<%BANNER%>

Optimization of Activated Carbon for Natural Color Removal

Permanent Link: http://ufdc.ufl.edu/UFE0025143/00001

Material Information

Title: Optimization of Activated Carbon for Natural Color Removal
Physical Description: 1 online resource (56 p.)
Language: english
Creator: Joiner, Matthew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This work focuses on the creation of powdered activated carbon (PAC) from sawdust and rice husks, two agricultural wastes, using both physical and chemical means. Physical means included a traditional charring stage using nitrogen followed by activation with either steam or carbon dioxide; chemical activation focused on the use of anhydrous potassium hydroxide as an activating agent. The performance of the activated carbon created was evaluated by its ability to decolorize commercially-available apple juice typically found at grocery stores, with an existing commercially-available PAC used as a control. The created activated carbon was tested for its ability to decolorize apple juice as well as Methylene Blue and Congo Red laboratory dyes, with selected samples analyzed for specific surface area, total pore volume, and average pore size using nitrogen adsorption. Generally, sawdust responded favorably to physical activation, with the rice husks responding more favorably to activation with potassium hydroxide. However, activation using raw precursors, which saves time and energy by avoiding the charring stage, proved unsuccessful.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Matthew Joiner.
Thesis: Thesis (M.E.)--University of Florida, 2009.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025143:00001

Permanent Link: http://ufdc.ufl.edu/UFE0025143/00001

Material Information

Title: Optimization of Activated Carbon for Natural Color Removal
Physical Description: 1 online resource (56 p.)
Language: english
Creator: Joiner, Matthew
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This work focuses on the creation of powdered activated carbon (PAC) from sawdust and rice husks, two agricultural wastes, using both physical and chemical means. Physical means included a traditional charring stage using nitrogen followed by activation with either steam or carbon dioxide; chemical activation focused on the use of anhydrous potassium hydroxide as an activating agent. The performance of the activated carbon created was evaluated by its ability to decolorize commercially-available apple juice typically found at grocery stores, with an existing commercially-available PAC used as a control. The created activated carbon was tested for its ability to decolorize apple juice as well as Methylene Blue and Congo Red laboratory dyes, with selected samples analyzed for specific surface area, total pore volume, and average pore size using nitrogen adsorption. Generally, sawdust responded favorably to physical activation, with the rice husks responding more favorably to activation with potassium hydroxide. However, activation using raw precursors, which saves time and energy by avoiding the charring stage, proved unsuccessful.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Matthew Joiner.
Thesis: Thesis (M.E.)--University of Florida, 2009.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025143:00001


This item has the following downloads:


Full Text

PAGE 1

1 OPTIMIZATION OF ACTIVATED CARBON FOR NATURAL COLOR REMOVAL By MATTHEW CHARLES JOINER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MAS TER OF ENGINEERING UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 M atthew J oiner

PAGE 3

3 To my parents, and also to my roommates, Kyle and Chad, w ho helped me when I needed help

PAGE 4

4 ACKNOWLEDGMENTS I would first and foremost lik e to thank Dr. David W. Mazyck, who allowed me the opportunity to study at this incredible department and who guided me throughout the course of this research Next I would like to thank my committee, Dr. James P. Heaney and Dr. Treavor H. Boyer, who took the time to review my work and offer helpful criticism, and also the people I had an opportunity to work with, including but not limited to Timothy English II, Vanessa Pineda, Kyle Ulmer, Alec Gruss, Heather Byrne, and Dr. Beau Kostedt. Special thanks to Paul Carpinone for his help with the electron microscope images.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .................... 4 LIST OF TABLES ................................ ................................ ................................ ................................ 6 LIST OF FIGURES ................................ ................................ ................................ .............................. 7 ABSTRACT ................................ ................................ ................................ ................................ .......... 8 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ......................... 9 2 ACTIVATED CARBON: A BRIEF OVERVIEW ................................ ................................ ... 11 Adsorption Process ................................ ................................ ................................ ...................... 11 Production Process ................................ ................................ ................................ ...................... 13 Physical Activation ................................ ................................ ................................ ..................... 14 Chemical Activation ................................ ................................ ................................ ................... 15 Surface Chemistry ................................ ................................ ................................ ....................... 17 Apple Juice Decolorization ................................ ................................ ................................ ........ 20 Sugar Decolorization ................................ ................................ ................................ ................... 22 3 EXPERIMENTAL PROCEDURE ................................ ................................ ............................ 24 Raw Materials ................................ ................................ ................................ .............................. 24 Charring ................................ ................................ ................................ ................................ ....... 24 Particle Processing ................................ ................................ ................................ ...................... 27 Activation ................................ ................................ ................................ ................................ .... 27 Carbon Characterization ................................ ................................ ................................ ............. 30 4 RESULTS ................................ ................................ ................................ ................................ .... 31 Charring ................................ ................................ ................................ ................................ ....... 31 Steam Activation ................................ ................................ ................................ ......................... 34 Chemical Activation ................................ ................................ ................................ ................... 38 Carbon Dioxide Activation ................................ ................................ ................................ ......... 44 Surface Charge Modification ................................ ................................ ................................ ..... 45 Surface Area Assessment ................................ ................................ ................................ ........... 46 5 CONCLUSIONS ................................ ................................ ................................ ......................... 51 LIST OF REFERENCES ................................ ................................ ................................ ................... 53 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ............. 56

PAGE 6

6 LIST OF TABLES Table page 4 1 Rice husks activated with steam ................................ ................................ ............................ 38 4 2 Chemical activations of sawd ust ................................ ................................ ........................... 43 4 3 Surface modification ................................ ................................ ................................ .............. 46 4 4 CA 50 p erformance ................................ ................................ ................................ ................ 47 4 5 BET c haracteri stics of selected s amples ................................ ................................ ............... 48

PAGE 7

7 LIST OF FIGURES Figure page 2 1 Parti cle transport mechanism ................................ ................................ ................................ 12 2 2 Aromatic bo nding ................................ ................................ ................................ ................... 13 2 3 Acidic and basic functio nal groups ................................ ................................ ....................... 18 2 4 Determinat ion of PZC ................................ ................................ ................................ ............ 19 2 5 Polyphenols in apple juice ................................ ................................ ................................ ..... 21 3 1 Furnace used for activated carbon production ................................ ................................ ...... 26 4 1 Sawdust mass loss du ring the charring process ................................ ................................ .... 32 4 2 Congo red and methylene blue adsorption for sawdust chars ................................ ............. 33 4 3 Mass loss of rice husks during the charring process ................................ ............................ 33 4 4 Steam activation of sawdust at 1.6 mL/min ................................ ................................ ......... 35 4 5 Images of activated carbon at high magnifications ................................ .............................. 36 4 6 Steam activation of sawdust at 0.4 mL/min ................................ ................................ ......... 37 4 7 Congo red and methylene blue adsorption for rice husks char ................................ ........... 38 4 8 Solution concentrations used for activation ................................ ................................ .......... 40 4 9 Potassium hydroxide to carbon ratios ................................ ................................ ................... 40 4 10 The impact of char and activation characteristics on apple color removal ......................... 41 4 11 Selected chemical activations ................................ ................................ ................................ 42 4 12 Chemical activations using sawdust ................................ ................................ ..................... 43 4 13 Carbon d io xide activation of s awdust ................................ ................................ ................... 44 4 14 Carbon dioxide activat ion of rice husk ................................ ................................ ................. 45 4 15 BET adsorption/desorption plots (a) CA 50 (B) Sawdust activated with steam for 2 hours at 850 C ................................ ................................ ................................ ....................... 47

PAGE 8

8 Abstract of The si s Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of M aster of E ngineering OPTIMIZATION OF ACTIVATED CARBON FOR NATURAL COLOR REMOVAL By M atthew J oiner December 2009 Chair: David W. Mazyck Major: Environmental Engineering Sciences This work focuses on the creation of powdered activated carbon (PAC) from sawdust and rice husks, two agricultural wastes, using both physical and chemical means. Physical means included a traditional char ring stage using nitrogen followed by activation with either steam or carbon dioxide; chemical activation focused on the use of anhydrous potassium hydroxide as an activating agent. The performance of the activated carbon created was evaluated by its abili ty to decolorize commercially available apple juice typically found at grocery stores, with an existing commercially available PAC used as a control. The created activated carbon was tested for its ability to decolorize apple juice as well as Methylene Blu e and Congo Red laboratory dyes, with selected samples analyzed for specific surface area, total pore volume, and average pore size using nitrogen adsorption. Generally, sawdust responded favorably to physical activation, with the rice husks responding mor e favorably to activation with potassium hydroxide. However, activation using raw precursors, which saves time and energy by avoiding the charring stage, proved unsuccessful.

