<%BANNER%>

One-Step Steam and Calcium Activation of Sawdust for Apple Juice Purification

Permanent Link: http://ufdc.ufl.edu/UFE0042798/00001

Material Information

Title: One-Step Steam and Calcium Activation of Sawdust for Apple Juice Purification
Physical Description: 1 online resource (60 p.)
Language: english
Creator: ENGLISH,TIMOTHY S,II
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ACTIVATED -- ACTIVATION -- APPLE -- CALCIUM -- CARBON -- CHEMICAL -- COLOR -- JUICE -- MESOPORE -- MESOPOROSITY -- REMOVAL -- STEAM
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: ONE-STEP STEAM AND CALCIUM ACTIVATION OF SAWDUST FOR APPLE JUICE PURIFICATION Timothy S. English II University of Florida Environmental Engineering Sciences Activated carbon was synthesized from a readily available agricultural byproduct, sawdust, using one-step steam activation and calcium impregnation with subsequent one-step steam activation. A carbon dose of 5,000 mg/L was necessary to reach the color removal capacity of 90% for a commercially available activated carbon, CA-50. Carbonized samples of sawdust exhibited minimal pore development, with methylene blue and congo red removals around 16% and 4% respectively. One-step steam samples had apple juice color removals as high as 79% and total pore volumes of up to 0.50 cc/g, as determined by nitrogen gas isotherm analysis. Calcium samples had similar apple juice color removal values (70-87% color removal), but significantly more mesopore volume (80-150% increase) and total pore volume (50% increase) over one-step steam samples. Further analysis of the Horvath-Kawazoe and density functional theory pore size distribution models found that the increase in the calcium samples? mesoporosity occurred due to increased general pore development and not at the expense of micropores. Linear correlations determined that increasing the pore volume of diameters greater than 40 ? led to increased apple juice color removal percentages.
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 TIMOTHY S ENGLISH.
Thesis: Thesis (M.E.)--University of Florida, 2011.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042798:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042798/00001

Material Information

Title: One-Step Steam and Calcium Activation of Sawdust for Apple Juice Purification
Physical Description: 1 online resource (60 p.)
Language: english
Creator: ENGLISH,TIMOTHY S,II
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ACTIVATED -- ACTIVATION -- APPLE -- CALCIUM -- CARBON -- CHEMICAL -- COLOR -- JUICE -- MESOPORE -- MESOPOROSITY -- REMOVAL -- STEAM
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: ONE-STEP STEAM AND CALCIUM ACTIVATION OF SAWDUST FOR APPLE JUICE PURIFICATION Timothy S. English II University of Florida Environmental Engineering Sciences Activated carbon was synthesized from a readily available agricultural byproduct, sawdust, using one-step steam activation and calcium impregnation with subsequent one-step steam activation. A carbon dose of 5,000 mg/L was necessary to reach the color removal capacity of 90% for a commercially available activated carbon, CA-50. Carbonized samples of sawdust exhibited minimal pore development, with methylene blue and congo red removals around 16% and 4% respectively. One-step steam samples had apple juice color removals as high as 79% and total pore volumes of up to 0.50 cc/g, as determined by nitrogen gas isotherm analysis. Calcium samples had similar apple juice color removal values (70-87% color removal), but significantly more mesopore volume (80-150% increase) and total pore volume (50% increase) over one-step steam samples. Further analysis of the Horvath-Kawazoe and density functional theory pore size distribution models found that the increase in the calcium samples? mesoporosity occurred due to increased general pore development and not at the expense of micropores. Linear correlations determined that increasing the pore volume of diameters greater than 40 ? led to increased apple juice color removal percentages.
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 TIMOTHY S ENGLISH.
Thesis: Thesis (M.E.)--University of Florida, 2011.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042798:00001


This item has the following downloads:


Full Text

PAGE 1

1 ONE STEP STEAM AND CALCIUM ACTIVATION OF SAWDUST FOR APPLE JUICE PURIFICATION By TIMOTHY S. ENGLISH II A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 T imothy S. English II

PAGE 3

3 To th ose who gave me a second chance

PAGE 4

4 ACKNOWLEDGMENTS I thank my friends and family for celebrating with me during the highs and consoling me during the lows I thank Dr. David Mazyck for his guidance and for giving me a chance. I appreciate the understanding of Drs. Paul Chadik and Ben Koopman for the hurried nature of my defense. I also thank the staff of the Environmental Engineering Sciences department for helping me file last minute paperwork and being fellow lab mates, Matt Joiner, Dr. Heathe r Byrne, Joe Roberts, and others, the data for this paper would not exist.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 2 LITERARY REVIEW ................................ ................................ ............................... 13 Introduction to Activated Carbon ................................ ................................ ............. 13 Introduction to Apple Juice Decolorization ................................ .............................. 13 Sugar Decolorization ................................ ................................ ........................ 14 Color Test Methods (Spectrophotometry) ................................ ......................... 14 Introduction to Activated Carbon Science ................................ ............................... 15 Adsorption ................................ ................................ ................................ ........ 15 Pore Structure and Adsorbate Diffusion ................................ ........................... 16 Gas Adsorption Isotherm Models ................................ ................................ ..... 16 Activated Carbon Production Methods ................................ ................................ .... 19 Physical Activation ................................ ................................ ........................... 19 Two s tep ................................ ................................ ................................ .... 20 One s tep ................................ ................................ ................................ .... 21 Chemical Activation ................................ ................................ .......................... 22 Introduc tion to c alcium ................................ ................................ ............... 22 Chemical i mpregnation/ a ctivation ................................ .............................. 23 3 EXPERIMENTAL METHODS ................................ ................................ ................. 25 Materi al Preparation ................................ ................................ ............................... 25 Two Step Steam Activation ................................ ................................ ..................... 25 Pyrolysis ................................ ................................ ................................ ........... 26 Steam Activation ................................ ................................ .............................. 27 One Step Steam Activation ................................ ................................ ..................... 28 Chemical Impregnation with One Step Steam Activation ................................ ....... 28 Acid W ashing ................................ ................................ ................................ ... 29 Impregnation ................................ ................................ ................................ .... 29 One Step Steam Activation ................................ ................................ .............. 29 De colorization Testing ................................ ................................ ............................ 30 Spectrophotometry ................................ ................................ ........................... 30 Batch Mixing ................................ ................................ ................................ ..... 30

PAGE 6

6 Pore Analysis ................................ ................................ ................................ .......... 31 4 RESULTS AND DISCUSSION ................................ ................................ ............... 34 CA 50 ................................ ................................ ................................ ...................... 34 Pyrolysis ................................ ................................ ................................ ................. 35 One Step Steam Activations ................................ ................................ ................... 36 Calcium Activations ................................ ................................ ................................ 37 Pore Ana lysis ................................ ................................ ................................ .......... 39 Statistical Analysis ................................ ................................ ................................ .. 40 5 CONCLUSION ................................ ................................ ................................ ........ 54 LIST OF REFERENC ES ................................ ................................ ............................... 57 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 60

PAGE 7

7 LIST OF TABLES Table page 3 1 Summary of chemical to wood ratios, masses, and solution volumes ................ 33 4 1 Summary of CA 50 decolorization and pore analysis tests ................................ 52 4 2 Summary of pore analysis for 650C, 1 h carbonized sample ............................ 53 4 3 Summary of 650C one step steam for 1 h and 750C one step steam at 1, 2, and 3 h decolorization and pore analysis tests ................................ ............... 53 4 4 Summary of 3:1 calcium samples B1, B2, B3, B6, and B7 decolorization and pore analysis tests ................................ ................................ .............................. 53

