Chemistry of Ascorbic Acid Reduction of Graphene Oxide

Chemistry of Ascorbic Acid Reduction of Graphene Oxide

Reduction of Graphene Oxide in Solution and Film

Pontus Cronqvist 13 september 2019

Master’s Thesis Examiner

Prof. Muhammet Toprak Supervisor

PhD Abhilash Sugunan

TRITA-SCI-GRU 2019:309 KTH Royal Institute of Technology

School of Information and Communication Technology

SE-100 44 Stockholm, Sweden

Abstract

Graphene is a 2D material with interesting properties. It is composed of carbon atoms in a hexagonal lattice, reminiscent of honeycomb. One way of synthesising graphene is by oxidis- ing graphite to create graphene oxide. Graphene oxide is graphene sheets that are covered in functional groups. These functional groups are often considered to be defects as they decrease or completely remove the often desired properties from the graphene. To eliminate the defects graphene oxide is reduced with a reducing agent to form reduced graphene oxide. Ideally, reduced graphene oxide is indistinguishable from pristine graphene.

Ascorbic acid, also known as vitamin C, has been used to successfully reduce graphene oxide.

The progression of the reduction has primarily been tracked using UV-Vis spectroscopy. FTIR and Raman spectroscopy has also been used as additional techniques to confirm the reduction and further characterise the reduced graphene oxide.

The reduction has been shown to occur both for graphene oxide in a water dispersion and graphene oxide in a poly(vinyl alcohol) film. The temperature dependence of the reduction has been investigated and it was found that the reduction is highly temperature dependent.

The level of reduction that was achieved after 168 h of reduction at 25

C was achieved within 6 minutes at 80

C. UV-Vis measurements indicate that the reduction is divided into two dis- tinct regimes of reduction, one for the reduction of C O-bonds and one for C O-bonds. This is supported by FTIR measurements of graphene oxide films that were reduced in ascorbic acid solutions. For both regimes the Arrhenius parameters have been calculated.

Keywords — Graphene Oxide; Reduced Graphene Oxide; Ascorbic Acid; Vitamin C; rGO; GO;

Activation Energy; Arrhenius parameters

Sammanfattning

Grafen ¨ar ett 2D material med intressanta egenskaper. Det best˚ar av kolatomer i ett hexago- nalt gitter som p˚aminner om h¨onsn¨at. Ett s¨att att syntetisera grafen ¨ar att oxidera grafit f¨or att skapa grafenoxid. Grafenoxid ¨ar grafen som har funktionella grupper bundet till sig. Dessa funktionella grupper anses ofta vara defekter d˚a de minskar, eller helt tar bort, de ofta ¨onskade egenskaperna hos grafen. F¨or att eliminera defekterna reducerar man grafenoxid f¨or att skapa reducerad grafenoxid. Idealiskt ska reducerad grafenoxid vara s˚a n¨ara felfri grafen som m¨ojligt.

Askorbinsyra, ¨aven k¨ant som vitamin C, har anv¨ants som reduceringsmedel f¨or att reduce- ra grafenoxid. Reduceringsf¨orloppet har prim¨art f¨oljts med UV-Vis spektroskopi. FTIR och Raman spektroskopi har ocks˚a anv¨ants f¨or att ytterligare styrka att reduktionen sker och f¨or enklare karakterisering av den reducerade grafenoxiden.

Reduktion har bevisats kunna ske b˚ade n¨ar grafenoxid ¨ar i vattendispersion och n¨ar grafe- noxid ¨ar i polyvinylalkoholfilm. Temperaturberoendet av reduktionen har utforskats och visat att reduktionen ¨ar mycket temperaturberoende. Den niv˚a av reduktion som uppn˚addes efter 168 h vid 25

C uppn˚addes efter 6 minuter vid 80

C. UV-Vis-m¨atningar tydde p˚a tv˚a distinkta omr˚aden av reduktion, en f¨or C O-bindningar och en f¨or C O-bindningar. Detta har st¨arkts med FTIR-m¨atningar. F¨or b˚ada dessa omr˚aden har Arrhenius parametrar ber¨aknats.

Nyckelord — Grafenoxid; Reducerad Grafenoxid; Askorbinsyra; Vitamin C; rGO; GO; Aktive-

ringsenergi; Arrhenius parametrar

Acknowledgements

I want to thank Dr. Abhilash Sugunan for being my supervisor during my thesis, for helping and supporting me, and many constructive conversations. I also want to thank Prof. Muhammet Toprak for being my examiner and taking the time to support me with my thesis.

RISE has been a fantastic place to do my thesis with colleagues that have been welcoming and

assisted me during my thesis. I want to extend special thanks to Mr. Wei Zhao for helping me with

general procedures in the lab and assisting me with lab instruments, and to Dr. Martin Andersson

for reading a draft of this thesis which has helped me write a better thesis. I want to thank Dr. Jens

Sommertune for doing the Raman spectroscopy measurements, and I want to thank Mr. Mikael

Sundin for assisting me with the FTIR.

Preface

This master’s thesis is done on behalf of BillerudKorsn¨as AB at RISE Research Institutes of Swe- den AB.

Problem Statement

The purpose of this thesis is to understand the chemistry of the reduction of graphene oxide using ascorbic acid both when graphene oxide is in a water dispersion and when the graphene oxide is in poly(vinyl alcohol)-films.

Delimitations

This project will be limited to the reduction of graphene oxide. The testing of properties of the

fabricated material is outside the scope of this report. The properties of the graphene oxide will not

be examined, but bought from trusted suppliers and rely on previously acquired knowledge from

Billerud Korsn¨as and RISE. Characterisation of the graphene oxide is also outside the scope of this

project.

