|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
INVESTIGATION OF PROPERTIES RESPONSIBLE FOR PHENOL REMOVAL VIA
TITANIA COATED ACTIVATED CARBON
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
I would sincerely like to thank Dr. David Mazyck for giving me the chance to work
on this challenging project. His guidance and constant encouragement enabled me to
work through and succeed during some frustrating times. I have gained a lot of
knowledge by working with him. I would like to thank my committee Dr. Paul Chadik,
Dr. Joseph Delfino and Dr. Joseph Geunes who were always willing to help and
provided me with a lot of suggestions throughout the duration of my research.
I would like to thank all the students in Dr. Mazyck's research group Ameena
Khan, William 'Beau' Kostedt, Morgana Bach, Jennifer Stokke and Heather Byrne. Their
suggestions and help at all the stages of my work helped me immensely. I consider
myself very fortunate to have worked with such a great group. Thanks also to Aly Byrne
and Gustavo Avila for their assistance on a number of experiments. I would also like to
acknowledge Gautam Kini for his help. I thank Rick and Matt at Engineering
Performance Solutions who were very efficient and quick in analyzing my samples.
I especially thank Sudeepti Southekal for being a great friend and a constant source
of encouragement. I hope to help her in the same way as she pursues her PhD.
Above all, I must thank my family for their love and support and for the principles
and ethics that they have instilled in me.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ................................................................................................. iii
TA BLE OF CON TEN TS................................................................ .......................... iv
LIST OF TABLES .............. ................. ........... ................... ........ vi
L IST O F FIG U R E S .... ...... ...................... ........................ .. ....... .............. vii
ABSTRACT ........ .............. ............. ...... ...................... ix
1 INTRODUCTION ............... ................. ........... ................. ... .... 1
2 LITERA TURE REVIEW .......................................................... ..............4
2.1 Activated Carbon ........................ .. .......... ........ .........................
2.1.1 Physical Properties of A C .................................. ............... ...................5
2.1.2 Chem ical Properties of A C ................................................ ...... ......... 7
2.1.3 A dsorption M echanism s.................................................................... .... .9
2 .2 P h oto cataly sis ................................................................................... 12
2 .2 .1 T itanium D ioxide........... ...... ...................................... .......... ... ......... 13
2.2.2 Mechanism of Photocatalysis ..................... ............... 14
2.2.3 Improvements in Photocatalysis................................ ...............16
2.3 Fermi Energies/Levels, Electrical Conductivity and Schottky Barrier ................18
2.4 Previous Synergy in AC-TiO2 Systems..... .......... ...................................... 21
3 M ATERIALS AND M ETHODS ........................................ ......................... 26
3.1. T target P ollutant ........ ............... ... ......... .... ....... .. ... .......... .. ......... .... 26
3.2 Carbon M materials ................. .............. ........ .. .. .... ...... .. ...............26
3.2.1 TiO2-AC Composites and Coating Procedures....................................26
3.2.2 Mass Based Coating (m/m) ................................................. 31
3.2.3 M olar Based Coating (m ol/m ol)...................................... ............... 31
3.2.4 Low Surface Area Carbon-TiO2 Composite.............................................33
3 .3 E xperim ental Setup ...................................................................... ...................34
3.3.1 Batch Reactor Configuration........................................................ ........ 34
3.3.2 A nalytical E quipm ent....................................................................... ... ...34
3.4 Experimental Procedure....................................................... .................. 35
3.4.1 Concentration Measurements ....................................... ............... 35
3.4.2 Batch Adsorption/Oxidation Tests with Activated Carbon......................36
3.4.3 Batch Adsorption/Oxidation Tests with Low Surface Area Carbons ........37
4 RESULTS AND DISCU SSION ........................................... .......................... 39
4.1 Evaluation of Carbon Coating Strategies .................................. ............... 39
4.2 A activated C arbon Studies ................................................. ........................ 41
4.2.1 Ash Analyses of Activated Carbons........... .......................41
4.2.2 SEM Images of Titania Coated ACs .................................. ............... 47
4.2.3 Activated Carbon Batch Studies.......................... ......................52
4.2.4. Determination of Existence of Phenol* (Phenol with Delocalized
Electron) in the Presence of UV ... ................. ..... ...............62
4.3 Low Surface Area Carbon Studies .................................................. 66
4.3.1 Metals Content of Low Surface Area Carbons................ .... ........... 68
4.3.2 Batch Studies with Low Surface Area Carbons ......................................71
5 SUMMARY AND CONCLUSIONS ............................................. ...............78
5 .1 S u m m a ry ......................................................................................................... 7 8
5 .2 C o n c lu sio n s ..................................................................................................... 8 1
6 CONTRIBUTIONS TO SCIENCE AND ENGINEERING .....................................82
6 .1 C contribution s to Science ............................................................ .....................82
6.2 Contributions to Engineering........................................... .......... ............... 83
A VIRGIN AC STUDIES IN PRESENCE OF UV ....................................................84
B C O L U M N ST U D IE S ....................................................................... .....................85
C C O L U M N STU D IE S II .............................................................................. ............86
LIST OF REFEREN CES ............................................................ .................... 87
B IO G R A PH IC A L SK E T C H ..................................................................... ..................91
LIST OF TABLES
2-1. Photocatalysis reaction m echanism s ............................................... ............... 16
3-1. Sum m ary of A C properties............................................... .............................. 26
3-2. Carbon contents of virgin and titania coated ((m/m) and (mol/mol)) activated
carbons and their surface areas........................................... .......................... 32