PAGE 9

9 CHAPTER 1 INTRODUCTION Activated carbon has long been used as a high quality ad sorbent for the removal of contaminants from the air and water phases, and is today used in a variety of settings. Activated carbon is typically manufactured from organic substrates in either a powder or granular form, used as an adsorbent for its intended purpose, and then disposed of or reclaimed using a regenerative process. The primary goal of this research is to use a renewable biomass to create a quality powdered activate d carbon (PAC), and to develop a consistent method for doing so in a laboratory s etting, with the final hope of adopting it for a larger industrial batch process. The produced activated carbon will be analyzed using traditional activated carbon testing, with the results then used to optimize the PAC for the removal of natural color fro m commercial grade apple juice in an effort to compete with commercially available PAC currently on the market. Removal of apple juice color was chosen as an available substitute for the sugar decolorization industry; activated carbon was first used in an industrial setting as part of the process of converting brown sugar to white sugar, and it remains a vital part of that industry. If the process developed via this research proves to be economically viable, it could later be incorporated directly into the sugar decolorization industrial procedure. For this project, two precursors were selected by the supplier: sawdust, which is a traditional precursor for activation, and rice husks, which are more novel. Rice husks are a by product of the rice milling indus try, and account for roughly 20% of the whole rice grain mass. The amount of rice husk waste produced in 1995 was approximately 500 million tons in developing countries, and this amount far exceeds any beneficial uses for the waste, thus posing a disposal problem (Guo et al., 2000 ). It was hoped that an activated carbon could be created

PAGE 10

10 using the rice husks as a precursor, thereby finding a beneficial use for an overly abundant waste product. Sawdust was also studied as an alternative agricultural waste pro duct. The following chapters of this thesis provide an overview of activated carbon, including its structure and traditional methods of production, followed by an overview of the experimental procedure for this researc h, the findings, and finally, the conc lusions that could be drawn.

PAGE 11

11 CHAPTER 2 ACTIVATED CARBON: A BRIEF OVERVIEW Activated carbons are a family of adsorbents formed by the partial combustion and thermal decomposition of carbonaceous materials. Activated carbon is used for a wide variety of a pplications including water purification and decolorization, air purification, deodorants, and industrial usages, such as food processes and the manufacture of pharmaceuticals. Activated carbon is prized for its high adsorptive capacity, due to its high su rface area to mass ratio, extensive microporous pore structure, and potential for surface reactivity. The surface area for a gram of commercially available activated carbon ranges from 800 to 1500 m 2 /g with carbons in the laboratory reaching as high as 30 00 m 2 /g (Bansal 1988) The first recorded use of activated carbon was by the Egyptians as early as 1500 B.C. It was later adapted for commercial use in 1900 for the sugar refining industry; however, wide scale usage did not begin until after the First Wor ld War, when activated carbon was necessary for the construction of gas masks to prevent casualties due to poison ous gas. Today, activated carbon can be made from a number of carbonaceous materials, particularly bituminous coals, lignite, wood, and coconut shells, with a new emphasis on creating activated carbon from renewable biomasses, such as seed husks, olive stones, and bamboo in the interest of sustainability (Warhurst Gonzlez, Hameed ). Adsorption Process Activated carbon owes much of its adsorptiv e capacity to its extensive pore structure. Activated carbon is composed of a series of layers, with each layer constructed from a matrix of aromatic rings. However, unlike graphite or other carbon allotropes, the layers within carbon are not arranged orde rly, with layers lying on different planes. T his disorder results in spaces

PAGE 12

12 between the carbon layers, causing the pores that make activated carbon so valuable (Marsh and Rodriguez Reinoso 2006 ) Contaminant adsorption is composed of four stages: the bulk transport stage, the film transport stage, the interparticle transport stage, and the adsorption attachment stage. Contaminants, either in the air or water phase, come into contact with an activated carbon particle and enter one of its many pores. The por e structure of activated carbon is cavernous, and contaminants will move farther into the pore network until they find a place to adsorb to the carbon surface ( Snoeyink and Jenkins, 1980 ) Figure 2 1. P article transport mecha nism (Chadik) Adsorption occurs at the carbon surface in either a physical or chemical manner. Physical adsorption occurs when contaminants are attracted to the carbon surface without an actual chemical bond being formed, and as a result these physical int ermolecular forces are weaker and can be reversed. Physical adsorption processes include weak intermolecular forces such as H bonding, dipole dipole forces, and van der Waal forces (Marsh and Rodriguez Reinoso 2006 ). Chemical adsorption between the contam inants and the carbon surface occurs less often, but is stronger than physical adsorption and irreversible. The majority of bonds between contaminants and carbon fall into two categories: covalent bonds, which occur at the edges of the

PAGE 13

13 carbon layers and re sult in a positive or negative surface charge, and aromatic bonding, which occurs on the sides of the carbon layer as aromatic molecules (Menendez et al., 1996 ). Aromatic bonding occurs when aromatic molecules, such as benzene or phenol, overlap aromatic m olecules within the carbon matrix. As the molecules approach each other, the initial repulsion and Sanders, 1 990 ). Figure 2 1 Aromatic b onding (Hunter and Sanders, 1990 ) Production Process Production of activated carbon is essentially a two step process. First, a carbonaceous material such as wood is heated under the absence of oxyge n in a step referred to as During the pyrolysis stage, volatile materials within the precursor are gasified and removed, leaving behind a porous carbon structure that can then be modified (Tennant 2003 ). Carbonization usually occurs at temperatures less than 800 C D uring the activation phase some additional volatilization of the carbon precursor will occur independently of activation (Bansal et al., 1988 ). Concerni ng the rate of heating, Ferraz ( 1988 ) found that slowly increasing the

PAGE 14

14 temperature of the furnace results in denser and harder chars, as gases created during volatilization have more time to escape the pore structure, allowing the precursor to shrink. Nex t, the carbon is oxidized at high temperatures (750 to 850 C) under a flow of steam and/ or carbon dioxide. Because the structure of the carbon is irregular, some carbon atoms within the matrix will react chemically while others do not: this selective reac tion allows the creation of a carbon with larger, more randomly distributed pores that have a much greater adsorptive capacity (Bansal et al., 1988 ). Physical Activation Once carbonization has taken place, the pores within the carbon matrix must be widene d in order for them to prove useful; it is this widening of existing pores and the creation of new pores steam, carbon dioxide gas, or some mixture of the two The mechanisms for the two processes are shown below ( Bansal et al., 1988 ). C + H 2 2 (2 1) C + CO 2 (2 2) Some of the carbon monoxide created during the steam activation process will be further oxidized by excess steam. All carbon removed by the activation process will leave the furnace as either CO or CO 2 CO + H 2 2 +H 2 (2 3) Both of the primary activation reactions are endothermic, meaning that a constant supply of energy must be provided by some sort of furnace or kiln. The reaction involving steam has a lower activation energy than the reaction invol ving carbon dioxide; as such, physical activation with steam occurs more quickly, widening existing pores and creating a wider pore size distribution throughout the carbon. Activation with carbon dioxide occurs more slowly, first