Abbreviations

AA ascorbic acid AR absorption ratio DHA dehydroascorbic acid E

activation energy

FTIR fourier-transform infrared spectroscopy GO graphene oxide

PVOH poly(vinyl alcohol) rGO reduced graphene oxide UV ultraviolet

UV-Vis ultraviolet-visible spectroscopy

Table of Contents

1 Theoretical Background 1

1.1 Graphene . . . . 1

1.2 Graphene Oxide . . . . 2

1.3 Reduced Graphene Oxide . . . . 5

1.3.1 Thermal Reduction . . . . 5

1.3.2 Chemical Reduction . . . . 6

1.4 Ascorbic Acid . . . . 7

1.5 Poly(vinyl alcohol) . . . . 9

2 Materials and Methods 11 2.1 Characterisation Techniques . . . 11

2.1.1 Visual Inspection . . . 11

2.1.2 Ultraviolet–Visible Spectroscopy . . . 11

2.1.3 Fourier-Transform Infrared Spectroscopy . . . 12

2.1.4 Raman Spectroscopy . . . 13

2.2 Experimental Techniques . . . 14

2.2.1 Solution Reduction . . . 14

2.2.2 Film Reduction . . . 15

2.2.3 Chemicals . . . 15

3 Results and Discussion 16 3.1 Finding a Way to Track Graphene Oxide Reduction . . . 16

3.1.1 Reduction Tracking by Absorption Ratio . . . 17

3.2 Variables Affecting the Reduction Speed . . . 19

3.2.1 The Temperature Dependence on Reduction Speed . . . 19

3.3 Reaction Regimes . . . 20

3.4 Calculating the Activation Energies . . . 22

3.4.1 Assuming First Order Reaction . . . 23

3.4.2 Assuming Second Order Reaction . . . 25

3.4.3 Significance of the Calculated Values . . . 26

3.5 Oxidation of Ascorbic Acid to Dehydroascorbic Acid . . . 26

3.6 Graphene Oxide-Film Reduction . . . 29

3.6.1 Tracking the Reduction of Graphene Oxide Films . . . 29

3.6.2 Raman Spectroscopy of Film Cross Section . . . 32

3.6.3 Reduction in Dry Films . . . 33

3.6.4 Reduction at Different Drying Temperatures . . . 34

3.6.5 Heat Treating a Dry Film . . . 35

4 Conclusions 37

5 Future Work 37

References 38

A Additional Results 43

List of Figures

Figure 1: The structure of graphene . . . . 1

Figure 2: sp

-hybridisation . . . . 2

Figure 3: sp

-hybridisation . . . . 2

Figure 4: An example of how the reaction can look like when going from graphite to reduced graphene oxide . . . . 4

Figure 5: An illustrative example of how a graphene layer distorts with the presence of a single epoxide on its basal plane . . . . 4

Figure 6: UV-Vis spectra of graphene oxide showing its brown colour in solution . . 5

Figure 7: The molecules of ascorbic acid and dehydroascorbic acid . . . . 7

Figure 8: The reduction mechanisms of ascorbic acid suggested by Gao et. al . . . . 8

Figure 9: UV-Vis absorption spectra of ascorbic acid . . . . 9

Figure 10: The structure of poly(vinyl alcohol) . . . . 9

Figure 11: UV-Vis spectra of poly(vinyl alcohol) . . . 10

Figure 12: A mixture of graphene oxide and ascorbic acid before and after reduction, the graphene oxide-concentration is 0.67 mg/ml . . . 11

Figure 13: A schematic of how ATR works . . . 13

Figure 14: FTIR spectrum of water . . . 13

Figure 15: Flocculation of rGO . . . 16

Figure 16: All measured absorptions for sample T25 . . . 16

Figure 17: UV-Vis absorption spectra of unreduced graphene oxide and reduced grap- hene oxide of the same sample . . . 17

Figure 18: UV-Vis absorption spectra of sample T25 . . . 18

Figure 19: The absorption ratio for sample T25 plotted against reduction time . . . 18

Figure 20: The absorption ratio for different temperatures plotted against reduction time 19 Figure 21: Absorption ratios of samples T25, T40, T65 and T80 . . . 20

Figure 22: Absorption ratios of samples T25, T40, T65 and T80 showing the separate regimes . . . 21

Figure 23: The transition absorption ratio plotted against temperature . . . 22

Figure 24: Plotting ln(AR fraction) against reduction time for sample T65, the linear fit is poor to the calculated values demonstrating the need for separation into two regimes . . . 23

Figure 25: Absorption ratios of samples T25, T40, T65 and T80 showing the separate regimes . . . 24

Figure 26: Absorption ratios of samples T25, T40, T65 and T80 showing the separate regimes . . . 25

Figure 27: The colour difference between ascorbic acid and dehydroascorbic acid af- ter two months . . . 27

Figure 28: The possible oxidation of ascorbic acid to dehydroascorbic acid . . . 27

Figure 29: UV-Vis absorption spectra of dehydroascorbic acid, ascorbic acid that has started breaking down to dehydroascorbic acid, and ascorbic acid. The DHA had oxidised from AA for two months . . . 27

Figure 30: An ascorbic acid-film that has turned to dehydroascorbic acid after 66 h in a 60

C . . . 28

Figure 31: The AR of sample T65 showing an increase at 24 h due to ascorbic acid

oxidation to dehydroascorbic acid . . . 28

Figure 32: Progression pictures of a GO:PVOH-film submerged in a 5 mg/ml ascorbic acid-solution. The time beneath each image is the time the film has been submerged . . . 29 Figure 33: The reduction and oxidation of carboxyl acid.

BY-SA 4.0, https://commons.wikimedia.org/w/index. php?curid=45373084

. . . 30 Figure 34: FTIR measurements of a GO:PVA-film at different reduction times . . . 31 Figure 35: FTIR Bands in detail showing increase over time . . . 31 Figure 36: How the cross section of a GO:PVOH-film was measured for Raman spectroscopy 32 Figure 37: The result of the Raman spectroscopy of a GO:PVOH-film cross section . . 33 Figure 38: A control film of graphene oxide without ascorbic acid (left) and an ascor-

bic acid-film cast on a graphene oxide-film (right) . . . 33 Figure 39: UV-Vis of ascorbic acid-film cast on graphene oxide-film and an unreduced

control-film . . . 34 Figure 40: GO:AA:PVOH-films that have dried at different temperatures . . . 35 Figure 41: UV-Vis absorption spectra of a dry GO:AA:PVOH-film, ratio 1:5:100, be-

fore and after being in a 60

C oven for 90 h . . . 36 Figure A1: The decrease of graphene oxide depending on centrifugation speed. Me-

asured absorption at 800 nm. The graphene oxide concentration was 0.04 mg/g . . . 43 Figure A2: The linear trend between concentration and absorption at 800 nm for grap-

hene oxide . . . 43

List of Tables

Table 1: The reaction conditions for the different samples . . . 15

Table 2: The chemicals used in this thesis . . . 15

Table 3: The transition AR between the first and second regime . . . 21

Table 4: An example of how to calculate the concentration fraction of sample T65 . 23 Table 5: Calculated Arrhenius parameters for a first order reaction . . . 24

Table 6: R

-values for the two regimes for a first order reaction . . . 24

Table 7: Calculated Arrhenius parameters for a second order reaction . . . 25

Table 8: R

-values for the two regimes for a second order reaction . . . 26

Table 9: FTIR bands with corresponding bonds . . . 30

Table 10: The AR of films after being dried at different temperatures . . . 35

1 Theoretical Background

This section contains a theoretical background of the materials and techniques used in this thesis

1.1 Graphene

Graphene is a 2D material that is made up of carbon atoms in a hexagonal lattice, depicted in Figure 1. It was first isolated as a stand alone material in 2004, proving that graphene is thermody- namically stable and can exist in ambient conditions [1]. The researchers were awarded the Nobel Prize in Physics for this discovery in 2010 [2].