PAGE 8

8 LIST OF FIGURES Figure page 2 1 A selection of color causing apple juice polyphenols ................................ .......... 24 3 1 Schematic detailing the layout of pipes for N 2 gas and water/steam delivery to the upright furnace. ................................ ................................ ......................... 32 3 2 Light absorbance curve for Hannaford apple juice from 300 nm to 500 nm wavelengths ................................ ................................ ................................ ........ 33 3 3 Apple juice light absorption ratio for varying CA 50 doses with 4 h of mixing time ................................ ................................ ................................ ..................... 33 4 1 Nitrogen gas volume adsorption isotherm for CA 50 ................................ .......... 41 4 2 Horvath Kawazoe micropore size distribution for CA 50 derived from a N 2 adsorption isotherm ................................ ................................ ............................ 41 4 3 Density functional theory mesopore size distribution for CA 50 derived from a N 2 adsorption isotherm ................................ ................................ ....................... 42 4 4 Carbonized sawdust congo red and methylene blue removal values at 0.3 mg/L dose for 4 h ................................ ................................ ................................ 42 4 5 Apple juice color removal with 1 h one step steam samples at varying temperatures ................................ ................................ ................................ ...... 43 4 6 Apple juice removal with 750C one step steam samples at varying times compared to a two step activated sample produced using 450C pyrolysis for 1 h foll owed by 750C steam for 2 h ................................ ................................ ... 43 4 7 Nitrogen gas volume adsorption isotherms ................................ ........................ 45 4 8 Horvath Kawazoe micropore size distributions for 750C one step steam for 1 h and 650C one step steam for 1 h samples derived from N 2 adsorption isotherm s ................................ ................................ ................................ ............ 46 4 9 Density functional theory mesopore size distribution for CA 50, 750C one step steam for 1 h and 650C one step steam for 1 h samples derived from N 2 adsorption isotherms ................................ ................................ ..................... 46 4 10 Apple juice removal with calcium impregnated samples of varying chemical to wood ratios activated with one step steam at 750C for 2 h ........................... 47 4 11 Nitrogen gas volume adsorption isotherms for several calcium impregnated samples ................................ ................................ ................................ .............. 49

PAGE 9

9 4 12 Horvath Kawazoe micropore size distributions for several 3:1 calcium impregnated samples, derived from N 2 adsorption isotherms ............................ 49 4 13 Density functional theory mesopore size distributio n for several 3:1 calcium samples, derived from N 2 adsorption isotherms ................................ ................. 50 4 14 Horvath Kawazoe micropore size distributions for CA 50, 750 ws 1hr, calcium sample 3:1 B1, and 650 ws 1 hr ................................ ............................ 50 4 15 Brunauer, Emmett, and Teller mesopore cumulative volume model for CA 50, calcium sample 3:1 B1, 750 ws 1 hr, and 650 ws 1 hr. ................................ 51 4 16 Density functional theory mesopore size distribution curves for CA 50, calcium sample 3:1 B1, 750 ws 1 hr, and 650 ws 1hr ................................ ......... 51 4 17 Percent apple juice color removal against mesopore volume for all samples with nitrogen gas adsorption isotherms ................................ .............................. 52 4 18 Percent apple juice color removal against mesopore volume for tested one step steam and calcium samples ................................ ................................ ........ 52

PAGE 10

10 LIST OF ABBREVIATION S BET Brunauer, Emme t t, and Teller method for estimating the surface area of porous materials BJH Barrett, Joyner, and Halenda method for calculating pore volume for pores with diameters greater than 20 DFT Density functional theory (non local) method for determining the micropore and mesopore size distribution of porous materials GAC Granular activated carbon. Activated carbon found in a solid, granular form. HK Horvath and Kawazoe method for determining the micropore size distribution of porous materials PAC Powdered activated carbon. Activated carbon that has been milled and passes a standard 325 mesh screen ws With steam. Used for labeling one step ste am activated samples

PAGE 11

11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering ONE STEP STEAM AND CALCIUM ACTIVATION OF SAWDUST F OR APPLE JUICE PURIFICATION By Timothy S. English II May 2011 Chair: David Mazyck Major: Environmental Engineering Sciences Activated carbon was synthesized from a readily available agricultural byproduct, sawdust, using one step steam activation and calcium impregnation with subsequent one step steam activation. A carbon dose of 5,000 mg/L was necessar y to reach the color rem oval capacity of 90 % for a commercially available activated carbon, CA 50. Carbonized samples of sawdust exhibited minimal pore development, with methylene blue and congo red removals around 16% and 4% respectively One step steam samples had apple juice color removals as high as 79% and total pore volumes of up to 0.50 cc/g, as determined by nitrogen gas isotherm analysis. Calcium samples had similar apple juice color removal values (70 87% color removal), but significantly more mesopore volume ( 80 150% increase ) and total pore volume ( 50% increase ) over one step steam samples Further analysis of the Horvath Kawazoe and den sity functional mesoporosity occurred due to increased general pore development and not at the expense of micropores. Linear correlations determined that increasing the po re volume of diameters greater than 40 led to increased apple juice color removal percentages.

PAGE 12

12 CHAPTER 1 INTRODUCTION Activated carbon is a highly porous material often employed as an adsorbent for the removal of trace contaminants from aqueous and gaseous systems. The high micro and mesoporosity of activated carbon allow it to remove contaminants at the molecular level, making it ideal for t carbon is typically produced through the gasification or chemical impregnation with subsequent heating of coal and lignin, and is often processed into granular or powdered media. The goal o f this research i s to produce a mesoporous and high volume (> 0.75 cc/g) p ow d ered a ctivated c arbon (PAC) from sawdust, a readily available agricultural byproduct, for the removal of the natural colors found in apple juice. If a PAC can be produced to effec tively remove the color in apple juice, it could then be applied to other industrial and environmental concerns, such as sugar decolorization and natural organic matter removal. A variety of activation methods were employed during this research, including steam activation and chemical activation with calcium salts. The products of t hese methods were compared to a high quality commercial PAC using a consistent decolorization protocol and nitrogen gas adsorption analysis. Sawdust was chosen as a precursor be cause of its low cost and availability. The proceeding chapters will review the characteristics and production of activated carbon, the experimental procedure and results of this research, and finally a conclusion based on the results and previous research

PAGE 13

13 CHAPTER 2 LITERARY REVIEW In tro duction to Activated Carbon Activated carbon has been used as an adsorbent for many centuries, but large scale industrial production of the material did not begin until World War I, when effective adsorbents were needed for protection against chlorine gas attacks (Yehaskel, 1978; Marsh, 2006). Early carbons were primarily created through pyrolysis, where wood or bones were treated under low oxygen conditions and high heat to create charcoals (Bansal, 1988 ; Bansal, 2005 ). After the war, manufacturing processes, carbon sources, and expanded uses were significantly researched and developed (Yehaskel, 1978). Activated carbon is commonly used when chemical constituents need to be removed from gaseous or liquid phases; t his includes the treatment of potable water, municipal and industrial wastewater, chemical purification, and solution decolorization (Bansal, 1988; Marsh, 2006; Hameed, 2007). Activated carbon is able to achieve effective constituent removal rates and trea tment efficiencies because it has a high adsorptive capacity, caused by its large specific surface area (generally between 400 1,500 m 2 /g), high surface reactivity, and extensive pore structures (Hassler, 1974; Bansal, 1988). Introduction to Apple Juice Decolorization Decolorizing apple juice for its sugar has been an important goal for the food and juice manufacturing industries for many years. Manufacturers currently use powdered activated carbon, gelatins, ion exchange media, chemical treatment, and me mbrane filtration to remove the color. These processes can be expensive and manufacturers are searching for more economic al solutions (Borneman, 1997)

PAGE 14

14 The brown color of apple juice is caused by a combination of phenols, polyphenols, tannic acids, flavono ids, and other organic complexes ( Borneman, 1997 ). Figure 2 1 shows a selection of organic complexes that cause apple juice color. While activated carbon is often used for solution and sugar decolorization, very little published research has specifically t argeted activated carbon apple juice decolorization. Sugar Decolorization Activated carbon has been used for over a century to decolorize sugar and sugar solutions, such as apple juice (Yehaskel, 1978; Bansal, 1988; Bernardo, 1997; Borneman, 1997; Bansal, 2005; Marsh, 2006). In order to accomplish this, the colored solution must be in direct contact with the activated carbon for a prescribed length of time (dependent on the solution mixture, quantity of activated carbon, and temperature) which is determined using batch reactor or breakthrough analysis (Reynolds, 1982 ; Bansal, 1988 ). Past studies have employed ful ly agitated mixing basins for a given duration that is held constant for the experiment (Bernardo, 1997; Hameed, 2007; Rafatullah, 2010). Since natu rally occurring molecules that cause color (e.g. tannic acids, flavonoids, etc.) tend to be relative ly large (8,000 15,000 Da), an activated greatly affects its decolorization capacity (Bansal, 2005). Color Test Methods (Spectrophotometry) Since the presence of dyes in waste effluents is an environmental and health concern, studies have been conducted using activated carbon to remove color causing agents before discharge (Hameed, 2007; Rafatullah, 2010). Hameed et al. (20 07) investigated using activated carbon to remove methylene blue dye from aqueous solution. This study found base readings from a spectroph otometer for a dyed solution calibrated with deionized water

