm 1. At the end of the sequence of measurements, an extra peak with a wavenumber of 982 cm 1 occurs (Figure3.8a). In the SO4-anti-symmetric stretch area we observe a change in the amount. At first there are three peaks: 1101 cm 1, 1132 cm 1 and 1153cm 1. This shifts towards one more broad peak at 1117 cm 1. With a small shift towards the left to 1115 cm 1 at the end of the sequence(Figure3.8b). In the SO4-symmetric bending area the peaks start at 452 cm 1 and 467cm 1, they shift towards 445 cm 1 and 459cm 1. At 1700 s, a small peak is formed at 452 cm 1(Figure3.8c). In the SO4-anti symmetric bending area, a change in the amount of peaks is observed as well. The sequence starts with peaks at 622 cm 1, 633 cm 1 and 649 cm 1(Figure

3.8d).

(16)

Smooth surface

0 5000 10000 15000 20000 970 980 990 1000 1010 1020 Inte nsity wavenumber(cm-1) Shi7 in Raman SO4 symmetric stretch peak during deliquescence of Na2SO4 T im e(s ) 0 1700 850 994 989 882 Time (s) T 850 982 989 994 0 200 400 600 800 1000 1200 1400 1600 1050 1070 1090 1110 1130 1150 1170 1190 Inte nsity Wavenumber(cm-1) Shi7 in Raman SO4 an;-symmetric stretch peak during deliquescence of Na2SO4 T im e(s ) 0 1700 850 1115₁₁₁₇ 1101 1132 1153 Wavenumber(cm-1) Wavenumber(cm-1) In te nsi ty In te nsi ty 100 300 500 700 900 1100 1300 1500 1700 1900 2100 500 550 600 650 700 Inte nsity wavenumber(cm-1) Shi7 in Raman SO4 an;-symmetric bending peak during deliquescence of Na2SO4 T im e(s ) 0 1700 850 618 629 T 649 633 622 0 500 1000 1500 2000 400 420 440 460 480 500 520 Inte nsity wavenumber(cm-1) Shi7 in Raman SO4 symmetric-bending peak during deliquescence of Na2SO4 452, 567 446, 459 45 2 T im e(s ) 0 1700 850 445 452 459 T 452 467 Wavenumber(cm-1) Wavenumber(cm-1) In te nsi ty In te nsi ty

Figure 3.8: Raman Shift during deliquescence of sodium sulfate, (a) SO4-symmetric stretch, (b)SO4-anti symmetric stretch ,(c)SO4-symmetric bending (d)SO4-anti symmetric bending

3.3 Conclusions

If we compare the shift in the wave numbers during deliquescence, to the reference spectra (Figure

3.1)and the literature values (Figure3.2), we can conclude that during deliquescence the sodium sulfate starts at thenardite. The first phase transition is the transition towards mirabilite. At the end of deliquescence there is a 1.4 molal sodium sulfate solution left. The start and end phase

(17)

are in agreement with the expectations: we put thenardite powder in the climate chamber and at the end of the deliquescence all the thenardite is dissolved. During deliquescence no other spectra then the fingerprint spectrum of mirabilite are shown. To confirm the hypotheses that we can identify the non-facetted crystals as mirabilite, one Raman spectrum is made (Figure 3.9). For this spectrum we used a wavelength of 532 nm, 20 x objective with long working distance objective and 100 % laser power. The laser is focused on one of the non-facetted crystals. This spectrum has a SO4-symmetric stretch peak at 989 cm 1 and all the other peaks are in agreement with the reference spectra and literature values as well [5, 1]. Using the information from the Raman

0 5000 10000 15000 20000 25000 0 200 400 600 800 1000 1200 1400 1600 1800 2000

In

te

ns

ity

Wavenumber(cm-1)

989

Raman spectrum deliquous crystal

(532 nm, 20x objecFve, 100% laser power)

1117, 1127 982 627, 634 457 50 um

rough

Raman spectrum non-face0ed crystal

Figure 3.9: Raman Spectrum of non-facetted Sodium Sulfate crystal, wavelength = 532 nm, 20 x long working distance objective, laser power = 100%

spectra during deliquescence and the spectrum of the facetted crystal we can identify the non-facetted crystals as mirabilite. This means that these entitites are well crystals and not droplets of a high saturated sodium sulfate solution.