3-3. Surface areas of virgin and coated low surface area carbons.................................33
3-4. Carbon contents of low surface area carbons ......................................................33
4-1. Sum m ary of coating techniques ........................................... .......................... 40
4-2. Ash contents of ACs and mass % of titania deposited (m/m) on AC surface ............42
4-3. Phenol removal as a function of AC coating strategy with and without irradiation ..63
4-4. Electrical conductivity values for low surface area carbons ....................................67
4-5. Elemental carbon and ash contents of low surface area carbons.............................68
4-6. Average densities of 3% (m/m) titanium dioxide coated carbons.............................74
LIST OF FIGURES
2-1. TiO 2 structure. A ) anatase, B ) rutile.................................. ....................... .. .......... 14
2-2. Photocatalysis m echanism .......................................................... ............... 15
2-3. Band gaps for (a) metals, (b) semiconductors and (c) insulators with Fermi level
(SF) in dictated .........................................................................19
2-4. Alignment of Fermi levels and formation of Schottky barrier ...............................20
2-5. Kinetics of phenol disappearance in the presence and absence (photolysis) of
various illum inated solids............................................... .............................. 22
3-1. Coating activated carbon with TiO2 via mechanical attachment (Theta Composer,
Tokuju Corp., Japan) .................. ............................. ........ .. ........ .... 31
3-2. Concentration vs. Absorbance correlation for phenol on UV spectrophotometer
m measured at 270nm w avelength ........................................ .......................... 35
4-1. EDS of A) ash residue of uncoated Bionuchar, B) Bionuchar coated with TiO2, C)
ash residue of uncoated F400, D) F400 coated with TiO2, E) ash residue of
uncoated HD4000, F) HD4000 coated with TiO2 ................... .... .......... 44
4-2. Titania coated Bionuchar with the TiO2 blocked pores circled..............................47
4-3. Bionuchar showing cellulose structure of the wood based AC...............................48
4-4. Titania coated on the surface of F400 ............................................. ............... 49
4-5. TiO2 agglomerate (shown circled) near a macropore on the surface of HD4000 ......50
4-6. Surface of HD4000 showing uniform coating of TiO2 which tends to form
agglom erates..................................... ............................... ........... 51
4-7. Batch adsorption studies of virgin ACs in the absence of UV..............................53
4-8. Batch adsorption studies of 3% (m/m) coated ACs in the dark compared to TiO2
slu rrie s ............................................................................... 5 4
4-9. Batch adsorption-photocatalysis studies of 3% (m/m) coated ACs in the presence
o f U V ..............................................................................................5 6
4-10. Batch adsorption-photocatalysis studies of 3% (m/m) coated ACs loaded to the
reactors on a volume basis in the presence of UV ................................................59
4-11. Batch adsorption studies: A) 3% (mol/mol) coated ACs in the absence of UV
loaded to the reactors by volume, B) Photocatalysis studies of 3% (mol/mol)
coated ACs in the presence of UV loaded to the reactors on a volume basis ..........60
4-12. Comparison of irradiated virgin ACs with virgin and coated ACs in the presence
and absence of U V ............................... .. .. .......... ...... .... ...... ...... 62
4-13. EDS scan of ash residue of uncoated anthracite coal ................... ....... .........69
4-14. EDS scan of ash residue of uncoated bituminous coal...........................................69
4-15. EDS scan of ash residue of pitch coke .... ............................................... 70
4-16. Adsorption studies of coated low surface area carbons and titanium dioxide
slurry .................................. ..................................71
4-17. Batch photocatalysis studies of carbons and titanium dioxide slurry (m/m)............73
4-18. Photocatalysis studies of coated low surface area carbons in the presence of UV
(m /v ) ...................................... .................................................... 7 5
4-19. Photocatalysis studies of low surface area carbons in the presence of UV (mol/v).76
A-1. Replicate data sets for virgin AC studies performed in the presence of UV
demonstrating the phenol* phenomenon observed in Section 4.2.4......................84
B-1. Recirculation column studies performed with F400 using different flowrates
(degrees of fluidization) in the presence and absence of UV..............................85
C-1. Adsorption runs of column tests performed with HD4000 in the presence and
absence of UV radiation measuring phenol concentration in the effluent ..............86
C-2. Concentration of phenol in effluent during regeneration runs of column studies .....86
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 Science
INVESTIGATION OF PROPERTIES RESPONSIBLE FOR PHENOL REMOVAL VIA
TITANIA COATED ACTIVATED CARBON
Chair: David Mazyck
Major Department: Environmental Engineering Sciences
Activated carbon has long been used to remove organic pollutants from water.
However, a major disadvantage is that it needs to be regenerated regularly, which is
primarily done offsite. The regeneration decreases the adsorption capacity of the
activated carbon and hence increases costs. Titanium dioxide, through the mechanism of
photocatalysis, mineralizes the organic pollutants to carbon dioxide and water. A
drawback of this technology is that when used in the form of a slurry, the separation of
the titanium dioxide from the treated effluent incurs costs. The use of a titanium dioxide
coated activated carbon would therefore result in the elimination of the separation of
titanium dioxide from water and would at the same time result in the in-situ regeneration
of activated carbon. It has however been found in previous research that the behavior of
the TiO2 coated activated carbons varies for different ACs.
The objective of this research was to identify a simple yet effective procedure to
coat the ACs with titanium dioxide and to identify properties of activated carbons that
influence a synergistic effect between the activated carbon and the titanium dioxide. On
the premise that conductivity due to the presence of metals influences photocatalysis in
the TiO2-AC composites, batch studies were performed comparing the adsorption and
photocatalysis of three activated carbons having different properties. Characterization
studies were also performed on the activated carbons comparing their surface
morphologies and compositions of their ash. It was observed that ACs having less acidic
groups on their surface resulted in better adsorption of the pollutant (i.e., phenol). It was
also observed that the coal based ACs had a rougher surface resulting in a more uniform
deposition of titanium dioxide on their surface.