PAGE 15

15 The solution was then put in contact with activated carbon, filtered, and measured with found by comparing readings from a spectrophotometer before and after contact with the carbon. Spectrophotometry is often used for determining color removal because its calibration curves are generally consistent, simple, and reproducible. Introduction to A ctivated C arbon Science As mentioned previously, activated carbon has been well studi ed since the early twentieth century; this has led to many scientific practices for the measurement and characterization of porous adsorbent materials. Marsh et al. (2007) outlines over 27 research has gone into adsorption concepts, surface chemistry interactions, and pore structure manipulation and measurement. Adsorption Activated carbon is an effective adsorbent because it has such an extensive pore structure and surface area Activated carbon is composed of layers of aromatic rings. Pores form in regions between layers that are not directly aligned, and have areas of disorder. It is within these pores that activated carbon achieves its surface area and active adsorption sites (Marsh, 2006). process. Physical adsorption occurs when a molecule surface through hydrogen bonding, dipole dipole moments, and van der Waals for ces (Reynolds, 1982; Marsh, 2006). These physical attractions are much weaker than chemical adsorption, which occurs when molecules covalent or aromatic bonds. Chemisorption is generally less frequent than physical

PAGE 16

16 adsor ption and is irreversible as the adsorbate becomes incorporated into the activated (Menendez, 1996). Pore Structure and Adsorbate Diffusion The pore structure of activated carbon is an important aspect to consider when evaluating its abili ty to adsorb chemicals. As mentioned previously, pores form when aromatic carbon ring stacks are misaligned (Marsh, 2006). Pores are classified by their diameter (d) as either micropores (d < 2 0 ) mesopores (2 0 < d < 50 0 ), or macropores (d > 50 0 ) ( Everett, 1972). Since the micropores contain the bulk of the surface area, most of the adsorption occurs within them (Reynolds, 1982; Fa ust, 1983; Marsh, 2006). For adsorbate molecules smaller than 20 the mesopore s and macropores tend to act a s a conduit to the micropores For molecules larger than the micropore diameter the mesopore surface area may be the site of adsorption (Bansal, 1988 ; Yang, 2003 ). After an adsorbate makes contact with the activated carbon, it move s to active adsorption site s by means of diffusion. This process is relatively slow compared to the adsorptive attachment stage ( Snoeyink, 1980 ) important factor in adsorptio n ; if the molecule adsorbs to a site in a pore of comparable size, it i s possible for the pore network past the adsorption site to be blocked for other molecules. I f the molecule is larger than a given pore size, that pore will not aid in adsorption ( Snoeyink, 1980; Yang, 2003 ). Gas Adsorption Isotherm Models The specific surface area (m 2 /g) of activated carbon is often determined using the Brunauer, Emme t t, and Teller (BET) method (Brunauer, 1938; Faust, 1983). This method uses gas adsorption over a range of relative pressures to determine the volume

PAGE 17

17 of nitrogen gas ( V m ) t hat would be required to cover the surface area of the material with a monolayer of the gas. This is found by plotting the adsorption isotherm of nitrogen gas and comparing it to the BET equation: (2 1 ) where P is the equilibrium pressure, P o is the saturation pressure of nitrogen, V is the volume of nitrogen absorbed, and C is a constant for the gas solid pair. V m is determined by calculating the slope of (C 1)/V m C With the volume of nitrogen gas known, the ge ometric dimensions of the molecule can be used to calculate the surface area (Brunauer, 1938; Faust, 1983). The Barrett, Joyner, and Halenda (BJH) method is a classical model for calculating the mesopor e size distribution of porous material s The BJH metho d incorporates the relationship between a nitrogen gas isotherm desorption curve and the pore volume for a given pore diameter (Barrett, 1951; Yang, 2003) This method assumes that all pore s are non and the Kelvin equation is applicable. The Kelvin equation is known to have problems with pore diameters that approach molecular sizes as it does not account for the effects of fluid wall interactions and methods incorporating it such as BJH, should not be used for micropore analysis (Yang, 2003). The Kelvin equation ( Equation 2 2 ) is used to determine the radius at which nitrogen condensation occurs for a given por e size at a particular relative pressure. The BJH equation ( Equation 2 3 ) can then be used to calculate the cumulative pore volume for a range of actual and Kelvin pore sizes. The derivative of this plot can be used to approximate the mesopore size distrib ution.

PAGE 18

18 (2 2 ) (2 3 ) where r k is the Kelvin pore radius, V p is the pore volume, r p is the actual pore radius, l is the pore length, is the nitrogen film thickness, and S is the pore area ( Barrett, 1951; Lowell, 2004). In order to measure micropore volumes and determine a micropore size distribution, the Horvath Kawazoe (HK) method can be applied to a nitrogen gas isotherm at varying relative pressures. This method corrects problems found in the Youn g Lapl ace and Kelvin equation s when measuring micropores, and can measure pores smaller than 10 at relative pressures less than 2 x 10 5 The basic form of the HK method relates the two equations below: (2 4 ) (2 5 ) w here l is the distance between the nuclei of two layers of carbon (pore), d a is the diameter of the adsorbate molecule (N 2 ), w is the known amount of nitrogen adsorbed, w is the maximum amount of nitrogen adsorbed, and w/w is a function of the Frenkel Halsey Hill theory. Finding the w value and relative pressure isotherm allows one to compare the above equations to find the volume for varying values of l resulting in a micropore size distribution (Horvath, 1983; Yang, 2003).

PAGE 19

19 The non local density fun ctional theory (DFT) method is a computer model that measures the volume of micro and mesopores using statistical mechanics. The DFT model assumes that the carbon has slit pore geometry, the length of the pore is far greater than the width, and that surfa ce functional groups can be disregarded (Lastoskie, 1993). DFT uses complex mathematical iterative modeling to create density profiles. These profiles are then compared and calibrated against a database of real isotherm data from non porous materials, as w ell as real physical data (Ustinov, 2006) Using these calibration curves, computer software determine s the best fit, from which pore volumes can be calculated and a pore s ize distribution can be m odeled ( Lastoskie, 1993 ). A ctivated C arbon Production Methods This review will highlight three varieties of steam activation: one step steam, two step pyrolysis and steam, and chemical impregnation with one step steam activation. The two step process, which consists of carbonization and oxidation stages, has traditionally been used for physical activation, while chemical impregnation has been around since the early 1930s and is often used when a carbon with very high surface area (>1 000 m 2 /g) is required (Yehaskel, 1978; Bansal, 1988). One step ac tivation skips the carbonization step, and activates the carbon directly with heat and steam (Laine, 1991; Warhurst, 1997; Alaya, 2000). The ultimate goal of these techniques is to remove disorganized carbon and develop pore networks within the carbon stru cture (Bansal, 1988). Physical Activation Physical activation is the process of exposing a carbonaceous material to high temperatures ( ca. 400 1 1 00C) and an oxidant (H 2 O, CO 2 etc.). The exposure has to

PAGE 20

20 be timed and temperatures well controlled for succes sful, consistent results. Many activated carbons are made using physical activation techniques ( Bansal, 1988; Wigmans, 1989; Marsh, 2006 ). Two S tep Step 1 Pyrolysis The first step of the two step activation process is to expose a carbonaceous material (e.g. wood, coal, peat, etc.) to temperatures between ca. 400 850C in an inert environment, such as nitrogen gas or more practically oxygen devoid ; this process is know n as carbonization or pyrolysis (Bansal, 1988; Rodriguez Reinoso, 1995). The purpose of pyrolysis is to remove volatile organics, non carbon chemical species and reduce tar formation, resulting in a rudimentary pore structure, with very little surface are a (5 20 m 2 /g) and pore volume (Bansal, 1988; Wigmans, 1989). Parameters that affect pyrolysis and the resulting char include the final temperature, duration of final temperature, and properties of the material (Bansal, 1988). Experimentation and char analy sis are necessary to find the proper pyrolysis methods for a given precursor material. Step 2 Steam In the two step process, the char is activated with gasification after pyrolysis. The purpose of activation is to increase the pore volume, pore size, an d surface area established through pyrolysis (Bansal, 1988; Wigmans, 1989). When an oxidant, such as steam, is combined with temperatures between 600 1100C the carbon atoms on the pore surfaces of the char can be gasified to carbon monoxide, as shown in E quation 2 6 ; the steam further interacts with gaseous carbon monoxide to produce carbon dioxide and hydrogen as shown in Equation 2 7 (Bansal, 1988; Marsh, 2006). C + H 2 2 29 kcal (2 6 )