Figure 1: The structure of graphene

Graphene has since then attracted a lot of attention due to its unique properties. It has a high theoretical carrier mobility of 10

cm

/(V s) [3], a high mechanical strength, which approaches the theoretical strength of defect-free solids and is the strongest material ever tested [4], a theoretical specific surface area of 2630 m

/g [5], high thermal conductivity [6], it is a good barrier material [7]

to name a few properties.

Graphite, which is the precursor used for production of graphene, is a material that ideally is 100 % carbon, which is an abundant element. In this regard, graphite is a sustainable material.

Graphite can be both mined (primarily mined in China and Brazil [8]) and synthetically produced.

As the need for graphene and graphite is increasing in the world, both mining and synthetic pro- duction of graphite is increasing [8], [9] and each has its own advantages, though there does not seem to be a consensus as to whether synthetic is more sustainable than mined graphite [9], though working in mines comes with occupational hazards and often the destruction of land.

The distinction between graphite and graphene is not precise, there is no distinct limit where it goes from behaving like a 2D material to a 3D material. The electronic structure of graphene quickly changes according to the number of layers, and when there are 11 or more layers the difference in band overlap compared to bulk graphite is less than 10 % [10]. For this reason it is often considered that graphene needs to have 10 or fewer layers. More than 10 layers and it will be considered graphite.

The carbon atoms that make the graphene lattice are sp

-hybridised, shown in Figure 2a.

The sp

-orbitals are angled 120

to each other, shown in Figure 2b, which demonstrate how the

graphene lattice is built by these orbitals. When a bond is made between two sp

-orbitals they form

what is called a σ-bond. In the sp

-hybridisation there are also two p-orbitals that are perpendicu-

lar to the sp

-orbitals, a bond between two p-orbitals is called a π-bond. The interaction between

two π-bonds (π—π-interaction) is the attractive force that keeps graphene layers attracted to each other in graphite.

(a) Side view (b) Top view (c) Two sp

-hybridised atoms bound to each other. Two p-orbitals create a π-bond and two sp

-orbitals create a σ -bond Figure 2: sp

-hybridisation

Carbon can also be sp

-hybridised, shown in Figure 3a. Each atom has four sp

-orbitals, which are tetrahedrally configured and are angled 109.5

to each other. A lattice of sp

-hybridised carbon atoms is not flat like graphene, shown in Figure 3b, it also lacks p-orbitals, which means that it cannot form π-bonds.

(a) Side view (b) A network of sp

-hybridised atoms Figure 3: sp

-hybridisation

1.2 Graphene Oxide

One of the major challenges with graphene, and some argue is the reason it has not found many commercial applications yet [11], is that it is difficult to synthesise at high quantity with satisfac- tory properties. For this reason a lot of research has gone into how to synthesise graphene.

One of the main drawbacks of graphene is that it is difficult to synthesise in large quantities

with high purity. One method is to chemically oxidise graphite. Graphite is simply put, many

layers of graphene. The oxidation leads to, amongst others, hydroxyl groups and epoxy groups to

bind to the edges of the graphite and between the individual layers of graphene that make up the

graphite crystal [12]. This process changes the forces between the graphene layers, which can then

be exfoliated to separate the individual layers of graphene oxide (GO). However, these GO-layers

have a lot of defects and different functional groups bound to its edges and surface, which reduce the desired properties of graphene.

Chemical oxidation of graphite is a promising methods of producing graphene in a large scale.

GO can then be reduced to what is known as reduced graphene oxide (rGO). The oxidation of graphite is shown in Figure 4a and 4b. Oxidising graphite is not a new procedure, it was first discovered in 1860 by Brodie [13]. Though it is relatively recently that the possibilities of pro- ducing rGO from graphite has been discovered. One of the main advantages of this route comes from the fact that GO is hydrophilic, while pristine graphene is hydrophobic [14]. This means that water dispersions of GO can have higher concentrations compared to pristine graphene dis- persions [15]. Because of this it is easy to reduce much GO in a relatively small volume. When the GO is reduced the rGO becomes more hydrophobic as it loses its functional groups [12], [16].

There are several methods of creating GO, Hummers’ method [17], Staudenmaier’s method [18]

and Brodie’s method [13]. The most common is using a modification of Hummers’ method.

The spacing between the graphite layers is 0.336 nm [19], whereas the thickness of a single GO-sheet is roughly 1 nm [20], [21], [22], shown as an increase in distance between the graphite layers in Figure 4b. This separation is caused by functional groups binding to the edges and on the graphene basal planes. When the distance between the layers of graphite are increased as in Figure 4b they can be exfoliated by ultrasonication. The resulting single layers of oxi- dised graphite are what is known as graphene oxide. The functional groups are epoxide and hydroxyl groups that most likely is bound to the basal planes while carbonyl groups are bound to the edges [23], [24], [25], [26], [27]. This is known as the Lerf-Klinowski model. In a study by Boukhvalov and Katsneslon on the oxidation of graphite they found that 75 % coverage of graphene flakes with functional groups is energetically favourable [28].

These defects in the form of functional groups distort the perfectly flat lattice and the carbon atoms can become sp

-hybridised, sp

-hybridisation is shown in Figure 3a. The sp

-hybridisation alters the van der Waals-forces between the graphene layers in graphite. This sp

-hybridisation together combined with functional groups makes it easier to separate individual layers of GO [29].

The distortion of the graphene lattice is illustrated by an epoxide on a graphene lattice in Figure 5b.

Without the epoxide the lattice is completely flat as in Figure 5a.

The typical C/O ratios of GO is in the range 2:1–4:1 [23], [30], [31]. The oxidation also leads to a destruction of the defect-free graphene lattice.

GO in a water dispersion is brown, a ultraviolet-visible spectroscopy (UV-Vis) measurement

of GO is shown in Figure 6.