To further isolate and study the mechanism of photocatalysis, low surface area
carbons were used in separate studies. These carbons differed in terms of ash composition
and electrical conductivity. These carbons were coated with titanium dioxide and tested
in batch for photocatalytic activity. The ash composition of these carbons was also
For activated carbons and low surface area carbons, results showed no noticeable
influence of conductivity on the photocatalytic activity of the ACs or the carbons.
Photocatalysis was observed as being a secondary mechanism compared to adsorption
which was clearly dominant.
Over the past 20 years, the planet's drinking water resources have entered into a
very fragile state. A number of organic pollutants, the sources of which are solvents,
fertilizers (e.g., NPK), pesticides and chlorophenols, among others, have resulted in
widespread groundwater contamination (Hoffman et al., 1995). To cope with the growing
pollution of the hydrosphere, a number of regulations and programs are being
implemented. The primary strategy being applied currently is the chemical treatment of
polluted drinking water, surface water, groundwater and wastewater. Pollutant removal in
drinking waters may involve flocculation, sterilization, filtration and disinfection
processes (Legrini et al., 1993). Technologies utilized may also include ultrafiltration, air
stripping, carbon adsorption and oxidation (via ozonation or hydrogen peroxide)
Recent advances in the field of water treatment have led to the development of
oxidative degradation techniques for organic compounds dissolved in water. These
processes are known as advanced oxidation processes (AOPs). Heterogeneous
photocatalysis is one type of advanced oxidation process that can be used to counter the
growing contamination of precious water resources due to organic compounds. This
process entails the use of a photocatalyst that is activated under the influence of UV
radiation which ultimately results in the oxidation of the organic contaminants to water
and carbon dioxide, resulting in an acceptably minimal waste stream. This lack of toxic
organic byproducts is a significant advantage of this technology over the contemporary
technologies mentioned earlier, which simply transfer the pollutant from one phase to
another. The technology can be very useful in applications for sustenance during long-
term NASA space missions. Due to the limited reserves of water that may be carried on
the space shuttle, a closed-loop photocatalytic system would theoretically be able to
extend the lifespan of a given finite reserve of water almost indefinitely. Additionally,
due to the depleted resources of drinking water here on Earth, it is of great interest to
adapt this technology for industrial purposes as well as for household water purification.
The use of photocatalysis has been widely studied with the photocatalyst (TiO2) in the
form of a slurry. However, a slurry based system used for industrial purposes or for
household use may result in a waste stream consisting of the photocatalyst itself.
Separation of TiO2 from the treated effluent could further complicate matters while
increasing the cost of the system.
Filters containing activated carbon and other treatment media are currently used in
many households as secondary treatment systems. However, these filters need to be
replaced from time to time due to the exhaustion of the activated carbon. Hence,
combining the theoretically infinite sustainability of the photocatalysis phenomenon with
the simplicity of an activated carbon filter, a compact treatment device is sought to be
developed with the photocatalyst (titanium dioxide) coated on the activated carbon. The
main challenges in the development of such a system as a viable application for industrial
and household use are as follows:
* To identify a simple technique by which the activated carbons may be coated with
titanium dioxide. Preferably, the coating technique would not incur any
considerable additional costs.
* To develop an AC-TiO2 composite that demonstrates both high capacity as well as
high photocatalytic efficiency.
Numerous coating techniques were tested. The techniques were evaluated on the
basis of simplicity and effectiveness. Making use of the coating technique that best fits
the need; efforts were concentrated on optimizing an AC-TiO2 composite that shows the
best performance with respect to photocatalysis of a model organic contaminant; in this
case phenol (C6H50H). Hence, it is imperative to isolate the intrinsic and physical
properties of activated carbons which aid in photocatalysis, assuming photocatalysis is
dependant on the intrinsic properties of the ACs. To do this, experiments were performed
on a variety of different ACs.
The need for greater scientific understanding of the phenomenon of photocatalysis
and the factors that affect it has provoked the author to concentrate part of the efforts on
the science behind photocatalysis. Focus has especially been placed on whether a synergy
exists between titanium dioxide and any particular activated carbon. This has been
investigated through the use of TiO2 coated carbonaceous materials with extremely low
adsorption capacity. Electrical conductivity of the ACs due to the presence of conductive
metals has been hypothesized to be a factor influencing the photocatalysis seen with the
AC-TiO2 composites. Identification of the factors affecting photocatalysis will in the
future allow us to further optimize a system.
In summary, this study is an effort to develop a system for in-situ regeneration of
activated carbon for household as well as industrial purposes with a concentration on
identifying a TiO2-AC composite that performs with the most synergy with respect to
adsorption and photocatalysis.
2.1 Activated Carbon
Activated carbons are adsorbents created from carbonaceous precursors that either
experience thermal or chemical activation to increase the internal porosity of the raw
material giving rise to high surface area carbonaceous materials (500-1500 m2/g).
Suitable raw materials can include wood, coal (anthracite, bituminous, and lignite), coke,
coconut shells, fly ash and even rice husk (Karthikeyan et al., 2005; Kinoshita, 1988).
Depending on the precursor, the ash content (which can include iron, aluminum, and
silica) of activated carbons can vary (e.g., 1-20% (w/w)).
Thermal or steam activation entails the oxidation of char, which is created during
the first phase of the activation process via pyrolysis, with the help of oxidizing agents
such as steam, air or carbon dioxide at temperatures between 800-1000 C. Chemical
activation involves heating the carbonaceous precursor and a dehydrating agent (ex.
phosphoric acid, carboxylic acid, sulfuric acid, nitric acid) to a temperature between 200-
6500C. The dehydrating agent may then be leached out and reused (Kinoshita, 1988).