PAGE 21

21 CO + H 2 2 + H 2 + 10 kcal (2 7 ) The removal of carbon atoms through gasification enlarges existing pores, creates new pores, or combines smaller pores to create larger ones with less surface area (Bansal, 1988). Since steam carbon gasification is an endothermic pro cess, an external heat source with accurate controls is necessary to maintain the reaction (Bansal, 1988; Wigmans, 1989; Marsh, 2006). During gasification, the oxygen atom from carbon monoxide or free hydrogen atoms may chemisorb to the char, forming surfa ce complexes. These surface complexes retard the gasification rate, as they tend to be more stable than the surrounding carbon (Bansal, 1988; Marsh, 2006). When using steam, hydrogen is a greater factor on the reaction than carbon monoxide because hydrogen surface complexes are generally more stable. While this inhibition increases activation times and temperatures, it may not necessarily have a negative impact, as slower reaction rates allow for better control of gasification and pore development (Marsh, 2 006). Also, Lopez et al. (2003) increased vinegar decolorization capacities and rates. One S tep Laine et al. (1991), Warhurst et al. (1997), and Alaya et al. (2000), all explored the potential o f one step steam activation. This process combines pyrolysis and activation. This can be advantageous because thermal activation is very energy intensive, and by reducing the amount of time the carbon must be in the furnace can reduce costs. The one step process has also been shown to increase the yield of the final product (Alaya, 2000).

PAGE 22

22 Alaya et al. (2000) was able to produce activated carbons with BET surface areas of 877 m 2 /g and 18% yields from palm branches, 1007 m 2 /g and 21% from palm leaves, and 77 5 m 2 /g and 19% from date pits. These carbons were produced using a one step constant steam process at temperatures ranging from 600 700C for 2 h The results show that one step activation produced a viable activated carbon product with adequate yield from plant based precursor materials. Warhurst et al. (1997) also had similar results activating seed husks at 700C for 30 60 minutes. The carbons had high phenol removal rates (90% removal in 30 minutes), but slightly lower yields of 12%. Chemical Activation Introduction to C alcium During the reactivation of granular activated carbon (GAC), calcium collected on the carbon modifies the pore structures (Cannon, 1994; Cazorla Amoros, 1996; Mazyck 2000; Juarez Galan, 2009). Cannon et al. (1994) found that calcium oxides act as a catalyst for carbon gasification during thermal treatment through the CaO to Ca(OH) 2 reaction, water/metal interactions, and direct reactions between water a nd the calcium species. Calcium infused GAC reactivated at 650 750C has greater micropore losses than if activated at 850 950C, due to the thermodynamic properties of carbon oxide speciation (Cannon, 1994). Cazorla Amoros et al. (1996) further explored the use of calcium catalysis for selective porosity development in plant derived biomass. Calcium impregnated chars had burn off rates that were 25 35% quicker than virgin char. The increased burn off rates also translated t o increased meso and macroporosity and a loss of microporosity. The results suggest that the calcium activation process can be used to create wider pores and greater mesopore volumes than typical steam activation.

PAGE 23

23 Juarez Galan et al. (2009) used calcium s pecifically for the development of mesoporosity in plant derived biomass. Nitrogen gas adsorption isotherm analysis had shown that the mesopore volumes of calcium impregnated samples increased up to 330% compared to the calcium free samples. The average 68 % reduction in BET surface area suggested that mesoporosity was increased at the expense of microporosity. Chemical I mpregnation/A ctivation In order to activate a carbon precursor with a desired chemical, the material must be impregnated by the chemical. C azorla Amoros et al. (1996), Juarez Galan et al. (2009), and Hu et al. (2001) thoroughly mixed organic precursors in solutions of the desired chemical species (e.g. CaCl 2 CaO, KOH, ZnCl 2 etc.) for a prescribed length of time ; some of the studies used a c onstant chemical solution protocol, while others varied the ratio of material to chemicals. Juarez Galan et al. (2009) used a constant concentration of CaCl 2 (7% by weight) and mass of precursor for 7 h at 85C. On the other hand, Hu et al. (2001) varied the amount of ZnCl 2 added to a constant volume of water in order to achieve ZnCl 2 to precursor ratios varying from 0.25 to 3 by weight. Cazorla Amoros et al. (1996) used a method similar to Juarez Galan et al. (2009), but the precursor was activated for 2 h at 850C under carbon dioxide and then mixed with calcium acetate. Following impregnation, the samples were then subjected to steam and carbon dioxide thermal re activation For these studies, a ctivation was achieved using carbon dioxide and steam as oxid ants Hu et al. (2001) introduced zinc chloride impregnated samples to 800C conditions for 2 and 3 h Following activation, the samples were rinsed successively

PAGE 24

24 with deionized water and a 0.1 mol/L solution of hydrochloric acid to remove excess zinc and chloride compounds. Juarez Galan et al. (2009) activated calcium chloride impregnated samples with carbon dioxide at a rate of 100 mL/min for a variety of temperatures and times (750 824C and 4, 6, and 12 h ). The activated samples were then washed with a 5% hydrochloric acid solution to remove any excess calcium. Cazorla Amoras (1996) used a nitrogen gas/steam mixture to activate calcium acetate impregnated carbon dioxide activated char Mixtures of 30%, 40%, and 50% steam at 100 mL/min were used at 850 C. Figure 2 1 A selection of color causing apple juice polyphenols (Borneman, 2001)

PAGE 25

25 CHAPTER 3 EXPERIMENTAL METHODS Pine wood sawdust a saw mill byproduct was used to create activated carbon using one step steam activation two step steam activation with pyrolysis and calcium impregnation with one step steam activation methods. The carbons were then pow d ered and tested for apple juice decolorization capacity using batch reactor mixing, filtering, and spectrophotometry analysis. A se lection of these carbons were run through nitrogen gas adsorption analysis, creating relative pressure isotherms, from which the BET surface area, total pore volume, average pore size, and pore size distributions (HK, BJH, and DFT) were determined. Materia l Preparation The saw dust was first passed through a 16 mesh sieve and hand sorted to remove any large pieces of wood or bark. The sieved sawdust was then dried in a 110C oven for at least 24 h before further treatment. This basic sawdust preparation was performed for all samples, regardless of the treatment process. Two Step Steam Activation Two step steam activation consists of two distinct stages : a pyrolysis stage and an activation stage. All activation and pyrolysis steps were performed in a vertical, upright furnace capable of accurately and consistently holding temperatures fro m 250 1200C. A 1 inch stainless steel pipe ran through the center of the furnace. The bottom opening of this p ipe was attached to tubing connected to variable area flow meters (rotameters) capable of precisely delivering nitrogen gas or water to the furnace pipe. When steam was needed, heating tape wrapped around the gas delivery tubing was