PAGE 15

15 creating pores before wide ning them, limiting the carbon to mainly microporous pores that are more uniform in size (Bach 2007 ). Physical activations are typically performed at temperatures ranging from 800 to 1100 C. Azargohar and Dalai (2008) used a furnace temperature ranging f rom 600 to 900 C for a period of 0.9 to 4 hours for the physical activation of spruce char. For activations using rice husks as the precursor, Malik (2003) created char at a temperature of 400 C and activated at a temperature of 600 C for 1 hour. Flows vary; when using carbon dioxide, anywhere from 80 to 200 mL/min seems acceptable ( Pastor Villegas Rodrguez Reinoso ). For steam flow is either measured as the actual steam flow, the amount of water required to create this steam flow, or the steam to carb on ratio. Mazyck and Cannon (2000) used 1.6 mL/min of liquid water as steam for conventional activation (8 g H 2 0/g carbon), and 9.5 The most important point to gather from the literature concerning steam activation is that pore s ize generally increases with activation time, and this pore size has a direct relationship on adsorption capacity (Tennant 2003 ). It has been proposed that organic contaminants will adsorb into pores that have an averag e pore diameter ranging from 1.5 to 2 times that of the contaminant. As steam exposure time increases and pore s gradually widen, the surface area of the carbon as well as the adsorptive capacity of the carbon increases. However, at a certain point the carbon becomes overactivated; pore wall s begin to collapse, and pore sizes exceed the desired size for adsorption (Bach 2007 ). Chemical Activation Chemical activation shares the same goals as physical activation but uses entirely different processes. The carbon precursor is mixed with a soluti on containing a chemical agent and then left to dry, impregnating the carbon with the chemical agent before being placed in a furnace or kiln to cause a chemical reaction. Traditionally, phosphoric acid or zinc chloride have been used

PAGE 16

16 for the chemical acti vation of lignocellulosic materials, while metal compounds such as KOH were used for the activation of coal precursors or chars ( Srinivasakannan and Bakar, 2004 ) Potassium hydroxide was chosen as the activating agent for this research because it is a rela tively inexpensive chemical that is readily available in the laboratory, and because the small amount of literature regarding the chemical activation of rice husks using KOH had been promising (Malik, Guo). Chemical activation is not nearly as well unders tood as physical activation, but it seems to work by dehydrating the carbon precursor and preventing the formation of tars during the activation process (Azargohar and Dalia, 2008 ). The chemical mechanism by which activation occurs using potassium hydroxid e consists of several steps; however, they can be condensed into two main reactions. 2 + 2 K 2 CO 3 (1 4) K 2 CO 3 2 O + CO 2 (1 5) The second reaction, in which potassium carbonate is converted to potassium oxide, occurs at a min imum temperature of 700 C and usually requires a minimum of one hour for completion (El Hendawy 2009 ) For the production of activated carbon from fir wood, Wu et al. (2005) used a KOH to char ratio varying from 0.5 to 6, and then activated the carbons i n a fur nace at 780 C for one hour For rice husks, Gou et al. (2000) used a rice husk charred at a temperature of 400 C for a KOH to char ratio of 4, and then activated the carbons at temperatures ranging from 650 to 800 C for a period ranging from 0.5 to 2 hours. Both approaches created carbons that were either very microporous or at the divide between microporous and mesoporous carbons, defined as 20

PAGE 17

17 Concerning the amount of water used to create the potassium hydroxide solutions, no definitive answe r is given by the literature, as all of the water should evaporate before the carbon is ready for activation and therefore does not play a role in the chemical activation process. El Hendawy (2009) enerally, as much water should be used to effectively mix the KOH solution and the carbon. Surface Chemistry Activated carbon does not consist solely of carbon atoms; while the carbon matrix that makes up the graphite layers will be entirely carbon, atoms other than c arbon will come to occupy edge sites. These atoms, referred to as heteroatoms, are typically oxygen, nitrogen, or sulfur (hydrogen is not included as its chemical bond energy is so low). These heteroatoms are responsible for giving the carbon a surface cha rge (L e i et al., 2002 ). Heteroatoms on the edge sites of activated carbon will combine to form a variety of different molecular structures, typically referred to as surface functional groups. Pure carbon is naturally amphoteric; that is, in solution it ca n act as either an acid or a base ( Marsh and Rodriguez Reinoso, 2006 ). Electrons flow freely through the carbon matrix, giving regions of the matrix a momentarily net positive or negative charge depending on the electron density at that given instant. In r egions where electrons have temporarily accumulated, the carbon will have a net negative charge. In response, this accumulation of electrons will produce a lack of electrons in the surrounding area, resulting in a net positive charge in those regions. In p ractice, however, because the vast majority of atoms that attach to carbon in solution are hydrogen, carbon is usually referred to in the literature as basic (Tennant 2003 ). Surface functional groups will give the edges of the carbon matrix a charge, and are grouped accordingly: acidic functional groups, such as carboxyl, lactone, hydroxyl, or carbonyl, will take hydrogen atoms from solution and give the carbon a positive charge. Basic functional

PAGE 18

18 groups, such as pyrone, amides, amines, or nitriles, will gi ve away protons in solution and result in a negative surface charge. W hen researchers refer to a carbon as being basic, they can either be trying to communicate that the carbon surface has basic surface functional groups, or that the carbon surface simply has a lack of functional groups. Figure 2 2 Acidic and basic functional groups ( Len and Radovic, 1994 ) Determination of the surface charge of a carbon is typically done by using either titration with a base or potentiometric t itration to calculate the point of zero charge (PZC) of the carbon. The PZC indicates the pH of the solution at which the surface charge of the carbon is negated. Figure 2 4 shows the relationship between the surface charge of the carbon and the pH of the changes in the pH. The PZC is defined as the pH at which the carbon has a neutral charge in solution; at a pH less than the PZC, the net surface charge of the carbon will be positive, while the carbon will have a negative net charge if the pH is greater than the PZC value.

PAGE 19

19 Figure 2 3 Determination of PZC (Barale et al., 2008 ) Functional groups containing the oxygen atom are alm ost always acidic, and these surface groups are undoubtedly the most important concerning surface charge. Oxygen has a relatively high electronegativity, meaning that in a molecule it will pull electrons towards itself and give the molecule a dipole (polar ity). More polarity within the functional groups results in a greater net surface charge. Generally speaking, a basic carbon is preferred for the sake of adsorption performance (Tennant 2003 ). Acidic functional groups, being polar, will attract other pola r molecules, particularly water. This hydrophilic property of the surface of the carbon results in water molecules crowding around and within pore openings, blocking other molecules and preventing target adsorption. Basic functional groups not only help pr event acidic groups from forming on the carbon, but research has also shown that the presence of basic fun ctional groups can aid in the ad sorption of phenolic compounds, as the negative carbon surface charge will attract the positively charged aromatic rin g portion of the phenols (Menendez et al., 1996 ). However, acidic function al groups can be beneficial for certain scenarios, specifically the removal of metals from solution (Monser and Adhoum, 2002 ).

PAGE 20

20 The formation of surface functional groups is controlle d primarily during activated carbon production by the maximum temperature of the activation furnace and the gases present during activation (Bach 2007 ). During heating of the carbon in an inert environment using either nitrogen or hydrogen gas, nearly all of the acidic functional groups will break down at a temperature of less than 900 C; p henolic and quino ne at approximately 700 C, and carbonide and lactone at 500 C. At roughly 900 C, oxygen on the surface of the carbon will bind into the aromatic mat rix, forming the basic functional group pyrone (Vidic et al., 1997 ). All functional groups, acidic or basic, on the surface of the carbon will have been destroyed at a temperature of 1200 C. Once the carbon has been cooled and reintroduced to ambient cond itions, however, acidic functional groups may begin to reappear at room temperature; hence the need to create a lasting basicity. It is for this reason that, when only considering the surface charge of the carbon as a target parameter, hydrogen is preferre d over nitrogen gas as the furnace atmosphere during heat treatment. While heat treatment using either gas produces a carbon without surface functionalization, use of hydrogen gas produces a carbon that will stay basic longer until ambient conditions due t o the formation of more stable C H bonds and the hydrogasification of loosely bound, highly reactive carbon atoms (Menendez et al., 1996 ). After activation has taken place, surface charge can be further modified using a variety of methods. At relatively l ow temperatures (400 700 C), treatment of the activated carbon with ammonium gas can form basic functional groups such as amides or nitriles (Chen et al., 2005 ). For the creation of acidic groups, contact between the carbon and a nitric acid or hydrogen p eroxide solution can catalyze surface oxidation ( Pereira et al., 2003 ). Apple Juice Decolorization The yellowish brown color of apple juice is caused by a number of different molecules that are collectively referred to as apple condensed tannins, phenolics phenolic compounds,

PAGE 21

21 antioxidants, flavonoids, apple peel powder, bioflavonoids, phytonutrients, and phytochemicals; in the context of this research, they will simply be referred to as apple polyphenols. For commercial reasons, the ability to remove these molecules from apple juice, thus making the juice more colorless, has been a particular goal of the juice industry for quite some time (Borneman et al., 1997 ). Figure 2 4 Polyphenols in apple juice ( http://www.applepolyphenols.com/ ) As seen in Figure 2 5, the molecular structure of these polyphenols is not so different from the natural organic matter that activated carbon is currently used to remove from municipal water sourc es: the molecules are a series of phenol and benzene aromatic rings, relatively large in size, with a number of protons that are likely to break off in solution, giving the molecules a slight negative charge. For nearly the last decade, the vast majority o f research concerning the removal of apple polyphenols has concentrated on ultrafiltration using polyethersulfo ne (PES), polyvinylpyrrolidone (PVP), or polyvinylpolyprrolidone (PVPP) membranes. Membrane technology has slowly begun to replace previous techn iques, which relied on coagulants such as gelatins or bentonite and a settling stage (Borneman et al., 1997 ).