Hence, chemically activated carbons typically have a low pH due to acidic groups on
their surface. This has resulted in loose definitions such as L-type activated carbons for
chemically activated carbons (due to 'low' pH) and H-type activated carbons for
physically activated carbons (due to 'high' pH). The difference in precursors and
activation processes result in activated carbons having varying physical and chemical
properties. As an aside, a physically activated carbon that is acid-washed to remove
organic or for other purposes may exhibit L-type characteristics.
Activated carbon is widely used in potable and wastewater treatment for the
removal of organic and inorganic pollutants. It may be used in its powdered form, known
as 'powdered activated carbon' (PAC) (particles pass through a 325 mesh sieve), or in the
granular form (GAC).
2.1.1 Physical Properties of AC
Surface area. The defining physical characteristic of activated carbon (AC) is its
well-defined pore structure which is one of the main reasons for its high internal surface
area. It is this internal surface area that is available for the adsorption of pollutants. Even
though properties such as pore size distribution, surface chemistry and adsorbate-
adsorbent interactions play a role, surface area is still considered the limiting factor for
adsorption of the target pollutant by practitioners. Hence, a greater surface area of the AC
will result in a greater potential for the adsorption of the pollutant. The typical surface
areas of activated carbons are between 500 m2/g-1500 m2/g.
Much of the carbon literature focuses on the importance of surface area but surface
chemistry is the dominant variable that controls adsorption in most cases. Therefore, the
role of surface chemistry will be addressed throughout this review and in the results and
discussions of the data.
The most common method for determining the surface area of an adsorbent is using
the Brunauer, Emmett, and Teller (BET) theory. The BET theory was initially developed
for characterization of physical adsorption. Hence its adaptation to characterize surface
area is not considered error free as it cannot universally be applied to all types of physical
adsorption (Dabrowski, 2001). The surface area analysis by the BET theory is performed
with N2 which may result in an inaccurate prediction of adsorption trends for other
molecules (Dabrowski, 2001). Hence, the size of the molecule being adsorbed must be
considered while characterizing the surface area of the adsorbent. For this reason, the
pore size distribution (PSD) of the adsorbent is an important factor in characterizing the
adsorption of activated carbon. The BET isotherm is given below.
1/ (W*((Po/P)-)) = 1/ (Wm*C) + (C-l)/ (Wm*C)*(P/Po)
W: Weight of gas adsorbed at P/Po
P: Pressure of overlying gas
Po: Saturation pressure of the gas
Wm: Adsorbate weight in monolayer coverage
C: Constant related to the energy of adsorption
Pore size distribution. The pores of activated carbon consist of three size ranges
according to IUPAC recommendations.
* Micropores: Less than 20 A (2 nm)
* Mesopores: Between 20 A (2 nm) and 500 A (50 nm)
* Macropores: Greater than 500 A (50 nm)
Nowadays, the word nanopore is used in general to define micropores and
mesopores (Dabrowski, 2001). The activation protocol to a large extent dictates the pore
sizes in the AC.
Micropores in activated carbon are usually comparable to the sizes of the adsorbate
molecules (e.g., size of phenol molecule is 6 A (Mazyck, 2000)). Therefore, all the atoms
of the adsorbent (activated carbon) can interact with the adsorbate species (Dabrowski,
2001) which is the main difference between adsorption mechanisms of micropores and
that of mesopores and macropores. Micropore adsorption is hence a pore filling process
in which the volume of the micropores is the controlling factor.
The walls of macropores are formed by a great number of adsorbent atoms or
molecules. The boundary of interfaces as well as the adsorbent surface area have a
distinct physical meaning (Dabrowski, 2001). Mono and multilayer adsorption
successively takes place on the surface of mesopores and their final fill is in accordance
with the mechanism of capillary adsorbate condensation (Dabrowski, 2001). Mesopores
participate in the transport of the pollutant molecules to the major adsorption sites in the
micropores. They are characterized mainly by their specific surface area and pore size
In the case of macropores, the action of adsorption forces does not occur
throughout their void volume but at a short distance from their walls. Like mesopores,
macropores are also diffusion pores in that they principally transport the pollutant
molecules to smaller pores.
During the adsorption process in activated carbon, four main steps occur:
1. Bulk diffusion or the diffusion of the molecule through the bulk liquid.
2. Film diffusion or the diffusion of molecules through the thin film (boundary layer)
surrounding the activated carbon.
3. Pore diffusion or the diffusion of the molecules through the pores (or along the
pore walls) of activated carbon.