With Raman spectroscopy alone it is not possible to explain the di↵erence in appearance on macro scale of the non-facetted crystals compared to the mirabilite crystals with facets. In the next chapter we will explain the study on the odd, deformable behavior of the non-facetted mirabilite crystals.

(18)

Chapter 4 Crystal Surface

The non-facetted mirabilite crystals show deformable behaviour. In the fiugres (Figure4.1) some black spots entrapped in the crystal can be observed. These are either unddissolved crystals or air bubbles. When these entrapped black spots dissolve or move towards the crystal wall, a hole in the crystal wall arises. The crystal wall seems to heal itself, by bending back till the hole is gone (Figure4.1).

T = 0 s _T = 4 s T = 9 s

T = 25 s T = 54 s

(19)

CHAPTER 4. CRYSTAL SURFACE

In this chapter the deformable behavior of the crystal wall of the non-facetted mirabilite crystals is described. The chapter starts with the measuring methods. Afterwards we will explain why this deformable behavior is not caused by the dissolution of the non-facetted mirabilite crystals. Furthermore the change in curvature of the holes in the crystal wall over time will be discussed. Eventually we will state out the conclusion of the chapter.

4.1 Methods

To study the deformable behavior of the non-facetted mirabilite crystals, we have measured the size of the holes in the crystal and their curvature over time, with the size of the mirabilite crystal. We did this on four di↵erent mirabilite crystals (Figure4.3). All the measurements are done using ImageJ. The mirabilite crystals are formed inside a glass climate chamber with wetted tissues inside to increase the relative humidity inside the climate chamber (Figure 4.2). The climate chamber was placed under a microscope with 40 x objective. All the pictures were taken using a camera and the PixelInk computer program.

Climate

chamber

Figure 4.2: Set up: microscope connected to computer. Deliquescence takes place in glass climate chamber

The dissolution rate is measured by taking the intersection of the crystal over time, or the distance between the crystal wall and a reference point over time. The depth and the width of the holes is measured over time as well,as is the curvature for all the four crystals. The depth of the whole is the distance between the extrapolation of the crystal wall and the lowest point of the whole (Figure4.3) The width of the whole is the widest gap in the crystal wall at the top of the hole.For calculating the curvature we used that Curvature = 1/Radius. We took the radius of the curvature by drawing a circle inside the hole and measuring the diameter of these circles(Figure

4.3) .

(20)

CHAPTER 4. CRYSTAL SURFACE 12,5 um 12,5 um 2 1 0 s 12,5um 0 s 12,5 um

Figure 4.3: Crystal 1, crystal 2, crystal 3, crystal 4

4.2 Dissolution rate of the crystal in comparison with the

deformation of the crystal surface

In this section we will compare the dissolution rate of the four di↵erent crystals, with the de-formation of holes in their crystal walls. We assume the crystal is dissolving homogeneously. If the deformation of the crystal wall inside the holes is caused by the dissolution of the crystal, we would expect the dissolution rate of the crystal to be of the same size as the change in depth of the hole of the crystal wall. We expect the change in the width of the hole to be twice as big as the dissolution rate of the crystal.

4.2.1 Results and Discussion

The change in size of the crystal (dissolution rate) over time together with the change in depth and width of the crystals are plotted in figures(4.4-4.8). These results show that the change in depth of the crystal is bigger then the dissolution rate of the crystal. Furthermore the change in width of the crystal is bigger then two times the dissolution rate of the crystal. From these results we can conclude that the movement performed by the crystal wall when a hole is closing, is not caused by the dissolution of the mirabilite crystal alone.