(a) Graphite

Oxidation

O H

O HO

OH COOH

OH COOH OH

O

COOH OH

(b) Graphite oxide Exfoliation

O OH

O

OH O

COOH COOH

OH OH

OH COOH O

H

(c) Graphene oxide Reduction

OH O

COOH OH

O H

(d) Reduced graphene oxide

Figure 4: An example of how the reaction can look like when going from graphite to reduced graphene oxide

(a) A graphene lattice (b) A graphene lattice with one epoxide

Figure 5: An illustrative example of how a graphene layer distorts with the presence of a single

epoxide on its basal plane

0 0,1 0,2 0,3 0,4 0,5 0,6

400 450 500 550 600 650 700 750 800

Absorption (a.u.)

Wavelength (nm)

Figure 6: UV-Vis spectra of graphene oxide showing its brown colour in solution

1.3 Reduced Graphene Oxide

To restore the graphene structure reduction of GO is performed to form rGO. Reduction ideally causes all functional groups to be removed and completely restores the hexagonal graphene struc- ture, which contains only sp

-hybridised carbon-carbon bonds. There are many ways of reduc- ing GO to rGO, some of which are thermal reduction [32], chemical reduction [33], solvother- mal reduction [34], photocatalytic reduction [35], microwave irradiation and hydrothermal reduc- tion [36]. Each reduction technique comes with their own advantages and disadvantages.

The two most common ways of reducing GO are chemical and thermal reduction [37]. Thermal reduction often requires high temperatures and the resulting rGO can be a brittle material, though the final material can have good barrier properties [38]. Chemical reduction involves reacting the GO with a reducing agent such as hydrazine [20] or ascorbic acid (also known as vitamin C) [39], [40]. This reduction process does not require high temperatures, but procedures and results vary greatly depending on what reducing agent was chosen. The big advantage of chemical reduction is that it is relatively easy to scale up to industrial production.

Using computational methods Boukhvalov et al. found that GO can be reduced from 75 % to 6.25 % relatively easy, the latter corresponding to a C/O-ratio of 16:1. However, a higher level of reduction is difficult to achieve [28].

1.3.1 Thermal Reduction

Thermal reduction of GO is the process of thermally annealing dry GO powder to remove the

functional groups from the GO-sheets. This is an energy consuming process as the temperature to

achieve a high C/O ratio is high, at least 1000

C [41], [42], [43].

There are advantages with thermal reduction over chemical reduction, one of them being that the rGO-sheets can have better barrier properties [38]. However, there is also contrary evidence, suggesting that thermally reducing GO leads to macropores due to escaping gases created by the reduction process [44]. Thermal reduction cannot be performed in a solution, so the GO needs to be placed on a substrate, preferably the substrate, which is to be the final application for the rGO.

There are studies reporting that thermally reduced GO also leads to a more brittle material [38]

when compared to chemical reduction. It has been suggested that this is due to rapid heating, which leads to gasses such as H

, CO and CO

being formed. These gases will in turn escape from the material and in the process raise the sheets of rGO, which stay in this raised position when dried [41].

There are other drawbacks with thermal reduction compared to other reduction techniques. It requires high temperatures, often for a long time. This makes it energy inefficient. The substrate needs to be able to handle these high temperatures, and if the substrate is necessary final application it needs to be able to withstand the heat treatment. This limits the number of substrates suitable for thermal reduction of GO.

One application for thermally reduced GO could be electronic devices as high conductivity due to n-type doping have been reported [41].

1.3.2 Chemical Reduction

Chemical reduction is a reduction of GO that is based on the interaction between GO and a re- ducing agent, which means that it can be performed in a great number of different ways. The most common technique is to reduce using hydrazine as it is a strong reducing agent, the reduc- tion mechanisms of which was first proposed by Stankovich et al. [20]. However, it is toxic and explosive [41], [44]. Hydrazine also has a risk of creating rGO that can become nitrogen-doped, which reduces the desired properties of graphene [45]. Because of this, many other reducing agents for GO have been investigated in recent research. Amongst them are reducing with ascor- bic acid (AA) [33], [39], [40], [46], [47], oxalic acid [48], green tea extract [49], chitosan [50], and TiO

[51], which can also be combined with ultraviolet (UV)-light to increase reduction effi- ciency [35], [36], [37].

An advantage with chemical reduction compared to thermal reduction is that it can be per- formed at ambient temperature or under mild heating. Depending on the reducing agent it can be selective in the reduction, it maintains the structure of the carbon plane [41] and if the rGO is stable in solution after reduction it is easy to create macroscopic structures by processes such as filtration [44]. One drawback with chemical reduction, if done in a water solution, is that graphene is more hydrophobic than GO. This can lead to aggregation of the synthesised rGO if it is not functionalised to make it more hydrophilic [12], [16], [41], [52].

It is of interest to have a shorter reduction time for chemical reduction, as longer reduction times can cause the final rGO to have more tears and be more crumpled [22].

Guex et al. has found that there are two distinct areas of reduction when chemically reducing

GO with sodium borohydride by measuring the reduction using electrical conductivity. Using X-

ray photoelectron spectroscopy (XPS) they could determine that the initial steep reduction was a

reduction of C O-bonds, during this time the amount of C O-bonds were constant. After some time of reduction the decrease in C O-bonds stopped and the amount of C O-bonds started decreasing instead [53].

1.4 Ascorbic Acid

Ascorbic acid (AA), also known as vitamin C, is a non toxic, natural antioxidant and a reducing agent that can be used in chemical reduction of GO [33], [41].

OH O O O H

H O H

O H

(a) Ascorbic acid

O O O O H

H O H

O

(b) Dehydroascorbic acid Figure 7: The molecules of ascorbic acid and dehydroascorbic acid

Fernandez-Merino et al. showed that AA can result in rGO that is reduced in the same time and has electrical conductivity of the same magnitude as GO reduced by hydrazine [33]. This result could indicate that it is unnecessary to have a lower redox potential than AA to reduce GO. AA has the standard redox potential E

= −0.39 V [54], which is roughly one fourth for that of hydrazine with E

= −1.49 V [55].

De Silva et al. have found two distinct regions of reduction of GO using AA. They tracked the reduction of GO using XPS and UV-Vis. Using these techniques they found that the first regime of reduction lead to a reduction in C O-bonds and that the second regime reduces the amount of C O-bonds [22]. This is in agreement with the results reported by Guex et al. [53].