4. Adsorption of the adsorbate.
The rate determining step is step 3 which is the diffusion through the pores of the
activated carbon, which is largely influenced by the pore size and the size of the diffusing
2.1.2 Chemical Properties of AC
According to Coughlin and Ezra (1968) the structure of activated carbon is
graphitic in nature with a large number of molecular layers. The layers contain carbon
atoms which are bound together by three a bonds and one n bond and result in sp2
hybridization. There is also the possibility of sp3 (tetrahedral) hybridization taking place
resulting in the cross linking of the graphite layers (Coughlin and Ezra, 1968; Hobbs,
2005). Van der Waals forces are responsible for holding together the carbon atoms within
the graphite layers resulting in the microcrystalline structure. Other atoms that are bound
in this structure may be present within the layers; known as basal planes, forming
heterocyclic rings or at the edges of the carbon molecules, known as edge sites. These
atoms form functional groups. Hence, as explained by Leon y Leon and Radovic (1994),
the surface of activated carbon consists of two types of sites: 1) basal planes and 2) edge
sites. The edge sites are considered to be more reactive with oxygen groups than the basal
planes due to the fact that edge sites contain a single unpaired electron (Radovic et al.,
Functional groups. Functional groups on the surface of activated carbons can be
classified as acidic and basic. A number of functional groups have been found on the
surface of activated carbon. These groups consist of carboxyl, lactonic, phenolic,
carbonyl, quinones and pyrones among others (e.g., peroxide). Oxygen functional groups,
despite being a small fraction of the overall carbon surface, are however very active and
exhibit a large influence on adsorption capacity (Nevskaia et al., 1998; Nevskaia et al.,
2004; Leon y Leon et al., 1994). The behavior of functional groups is governed by the
variable electronegativity of the base atoms. Hence, atoms such as oxygen with higher
electronegativity pull electrons towards its nucleus, resulting in a negative charge on that
atom and a positive charge on the remaining atoms. This results in the groups becoming
polar, hence affecting their interactions with the adsorbing molecules. Hence, bound
oxygen functionality plays a crucial role in adsorption behavior of the activated carbon
towards the pollutant.
The surface oxides may either be acidic or basic as explained earlier. Acidic surface
oxides result in a positive surface charge due to protonation whereas basic surface oxides
result in a negative surface charge due to deprotonation of the functional groups. The
source of a positive surface is illustrated by the following equation:
C, + 2H20 CH30+ + OH-
The formation of acidic or basic surface oxides depends upon the temperature to
which the carbon is exposed to during its activation process. Exposing the carbon to
temperatures between 200-5000C results in the formation of acidic oxides carboxylicc,
lactonic and phenolic). Basic surfaces on the carbon are formed when it is heated in an
inert environment to remove initial oxides and cooled prior to exposing it to oxygen. It
has also been found that the catalytic properties of ACs can be enhanced by the
elimination of some of the acidic functional groups and introduction of basic functional
groups on the carbon surface (Salame et al., 2003; Tessmer et al., 1997).
2.1.3 Adsorption Mechanisms
The topic of adsorption mechanisms seen in activated carbon has been one of much
debate. There are two schools of thought the 7t-7t bonding theory proposed by Coughlin
and Ezra (1968) and the electron donor acceptor theory (EDA) proposed by Mattson
(1969). The two theories have been explained as follows:
7r-7r bonding theory. This theory proposes that the 7t bond of the aromatic ring of
carbon covalently bonds with the aromatic ring of the adsorbate or in other words that 7t
bonds occur between two p-orbitals. The presence of electron withdrawing groups such
as oxygen on the carbon surface results in a reduction in the electron density in the 7t
system of the carbon basal plane. Hence, in theory the presence of electron withdrawing
groups should weaken the dispersive adsorption forces between the adsorbate 7t electron
and the carbon basal planes by creating positive holes in the conduction band of the 7n
electron system (Coughlin and Ezra, 1968; Radovic et al., 2001). Adsorption at the basal
planes is hence due to 7t-7t interactions and is typically weaker whereas adsorption at the
edge sites is typically stronger (Coughlin and Ezra, 1968). In fact species bonded to the
edge sites may cause the disturbances in the electron density of the basal plane, resulting
in decreased interactions and reduced adsorption of organic compounds. The presence of
oxygen groups or electron withdrawing groups in general is hence thought to be one of
the primary reasons for reduced adsorption of organic by activated carbon. It should be
noted that the opposite is also true, where the removal of electron withdrawing groups
from the surface of ACs results in an increase in adsorption potential.
Electron donor theory. The electron donor theory was proposed by Mattson et al.
(1969) and refuted Coughlin and Ezra (1968) by suggesting that an exchange of electrons
takes place during the adsorption process in which the carbonyl oxygen serves as the
donor and the aromatic ring serves as the acceptor. Mattson et al. (1969) suggest that the
carbonyl group serves as the major electron donor in the donor-acceptor complex and is
hence responsible for an increase in phenol adsorption. The subsequent decrease in
adsorption has been credited to carboxyl functional group formation (Mattson et al.,
1969). Subsequent findings by numerous research groups has shown that an increase in
oxygen functional groups results in the increased formation of both carboxyl as well as
carbonyl groups, hence refuting the findings of Mattson et al (1969).
Adsorption of phenol. Phenol adsorption is dependent mainly on the porosity and
the surface chemistry of the activated carbon. Other factors in its adsorption include the
surface area, ash content of the AC and the concentration of carbon atoms in the matrix.
The adsorption of aromatic compounds from solution has been studied extensively
(Radovic et al., 1997; Mattson et al., 1969; Coughlin and Ezra, 1968; Coughlin et al.,
1968; Moreno-Castilla et al., 1995; Nevskaia et al., 2004). It has been found by Leng and
Pinto (Leng and Pinto., 1997: as reported by Salame et al., 2003) that the uptake of
phenol is due to a combination of the interaction of phenol with the basal planes
(physisorption) as well as by surface polymerization. It has also been found that the
phenol uptake increases as the carboxylic functional groups are removed from the carbon
surface. The presence of carboxylic functional groups (and hence oxygen groups) which
are primarily formed during activation, results in the AC surface becoming more polar
resulting in water behaving as a competing species for the adsorption sites. In addition to
the decreased polarity of the AC surface, the removal of carboxylic functional groups
also enhances 7t-7t interactions. Furthermore, activated carbons, under oxic conditions are
prevented from adsorbing phenol in the presence of functional groups (Salame et al.,
2003; Nevskaia et al., 1998, Nevskaia et al., 2004). This is due to the fact that the ability
for the adsorption of phenol via oxidative coupling decreases as a result of the functional
groups. Therefore, the catalytic properties of AC can be enhanced by the introduction of
basic functional groups and the elimination of some of its acidic groups. Hence, at low
pH, the amount of phenol adsorbed increases slightly with an increase in pH until a
certain point after which a decrease in uptake is noticed.