PAGE 22

22 Very little literature exists concerning activated carbon and polyphenol removal in apple ( 2005) ha ve shown mil d success using powdered activated oval of nearly 40% Sugar Decolorization Although the juice industry does not currently use activated carbon as the primary tool for removing color from its juices, activated ca rbon has long been used as an ad sorbent for the removal of color from the clarified juice in sugar refineries. In 1941, Hertzog and Broderick published a paper detailing the process of using steam activated coal to remove the color from raw sugar solutions. At the time, the activation of carbon was not fully understood, and the sugar industry was using granulated bone char for the color removal process. The raw sugar, deep brown in color, was mixed into solution with powdered activa ted carbon and filter pressed. This process was repeated twice to remove nearly all of the color from the sugar, leaving it nearly white. The most important observation gathered from their work is that carbons with larger pore diameters were more effective in removing the sugar color, and that an increase in pore diameter resulted in a greater carbon mass loss overall (Hertzog and Broderick, 1941 ). Today, activated carbon is routinely used in the sugar industry. During the sugar refining process, brown sug ar is typically dissolved to make a syrup and then crystallized under vacuum to produce white sugar. Any remaining syrup that can no longer be economically converted back into sugar is referred to as molasses, and is a common waste product of the industry. Its color is due in large part to the presence of melanoidin, a brown polymer formed when amino acids react with sugar in a process referred to as the Maillard reaction. This colored polymer is considered a major pollutant; untreated wastewaters from suga r refineries are often so opaque that sunlight cannot reach aquatic vegetation beneath the surface of the water (Bernardo 1997 ).

PAGE 23

23 Because melanoidin and polyphenols are both responsible for the yellowish brown color of apple juice (polyphenols for the yel lowish color, melanoidin for gradual browning), it was hoped that any research on the removal of polyphenols could also relate to the use of activated carbon in the sugar decolorization industry.

PAGE 24

24 C HAP TER 3 EXPERIMENTAL PROCEDURE Activated carbon for this series of experiments was created using a traditional two step approach, incorporating a pyrolysis stage followed by activation with either steam, potassium hydroxide, or carbon dioxide, with renewable biomasses chosen as the starting material. Raw Materi als Sawdust and rice husks, two waste products from the agricultural industry, arrived in two distinct conditions: one in a natural state, which will henceforth be regarded to by the term The samples were provided by an independent supplier operating out of China. The pre char sawdust and rice husk samples were previously exposed to an inert high temperature environment by a third party. Details such as the temperatur e of the furnace used in the charring process and the time spent in the furnace by the samples are unknown. Before any other preparation took place, the raw sawdust and rice husk samples were placed in an oven overnight at approximately 110 Celsius to rem ove any moisture; pre charred samples were not placed in the oven as experiments had shown mass loss due to moisture loss to be less than 2% for these samples, and deemed inconsequential. Charring After the initial drying process, the raw samples were plac ed in a vertical tube furnace for charring. Two methods were used to char the samples in the furnace. In the first or traditional, method, the sample was placed on top of a frit in either a stainless steel or quartz tube and then placed in the furnace at room temperature (Figure 3 1 ). The clam shell furnace was closed and then turned on; the temperature of the furnace rose gradually (about 20 C per minute) to the desired temperature, remained at that temperature for the desired time, and then descended

PAGE 25

25 gr adually to room temperature once the desired time was reached and the furnace was turned off. T his method, with some variations, seems to be preferred by the majority of activated carb on researchers (Bernardo, Bach, Wu). However, this method had the disadv antage of being very time consuming: for higher temperatures, it can take several hours for the furnace to reach the desired temperature, and several more once the experiment has concluded. In an effort to be more efficient at charring samples, a new metho d was devised, which is henceforth referred to as a stainless steel or quartz tube was placed in the furnace without the sample (i.e., activated carbon precursor) inside of it, and the furnace was turned on. Once the furnace reached the desired temperature, the sample was placed in a small, chromium wire. Wire mesh was placed at both the top and bottom of the bucket, a llowing it to be porous while still prohibiting most of the activated carbon from escaping. The sample was suspended in the furnace for the allotted time and then removed from the furnace. It was beneficial to suspend the bucket near the top of the furnace tube before completely removing it in order to allow it to partially cool before exposure to air. Therefore, the heating and cooling times for samples activated using the bucket method are considerably lower, probably occurring within a matter of a few mi nutes as opposed to over an hour for samples activated using the traditional method. Although it is still not fully known how heating and cooling the samples so rapidly may have affected the results, results from using the traditional and bucket method wer e comparable. Herein, it will be noted whether the traditional or bucket method was used for a particular experiment.

PAGE 26

26 Figure 3 1. Furnace used for activated carbon production For the charring process, the raw samples were placed in a vertical tube furn ace at varying temperatures, ranging from 350 to 850 Celsius, for times ranging from 30 to 120 minutes, under nitrogen gas (either High Purity or Ultra High Purity Nitrogen gas was used) at a flow rate of approximately 2500 mL/min. It is not expected tha t high purity or ultra high purity nitrogen would impact the results, for the objective was to maintain an inert environment. For the charring process, the bucket method was used, with approximately five grams of raw sample being charred during a single ex periment. The pre char samples were excluded from this step, as they had already been charred, albeit at unknown conditions. After charring, the sample was cooled and weighed to determine the mass loss. For these experiments, temperature reading s were reco rded by lowering a two foot thermocouple into the tube furnace, with the temperature of the furnace adjusted manually. The thermocouple measured temperatures at its tip; however, for a tube furnace with an opening at the top, such as that used herein, temp erature gradients on the inside of the furnace can be very high, ranging to as much as a 100 Celsius every few inches. When placing the sample inside the tube furnace, the tube was positioned such that the bottom of the sample coincided with the tip of

PAGE 27

27 th e thermocouple. Therefore, any temperature given in this paper should be considered as a temperature range and not a precise temperature existing throughout the entire furnace. Particle Processing After the charring process, samples were ground using a ce ramic mortar and pestle and sieved through a 325 subjected to dye testing or reserved for activation by being placed in glassware and kept in an oven at approximately 110 Celsius. For dye testin g, 30 mg of sample was placed in a glass vial containing 100 mL of either Methylene Blue or Congo Red dye at a concentration of 10 mg/L. HAWP cellulose ester filte r using a vacuum pump. Dye adsorption was assessed using a Hach 4000U Spectropho tometer at a wavelength of 650 nm and 487 n m for Methylene Blue and Congo Red, respectively. Results were calculated as a percentage removal of the dye, with nano filtrated wat er used for calibration. The filtering process did not remove any appreciable concentration of either dye. Activation After being charred, the samples not used for dye testing were activated using two distinct methods, one physical, using steam (and in ver y rare cases, carbon dioxide), and the other chemical, impregnating the char with an aqueous solution of potassium hydroxide (KOH) prior to heating. For physical activation, either the traditional or the bucket method was used, with the traditional method usually reserved for samples larger than five grams. The physical activation process was nearly identical to the charring process only that instead of using a nitrogen flow steam was introduced through the furnace tube at a constant rate. Samples ranging from 1.0 to 5.0 grams were activated for varying furnace temperatures (typically greater than 750 C), varying times (30 minutes to 4 hours) in the furnace, and varying steam flows, with steam flows