After the AA has reduced GO to rGO the remaining molecule is not AA but dehydroascorbic acid (DHA), Figure 7b. The suggested reduction paths of hydroxyl and epoxide groups present in GO using AA is shown in Figure 8.

AA greatly absorbs ultraviolet-light, evident from Figure 9, but does not absorb visible light

making it a transparent liquid.

C O

C +

OH O O O H

H O H

O H

O

C C

OH O O

O H H O H

O H

-H

O

O C C

O O O H

H O H

O

C C +

O O O O H

H O H

O

(a) The reduction of an epoxide group

C OH

C OH

+

OH O O O H

H O H

O H

-H

O

O

C C

OH O O

O H H O H

O H

-H

O

O C C

O O O H

H O H

O

C C +

O O O O H

H O H

O

(b) The reduction of two OH groups

Figure 8: The reduction mechanisms of ascorbic acid suggested by Gao et. al. Adapted from [40,

Fig 6]

0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2

200 300 400 500 600 700 800

Absorption (a.u.)

Wavelength (nm)

Figure 9: UV-Vis absorption spectra of ascorbic acid

1.5 Poly(vinyl alcohol)

Poly(vinyl alcohol) (PVOH) is a common water based binding agent that is used in the formulation of films that has many applications. PVOH is non toxic and environmentally friendly [56], [57].

PVOH is a polymer that consists mainly of 1,3-diol units, its chemical structure is shown in Figure 10. It is a crystalline material that is highly soluble in water. The two main properties of PVOH that affect the solubility in water is the degree of polymerisation and degree of hydrolysis.

The lower the polymerisation, the more soluble the PVOH is in water. The higher the level of hydrolysis the more soluble the PVOH is in water [58].

O H













n

Figure 10: The structure of poly(vinyl alcohol)

0 0,005 0,01 0,015 0,02 0,025

400 450 500 550 600 650 700 750 800

Absorption (a.u.)

Wavelength (nm)

Figure 11: UV-Vis spectra of poly(vinyl alcohol)

2 Materials and Methods

2.1 Characterisation Techniques

Several different characterisation techniques can be used to determine the reduction of GO to rGO.

Presented below are the techniques that have been used in this thesis, how they work and what information they provide.

2.1.1 Visual Inspection

The easiest way of quickly determining if GO has started reducing is by visual inspection. Dis- persing GO in a liquid gives the dispersion a dark brown colour as in Figure 12a. When GO is reduced to rGO the dispersion turns black, which can be seen in Figure 12b. This can be used as a quick indicator of whether a reduction reaction has taken place or not.

(a) Unreduced (b) Reduced

Figure 12: A mixture of graphene oxide and ascorbic acid before and after reduction, the graphene oxide-concentration is 0.67 mg/ml

2.1.2 Ultraviolet–Visible Spectroscopy

Ultraviolet-visible spectroscopy (UV-Vis) is a characterisation technique that analyses samples by the absorbance of light in the ultraviolet- and visible range. It works by transmitting a beam of light through a sample. The light used is over a broad range of wavelengths, which covers the visible and ultraviolet ranges. The light that has passed through the sample is detected and analysed.

Depending on what wavelengths were absorbed, and how much they were absorbed by the sample it can be characterised.

For a sample to be able to be analysed by UV-Vis it needs to be somewhat translucent to light.

that no transmitted light reaches the detector. If the sample is a dispersion containing particles, the particles should be small enough not to scatter the incident light.

A = log(I

/I) = εcL (1)

UV-Vis characterisation typically follows the Beer-Lambert Law, Equation 1. Here, A is the absorbance, I

is the intensity of the incident light, I is the intensity of the transmitted light, ε is the extinction coefficient, which is a materials property, c is the concentration of the absorbing sample and L is the path length through the sample. As ε is a constant and L is known from the sample we find that the absorbance is directly proportional to the concentration of the sample. By analysing samples of known concentrations a calibration curve can be created using Equation 1. A calibration curve for GO in water dispersion has been performed in this thesis and can be found in Appendix A.

UV-Vis can be used to analyse the reduction of GO by looking at two characteristic GO-peaks at 230 and 300 nm. The peak at 230 nm appears due to π—π

interactions from C C-bonds. The shoulder peak at 300 nm is due to n—π

from C O-bonds [46]. The reason these peaks appear in UV-Vis is due to plasmon resonance frequencies. As reduction progresses these peaks decay and eventually disappear, and are replaced with a peak around 264 nm [59]. This peak appears due to the graphene network being restored and changing the electronic properties, which in turn changes the plasmon resonance frequency [60]. However, AA absorbs greatly in the UV-range, see Figure 9, and overshadows the characteristic peaks for GO and rGO. Because of this, it was decided to limit the measured absorption between 800 and 400 nm.

The instrument used for the UV-Vis-measurements was a LAMBDA 650 UV/Vis Spectropho- tometer from PerkinElmer. For all UV-Vis measurements of solutions it was calibrated against Milli-Q to remove the absorption of water.

2.1.3 Fourier-Transform Infrared Spectroscopy

Fourier-transform infrared spectroscopy (FTIR) is a characterisation technique that utilises infrared (IR) radiation. Just as with UV-Vis, this technique relies on measuring the absorption of light at different wavelengths. An IR emitter sends a beam into a crystal at such an angle so that the radiation is totally reflected inside the crystal. A sample is placed directly on top of the crystal.

Even though there is total reflection inside the crystal an evanescent wave is formed that can be used to analyse the sample. This is the foundation of FTIR in attenuated total reflectance (ATR) mode and is schematically shown in Figure 13.

d

= λ

2π(n

sin

θ − n

)

(2)

The depth that the evanescent waves penetrate into the sample is given by Equation 2. d

is the

depth of penetration, λ is the wavelength, n

are the refractive indices of the crystal and sample,

and θ is the angle of incident to the normal of the sample. Typically d

is 0.1 λ, though by changing

the incident angle and refractive index of the crystal the depth of penetration can range between

0.12–0.05 λ [61].

Figure 13: A schematic of how ATR works

A major drawback with FTIR in ATR mode is that samples containing water, or are in water solutions, are difficult to analyse as water absorbs strongly over many bands, see Figure 14.