The pH of the phenol solution is not believed to play a significant role in its
adsorption onto activated carbon. Phenol is considered to be a weak acid. The loss of a
hydrogen ion results in the formation of a phenolate anion, the negative charge of which
is delocalized around the ring. The negative charge gets spread out over the entire ion due
to the overlapping of one of the lone pairs of electrons on the oxygen with the delocalized
electrons on the benzene ring. Hence, the phenolate ion becomes relatively more stable.
However, since oxygen is the most electronegative element on the anion, the delocalized
electrons will be drawn towards it which results in the hydrogen ion being attracted
towards it again rendering solution pH a less important factor in the adsorption of phenol.
The phenomenon of photocatalysis was discovered by Fujishima and Honda in
1972 (Linsebigler et al., 1995). Since then it has been extensively studied as an option for
the treatment of contaminants in water. Photocatalysis results in the mineralization of
organic to CO2, water and the corresponding mineral acids as the final products.
Photocatalytic treatment of water requires the use of the titanium dioxide (TiO2)
photocatalyst in the form of a slurry or the immobilization of the titanium dioxide on a
substrate. The use of a slurry is beneficial due to the large surface area of the catalyst
which would be available for the photocatalysis reaction (Byrne et al., 1997). However,
due to the size of particles used in the slurry, this method of treatment requires
centrifugation or microfiltration techniques to remove the fine particles from the treated
liquid, resulting in technical and economical problems (Legrini et al., 1993). The use of
titania as a slurry also results in an impedence to the penetration of light through the
solution due to increased opacity caused by the TiO2 particles. This however only takes
place after a certain optimum titanium dioxide loading which is system dependent.
Alternatively, the separation procedure problem can be overcome by immobilizing
the TiO2 on a substrate (Dijkstra et al., 2001; Ollis et al., 1991; Ray et al., 1997). A
number of techniques have been used such as immobilization on beads, glass pearls
(Trillas et al., 1996) inside tubes of glass or Teflon (Dijkstra et al., 2001) fiberglass or
woven fibers, silica-titania composites and immobilization onto activated carbon
(Dijkstra et al., 2001). This procedure has had varying results depending on the substrate
onto which the TiO2 has been immobilized. Historically, it has been found that a
suspended slurry is more efficient due to the absence of mass transfer limitations and
larger specific surface area available for photocatalysis (Dijkstra et al., 2001). However,
it has also been shown that a TiO2-activated carbon composite particle has resulted in a
synergy by which the degradation of a dye (methylene blue) was increased (Khan, 2003;
Khan et al., 2006).
2.2.1 Titanium Dioxide
Titanium dioxide is the most commonly used photocatalyst due to its properties of
being relatively inert, corrosion resistant and low cost as compared to other
photocatalysts such as ZnS, WO3 and SrTiO3. Along with ZnO, it also shows the most
photocatalytic activity as compared to other photocatalysts. It is found in three major
forms: anatase (octahedral) (Figure 2-1 A), rutile (tetragonal) (Figure 2-1 B) and brookite
(orthorhombic). Anatase and rutile are the forms that show photocatalytic activity with
anatase showing higher activity than rutile (Ohtani et al., 1993). The synthesis of anatase
from rutile titania can usually be achieved by heat treatment at temperatures between
3000C-6000C. Pure grades of titanium dioxide are seldom found naturally as it is often
completed with other minerals and metals. Recent studies have however shown that a
combination of anatase (70-75%) and rutile (30-35%) are more active than pure anatase
(Bacsa and Kiwi., 1998; Muggli et al. 2001). The grade of titanium dioxide used herein
was commercially available, synthetically manufactured Degussa P25. Degussa P25 has a
structure which is a combination of anatase (70%) and rutile (30%). It has repeatedly
demonstrated good photocatalytic degradation.
Figure 2-1. TiO2 structure. A) anatase, B) rutile
2.2.2 Mechanism of Photocatalysis
When UV light energy is incident on TiO2, a valence band electron is excited to the
conductance band; on the condition that the incident photons have a higher energy (X <
380 nm for anatase) than the band-gap energy. When this occurs, it results in the
formation of a hole in the valence band and hence an electron-hole (e-h) pair. A high
quantum yield of the photocatalytic reaction can be achieved by the prevention or
prolongation of the recombination of the electron-hole pair. This can be done with the
help of an electron scavenger. An electron scavenger which may also be called a 'hole
trap' may consist of adsorbed hydroxide ions (OH-), oxygen or water molecules which
react with the electron-hole to create a hydroxyl radical (OH*). An illustration of the
mechanism of photocatalysis is shown in Figure 2-2 (adapted from Linsebigler et al.,
1995). The presence of water or hydroxide ions is believed to be essential for the
complete oxidation of organic molecules to CO2 and H20 since their absence has resulted
in incomplete oxidation of organic molecules (Turchi and Ollis., 1990). Dissolved
oxygen is believed to play an important role as an electron acceptor, hence preventing the
electron-hole pair from recombining. In the absence of an electron acceptor,
recombination may take place in picoseconds, resulting in no photocatalytic degradation.
Studies performed (Chhor et al., 2004; Dijkstra et al., 2001) have shown higher
efficiencies in photocatalytic activity for phenol, salicylic acid and formic acid with an
increase in the presence of dissolved oxygen. Semiconductors such as TiO2 however have
an advantage in that their recombination times are slower than those for conductors
(metals) due to the absence of a continuum of interband states which would aid in the
recombination of the electron-hole pair (Carp et al., 2004). This ensures that the electron-
hole pair has a sufficiently long time to diffuse to the surface of the catalyst and initiate
the redox reaction (Carp et al., 2004).