(21)

CHAPTER 4. CRYSTAL SURFACE -1,5 -1 -0,5 0 0,5 1 1,5 2 2,5 3 3,5 0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 siz e (um ) *me(s)

**change in depth and width of the hole compared to dissolu*on**

crystal

depth hole (um) width hole (um) dissolu:on crystal (um)

Figure 4.4: Change in depth and width of hole compared to dissolution rate crystal 1

-2 -1 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 8 siz e (um ) *me(s)

**change in depth hole 1, hole 2 over *me**

depth hole 1(um) dissolu9on crystal (um) depth hole2 (um)

Figure 4.5: Change in depth of hole1 and hole2 compared to dissolution rate crystal 2

(22)

CHAPTER 4. CRYSTAL SURFACE -1 0 1 2 3 4 5 0 1 2 3 4 5 6 7 8 siz e (um ) *me(s)

**Change in width hole 1, hole 2 over *me**

width hole1 (um) dissolu9on crystal (um) width hole2 (um)

Figure 4.6: Change in width of hole1 and hole2 compared to dissolution rate crystal 2

-4 -2 0 2 4 6 8 10 0 2 4 6 8 10 12 ch an ge s ize ( um) /me(s) change depth hole (um) change width hole(um) dissolu;on crystal (um)

(23)

CHAPTER 4. CRYSTAL SURFACE -1 -0,5 0 0,5 1 1,5 2 2,5 3 3,5 4 0 0,5 1 1,5 2 2,5 3 3,5 4 siz e (um ) *me(s) change in depth hole (um) change in width hole (um) dissolu<on crystal (um)

Figure 4.8: Change in depth and width of hole compared to dissolution rate crystal 4

(24)

CHAPTER 4. CRYSTAL SURFACE

4.3 Curvature of the crystal surface

In this section the e↵ect of the size of the entrapped black spots on the velocity of the deformation of the holes in the crystal surface will be explained. We compared the change curvature of the holes over time for entrapped black spots with di↵erent sizes.

4.3.1 Results and Discussion

The deformation speed is higher for holes with a smaller curvature(Figure4.9). We observe that the curvature of the hole in the crystal wall going to zero during the deformation. We took the point where the crystal is just closed again as t0 and all the events before as t. The plot of (t0 t) against the size of the curvature shows that the curvature goes to zero as a finite time singularity (Figure4.10).

Finite time singularities are moments in time where something changes radically. One phase goes into the other or one structure goes into another structure [3]. All the functions have a power between 0,2 and 0,5 (Figure 4.10). The power seems to go more close to 0.5 if the influence of the dissolving crystal is lower and more measuring points are taken. The power is caused by a competition between forces [3]. We do not know yet which forces are in competition when the holes are closing. More measurements with less influence of the dissolving crystal need to be done to take further conclusions.

(25)

CHAPTER 4. CRYSTAL SURFACE 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0 1 2 3 4 5 6 7 8 9 10 Cu rva tu re( um-1 ) .me(s) Curvature hole in crystal wall over .me, comparison different crystals sorted by size 3,3 um 3,4 um 4,8 um 5,0 um 7,0 um -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0 0,1 0,2 0 1 2 3 4 5 6 7 8 9 10 de lta Cur vatur e (um -1) Time (s) Change in urvature hole in crystal wall over :me, comparison different crystals sorted by size 31 um 30 um 44 um 45 um 64 um Change in Curvature holes in crystal wall over 3me, Comparison different crystals sorted by size

Figure 4.9: (a)Curvature over time, (b) Change in Curvature over time

(26)

CHAPTER 4. CRYSTAL SURFACE y = 0,5676x0,2891 0,1 1 0,1 1 10 Cu rva tu re( um-1 ) Time (s) Curvature crystal 1 over 5me y = 0,2372x0,3733 y = 0,2353x0,4822 0,1 1 0,1 1 10 cu rva tu re ( um-1 ) .me(s) Curvature crystals 2.1 and 2.2 over .me curvature1(1/R) (um) curvature 2 (1/R) (um) y = 0,1649x0,4822 0,01 0,1 1 0,1 1 10 Cu rva tu re( um-1 ) Time (s) Curvature crystal 3 over 6me y = 0,5638x0,4213 0,1 1 0,1 1 10 Cu rva tu re( um-1 ) Time(s) Curvature crystal 4 over 6me