PAGE 26

26 heat ed to approximately 250C to flash boil the water. Figure 3 1 illustrates the tubing and furnace pipe set up. In order to accurately measure the temperature within the furnace pipe, a 24 inch thermocouple was lowered into it Af ter the furnace had heated to within 50C of the desired temperature, the gas or steam was turned on at a constant flow rate (1,000 mL/min for nitrogen gas and 1.1 mL/min for water/steam) The temperature controls on a digital temperature control unit were then adjusted until the desir ed temperature was held constant for at least 10 minutes. Samples were lowered 24 inches into the furnace tube using a stainless steel basket and metal wire ; it was important that the length of the basket and wire was the same as the thermocouple, as tempe ratures inside the furnace tube vary greatly with distance The basket was equipped w ith 120 mesh frets to allow gas or steam to flow through the sample while minimizing losses. Pyrolysis The first step for pyrolysis was to pre heat the furnace to the desi red charring temperature (450 C, 550C, 650C, or 750C) with a 1,000 mL/min nitrogen gas flow The high gas flow ensures that the environment within the furnace tube will be inert and oxidant free. For each sample, 6 g of the dried sawdust was massed and lightly packed into the furnace basket. Two 120 mesh frets were installed on the top and bottom of the basket to prevent the sawdust from blowing out of the basket while allowing the nitrogen gas to pass through the sample. After the sawdust was loaded into the basket, it was lowered 24 inches into the pre heated furnace with a metal wire. The sample was then allowed to carbonize for 1 h After pyrolysis, the basket was raised to the top of the furnace tube, where tem peratures were much cooler (~1 5 0 C), and allowed to cool under the nitrogen gas

PAGE 27

27 flow for 15 minutes. The basket was then removed from the furnace tube and cooled at room temperature for 15 minutes. During the room temperature cooling stage, the thermocoupl e was placed back in the furnace tube and the temperature checked; if the measured temperature was 10 C of the desired pyrolysis temperature, the sample was saved, otherwise, it was discarded. Saved samples were funneled into labeled test tubes for storage in a vacuum desiccator until needed for testing or activation Steam Activation After pyroly sis, the furnace temperature was raised to the activation temperature (750C) under 1.1 mL/min of s team. The carbonized wood was massed and loaded into the basket which was then lowered 24 inches into the pre heated furnace tube. Two step activat ed s amples were held under constant s team flow and temperature for 2 h After this time the steam was stopp ed and replaced with a nitrogen gas flow of 1,000 mL/min T he basket was then raised in the furnace tube to the cooler upper region of the furnace, where it was cooled under nitrogen gas for 15 minutes. The basket was then cooled at room temperature, while the furnace temperature was checked with the 24 inch thermocouple. As with pyrolysis, samples with temperature variations greater than 10C were discarded. After activation, the cooled sample was massed and then milled with a mortar and pestle until it w as a fine powder. This powder was separated with a 325 mesh sieve ; the sample that passed through was, by definition, a pow d ered activated carbon (PAC) This PAC was funneled into labeled vials, and stored in a vacuum desiccator until needed for testing

PAGE 28

28 O ne Step Steam Activation One step steam activation used a process similar to two step activation, without the first, pyrolysis step. The furnace was pre heated to the desired temperature (450 C 550 C 650 C 750 C or 800C) with steam at 1.1 mL/min. Approximately 6 g of sieved, dried wood was lightly packed into the basket which was lowered 24 inches into the pre heated furnace tube. The temperature and steam flow were held constant for a prescribed length of time (1 h or 2 h ). After time was reached, the steam was turned off and replaced with 1,000 mL/min of nitrogen gas, and the sample was cooled in the upper region of the furnace tube for 15 minutes. The basket was then cooled at room temperature, while the furnace temperature was checked; samples w ith temperature variations greater than 10C were discarded. After activation, the cooled sample was massed and then milled with a mortar and pestle until it was a fine powder. This powder was separated with a 325 mesh sieve The PAC that passed through t he sieve was funneled into labeled vials, and stored in a vacuum desiccator until needed for testing. Chemical Impregnation with One Step Steam Activation The chemical samples had significant pre treatment prior to activation including sorting, acid washi ng, chemical impregnation, and several drying steps This process was streamlined into sequential batches to minimize variations in the protocol ; wood was acid washed in larger quantities, capable of creating 5 individual samples, while these samples were impregnated in sequential batches with enough time between batches to allow for furnace activation. After post activation treatment, the samples were stored and tested at the same time.

PAGE 29

29 A cid W ashing Before impregnation, the sieved, sorted, and dried sawdus t was washed in a 10% by volume sulfuric acid solution. Enough wood was acid washed in each batch to create 5 samples at 5 g each (~28 g per acid wash batch). 50 mL of acid solution was used for each gram of wood; therefore each batch had 1,400 mL of acid solution. The massed wood was added directly to the measured acid solution and quickly mi xed on a stir plate for 1 h After mixing, the acid washed wood was filtered from the acid solution with a very fine mesh. The wood was then rinsed with several liters of deionized water Each batch of acid washed and rinsed wood was allowed to fully dry in a 110C oven for at least 24 h prior to further treatment. Impregnation For calcium impregnations, a 7% by weight solution of calcium chloride was created in 500 g b atches by mixing 35 g of CaCl 2 2H 2 O solid with 465 mL of Nanopure deionized water. Samples were impregnated by mixing this solution with 5 g of acid washed sawdust for 8 h on an 85C stir plate. The chemical to wood ratio was varied by using different volumes of calcium solution as shown in Table 3 1 After mixing, the slurry was transferred to a 110C oven, where it was dried for 24 h before activation ; this result ed in 3 2 h of chemical contact time. One Step Steam Activation After drying, the calcium impregnated sawdust was loaded into the basket and activated using a process similar to the one step steam activation method previously mentioned. The furnace was pre heated to 750C with a steam flow of 1.1 mL/min. The

PAGE 30

30 samples were activated for 2 h and cooled under nitrogen gas. The temperature was checked, and samples with temperature variations greater than 10C were discarded. After activation and cooling, the individual samples were washed for 30 minutes in 200 mL of 5 % by volume hydrochlo ric acid solution to remove the calcium and other ions. The samples were then rinsed with approximately 2 L of deionized water in a membrane vacuum filter until a pH of 6 8 was reached. The sample was then dried in a 110C oven for 24 h After drying the sample was massed and then milled with a mortar and pestle until it was a fine powder. Th is powder was separated with a 325 mesh sieve funneled into labeled vials, and stored in a vacuum desiccator until needed for testing. Decolorization Testing Spectrophotometry The apple juice used in these experiments was Hannaford Apple Juice from conce ntrate. It was tested in a HACH DR/4000U Spectrophotometer from wavelengths of 300 nm to 500 nm, as shown in Figure 3 discernable peaks, the value of 330 nm was chosen as a reference for color removal. Percent col or removal was determined using the following equation: (4 1) where A 0 is the absorbance of the untreated apple juice measure d at 330 nm, and A is the measured absorbance of the treated (decolorized) apple juice at 330 nm. Batch Mixing apple juice color adsorptive capacity. A commercially available phosphoric acid

PAGE 31

31 activated PAC, Carbochem CA 50, was used as a reference and benchmark for setting the decolorizat ion testing protocol. In order to determine the PAC dose for batch analysis of color removal, varying masses of CA 50 were added to 100 mL samples of apple juice. The samples were then tumbled and mixed for 4 h w ith a control of untreated juice. After mixi ng, the PAC was removed using a 0.45 m cellulous membrane vacuum filter and the apple juice was tested with the spectrophotometer at 330 nm. A s shown in Figure 3 3 a dose of 5,000 mg/L was chosen for maximum decolorization capacity batch testing. This PA C dose and testing method was used as the testing protocol for produced samples using 200 mg of PAC in 40 mL of apple juice. Methylene blue and congo red dye removal tests were also perform ed on the carbonized samples to give a basic analysis of micro and mesopore development. Solutions of 10 mg/L dyes were made; 100 mL of the solution was mixed with 30 mg of carbonized wood for 4 h The mixtures were vacuum filtered with a 0.45 m membrane a nd the m ethylene blue removal was tested in the spectrophotometer at 650 nm while congo red was tested at 487 nm. Pore Analysis A Quantachrome NOVA 2200e surface area and pore size analyzer was used for characterizin g the CA 50 and select samples. For ea ch tested sample, n itro gen gas adsorption isotherms were developed using volumes for 20 adsorption and 10 desorption relative pressure values ranging from 0.013 to 0.99 NovaWin version 9.0 software was used to calculate the BET surface area, total pore v olume, average pore size, and the pore size distributions (HK, BJH, and DFT). BET surface area was calculated using the multipoint method, wh ere the low relative pressure points (P/P 0 < 0.15) are used in the BET equations to find an

PAGE 32

32 estimated total surface area. The total pore volume was found by calculating the total amount of N 2 gas adsorbed at the highest relative pressure point (~0.98) The average pore size, HK, BJH and DFT pore size distributions we re calculated through analysis of the N 2 isotherm. Figure 3 1 Schematic detailing the layout of pipes for N 2 gas and water/steam delivery to the upright furnace 15 psi