Figure 2-2. Photocatalysis mechanism
The following four cases of hydroxyl attack are typically considered which are also
illustrated in Table 2-1 (adapted from Turchi and Ollis, 1990).
* Reaction occurs when the radical and the pollutant are adsorbed.
* The radical is not bound and it reacts with the organic pollutant molecule.
* The adsorbed radical reacts with a free organic molecule moving past the surface.
* Both the radical as well as the organic molecule are unbound and they react with
each other in the fluid phase.
Table 2-1. Photocatalysis reaction mechanisms
Excitation TiO2 + hv e- + h+
Recombination e- + h+ heat
Adsorption 02 + TiV OH-
TiIv + H20 Tiv H20
Trapping Tiv OH- + h+ Ti" OH*
TiV H20 + h+ TiV OH* + H
Casel Ti" OH* + R1,ads TiIv + R2,ads
Case 2 OH* + R1,ads R2,ads
Case 3 Ti" OH* + R1 Ti" + R2
Case 4 OH* + R1 R2
2.2.3 Improvements in Photocatalysis
A number of techniques have been attempted to enhance the efficiency of
photocatalytic systems. One of these techniques is the doping of TiO2 particles with
metals. Doping of the semiconductor with various transition metal ions has in the past led
to an enhanced efficiency of photocatalytic systems. TiO2 particles can be substituted or
interstitially doped with different cations or can form mixed oxides. However, this
technique also poses a complex problem. The net induced alteration of the photocatalytic
activity is made up from the sum of the changes which occurs due to the following
* The light absorption capability of the TiO2 photocatalyst
* Adsorption capacity of the substrate (activated carbon) molecules at the catalyst
* Interfacial charge transfer rate
As the concentration of the dopant increases, the space charge region becomes
narrower resulting in the efficient separation of the electron-hole pairs within the region
by the dielectric field before recombination. An increase in dopant concentration
however, results in the narrowing of the space charge region. When this occurs, the
penetration of light into the TiO2 greatly exceeds the space charge layer. Due to the
absence of a sufficient driving force dopantt), to separate them, electron-hole
recombinations increase due to the absence of a driving force to separate them. Hence, an
optimum concentration of dopant ions is required which would make the thickness of the
space charge layer similar in magnitude to the light penetration depth. (Carp et al., 2004;
Hoffman et al., 1995; Linsebigler et al., 1995).
The recombination of photogenerated electrons and holes are often times
influenced by doping ions. In most cases the recombination times are enhanced by the
doping ions, hence impeding the progress of the reaction. For example, in both p-type
doping cationss having valencies lower than that of Ti+4 such as Al+3, Cr+3 and Ga+3) and
n-type doping cationss having valencies higher than Ti+4 such as Nb+5, Ta+5 and Sb+5) an
inhibition effect which is ascribed to the recombination of the electrons and holes is seen.
P-type dopants act as acceptor centers for photoelectrons whereas n-type dopants act as
donor centers of photoelectrons (Carp et al., 2004; Hoffman et al., 1995; Linsebigler et
al., 1995; Matos et al., 1998). Metallic ions such as Fe3+ and Ru3+ however behave
differently than P-type and N-type dopants. This is due to their half-filled electronic
configuration which results in them being more stable. This electronic configuration is
destroyed when metallic ions are trapped, leading to a decrease in their stability.
Moreover, a transfer of the trapped electrons to oxygen adsorbed to the surface of the
catalyst may take place with the metallic ions returning to the original stable half filled
electron structure resulting in the transfer of charge and efficient separation of the
electrons and holes by trapped electrons. Hence, co-doping may be a viable technique to
improve charge separation, thereby improving photocatalytic efficiency. In fact, studies
by Leng et al. (2002) have shown a synergistic effect resulting in the increased
photocatalytic degradation of chloroform in solution to as much as 5 times the original
rate when TiO2 has been doped with Fe+3 and Eu+3. In this case, Fe+ serves as a hole trap
and Eu+3 serves as an electron trap (Carp et al., 2004). It can therefore be summarized
that the process of photocatalytic oxidation through a semiconductor such as titanium
dioxide can be modified and controlled to a certain extent depending on the type of
metals, the extent to which they are doped on the titania and the combinations of metals
used in the doping process.
2.3 Fermi Energies/Levels, Electrical Conductivity and Schottky Barrier
Fermi level is a term used to describe the top of the 'sea' of electrons (electron
energy levels) at a temperature of absolute zero. At absolute zero, electrons pack into the
lowest available energy states to form this collection of energy states. Hence, at absolute
zero, no electron will have enough energy to rise above the surface of the Fermi 'sea'.
Ordinary electrical processes involve energies in the range of a fraction of an electron
volt. However, the Fermi energies in the case of metals are on the order of electron volts,
implying that most of these electrons cannot receive sufficient energy from those
The Fermi energy is also important in characterizing the electrical conductivity of
metals. The velocities of the electrons participating in conduction can be calculated from
the Fermi energy hence providing us with the drift velocities of electrons encountered in
the metals. Also crucial to the conduction process is the 'Band theory of solids' which is
useful in visualizing the difference between conductors, semiconductors and insulators.