Figure 4.10: Curvature over time (a) Crystal 1 y = 0, 5676(t0 t)0,2891 _{(b) Crystal 2.1 y =}

0, 2372(t0 t)0,3733 _{and 2.2 y = 0, 2353(t0} _t)0,4822 _{(c) Crystal 3 y = 0, 1649(t0} _t)0,4822 _(d)

Crystal 4 y = 0, 5638(t0 t)0,4213

4.4 Conclusions

In this section we have seen that the deformation of the crystal wall, caused by the movement or dissolution of entrapped black spots, is not only caused by the dissolution of the crystal. Furthermore the deformation speed is higher for holes with smaller curvature. We have seen that the function of the deformation of the holes in the crystal walls, is finite time singularity. There is a need for more research on which forces are in competition to cause the power around 0,5 in the function of the curvature over time.

(27)

Chapter 5 Conclusion and Outlook

In this thesis we looked at the kinetics of non-facetted crystals that are formed during the deli-quescence of sodium sulfate. Our results show that they emerge at a relative humidity of 96%. The crystals do not have facets and show deformable behavior. We identified these non-facetted crystals by using Raman spectroscopy. Furthermore we studied the deformable behavior of the non-facetted crystals.

By using Raman spectroscopy, we saw a shift from the fingerprint spectrum of thenardite towards the fingerprint spectrum of mirabilite during deliquescence. The deliquescence ends with the fingerprint spectrum of a 1.4 molal solution of sodium sulfate. By taking one spectrum focused on the non-facetted crystals we could conclude that we can identify the crystals as mirabilite.

Furthermore the dissolution rate of four crystals was compared, to the depth and the width of the holes in the crystal walls. The change in depth and width of the holes over time is bigger then the dissolution rate of the crystal. From this results we could say that the surface deformation is not caused by the dissolution of the non-facetted crystal alone. It is recommended to investigate wether the crystals dissolve homogeneously.

By plotting the curvature and the change in curvature of the holes over time, we can conclude that for holes with larger curvature, the curvature returns back to zero faster. An indication that forces on the crystal surface are involved in the closing of the holes, is the fact that the function of the curvature over time is a finite time singularity. When the dissolution rate is small, the power in the function seems to go to 0,5. This power is caused by a competition of forces. We do not know yet which forces are in competition when the crystal wall is closing. There is a need to do more research on the forces that act on the crystal wall. A method needs to be found to measure the surface tension on the surface of the non-facetted mirabilite crystals.

It would be interesting to see if this behavior of the highest hydrate is something typically for sodium sulfate or if other salts show similar behavior. We have indications that magnesium sulfate behaves in the same way for the heptahydrate (Appendix). As the heptahydrate is the highest hydrate of magnesium sulfate this could be an indicator that the highest hydrates of sulfate crystals show non-facetted crystals with deformable behavior in equilibrium with the salt solution at the limit of the solubility concentration.

(28)

Chapter 6 Acknowledgements

First of all I would like to thank dr. Freek Ariese and his students Pien Bouman and Bram Mooij, for letting me use the Raman microscope at the VU and for helping me out with the use of the microscope.Furthermore I would like to thank dr. Freek Ariesenfor his help with the interpretation of the Raman spectra and his enthusiasm about the subject. I would like to thank prof. dr. Sander Woutersen for bringing me into contact with dr. Freek Ariese. I like to thank the Soft Matter Group for their humor, support, help and for the fact that they let me feel welcome. Finally I would like to thank dr. Noushine Shahidzadeh for her trust, her enthusiasm, the critical questions and for being the first female rolemodel that I have ever had on this university.