PAGE 28

28 (0.4 to 1.6 mL/min volumetric flow rate). Steam for the p hysical activation was created by using a 400 mL glass bottle, filled with distilled water, on top of a laboratory hot plate and connected to the lower end of the tube furnace. Laboratory heating tape was used at the top of the bottle to prevent steam from condensing before entering the furnace and returning to the bottle as liquid water. This system, although slightly unorthodox, created a consistent, measurable flow of with other methods. For physical activation using carbon dioxide, the bucket method was used, with carbon dioxide used in the tube furnace instead of steam. The flow rate for the carbon dioxide was approximately 200 mL/min. Once the samples were physically activated, they were removed from the oven and allowed to cool to room temperature under steam flow. After cooling, their mass was taken and they were placed in a glass vial and kept in a desiccator until being used for testing. For chemical activation, s amples were first mixed with an aqueous solution containing varying concentrations (15 to 50% by mass) of potassium hydroxide. Anhydrous potassium hydroxide (KOH) was mixed by hand with water in a glass beaker under a fume hood at a given ratio (0.25:1 to 6:1 KOH:C). Carbon that had been charred, ground, and sieved was then added to the solution; laboratory parafilm was used to cover the beakers to prevent loss and the beakers were then mixed from two to eight hours using a magnetic bar and mixing plate. Af ter mixing for the allotted time, samples were dried in an oven at 110 Celsius until visibly dry. It is extremely important that the sample be completely dry before being placed in the furnace, for potassium can be combustible when allowed to react with w ater at high temperatures. After drying, the mass of the samples was taken and they were then lowered into the tube furnace under nitrogen gas flow using the bucket method in the same manner as the charring process, with a nitrogen

PAGE 29

29 gas flow of approximatel y 2500 mL/min. After being chemically activated and allowed to cool to room temperature, the samples were ground using a ceramic mortar and pestle and sieved through a 325 then washed in a 2500 mL glass beaker with distilled water to remove residual potassium and paper should not be used if the sample is to be retained because it has a tend ency to decompose upon repeated use ) using a vacuum pump. After washing, the samples were placed in an oven overnight at 110 Celsius to dry. After physical and chemical activations the mass of the sample was weighed to calculate the mass loss from activa tion. Once samples were activated and washed (only for the chemical activations), they were subjected to Methylene Blue and Congo Red synthetic dye testing in the manner described above. Congo Red dye testing is typically a better indicator of apple juice testing success than Methylene Blue, as the Congo Red molecules are larger. Samples that performed exceptionally well in the synthetic dye testing (> 50% Congo Red removal) were then subjected to apple juice testing. For apple juice testing, 100 mg of samp le was mixed with 20 mL of refrigerated Hannaford Apple Juice, 100% juice, using a rotating mixer for four hours and then filtered through a 0 ster filter using a vacuum pump. Apple juice color apple juice is not a simple solution containing only one solute, such as the Congo Red dye, it does not have a single spectrophotometer absorption peak that can be assessed for removal; as such, the 330 n m value used is somewhat arbitrary. Value s as low as 270 n m could still be used for the spectrophotometer, with comparable results. Results were calculated as a percentage rem oval of the apple juice at 330 n m, with nano filtrated water used for calibration.

PAGE 30

30 Carbon C haracterization Samples that performed particularly well in the apple juice color removal tests were analyzed using a NOVA 2200e, manu factured by Quantachrome Instruments, to determine the specific surface area (m 2 /g) and the pore size distribution of the PAC. Approximately 200 mg of activated carbon was outgassed overnight under vacuum at 110 Celsius, and then dosed with liquid nitroge n to determine the isotherm. The BJH method was used by the provided software to calculate pore size distribution. For surface charge adjustment, selected samples were placed in a 250 mL glass beaker filled with 200 mL of a 5 M solution of nitric acid and mixed by hand. Samples were refluxed (boiled with no steam loss) for four hours, and then washed with distilled water until the wash water had a negligible pH change (for distilled water, pH =5). The samples were filtered through r paper using a vacuum pump, and then dried overnight in an oven at 110 Celsius. The samples were then tested for apple juice color removal, using the method described earlier.

PAGE 31

31 CHAPTER 4 RESULTS Charring For the production of activated carbon, the firs t step was to create a charred material from the raw precursor by using heat in an inert environment. Some researchers skip this step and simultaneously char and activate the carbon in a single one step process (Warhurst et al., 1996 ). However, this was av oided in order to gather information about the activation process and also to create as much charred carbon as possible for both the physical and chemical activations. As the raw carbon source was heated, volatile organic matter is gasified and removed, le aving behind a mostly carbon material that resembles what could be considered flaky charcoal. Any ash in the sample (i.e. charred carbon) is a sign that the carbon had exposure to oxygen at some point while the temperature was greater than about 400 C and samples containing ash should be discarded. Figure 4 1 shows mass loss for sawdust as it undergoes the charring process. Samples of raw sawdust were lowered into the furnace using the bucket method under nitrogen flow. The legend on the right of the graph shows the temperature of the furnace while charring was taking place; the x axis shows what amount of time the sample was exposed to heat. Because the raw sawdust is dried in an oven overnight at 110 C, the mass loss presented on this slide should not in clude mass loss due to moisture loss. Because samples charred at 550, 650, and 750 C have approximately the same mass losses, the mass loss for the sawdust seems to occur in three distinct phases: one group of organic matter seems to volatize at a tempera ture lower than 450 C, another group at some temperature between 450 and 550 C, and a third group at some temperature between 750 and 850 C. Also, the graph shows that mass loss increases marginally with an increase in charring time.

PAGE 32

32 Figure 4 1 Saw dust mass loss during the charring process After charring, dye adsorption tests were carried out (Figure 4 2 ). At this point, the carbon had a capacity to adsorb dye molecules from solution because of pores created in the carbon structure as volatile organ ic matter gasified; however, it has not yet been technically activated, as an activation step has no t yet been completed. Figure 4 2 demonstrate s a general relationship between charring temperature (and hence mass loss) and adsorptive capacity, with seemin gly minor differences in mass loss resulting in very different dye removals: raw sawdust exposed to nitrogen at 850 C for two hours loses only seven percent more mass than the same sawdust exposed to nitrogen at 450 C for 30 minutes, and yet removes near ly 4.5 times the amount of methylene blue dye from solution. Also, as noted earlier, removal performances for the carbons are lower for the Congo Red dye as opposed to the Methylene Blue, as the Congo Red molecules are larger and require larger pores that would not be created during charring, but rather during activation. Because of the relatively low dye adsorption, it can be assumed that these charred sawdust samples would not remove much apple juice color, as the molecules that give apple juice its color tend to be larger and thus harder to remove.

PAGE 33

33 Figure 4 2 Congo red and me thylene blue adsorption for sawdust chars ( Figure 4 3) is a similar graph for the mass loss of raw rice husks during the charring process, which were performed in the sa me manner as the sawdust samples The primary difference between the rice husks and the sawdust during the charring process is that mass loss for the rice husks is far less dependent on time; as opposed to the sawdust, mass loss for the rice husks occurs w ithin the first thirty minutes of charring. It seems that the rice husks, being less dense, lose their volatile organic matter much faster than the sawdust. As another consequence of their lower density, the rice husks lose less mass overall, losing roughl y 62% of their mass after 30 minutes at 450 C, as opposed to 74% for the sawdust under the same conditions. Figure 4 3. Mass loss of rice husks during the charring process

PAGE 34

34 Steam Activation Once the raw sawdust and rice husks have been charred, activat ion is necessary to widen pores in the carbon structure to increase the adsorptive capacity of the carbon; it is during this research concentrated on physical activation using steam as opposed to carbon dioxide (which is covered later). For steam activation, the sample was exposed to a steam flow (0.4 to 1.6 mL/min) while at temperatures greater than 750 C The graph below (Figure 4 4 ) provides data for charre d sawdust samples that were activated using steam. For this group of data, the bucket method was used, with a relatively high flow rate of approximately 1.6 mL/min (steam) to ensure that steam would have the velocity to enter the bucket, as opposed to circ umventing it. All of these activations occur at a furnace temperature of 850 C. It should be noted that apple color removal performance must be used as an indicator of success for these tests; all of the samples below had a Congo Red and Methylene Blue sy nthetic dye adsorption of over 99%. The graph below demonstrates two important points regarding steam activation of charred sawdust: one, that all the samples subjected to steam for two hours performed worse than the samples subjected to steam for only one hour, and two, that the char times of the samples before being activated is largely irrelevant. For example, the sample that had been charred for two hours and then subjected to steam for one hour performed roughly as well as the sample that had been char red for only 30 minutes before being subjected to steam for one hour.