0 10 20 30 40 50 60 70 80 90 100

400 800 1200 1600 2000 2400 2800 3200 3600 4000

% T ransm issio n

Wave number (cm^-1) O-H

Streching

O-H Scissoring

Figure 14: FTIR spectrum of water

The instrument used for the FTIR-measurements was a Spectrum One FT-IR Spectrometer from PerkinElmer. The ATR crystal used for this model was a diamond. The spectra presented in this thesis have been processed by an automated data processing technique called ”Data Tune up”

in the software for the Spectrum One. This process corrects distortions caused by the ATR.

2.1.4 Raman Spectroscopy

Raman spectroscopy is a characterisation technique that uses IR radiation to characterise the chem- ical bonds in a sample. Even though Raman and FTIR both analyse a sample using IR radiation, they are fundamentally different. Raman spectroscopy is dependent on the molecules in a sam- ple inelastically scattering the IR radiation. FTIR is dependent on the absorption of IR radiation.

This difference leads to the two techniques measuring two different properties of samples. Raman

spectroscopy measures the change in Raman polarisability of a sample, while FTIR measures the

When GO is analysed using Raman spectroscopy there are two characteristic bands, called the D-band and G-band. The D-band is centred at ∼ 1350 cm

and the G-band is centred at

∼ 1582 cm

. These bands are related to the structure of graphene. The G-band appears due to sp

-domains, the structure of pure graphene, while the D-band is due to sp

-domains, typical for GO. By looking at the intensities of these bands denoted as I

and I

the reduction of GO can be followed [20], [62].

The instrument used for Raman spectroscopy was a Witec Alpha300 RAS.

2.2 Experimental Techniques

The reduction of GO was first performed in a water dispersion, then as reduction of GO in PVOH- films. This was done as the reduction in water dispersion is quicker than in films and therefore more useful to find reaction kinetics. The parameters that were thought to affect the reduction of GO were altered and the reduction was tracked in dispersion reduction.

The parameters that were altered were the reaction temperature, the ratio between the GO and AA, and the pH of the reaction. Previous studies have found that having an alkaline solution when reducing the GO should keep the GO and the rGO more stable in dispersion before and after reduction, respectively [33].

All ratios that are presented in this thesis have been calculated gravimetrically.

2.2.1 Solution Reduction

The solution reduction of GO into rGO was performed by mixing GO dispersion and AA solution in glass reaction vessels containing a magnetic stir bar. The vessels were of varying sizes, from 15 ml up to 500 ml, depending on how many samples were planned to be extracted. If more samples was to be extracted, a bigger vessel was used so that the overall reaction would not to be affected by the extraction of a sample. During the reactions the vessels were sealed with a plastic lid to prevent evaporation.

When taking a sample of the GO:AA-dispersion for measurement with UV-Vis 300 µl of dis- persion was pipetted out and mixed with 9.7 ml Milli-Q water. This diluted sample was then centrifuged at 3000 rpm to sediment the rGO leaving GO in the liquid, as GO is stable in water dispersion. The supernatant was pipetted out and analysed using UV-Vis to determine the level of unreduced GO.

For the reactions at elevated temperatures, the solutions were mixed at room temperature and the reaction vessel was then submerged in a water bath keeping the desired temperature.

Four samples were prepared for solution reduction, the details of which are presented in Table 1.

Table 1: The reaction conditions for the different samples Sample GO:AA-ratio Temperature (

C)

T25 1:10 25

T40 1:10 40

T65 1:10 65

T80 1:10 80

2.2.2 Film Reduction

Two different types of films were cast for this thesis. One film was a mixture of GO and PVOH, which was allowed to dry and then submerged in an AA-solution. The other type of film was a mixture of GO, AA and PVOH. The films were then analysed using UV-Vis, FTIR and Raman spectroscopy.

2.2.3 Chemicals

The chemicals used in this thesis are presented in Table 2. All water used for preparing dispersions and solutions was Milli-Q.

Table 2: The chemicals used in this thesis

Chemical Abbreviation Supplier Additional information

Graphene oxide GO Graphenea Water dispersion

L-Ascorbic acid AA Sigma-Aldrich Reagent grade

Poly(vinyl alcohol) PVOH Fluka M

= 22000, Hydrolysation: 97.5–99.5

3 Results and Discussion

3.1 Finding a Way to Track Graphene Oxide Reduction

The solution reduction of GO started by simply mixing a water dispersion of GO and a water solution of AA to make sure that AA in fact can reduce GO. It was found that AA could reduce GO, which is in agreement with previous research. However, the resulting rGO flocculated.

Figure 15: Floccu- lation of rGO Pure rGO is hydrophobic [14] and the interaction between the hydropho-

bic rGO-flakes cause flocculation, as shown in Figure 15. These flocculates scatter light when measured with UV-Vis. A sample of 300µl was extracted from the reaction vessel, diluted with 9.7 ml Milli-Q and centrifuged in or- der to sediment the flocculated rGO. Since GO is hydrophilic and rGO is hydrophobic, the rGO should sediment and the GO should stay in dispersion after centrifugation. The supernatant was then pipetted out and measured with UV-Vis. Unfortunately the rGO was so light that not all sedimented dur- ing centrifugation and some stayed in suspension after centrifugation. This caused some rGO to be pipetted out with the supernatant that should contain only GO without any rGO.

The rGO is black, meaning that it absorbs all visible wavelengths of light equally, and when the supernatant was measured with UV-Vis it raised the absorption equally over the measured range. This effect is seen in Figure 16 where the absorption for 93 h reduction had the highest value at 800 nm. For this reason it was necessary to find a different way of tracking the reduction that is unaffected by rGO particles being pipetted out with the supernatant.

0 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0,45 0,5

400 450 500 550 600 650 700 750 800

Absorption (a.u.)

Wavelength (nm)

0 h 1 h 2 h 3 h 24 h 93 h 117h 168 h

Figure 16: All measured absorptions for sample T25

3.1.1 Reduction Tracking by Absorption Ratio

For the reasons explained in section 3.1, a different method than simply looking at the absorption at a single wavelength was necessary. The absorption spectra of pure GO is presented in Figure 17a and it is clearly seen that there is a curvature to the absorption spectra, this corresponds to the brown colour of the GO. The absorption spectra for rGO, shown in Figure 17b, is close to being completely flat, the reason for this is that rGO is black, which means that it absorbs all wavelengths equally in the visible range.

0 0,1 0,2 0,3 0,4 0,5 0,6

400 450 500 550 600 650 700 750 800

Absorption (a.u.)