PAGE 33

33 Figure 3 2 Light absorbance curve for Hannaford apple j uice from 300 nm to 500 nm wavelengths Figure 3 3 Apple juice light absorption ratio for varying CA 50 doses with 4 h of mixing time at 330 nm wavelength light Table 3 1. Summary of chemical to wood ratios, masses, and solution volumes CaCl 2 2H 2 O:Wood Wood mass (g) CaCl 2 2H 2 0 mass (g) Vol. of 7% solution (mL) 2:1 5 10 143 3:1 5 15 215 4:1 5 20 286

PAGE 34

34 CHAPTER 4 RESULTS AND DISCUSSI ON CA 50 The commercially available Carbochem CA 50 PAC was used as a reference and benchmark for the following set of experiments. As mentioned previously, CA 50 was used to find that the maximum adsorbed color removal w as achieved with PAC doses of 5,000 mg/L or greater. Running several apple jui ce color adsorption tests at 5,000 mg/L PAC for 4 h mixing time found that CA 50 had consistent color removal percentages around 9 0 %, with a mean of 92%, minimum of 88% and maximum of 94%. While the 5,000 mg/L concentration effectively removes color from apple juice, it is a very high dose and producing lower cost activated carbons would be very beneficial to industries requiring decolorized apple juice. Also, CA 50 may contain residual phosphorus as a result of its chemical activation methods, which could be leached into the apple juice solution; a s a result of this, a more benign activation chemical (e.g. calcium) would be preferable A nitrogen gas adsorption isotherm for CA 50 (Figure 4 1) was used to determine the BET surface area, average pore si ze, and total pore volume, which are shown in Table 4 1. The total pore volume suggests that CA 50 has a well developed pore structure the average pore size indicating that there is substantial mesoporosity, and the BET surface area shows that there is al so significant microporosity Further analysis of the N 2 isotherm using the HK, BJH, and DFT pore size distribution models give more detail into the physical pore structure of CA 50. Figure 4 2 shows that CA 50 has a peak specific pore volume differential at 10 with a value of 0.045 cc/ /g that subsequently tapers off to 0.013 cc//g as it approaches 20 The

PAGE 35

35 micropore volume was found to be 0. 5 0 cc/g or 4 0% of the total volume. It is within located. The DFT mesopore size distribution, Figure 4 3, shows that the peak pore volume differentials are much less than in the micropore range, but the overall volume (area under the curve) is much greater. The BJH method found that the mesopore volume i s around 60% of the total pore volume (0.7 4 cc/g). Table 4 1 summarizes the results of the various tests for CA 50. Pyrolysis Pyrolysis is used to remove non carbon chemical species and volatile organics from the sample and create a rudimentary pore struct ure prior to activation. It is the first step in the traditional two step activation process. For this experiment, prepared sawdust was carbonized and analyzed for basic pore development Four samples were made under nitrogen gas flow at 450C, 550C, 650 C, and 750C and tested using methylene blue and congo red adsorption analysis, the results of which are shown in Figure 4 4. Dye adsorption was studied to provide quicker insight into the porosity of the activated carbons, for nitrogen adsorption requires about 30 h per sample. The methylene blue removal values suggest that there is some microporosity and surface area, but methylene blue removal with the same concentration of CA 50 was 100%. The negligible congo red removals suggest tha t there is very little mesopore surface area. Table 4 2 further supports these analyses with the isotherm analysis of the 650C sample; there is some surface area (200 m 2 /g) and microporosity (20 ), but very little pore development (0.09 cc/g).

PAGE 36

36 One Step S team Activations One step steam activations were performed by placing the prepared sawdust samples directly into the preheated furnace under 1.1 mL/min steam flow at a variety of temperatures and times. Figure 4 5 highlights the apple juice color removal c apacities of PACs activated at 450C, 550C, 650C, 750C, and 800C for 1 h At 450C and 550C there is no substantia l color removal; this is due to a lack of porosity development. Since the conversion of carbon to CO and CO 2 using steam is endothermic it is apparent that these samples required more activation energy than that available at 450C and 550C. On the other hand, samples activated at 650C and 750C had significant increases in ap ple juice color removal values The 800C sample shows a very l arge drop in removal capacity; this is most likely due to pore collapse. When the surface of the micropo res is gasified, it expands them until they combine to form larger pores. If enough pores rapidly combine in this manner, the physical matrix of the saw dust can break down, leading to pore collapse. Since the 750C 1 h sample had the highest color removal, 750C was explored at times of 2 and 3 h These samples al l had similar removal values, around 78%. Figure 4 6 compares the different times, as well as a sample made using two step steam activation, where the pyrolysis step was performed at 450C for 1 h and activation at 750C for 2 h This comparison shows that results of the one step activation method are comparable to two step and that greater times more color removal. Several of the one step steam activated samples were analyz ed with nitrogen gas adsorption: 650C for 1 h (650 ws 1hr), 750C for 1 h (750 ws 1hr), 750C for 2 h (750 ws 2hr), and 750C for 3 h (750 ws 3hr). The N 2 isotherms in Figure 4 7 show that 750

PAGE 37

37 ws 1 hr and 750 ws 2hr have the most pore volume and overall porosity; the steep initial slope (P/P 0 < 0.15) for 750 ws 2 hr suggests that it has more microporosity than the other samples. The 650C 1 h sample (650 ws 1 hr) has the least pore development as indicated by its lower adsorption volume s HK analysis verifies that 750 ws 2hr does have the most microporosity, while 650 ws 1hr had the lea st. The HK micropore volumes for the samples are summarized in Table 4 3. The micropore volumes do not vary greatly (0.23 0.30 cc/g) and the percentage of microporosity compared to the total pore volume was fairly close as well (47 60%). The mesoporosity w as modeled with the BJH and DFT methods. Figure 4 9 shows that the samples are all fairly close at the peak pore volume differentials. However, 750 ws 1hr has the most overall mesopore volume, while 650 ws 1hr and 750 ws 3hr both have less differential vol ume between 40 50 As summarized in Table 4 3, 750 ws 1hr had the most mesopore volume, which was 51% of the total volume. The 750C samples all had more mesopore volume than 650 ws 1hr and better removal values (78 79% for 750C compared to 67% for 650 ws 1hr). Calcium Activations Calcium samples were acid treated and impregnated with calcium chloride prior to one step steam activation. The ratio s of calcium chloride to saw dust used were 2:1, 3:1, and 4:1. Two samples of the 2:1 ratio were created, while seven were created for 3:1. The results of the apple juice decolorization tests are shown in Figure 4 10; for the 2:1 and 3:1 samples, the removal value was calculated from the mean of the samples, while the error bars represent the maximum and minimum re moval values. The figure shows that mean color removals are around 80% for 2:1 and 3:1 and 74% for 4:1. The 2:1

PAGE 38

38 samples had a range of 74% and 82%, while the 3:1 samples had a larger range of 70% to 87%. This variability is most likely a result of differen ces in the initial precursor caused by the small sample size and slight variations in the acid washing, impregnation, and activation protocols. Several of the 3:1 samples (labeled as B1 B7) were analyzed with N 2 adsorption. The isotherms for these samples are shown in Figure 4 11. Batch 1 has the most pore volume, while batch 7 has the least The sharp rise in the isotherms at relative pressure s greater than 0.8 suggests that these carbons have some meso and macropore structures. The HK micropore size distributions were determined for several of the 3:1 samples and plotted in Figure 4 12. This plot shows that the calcium samples all had similar micropore structures with little variability The average micropore volume for the samples is 0 .28 cc/g, which is about 40 % of the total pore volume for samples B1 and B2 and 45% for samples B3, B6, and B7 T able 4 4 shows this information in detail for each sample. The DFT mesopore size distribution was found for the tested samples and is shown in Figure 4 13. All of the calcium samples had pore volume differential peaks at 35 45 and 80 The sample B7 had the least mesoporosity and lowest apple juice decolorization value. Using the BJH method, sample s B1 and B2 had the most mesoporosity at 0 .44 cc/g, or 60% of the ir total pore volume s ; sample B7 had the least mesopore volume at 0.25 cc/g, or 50% of its total pore volume. The mesopore volumes for all of the tested samples are summarized in Table 4 4.