PAGE 35

35 Figure 4 4 Steam activation of sawdust at 1.6 mL/min The fact that all of the samples subjected to steam for two hours performed worse than their counterparts subjected to steam for only one hour proves that, for activation using steam, there is an optimum activation time for specific steam flows at specific temperatures (and for specific materials). As the charred s ample is heated under steam pores within the carbon structure widen as select carbon molecules oxidize; however, the pores can only widen so much befor e the pore walls break down and the pores collapse, lowering adsorption potential. Figure 4 5 is comprised of two images taken of an activated carbon created during this pr oject using steam activation of sawdust. The images, which were taken using a Field Emission Scanning Electron Microscope (SEM) at magnifications of 43,000X (left) and 80,000X (right) show the cavernous pore structure that develops on the surface and with in the activated carbon, giving large surface area and adsorptive capacity.

PAGE 36

36 Figure 4 5. Images of activated carbon at high magnifications To activate larger samples of sawdust char, the bucket method was abandoned in favor of the traditiona l method as it was feared using large samples in the bucket would encourage flow to circumvent the bucket altogether To keep sawdust char from exit i ng the vertical tube furnace, a lower steam flow rate was adopted, that of approximately 0.4 mL/min. Lower steam rates will change the steam to carbon ratio during activation and have an impact on the results. Samples sizes for this set of experiments were approximately 10 grams, as opposed to approximately one gram for the bucket method. Three of the samples were mixed approximately every 15 minutes by lowering a twisted nickel chromium wire into the tube furnace; as shown, mixing has no noticeable effect on the performance of the activated carbon created. The graph (Figure 4 6 ) shows that, as opposed to havin g an optimum activation time of roughly one hour for higher steam flows, a lower steam flow shifts the optimum steam time to approximately two to three hours and allows for slightly greater adsorptive performance. The fact that the adsorptive capacity of t he carbons did not increase dramatically when the activation time was increased from 2 to 3 .5 hours tends to favor the hypothesis that sawdust may have an adsorptive limit of approximately 90% or less. Also, there seems to be no observable peak followed by a decrease in

PAGE 37

37 adsorptive capacity for this lower flow rate; it would be interesting to see if pore breakdown occurred with activation times of four hours or more. Figure 4 6 Steam activation of sawdust at 0.4 mL/min Steam activation of charred rice hu sks is very similar to that of charred sawdust. The key difference between the two is that, while the optimum amount of steaming time for charred sawdust at a flow of 1.6 mL/min is approximately one hour, charred rice husks activated with steam at the same flow rate seemed to perform best when exposed to steam for less than 30 minutes. Figure 4 7 show s Congo Red and Methylene Blue dye adsorptions for the rice husks activated with steam; as expected, adsorption removal percentages for Congo Red are lower com pared to that of Methylene Blue.

PAGE 38

38 Figure 4 7. Congo red and methylene blue adsorption for rice husks char Further experiments showed that the optimum way to create activated carbon from rice husks using steam at a flow rate of 1.6 mL/min was to cha r the sample for 20 minutes and then expose the sample to steam for 20 minutes. It seems that longer exposure to steam lowers the adsorptive capacity of the carbon, probably due to pore breakdown as a result of excessive widening. The sample below (Table 4 1 ) that achieved a 91.2% removal of Congo Red dye also achieved an apple juice color removal of nearly 82%. Table 4 1 Rice husks activated with steam Charring Time (min) Activation Time (min) Meth. Blue % Rem. Congo Red % Rem. 20 20 100 91.2 30 30 97. 9 30.9 30 30 99.2 25.3 Chemical Activation From very early on in the research, it was decided that chemical activation would be achieved using potassium hydroxide (KOH), as potassium hydroxide is a relatively inexpensive strong base, and because there i s a strong record in the literature of researchers achieving promising results with its usage (Azargohar, Malik). It was also decided very early that the majority of the chemical activation research would concentrate on using rice husks as a

PAGE 39

39 precursor, bot h because this was the express ed intent of the supplier and because early studies using sawdust as a precursor for chemical activation were not very promising. The primary problem associated with chemical activation is the large number of variables that pl the characteristics of the char material be considered, but also the concentration of the potassium hydroxide solution used, the ratio of KOH to carbon once the charred material is mixed with the solution, the temperature of the furnace used for activation, and the time allotted for activation. As such, it was decided early in the research to isolate certain variables to assess their impact. Once an optimum value for a v ariable was chosen, this value was used for all future experiments. Figure 4 8 was used to determine the optimum solution concentration for activating rice husk char. Rice husks were charred for 30 minutes at a temperature of either 450 C or 650 C and th en mixed with various solutions of aqueous potassium hydroxide. All of the samples have a ratio of anhydrous KOH to carbon of 3:1 by mass. F or example, a 15% solution with a 3:1 KOH to carbon ratio would contain 1 gram of charred carbon, 3 grams of potassi um hydroxide, ( meas ured by mass in the solid state) and 17 mL of distilled water. The samples were mixed, allowed to dry overnight in an oven, and then activated in a furnace at 800 C for three hours. Higher temperatures were chosen for the chemical acti vation studies because important reactions occur between the potassium and carbon atoms at temperatures above 700 C, namely the conversion of potassium carbonate to potassium oxide and carbo n dioxide. As Figure 4 8 shows, a 15% solution concentration is optimum. It is assumed that, with more water available, a greater degree of mixing occurs once the carbon comes in contact with the potassium hydroxide solution.

PAGE 40

40 Figure 4 8 Solution concentrations used for activation A 3:1 ratio of KOH to carbon by mas s was initially chosen due to the supporting literature (Gou). Figure 4 9 below supports the decision. For the set of data below, pre charred rice husks were mixed with two potassium hydroxide solutions, one at 15% and the other at 50%, at KOH to carbon ra tios of 2:1, 3:1, and 4:1. The samples were allowed to dry overnight in an oven and then activated in a furnace at 800 C for three hours. It seems that, at a 3:1 ratio of KOH to carbon there is enough potassium available to create a porosity in the carbo n without overactivation occurring. Figure 4 9 Potassium hydroxide to carbon ratios

PAGE 41

41 Once it was decided that a 15% solution strength would be used in conjunction with a 3:1 KOH to carbon ratio, the optimum charring temperature and time, as well as the optimum activation temperature and time, had to be determined. In contrast to solution strength and the carbon to KOH ratio, the impact of changes in furnace temperatures and the time spent in the furnace are less definitive. For Figure 4 10 rice husks ch arred at various temperatures and for various times were mixed with a 15% KOH solution at a carbon to KOH ratio of 1:3, allowed to dry overnight in an oven, and activated in a furnace at various temperatures and for various times. When reading the legend, axis value of 450 should be interpreted as a rice husk sample charred for 30 minutes at a temperature of 450 C, mixed with a potassium hydroxide solution, dried, and activated at 800 C for 100 minutes. Figure 4 10 The impac t of char and activation characteristics on apple color removal As the Figure 4 10 shows, there is very little statistical difference in apple juice color removal performance between samples charred for 30 minutes as opposed to 60 minutes, with samples cha rred for 120 minutes actually performing marginally worse. This coincides with data presented earlier in this paper showing that rice husks lose the vast majority of their volatile mass

PAGE 42

42 within the first 30 minutes of charring. As for charring temperature, the data shows a very slight increase in performance with an increase in charring temperature after 450 C. It was decided that a charring temperature of 450 C would be used in the interest of energy consumption, with future research divided between sampl es charred at 450 C and those charred at 650 C. Concerning activation, a minimum activation temperature of 800 C was selected, as well as an activation time of three hours. Once optimum variable values were selected, several exceptional samples were cr eated using chemical activation of rice husks, with apple juice color removal percentages appr oaching 90%. For Figure 4 11 (450 C, 30, 800 C, 180) should be interpreted as rice husk charred at 450 C for 30 minutes, mixed with a 15% potassium hydroxide solution at a 1:3 carbon to KOH ratio, and activated at 800 C for 180 minutes. Figure 4 11 Selected chemical activations Production of activated carbon using potassium hydroxide and sawdust was done in the same fashion as the chemical activation of ri ce husks. As stated earlier, only a small portion of this project was devoted to the chemical activation of sawdust due to a pre determined emphasis on rice husks. Unfortunately, our laboratory has had little success achieving significant results using the chemical activation of sawdust. Figure 4 12 shows the Congo Red and Methylene Blue