Wavelength (nm)

(a) Graphene oxide

0 0,1 0,2 0,3 0,4 0,5 0,6

400 450 500 550 600 650 700 750 800

Absorption (-)

Wavelength (nm)

(b) Reduced graphene oxide

Figure 17: UV-Vis absorption spectra of unreduced graphene oxide and reduced graphene oxide of the same sample

This difference between GO and rGO can be used to track the reduction of GO. For this purpose absorption ratio (AR) was created for this thesis and is defined as in Equation 3.

Absorption ratio (AR) = Absorption

nm Absorption

nm

(3)

To remove the dependence of how much rGO was pipetted out with the supernatant after cen-

trifugation, all absorption spectra needed to be vertically shifted to a common absorption value at

800 nm, demonstrated in Figure 18b.

0 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0,45 0,5

400 450 500 550 600 650 700 750 800

Absorption (a.u.)

Wavelength (nm)

0 h 1 h 2 h 3 h 24 h 93 h 117h 168 h

(a) Absorption over time

400 450 500 550 600 650 700 750 800

Absorption (a.u.)

Wavelength (nm)

0 h 1 h 2 h 3 h 24 h 93 h 117 h 168 h

(b) Absorption over time with vertically sepa- rated curves

Figure 18: UV-Vis absorption spectra of sample T25

1 2 3 4 5 6 7 8

0 20 40 60 80 100 120 140 160

Absorption ratio

Reduction time (h)

Figure 19: The absorption ratio for sample T25 plotted against reduction time

By then plotting the AR against reduction time, shown in Figure 19, the progression of the reduction is seen. When full reduction is achieved the AR should be have the value 1 as this would be a completely flat line. AA and water have no absorption in the visible range, the rGO should be centrifuged to the bottom and not be extracted when the supernatant is pipetted out and the GO should be consumed. All of this means that when the supernatant is measured with UV-Vis when the reduction is complete should be a flat line, which equals an AR of 1.

As it was found that AR is a viable way of tracking the solution reduction of GO more samples

were prepared. The two parameters that were suspected to have the greatest effect on the reduction

time was the reaction temperature and the ratio between GO and AA.

3.2 Variables Affecting the Reduction Speed

The reduction of GO is dependent on variables apart from time, amongst these are reaction tem- perature, the ratio between GO and AA [45], and the pH of the reaction.

The ratio between the GO and the AA was investigated by tracking the reduction using AR.

Three samples were prepared with GO:AA-ratios 1:5, 1:10 and 1:100 and were reduced at 25

C.

There was a difference in reduction rate, higher concentration of AA lead to a faster reduction.

Though, this effect was found to be very small compared to the increase in reduction rate that came with increasing the reaction temperature. For this reason a GO:AA ratio of 1:10 was chosen and emphasis was placed on the temperature dependence.

Tests regarding the stability increase of GO dispersions at an alkaline pH-value was also inves- tigated. By centrifugation of GO dispersions at alkaline pH it was found that the higher pH lead to a decrease in stability of GO water dispersion. This is not desirable, therefore the variable of pH was not investigated further. The result of the centrifugation can be found in Appendix A.

3.2.1 The Temperature Dependence on Reduction Speed

The reduction of GO is highly temperature dependent, which is evident from Figure 20 and Fig- ure 21. This is in agreement with literature [45]. When reducing at 25

C it reaches an AR of 1.282 after 168 h, seen in Figure 21a. The same AR (1.287) is achieved after 6 minutes at 80

C, seen in Figure 21d.

Sample T40 had a different starting AR compared to the other samples. Since all samples have the same concentration and ratio of chemicals this should not happen, the cause of this is unknown.

1 2 3 4 5 6 7 8

0 20 40 60 80 100 120 140 160

Absorption ratio

Reduction time (h)

T25 T40 T65 T80

Figure 20: The absorption ratio for different temperatures plotted against reduction time

1 2 3 4 5 6 7 8

0 20 40 60 80 100 120 140 160

Absorption ratio

Reduction time (h)

(a) Absorption ratio of T25

1 2 3 4 5 6 7 8

0 5 10 15 20 25

Absorption ratio

Reduction time (h)

(b) Absorption ratio of T40

1 2 3 4 5 6 7 8

0 0,5 1 1,5 2 2,5 3

Absorption ratio

Reduction time (h)

(c) Absorption ratio of T65

1 2 3 4 5 6 7 8

0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0,16 0,18

Absorption ratio

Reduction time (h)

(d) Absorption ratio of T80 Figure 21: Absorption ratios of samples T25, T40, T65 and T80

3.3 Reaction Regimes

When trying to fit the reduction of GO as a textbook reaction, zeroth, first or second order, it became apparent that the reactions did not fit any of the textbook reactions. This lead to the conclusion that it was in fact not one reaction that is taking place. Instead there were two, distinct

”regions” of reaction.

The data obtained in this thesis has an ”initial reaction” before the ”first regime” starts, clearly seen in Figure 22(a-c). In this reaction the AR increases for one sample and decreases for other, however it then returns to the initial value. The author of this thesis has not been able to find any publication with similar results. The first regime is due to the reduction of C O-bonds, in Figure 8 it is clearly shown that there are two different reduction mechanisms. One mechanism for OH-groups and one mechanism for epoxides, both of these are reduction of C O-bonds. However, after the first reduction step for OH-groups the following reduction mechanics are exactly the same.

The initial reaction could be the the first step in the reduction mechanism of two OH-groups, which is the reduction of one OH-group to release a water molecule.

The AR for the transition between the first and ”second regimes” are temperature dependent,

the AR for the transitions are presented in Table 3 and is plotted in Figure 23. The reason for

why this is interesting is that all samples have the same GO source, which means that the fraction

between C O and C O-bonds is the same for all sample. Guex et al. [53], and De Silva et

al. [22] have independently shown that the first regime is the reduction of C O-bonds while the second regime is reduction of C O-bonds. The data presented in Table 3 suggests that the amount of C O-bonds is temperature dependent since the transition AR is different depending on the reaction temperature. This is of course not true, the ratio between C O and C O is constant for all samples.

This finding indicates that the rate of reduction of C O-bonds are more temperature dependent than C O-bonds. This effect is previously unreported and has a great effect on the speed of reduction of GO. This could mean that the reduction temperature could be chosen in such a way that the final rGO has more or less C O-bonds, which would change the properties of the produced rGO. Reduction at a lower temperatures would then be preferable if it is desired to keep the edge functionalities of the GO in the rGO.

In Figure 23 the transition AR is plotted against the temperature at which the reduction was performed. The data is however not sufficient to define the relationship and more experiments needs to be performed to find the relationship between these variables.