PAGE 39

39 Pore Analysis The HK, BJH, and DFT pore ana lyses for CA 50, 650 ws 1hr, 750 ws 1hr, and 3:1 B1 were compared against ea ch other. Figure 4 14 shows that CA 50 has much greater micropore differential volumes than the other samples; this is why CA 50 has such a relatively large BET surface area. Calci um s ample 3:1 B1 which has a micropore size distribution that is representative of all the tested calcium samples, has microporosity similar to that of the one step steam samples, meaning that t he increased pore volume of the calcium samples does n o t come at the expense of microporosity. The BJH mesopore volume comparisons (Figure 4 15) show that CA 50 has substantial mesopore volume compared to the other samples. The calcium sample had almost twice the cumulative volume than that of the best one step stea m sample. The slope of the curves show that the majority of the mesopore development for all the samples occurs with pore diameters under 12 0 ; the decreased slope after that point suggests that there is no t significant pore development in terms of pore volume, beyond this point. The DFT pore size distributions (Figure 4 16) show that a relatively large portion of the mesopore volume for CA 50 comes from the pore s with diameters between 20 75 After 75 match that of CA 50 I t was observed in Figure 4 13, that all of the tested calcium activations had this peak at 75 The one step steam samples do not exhibit this increased mesoporosity at p ore diameters greater than 75 suggesting that this region of pore development is related to the calcium catalyzed activation process. An a nalysis of the data was conducted to find any relations that might exist between the mesopore volume and apple juice color removal. Figure 4 17 shows a plot

PAGE 40

40 of color removal pe rcentage against total mesopore volume for all of the tested samples There is a moderate linear correlation (R 2 = 0. 69 ) between increasing mesoporosity and mesopore volume. However, w hen this analysis is done with just the steam or calcium samples, the co rrelation is very strong with R 2 value s of 0.99 and 0.95 respectively as shown in Figure 4 18 The pore size distributions for the calcium sample (Figures 4 12 and 4 13) show that they all have very similar differential volumes under 40 suggesting that differences in color removal due to pore structure are caused by the variations in differential volumes of pores with diameters greater than 40 The bulk of the steam iameters less than 75 (Figure 4 16). Statistical Analysis The differences between the apple juice color removal and pore structures of one step steam and calcium samples were verified using a statistical t test with a 95% confidence interval. The one ste p steam 750C s amples at 1, 2, and 3 h (3 samples ) were tested against the 3:1 calcium sample s (7 samples for color removal and 5 for pore structure values ) For percent color rem oval, the two tailed P value is 0.63, meaning that the sample sets are not considered statistically different. The BET surface area was also not statistically different (P = 0.39). For t he average pore size mesopore volume, and total pore volume, P equals 0.005 0.01, and 0.02 respecti vely, which are all considered very statistically different. This analysis verifies that the mesopore volume, total pore volume and average pore size are greater for calcium activated samples, while the BET surface area and color removal values are not st atistically different. Thus, the calcium samples had greater mesopore development but this did not come at a loss of microporosity or BET surface area.

PAGE 41

41 Figure 4 1 Nitrogen gas volume adsorption isotherm for CA 50 Figure 4 2 Horvath Kawazoe micropore size distribution for CA 50 derived from a N 2 adsorption isotherm

PAGE 42

42 Figure 4 3 Density functional theory mesopore size distribution for CA 50 derived from a N 2 adsorption isotherm Figure 4 4 Carbonized sawdust congo red and methylene blue removal values at 0.3 mg/L dose for 4 h

PAGE 43

43 Figure 4 5. Apple juice color removal with 1 h one step steam samples at varying temperatures Figure 4 6 Apple juice removal with 750C one step steam samples at varying times compared to a two step activ ated sample produced using 450C pyrolysis for 1 h followed by 750C steam for 2 h

PAGE 44

44 Figure 4 7 Nitrogen gas volume adsorption isotherms. A) One step steam activation at 650C for 1 h (650 ws 1hr). B) One step steam activation at 750C for 1 h (750 ws 1hr). C) One step steam activation at 750C for 2 h (750 ws 2hr). D) One step steam activation at 750C for 3 h (750 ws 3hr).

PAGE 45

45 Figure 4 7 Continued

PAGE 46

46 Figure 4 8 Horvath Kawazoe micropore size distributions for 750C one step steam for 1 h and 650C one step steam for 1 h samples derived from N 2 adsorption isotherms Figure 4 9 Density functional theory mesopore size distribution for CA 50, 750C one step steam for 1 h and 650C one step steam for 1 h samples derived from N 2 adsorption isotherms

PAGE 47

47 Figure 4 10 Apple juice removal with calcium impregnated samples of varying chemical to wood ratios activated with one step steam at 750C for 2 h Figure 4 11 Nitrogen gas volume adsorption isotherms for several calcium imp regnated samples. A) 3:1 calcium sample, B1. B) 3:1 calcium sample, B3. C) 3:1 calcium sample, B6. D) 3:1 calcium sample, B7.

PAGE 48

48 Figure 4 11 Continued

PAGE 49

49 Figure 4 11. Continued Figure 4 12 Horvath Kawazoe micropore size distributions for several 3:1 calcium impregnated samples, derived from N 2 adsorption isotherms

PAGE 50

50 Figure 4 13 Density functional theory mesopore size distribution for several 3:1 calcium samples, derived from N 2 adsorption isotherms Figure 4 14 Horvath Kawazoe mic ropore size distributions for CA 50, 750 ws 1hr, calcium sample 3:1 B1, and 650 ws 1 hr

PAGE 51

51 Figure 4 15 Brunauer, Emmett, and Teller mesopore cumulative volume model for CA 50, calcium sample 3:1 B1, 750 ws 1 hr, and 650 ws 1 hr. Figure 4 16. Density functional theory mesopore size distribution curves for CA 50, calcium sample 3:1 B1, 750 ws 1 hr, and 650 ws 1hr

PAGE 52

52 Figure 4 17 Percent apple juice color removal against mesopore volume for all samples with nitrogen gas adsorption isotherms Figure 4 18 Percent apple juice color removal against mesopore volume for tested one step steam and calcium samples Table 4 1 Summary of CA 50 decolorization and pore analysis tests Apple juice decolorization BET surface area (m 2 /g) Avearge pore size ( ) Micropore volume (cc/g) Mesopore volume (cc/g) Total pore volume (cc/g) 92% 1,500 34 0.5 0.74 1.30

PAGE 53

53 Table 4 2 Summary of pore analysis for 650C, 1 h carbonized sample BET surface area (m 2 /g) Average pore size ( ) Total pore volume (cc/g) 200 20 0.09 Table 4 3 Summary of 650C one step steam for 1 h and 750C one step steam at 1, 2, and 3 h decolorization and pore analysis tests Activation temp. (C) Activation time (h) Apple juice decolorization BET surface area (m 2 /g) Avg. pore size () Micropore vol. (cc/g) Mesopore vol. (cc/g) Total pore vol. (cc/g) 650 1 67% 600 25 0.23 0.15 0.39 750 1 79% 630 30 0.23 0.25 0.49 750 2 78% 820 25 0.30 0.18 0.50 750 3 78% 710 25 0.25 0.19 0.45 Table 4 4 Summary of 3:1 calcium samples B1, B2, B3, B6, and B7 decolorization and pore analysis tests 3:1 Sample Apple juice decolorization BET surface area (m 2 /g) Avg. pore size () Micropore vol. (cc/g) Mesopore vol. (cc/g) Total pore vol. (cc/g) B1 87% 710 42 0.29 0.45 0.75 B2 83% 710 41 0.29 0.44 0.74 B3 83% 630 41 0.26 0.39 0.65 B6 80% 690 37 0.28 0.36 0.64 B7 70% 600 33 0.24 0.25 0.50

PAGE 54

54 CHAPTER 5 CONCLUSION The goal of this research was to produce a mesoporous and high volume (> 0.75 cc/g) PAC from sawdust for the removal of the color f ound in apple juice Apple juice color adsorption tests were performed at the lowest, maximum color removal dose of 5,000 mg/L. This high of a dose can be come very costly when used to treat large quantities of app le juice Pyro lysis and subsequent m ethylene blue and congo red dye adsorption and ni trogen gas adsorption tests verified that pyrolysis at varying temperatures under nitrogen gas led to a very rudimentary pore structure with little BET surface area ( 650C : 200 mg/g) and pore volume ( 650C: 0.09 cc/g). One step steam activation at a temperature range of 450 C 800C for 1 h had a pple juice color test results that show ed that there was not enough energy at 450C and 550C to produc e significant porosity The 650C and 750C samples had color removals of 67% and 79% respectively suggesting significantly more porosity At 800C, the removal value drops significantly (28%), most likely due to pore collapse. S team s amples created at 750C for 2 and 3 h had color removals near 80%. A two step, carbonized and steam activated sample tested at similar color removal levels, supporting the claim that one step steam activation produces activated carbon with comparable color adsorption capacity to that of two step steam a ctivation. N itrogen gas adsorption tests found that the micropore volumes do not vary greatly amongst the tested steam sample s and that 750C produced more mesopore volume than the 650C.