PAGE 43

43 dye adsorption values for a few of the samples created using chemical activation of sawdust A s expected, Congo Red dye removal is lower than the corresponding Methylene Bl ue removal, suggesting the activated carbons created are microporous. For these experiments, an emphasis was put on activating the sawdust and KOH mixtures at relatively low temperatures (< 800 C) in an effort to develop a procedure that would minimize en ergy costs Figure 4 12 Chemical activations using sawdust Later, towards the end of the project, chemical activation of sawdust was briefly revisited, this time using much higher carbon to KOH ratios as well as higher activation temperatures, both of which are suggested by the literature. For the samples below, sawdust char was created at a temperature of 450 C for one hour and mixed at three distinct carbon to KOH ratios: 1:0.25, 1:0.5, and 1:1. For these particular experiments, the amount of water c alculated for the 1:1 ratio sample was used for all three samples in an effort to maintain similar mixing and drying. As shown in Table 4 2, these samples also performed unfavorably. Table 4 2. Chemical activations of sawdust Carbon to KOH Ratio Solution Strength Apple Color Removal 1:0.25 3.7% 32.1% 1:0.5 7.5% 36.2% 1:1 15% 24.5%

PAGE 44

44 Carbon Dioxide Activation Carbon dioxide was used as a replacement for steam in an effort to learn more about the process for future research. For these activations, pre cha rred sawdust from our supplier was sieved through a 325 mesh sieve and activated under carbon dioxide flow at approximately 200 mL/min using the bucket method. Unlike steam activation, in which it can be assumed that a significant portion of the sample is being lost during the process, the carbon dioxide flow is not mass loss during activation and adsorptive capacity can be created. All of the samples were acti vated at a temperature of 850 C; Rodrguez Reinoso (1995) have shown a minimum temperature of 800 C to be necessary. Figure 4 13 Carbon dioxide activation of sawdust Figure 4 13 demonstates a direct relationship between mass loss during carbon dioxid e activation an d resulting adsorptive capacity. I t also shows that carbon dioxide activation is promising for mesoporic pore creation, at least for sawdust. Concerning rice husks, Figure 4 14 demonstrates a peak color removal of 70% after an activation per iod of roughly two hours, with pore breakdown occuring afterward.

PAGE 45

45 Figure 4 14. Carbon dioxide activation of rice husk A key issue needs to be addressed before carbon dioxide is used as a replacement for steam for physical activations First and foremo st, carbon dioxide activation seems, at least in the context of this research, extemely volatile; not shown in Figure 4 14 is that several samples were completely gasified entirely and without explanation during the process, including numerous samples acti vated for over 4 hours that would have expanded the graph s Secondly, when activating rice husks with carbon dioxide, there were several instances where the apple juice samples were visibly clear, yet performed poorly when analyzed with the photospectromet er. This underperformance did not occur with rice husks activated with steam or sawdust in general, and is presently unexplained. For these reason s further research was not conducted. Surface Charge Modification Lastly, an effort was made to modify the su rface charge of an activated carbon sample in an attempt to further improve adsorption. As stated earlier, the surface of the PAC created using the standard two step steam method is basic in nature and has relatively little surface functionalization. It wa s hoped that by creating acidic functional groups on the surface of the

PAGE 46

46 PAC, a net positive charge could be produced, thereby attracting the negatively charged color molecules in the apple juice and increasing adsorption. For this experiment, a sample of P AC that had been produced using steam was allowed to boil in a nitric acid solution for four hours. After washing and drying to remove the nitric acid, the apple juice color removal test was repeated. The results shown in Table 4 3 indicate that surface mo dification with nitric acid can improve the adsorptive capacity of the activated carbon to some degree. Table 4 3 Surface m odification Removal before acid treatment 71% Removal after acid treatment 84% Although further research obviously needs to be d one, especially on higher performing samples and on samples created using chemical activation, it seems that treatment with nitric acid may give the carbon surface a positive net surface charge, which could contribute to the adsorption of the negatively ch arged polyphenol molecules. Surface Area Assessment Selected samples were chosen to be analyzed by the NOVA Surface Analyzer in order to determine their specific surface areas and pore size distributions. The NOVA uses nitrogen gas at a temperature of 77 K as an adsorbate, and uses the amount of nitrogen gas adsorbed onto the carbon at various pressures to calculate the specific surface area, total pore volume, and average pore volume of the activated carbon sample. For surface area assessment, two samples were chosen: CA 50, a commercially available carbon created using chemical activation of sawdust with phosphoric acid and a physically activated carbon derived from sawdust that performed very well. Dye and apple color removal performance for CA 50 is pr ovided in Table 4 4; its exceptional performance was the basis for it being chosen as a test control for this research

PAGE 47

47 Table 4 4. CA 50 p erformance Congo Red Removal (%) 98 Methylene Blue Removal (%) 100 Apple Color Removal (%) 92 Figure 4 15 show s the actual data points that the NOVA machine uses to calculate the values in the tables above. Adsorption points, which are taken as nitrogen gas molecules attach themselves to the carbon surface, are colored blue, with desorption points, taken as the nitrogen is detaching, are colored red. While the two curves have roughly the same shape, CA 50 has a sharper curve in the high pressure range, which can be attributed to its larger average pore diameter, as well as a greater pore volume overall. Figure 4 15. BET adsorption/desorption p lots (a) CA 50 (B) S awdust a ctivated with s team for 2 hours at 850 C B A

PAGE 48

48 D ata for five samples is shown below in Table 4 5 The five samples include : two samples created using steam activation of sawdust, one sample cr eated using chemical activation of rice husks, one sample created using carbon dioxide activation of sawdust, and CA 50, a commercially available powdered activated carbon. The key point to be gained from the BET data is that there is a direct correlation between the average pore size of the PAC and its adsorptive capacity ; this conclusion has been previously expressed (Tennant, 2003). It could be said that the pore size of the PAC is the defining factor when relating its molecular structure to adsorption. This statement holds especially true for larger adsorbate molecules, as is the case with this research. Samples 1 and 2 were both created using identical procedures; the only difference between the two is the amount of time that the samples were exposed to steam flow. As Sample 1 had a longer exposure to steam, it had more time for its pore structures to widen, creating larger pores and thus a higher capacity for adsorption. Also, Sample 3, which was created using chemical activation, is similar to Sample 1 in every parameter, showing that physical activation and chemical activation of two different organic precursors can achieve similar results Table 4 5 BET characteristics of selected s amples Sample 1: Charred Sawdust with Steam, 2 hr Apple Color Remo val 79 % Specific Surface Area 133 1 m2/g Total Pore Volume 0.84 8 cc/g Average Pore Volume 25.5

PAGE 49

49 Table 4 5 Continued Sample 2: Charred Sawdust with Steam, 1 hr Apple Color Removal 31 % Specific Surface Area 520.2 m2/g Total Pore Volume 0.51 9 c c/g Average Pore Volume 20.6 Sample 3: Rice at 450 C, 60 min, 15%, 850 C, 180 min Apple Color Removal 84 % Specific Surface Area 148 7 m2/g Total Pore Volume 0.946 cc/g Average Pore Volume 25.5 Sample 4: Sawdust with Carbon Dioxide, 4 hr Apple Color Removal 84 % Specific Surface Area 1096 m2/g Total Pore Volume 0.62 1 cc/g Average Pore Volume 22.7

PAGE 50

50 Table 4 5 Continued Sample 5: CA 50 Apple Color Removal 92 % Specific Surface Area 148 9 m2/g Total Pore Volume 1.255 cc/g Av erage Pore Volume 33.7