Table 3: The transition AR between the first and second regime

T25 T40 T65 T80

Transition AR 4.01 1.74 1.72 1.29

1 2 3 4 5 6 7 8

0 20 40 60 80 100 120 140 160

Absorption ratio

Reduction time (h)

Initial reaction 1:st regime 2:nd regime

(a) Regimes of T25

1 2 3 4 5 6 7 8

0 5 10 15 20 25

Absorption ratio

Reduction time (h)

Initial reaction 1:st regime 2:nd regime

(b) Regimes of T40

1 2 3 4 5 6 7 8

0 0,5 1 1,5 2 2,5 3 3,5

Absorption ratio

Reduction time (h)

Initial reaction 1:st regime 2:nd regime

(c) Regimes of T65

1 2 3 4 5 6 7 8

0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0,16 0,18

Absorption ratio

Reduction time (h)

Initial reaction 1:st regime 2:nd regime

(d) Regimes of T80

Figure 22: Absorption ratios of samples T25, T40, T65 and T80 showing the separate regimes

0 0,5 1 1,5 2 2,5 3 3,5 4 4,5

290 300 310 320 330 340 350 360

Transition absorption ratio (a.u.)

Temperature (K) T25

T65

T80 T40

Figure 23: The transition absorption ratio plotted against temperature

3.4 Calculating the Activation Energies

The reduction of GO is temperature dependent, this is shown in Section 3.2.1. When there is a temperature dependence of a reaction the activation energy (E

) can be calculated. The first step to calculate the E

is to find the rate constant, k. The rate constant is calculated in different ways if the reaction is zeroth, first or second order. The author of this thesis has not been able to find any literature on the reaction order of GO reduction.

Kim et al. has found the activation energy of GO reduction to be 70.58 kJ [63], though they have not stated wether reaction is a zeroth, first or second order. The activation energy for oxidation of AA is 48.2 kJ/mol [64].

To calculate the E

the Arrhenius equation is used, Equation 4.

k = Ae

(4)

Taking the natural logarithm of Equation 4 gives Equation 5.

ln k = ln A − E

R

1

T (5)

Equation 5 is used to calculate the E

and pre-exponential factor, A. T is the absolute temper- ature of the reaction and R is the gas constant. The pre-exponential factor is a number that shows the reactivity of the reaction. The higher A is, the more often molecules have the correct energies and orientation for the reaction to occur.

As the author of this thesis has not been able to find any information on whether the reduction

of GO is a first or second order reaction, the E

and A-values have been calculated assuming both first order and second order reactions.

3.4.1 Assuming First Order Reaction

To calculate the rate constant for a first order reaction the concentration fraction is plotted against time, in the case of this thesis the concentration has been substituted with AR. An example of this calculation is presented in Table 4. The AR fraction is calculated by dividing the AR at the desired time with the AR at time 0 h.

Table 4: An example of how to calculate the concentration fraction of sample T65

0 h 0.5 h 1 h 1.5 h 2 h 2.5 h 3 h

AR 7.160 6.814 3.051 1.717 1.565 1.307 1.212

AR fraction 1 0.952 0.426 0.240 0.219 0.183 0.169 ln(AR fraction) 0 -0.049 -0.853 -1.428 -1.521 -1.701 -1.776

Plotting the ln(AR fraction) against reduction time gives a slope, shown in Figure 24. The negative of this slope is the rate constant, k, for the whole T65 reaction. However, it is obvious that the linear fit does not follow the calculated values in Figure 24. As discussed in section 3.3 the reaction is divided into two regimes.

y = -0,6643x - 0,0505 R² = 0,8932

-2 -1,8 -1,6 -1,4 -1,2 -1 -0,8 -0,6 -0,4 -0,2 0

0 0,5 1 1,5 2 2,5 3 3,5

ln(AR fraction)

Reduction time (h)

Figure 24: Plotting ln(AR fraction) against reduction time for sample T65, the linear fit is poor to the calculated values demonstrating the need for separation into two regimes

The final step is to calculate ln(k) and the reciprocal of the absolute temperature at which

the reaction was performed. For the T65 reaction the ln(k) is 0.321 and 1/T is 1/(65 + 273) =

0.00295 1/K. By plotting these values for all reactions at different temperatures the E

is found by multiplying the slope of this curve by R, where R is the gas constant.

y = -0,0284x + 0,1669 R² = 1

y = -0,0084x - 0,3112 R² = 0,9118

-2 -1,8 -1,6 -1,4 -1,2 -1 -0,8 -0,6 -0,4 -0,2 0 0,2

0 20 40 60 80 100 120 140 160

ln(AR fraction)

Reduction time (h)

Initial reaction First regime Second regime

(a) Regimes of T25

y = -0,3899x + 0,8185 R² = 0,9773

y = -0,0174x - 1,0273 R² = 1

-1,6 -1,4 -1,2 -1 -0,8 -0,6 -0,4 -0,2 0

0 5 10 15 20 25 30

ln(AR fraction)

Reduction time (h)

Initial reaction First regime Second regime

(b) Regimes of T40

y = -1,3787x + 0,6019 R² = 0,9909

y = -0,245x - 1,0553 R² = 0,9748

-2 -1,8 -1,6 -1,4 -1,2 -1 -0,8 -0,6 -0,4 -0,2 0

0 0,5 1 1,5 2 2,5 3 3,5

ln(AR fraction)

Reduction time (h)

Initial reaction First regime Second regime

(c) Regimes of T65

y = -23,766x + 0,8944 R² = 0,9221

y = -2,1404x - 1,5105 R² = 0,6333

-2 -1,8 -1,6 -1,4 -1,2 -1 -0,8 -0,6 -0,4 -0,2 0

0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0,16 0,18

ln(AR fraction)

Reduction time (h)

Initial reaction First regime Second regime

(d) Regimes of T80

Figure 25: Absorption ratios of samples T25, T40, T65 and T80 showing the separate regimes

Table 5: Calculated Arrhenius parameters for a first order reaction First regime Second regime

E

96.1 kJ 88.2 kJ

A 2.42 ∗ 10

1/s 1.52 ∗ 10

1/s

R

0.940 0.957

Table 6: R

-values for the two regimes for a first order reaction

Sample k-value first regime R

first regime k-value second regime R

second regime

T25 0.0283 1 0.00839 0.912

T40 0.390 0.977 0.0174 1

T65 1.379 0.991 0.245 0.975

T80 23.77 0.922 2.140 0.633