(29)

Bibliography

[1] Kawther Ben Mabrouk, Thomas H Kau↵mann, Hassen Aroui, and Marc D Fontana. Raman study of cation e↵ect on sulfate vibration modes in solid state and in aqueous solutions. Journal of Raman Spectroscopy, 44(11):1603–1608, 2013. 8,12

[2] Steve J Chipera and David T Vaniman. Experimental stability of magnesium sulfate hydrates that may be present on mars. Geochimica et cosmochimica acta, 71(1):241–250, 2007. 1

[3] Jens Eggers and Marco Antonio Fontelos. Singularities: formation, structure, and propaga-tion, volume 53. Cambridge University Press, 2015. 19

[4] Sanne JC Granneman, Noushine Shahidzadeh, Barbara Lubelli, and Rob PJ van Hees. E↵ect of borax on the wetting properties and crystallization behavior of sodium sulfate. CrystEng-Comm, 19(7):1106–1114, 2017. 3

[5] Andrea Hamilton and Robert I Menzies. Raman spectra of mirabilite, na2so4· 10h2o and the rediscovered metastable heptahydrate, na2so4· 7h2o. Journal of Raman Spectroscopy, 41(9):1014–1020, 2010. 8, 12

[6] EG Hammerschmidt. Formation of gas hydrates in natural gas transmission lines. Industrial & Engineering Chemistry, 26(8):851–855, 1934. 1

[7] Sang-Yong Lee and Gerald D Holder. Methane hydrates potential as a future energy source. Fuel processing technology, 71(1-3):181–186, 2001. 1

[8] Nadine Lindstr¨om, Tanya Talreja, Kirsten Linnow, Amelie Stahlbuhk, and Michael Steiger. Crystallization behavior of na2so4–mgso4 salt mixtures in sandstone and comparison to single salt behavior. Applied Geochemistry, 69:50–70, 2016. 1

[9] Alexei V Milkov. Global estimates of hydrate-bound gas in marine sediments: how much is really out there? Earth-Science Reviews, 66(3-4):183–197, 2004. 1

[10] Nuri Oh and Ji-Ho Park. Endocytosis and exocytosis of nanoparticles in mammalian cells. International journal of nanomedicine, 9(Suppl 1):51, 2014. 3

[11] Carlos Rodriguez-Navarro, Eric Doehne, and Eduardo Sebastian. How does sodium sulfate crystallize? implications for the decay and testing of building materials. Cement and concrete research, 30(10):1527–1534, 2000. 1

[12] Helena W Ruben, David H Templeton, Robert D Rosenstein, and Ivar Olovsson. Crystal structure and entropy of sodium sulfate decahydrate. Journal of the American Chemical Society, 83(4):820–824, 1961. 1, 2

[13] E Ruiz-Agudo, F Mees, Patric Jacobs, and C Rodriguez-Navarro. The role of saline solution properties on porous limestone salt weathering by magnesium and sodium sulfates. Environ-mental geology, 52(2):269–281, 2007. 1

[14] Daniel V Schroeder. An introduction to thermal physics, 1999. 6

(30)

BIBLIOGRAPHY

[15] Dustin W. Shipp, Faris Sinjab, and Ioan Notingher. Raman spectroscopy: techniques and applications in the life sciences. Adv. Opt. Photon., 9(2):315–428, Jun 2017. 7,8

[16] Michael Steiger and S¨onke Asmussen. Crystallization of sodium sulfate phases in porous materials: The phase diagram na2so4–h2o and the generation of stress. Geochimica et Cosmochimica Acta, 72(17):4291–4306, 2008. 1, 2,3,4,5

[17] Nicholas Tsui, Robert J Flatt, and George W Scherer. Crystallization damage by sodium sulfate. Journal of cultural heritage, 4(2):109–115, 2003. 6

(31)

Appendix

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 400 600 800 1000 1200 1400 1600 Inte nsity wavenumber(cm-1) Raman spectrum MgSO4 deliquous crystal (heptahydrate) 984

Deliquescence

MgSO

₄

, RH =

75%, T = 21°C

25 um

(a) (b) (c)

First observa,ons magnesium sulfate

100 um

Figure 6.1: First observations on magnesium sulfate. (a) porous structure of the dried heptahydrate, made with high resolution scanning microscopy (b) non-facetted crystals with entrapped crystals inside. (c) Raman spectrum with so4-stretching peak at 984 cm 1, the peak of the magnesium sulfate heptahydrate.