PAGE 51

51 CHAPTER 5 CONCLUSIONS The goal of this research was to create a method by which sawdust or rice husks, two agricultural waste products, could be used to create a powdered activated carbon by either physical or chemical means, with the latter concentrating on the use of potassium hydroxide in solution. The litmus test for this research would be the optimization of the activated carbon created for the removal of natural color from commercially available apple juice, a task that w ould provide a unique set of challenges. Physical activation of charred sawdust, either by using steam or carbon dioxide, can consistently create a PAC that will remove 80 to 85% of the natural color from the particular brand of apple juice used (using 330 n m wavelength absorbance as a measurement). However, chemical activation has yet to prove its worth concerning sawdust. As for rice husks, steam activation can perform just as well as with the sawdust, with results thus far reaching a maximum removal of 8 2%. For rice husks, however, the emphasis was on chemical activation with potassium hydroxide, and r ice husks truly outshined its sawdust counterpart in this regard By the end of the experiment, powdered activated carbon samples created using KOH and rice husks were consistently removing 85 to 90% of the natural apple juice color, nearly outperforming the commercially available powdered activated carbon, CA 50. T he failure of the activated carbon created in our laboratory to outperform the commercially ava ilable carbon, CA 50, casts somewhat of a shadow on this research. It would be convenient to assume that surface chemistry or some other unaccounted for property of the CA 50 gave it an advantage H owever, the fundamental reason why CA 50 outperforms the a ctivated carbon created by our laboratory is that the carbon possesses a particularly large average pore size, and our methods were unable to recreate these mesoporous pore structures in

PAGE 52

52 either the sawdust or rice husk char. Had our laboratory been able to create an activated carbon sample with an average pore size of at least 30 I have no doubt that our samples would have outperformed the competition. It is unknown at this time if the failure lies with the research or the original source materials; it c ould be that it is simply too difficult to create a highly mesoporous activated carbon from either sawdust or rice husks. However, with so many samples performing so well, it is hoped that eventually the threshold for these materials could be broken, and t he activated carbon created could be truly optimized for natural color removal.

PAGE 53

53 LIST OF REFERENCES Apple poly phenols.com. Accessed April 15, 2009. < http://apple poly phenols.com/ >. F.N., Kar F., Arslan, N., 2005. Adsorption of dark coloured compounds from peach pulp by using powdered activated carbon. Journal of Food Engineering 71 (2), 156 163. Azargohar R., Dalai, A.K., 2008. Steam and KOH activation of biochar: Experimental and modeling studies. Microporous an d mesoporous materials 110 (2 3), 413 421. Bach, M.T., 2007. Impact of surface chemistry on adsorption tailoring of activated carbon. Dissertation (Ph. D.). University of Florida. Bansal R.C., Donnet, J., Stoeckli F., 1988 Active Carbon New York and Ba sel: Marcel Dekker, Inc.; New York Barale M., Lefvre, G., Carrette, F., Catalette, H., Fdoroff, M., Cote, G., 2008. Effect of the adsorption of lithium and borate species on the zeta poten tial of particles of cobalt ferrite, nickel ferrite, and magnetite Journal of Colloid and Interface Science 328 (1) 34 40. Bernardo E. C. Egashira, R., Kawasaki J., 1997. activated carbon. Carbon 35 ( 9 ), 12 17 1221. Borneman, Z., Gtkmen, V., Nijhuis, H.H., 1997. Selective removal of polyphenols and brown colour in apple juices using PES/PVP membranes in a single ultrafiltration process. Journal of Membrane Science 134 (2), 191 197 Chadik, P., 2009 Class Not es, Advanced Water Trea tment Process Design, GAC. 5 19. Chen, W., Cannon, F.S., Rangel Mendez, J.R., 2005. Ammonia tailoring of GAC to enhance perchlorate removal. I: Characterization of NH 3 thermally tailored GACs with steam. Carbon 43 (3), 573 580 El H endawy, A., 2009. An insight into the KOH activation mechanism through the production of microporous activated carbon for the removal of Pb 2+ cations. Applied Surface Science 255 (6), 3723 3730. Ferraz M.C., 1988. Preparation of activated carbon for air pollution control. Fuel 67 (9), 1237 1241. Gonzlez, M.T., Rodrguez Reinoso, F., Garca, A.N., Marcilla, A., 1997. CO 2 activation of olive stones carbonized under different experimental conditions. Carbon 35 (1 ) 159 162. Guo Y., Yu, K., Wang, Z., Xu H., 2000 Effects of activation conditions on preparation of porous carbon from rice husk. Carbon 41 (8), 1645 1687.

PAGE 54

54 Hameed B.H., Din A.T.M., Ahmad A.L., 2007. Adsorption of methylene blue onto bamboo based activated carbon: Kinetics and equilibrium studies Journal of Hazardous Materials 141 (3 ), 819 825 Hertzog, E.S., Broderick S.J. 1941. Activated carbon for sugar decolorization. In. Eng Chem 33 ( 9 ), 1192 1198. Hunter C.A., Sanders J., 1990 The nature of .pi. .pi. interactions. J. Am Chem. Soc. 1122 (14), 5525 5534. Khan A.Y., 2006. Modification of activated carbon to improve aqueous manganese removal. Dissertation (Ph. D.). University of Florida. Lei, L., Quinlivan, P.A., Knappe D. 2002. Effects of activated carbon surface chemistry and pore structure on the adsorption of organic contaminants from aqueous solution. Carbon 40 ( 12 ), 2085 2100. Len y L en C Radovic L. 1994. Interfacial Chemistry and Electrochemistry of Carbon S urfaces. In: Chemistry a nd Physics of Carbon, Volume 24. Dekker; New York Malik, P.K., 2003. Use of activated carbons prepared from sawdust and rice husk for adsorption of a cid dyes: a case study of Acid Yellow 36. Dyes and Pigments 56 (3), 239 249. Marsh, H., Rodriguez Reinoso, F., 2006. Activated Carbon. Elsevier Ltd.; New York. Mazyck D.W., Cannon, F.S., 2000. Overcoming calcium catalysis during the thermal reactivation o f granular activated carbon: Part I. Steam curing plus ramped temperature N 2 treatment. Carbon 38 (13), 1785 1799. Menendez, J.A., Phill ips, J., Xia, B., Radovic, L.R., 1996. On the mo dification and characterization of chemical surface properties of activated carbon: In the search of carbons with stable basic properties. Langmuir 12 (18), 4404 4410. Monser L., Adhoum N. 2002. Modified activated carbon for the removal of copper, zinc, chromium and cyanide from wastewater. Separation and Purification Technology, 26 (2 3), 137 146. Pastor Villegas J., Duran Valle, C.J., 2002. Pore structure of activated carbons prepared by carbon dioxide and steam activation at different temperatures fr om extracted rockrose. Carbon 40 ( 3 ), 397 402. Pereira, M.R., Soares, S.F., Orfao, J.M., Figueiredo, J.L., 2003. Adsorption of dyes on activated carbon: Influence of surface chemical groups Carbon 41 (4), 811 821. Rodrguez Reinoso, F., Molina Sabio, M., Gonzlez, M.T., 1995. The use of steam and CO 2 as activating agents in the preparation of activated carbons. Carbon 33 (1), 15 23. Snoeyink, V.L., Jenkins, D., 1980. Water Chemistry. Wiley; New York.

PAGE 55

55 Srinivasakannan C., Bakar M.A., 2004. Production of activated carbon from rubber wood sawdust. Biomass and Bioenergy 27 (1), 10 19 Tennant M., 2003. Steam pyrolysis activation of wood char for superior odorant removal. Carbon 41 (12), 2195 2202. Vidic R.D., Tessmer, C.H., Uranowski, L.J., 1997. Impact of surface properties of activated carbons on oxidative coupling of phenolic compounds. Carbon 35 (9), 1349 1359. Warhurst, A.W., McConnachie G.L., Pollard S.J.T., 1996. The production of activated carbon for water treatment in Malawi from the waste seed h usks of Moringa oleifera. W ater Q uality I nternational '96. P art 7: Agro I ndustries W aste M anagement; Appropriate Technologies and Waste S tabilization P onds ; W astewater R eclamation and R euse Water Science & Technology 34 ( 11 ) Wu F., Tseng, R., Juang, R. 2005. Preparation of highly microporous carbons from fir wood by KOH activation for adsorption of dyes and phenols from water. Separation and Purification Technology 47 (1 2), 10 19.

PAGE 56

56 BIOGRAPHICAL SKETCH egree in Civil Engineering at the University of Flo egree in Environmental Engineering in 2009 at the same university. After graduating he moved to Austin, TX to work in the water treatment industry