PAGE 55

55 The calcium samples activated under steam at 750C for 2 h with vary ing chemical to wood ratios had average color removals similar to that of the 750C one step steam samples. The 3:1 chemical to wood samples had a range of 70% to 87% Compared to CA 50, some of the better performing 3:1 samples had color removal differenc es of only 5 8%. Pore analysis found that on average, the calcium samples had a 40 % increase in total pore volume and 8 0 % increase in mesopore volume over 750C one step steam activated samples. The calcium samples all had peaks in the DFT mesopore size distribution at 75 90 that were unique to those samples as these pores were created This peak is where most of the increased mesoporosity comes from. The color removal and BET sur face area were not considered statistically different. The statistically substantial increase in mesopore and total pore volumes in the calcium did not come at the expense of microporosity. For the CA 50, one step steam, and calcium samples, i ncreasing mesopore volume was linearly correlated to increased apple juice color removal with an R 2 = 0.69 Since the calcium and one step steam samples have very similar pore size distributions before 40 differences in color removal due to pore structure are most likely caused by the variations in the differential volumes of pores with diameters greater than 40 This research was able to produce a high pore volume PAC with well developed mesoporosity that also had significant apple juice color removal values (up to 87% ) using calcium impregnation combined with one step steam activation With further refinement of the chemical activation protocol a production cost analysis and an operational cost analys is, it could be possible to produce a consistent calcium activated

PAGE 56

56 PAC that would perform as well as or better than the CA 50 and other similar more costly activated carbons.

PAGE 57

57 LIST OF REFERENCES Alaya, M. N., Girgis, B. S., & Mourad, W. E. (2000). Activated carbon from some agricultural wastes under action of one s tep steam pyrolysis. Journal of Porous Materials, 7 (4), 509. Bansal, R. C., Donnet, J., & Stoeckli, F. (1988). Active carbon New York, NY: Marcel Dekker, Inc. Bansal, R. C. & Goyal, M. (2005). Activated Carbon Adsorption. Boca Raton, FL: CRC Press. Barr ett, E. P., Joyner, L. G., & Halenda, P. P. (1951). The determination of pore volume and area distributions in porous substances. I. Computations from nitrogen isotherms. Journal of the American Chemical Society, 73(7) 373 380. Borneman, Z., Gokmen, 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 (1), 191. Borneman, Z., Gokmen, V., and Nijhuis, H. H. (2001). Selective removal of polyphenols and brown colour in apple juices using PES/PVP membranes in a single ultrafiltration process. Separation and Purification Technology, 22 23, 53 61. Brunauer, S., Emmett, P. H., & Teller, E. (1938) Adsorption of gases in multimolecula r layers. Journal of the American Chemical Society 60 (1) 309 319. Cannon, F. S., Snoeyink, V. L., Lee, R. G., & Dagois, G. (1994). Reaction mechanism of calcium catalyzed thermal regeneration of spent granular activated carbon. Carbon, 32 (7), 1285 1301. Cazorla Amoros, D., Ribes Perez, D., Roman Martinez, M. C., & Linares Solano, A. (1996). Selective porosity development by calcium catalyzed carbon gasification. Carbon, 34 (7), 869. Everett, D. H. (1972). Manual of symbol and terminology for physiochemic al quantities and units, part I. Pure Applied Chemistry, 31(4) 579 638. Faust, S. D. & Aly, O. M. (1983). Chemistry of Water Treatment Boston : Butterworth 130 131. 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. Hassler, J.W. (1974) Purification with Activated Carbon: Industrial, Commercial, Environmental. New York, NY: Chemical Publishing Co., Inc.

PAGE 58

58 Horvath, G. & Kawazoe, K. (1983). Method for the calculation of effective pore size distribution in molecular sieve carbon. Journal of Chemical Engineering of Japan, 16(6), 470 475. Hu, Z., & Srinivasan, M. P. (2001). Mesoporous high surface area activated carbon. Microporous and Mesoporous Materials, 43 267 275. Juarez Galan, J. M., Silvestre Albero, A., Silvestre Albero, J., & Rodrigues Reinoso, F. (2009). Synthesis of activated carbon with highly developed "mesoporosity". Microporous and Mesoporous Mat erials, 117 519 521. Lastoskie, C., Gubbins, K. E., & Quirke, N. (1993). Pore size heterogen eity and the carbon slit pore: A density functional theory model. Langmuir, 9 (10) 2693 2702. Lowell, S., Shields, J. E., Thomas, M. A., & Thommes, M. (2004). Cha racterization of Porous Solids and Powders: Surface Area, Pore Size and Density. Dordrecht, Netherlands: Kluwer Academic Publishers. Marsh, H., & Rodriguez Reinoso, F (2006). Activated carbon New York, NY : Elsevier. Mazyck, D. W., & Cannon, F. S. (2000). Overcoming calcium catalysis during the thermal reactivation of granular activated carbon. part I. steam curing plus ramped temperature N treatment 2. Carbon, 38 1785 1799. Menendez, J.A., Phillips, J., Xia, B., & Radovic, L.R. (19 96). On the modification and characterization of chemical surface properties of activated carbon: In the search of carbons with stable basic properties. Langmuir, 12, 4404 4410. Laine, J., Simoni, S., & Calles, R. (1991). Preparation of activated carbon fr om coconut shell in a small scale cocurrent flow rotary kiln. Chemical Engineering Communications, 99 (1), 15 23. Rafatullah, M., Sulaiman, O., Hashim, R., & Ahmad, A. (2010). Adsorption of methylene blue on low cost adsorbents: A review. Journal of Hazard ous Materials, 177 70 80. Reynolds, T. D. (1982). Unit Operations and Processes in Environmental Engineering. Monterey, CA: Brooks/Cole 187 211. Rodriguez Reinoso, F., Molina Sabio, M., & Gonzalez, M. T. (1995). The use of steam and CO2 as activating ag ents in the preparation of activated carbons. Carbon, 33 (1), 15 23. Snoeyink, V. L., & Jenkins, D. (1980). Water Chemistry. New York, NY: John Wiley & Sons, Inc. Ustinov, E. A., Do, D. D., & Fenelonov, V. B. (2006). Pore size distribution analysis of acti vated carbons: A pplication of density functional theory using nongraphitized carbon black as a reference system. Carbon 44(4), 653 663.

PAGE 59

59 Warhurst, A. M., McConnachie, G. L., & Pollard, S. J. T. (1997). Characterisation and applications of activated carbon produced from moringa oleifera seed husks by single step steam pyrolysis. Water Research, 31 (4), 759 766. Wigmans, T. (1989). Industrial aspects of production and use of activated carbons. Carbon, 27 (1), 13 22. Yang, R. T. (2003). Absorbents: Fundamentals and Applications. Hoboken, NJ: John Wiley & Sons, Inc. Yehaskel, A. (1978). Activated carbon: Manufacture and regeneration Park Ridge, NJ: Noyes Data Corporation.

PAGE 60

60 BIOGRAPHICAL S KETCH Timothy English II has a Bachelor of Science in Environmental Engineering Sciences from the University of Florida. He received a Master of Engineering in the same program and university. After finishing his degree, he plans on working as a consulting engineer in the fields of water/wastewater treatment and water resources in the state of Florida.