The primary requirement for the conduction process is the presence of electrons in
the conduction band. In insulators, electrons in the valence band are separated from the
conduction band by a very large gap whereas in conductors, the valence band and the
conduction band overlap. The case of semiconductors (such as titanium dioxide) is
unique in that the intrinsic Fermi level is half way between the bottom of the conduction
band and the top of the valence band. Therefore, even a small amount of doping of the
semiconductor can result in a large increase in conductivity. Hence, as the Fermi level for
semiconductors approaches the conduction band, it denotes an n-type whereas when the
Fermi level is closer to the valence band, it denotes a p-type semiconductor. Figure 2-3
and Figure 2-4 illustrate the concept of Fermi levels of metals, semi-conductors and
insulators along with the associated band gap.
- - - - -
Figure 2-3. Band gaps for (a) metals, (b) semiconductors and (c) insulators with Fermi
level (SF) indicated
The effect of metals added to semiconductors such as TiO2 is as follows. The
dopant metals change the photocatalytic properties of the semiconductor by changing the
distribution of electrons. Hence, when a metal and a semiconductor are electrically
connected, electron migration takes place from the semiconductor to the metal until the
two Fermi levels are aligned as shown in Figure 2-4 (Linsebigler et al., 1995; Carp et al.,
2004; Hoffman et al., 1995) (figure adapted from Linsebigler et al., 1995). This
phenomenon forms what is known as the Schottky barrier. It must be noted that the
particular case mentioned above takes place when the work function (
higher than that of the semiconductor. The work function is the least amount of energy
required to remove an electron from the surface of a conducting material to a point just
outside the metal.
E 7-b ---CB
Metal Semiconductor (n-type)
Figure 2-4. Alignment of Fermi levels and formation of Schottky barrier
The Schottky barrier produced at the interface of the metal and the semiconductor
acts as an electron trap, hence hindering electron-hole recombination which results in
increased photocatalysis (Linsebigler et al., 1995; Carp et al., 2004). This scenario is
what will be referred to as a favorable difference in Fermi energies during the progress of
2.4 Previous Synergy in AC-TiO2 Systems
Activated carbon is a well known adsorbent especially for systems dealing with
organic. It is also well known that TiO2 is capable of oxidizing organic to water and
carbon dioxide. Hence, it can be hypothesized that a combination of these two materials
would result in a combination of adsorption and degradation which in theory would result
in a synergy. This would effectively mean that the pollutant is adsorbed onto the
activated carbon and is then degraded due to the TiO2 photocatalyst in the presence of UV
radiation resulting in the theoretically infinite life of AC.
However, a number of variables are involved in the process which may limit the
effect of photocatalysis on the adsorbed pollutant. Properties of the AC such as pore size
distribution, surface area, surface functional groups and the acidic-basic characteristics of
the AC may make this theory extremely system specific when applied. Considerable
research has been conducted in the field of TiO2-AC composites (Matos et al., 1998;
Matos et al., 2001; Khan., 2003; Khan et al., 2006).
Matos et al. (1998) and Khan et al. (2006) both found that titania seems to show a
synergistic effect with activated carbon. To elaborate, the presence of activated carbon
alone in a system for the removal of a pollutant has been shown to be less effective than a
combination of the activated carbon and titania in the presence of UV. Similarly, the
presence of the titania and AC (in the presence of UV) in systems has shown better
performance than a system consisting solely of titania.
A system tested by Matos et al. (1998) consisted of 50 mg of TiO2 combined with
10 mg of AC in the form of a slurry with the target pollutant being phenol. This system
was compared to other systems consisting solely of either titania or activated carbon. The
experiments were conducted such that the systems were not exposed to UV radiation
until equilibrium had been established with respect to adsorption. Once equilibrium was
established, the systems were exposed to UV radiation (340 nm). The following figure
(Figure 2-5, adapted from Matos et al., 1998) illustrates the results of the kinetics of
phenol disappearance under the various systems tested.
%0 0 2
a 0 TiO2
| o o Photolyti
S TO D 1 D T02 -A
-60 0 60 126 180 240 300 360 420 4s0
Figure 2-5. Kinetics of phenol disappearance in the presence and absence (photolysis) of
various illuminated solids
The results of Matos et al. (1998) show that the apparent rate constant for the
system consisting solely of titania is 5.6 x 10-3/min whereas the apparent rate constant for
the TiO2 AC system is 1.39 x 10-2/min. Hence it has been concluded that the
photocatalytic efficiency seen in the case of the titania-AC slurry is 2.5 times that seen by
the system with titania alone. The authors ascribe this result to the adsorption of phenol to
the AC which is followed by a mass transfer to the photoactive titania. In effect, the
essence of the synergy in the case of the titania-AC system is the common interface
between the AC and TiO2 which has been postulated to consist of almost half of the total
surface of exposed titania.
However, the rate constants would be calculated from a point at which adsorption
of phenol has taken place and the systems have reached equilibrium. In the case that the
rate constants of the two systems are calculated solely from the time of irradiation, it may
be noticed that the AC-titania system actually has a slower rate than the system
containing only titania. Hence, although a beneficial effect is seen to be taking place with
the addition of AC, it does not agree with the definition of synergy as proposed by this
author. From the arguments made above, it is clear that a synergy was not prevalent in the
titania-AC system tested by Matos et al. (1998).
A subsequent paper by the same authors (Matos et al., 2001) investigated the role
of the type of AC on the photocatalytic degradation of organic pollutants. Similar tests to
the work described previously have been performed although emphasis has been given to
the surface properties of the AC and their influence on photocatalysis. It was concluded
by Matos et al. (2001) that H-type ACs (High pH/ High temperature activated) when
added to a titania slurry show a beneficial trend in photocatalysis while the opposite is
true for L-type ACs (Low pH/Low temperature activated). Although the terms H-type
and L-type carbons can only be used loosely, the veracity and extent to which the
conclusion by Matos et al. (2001) may apply has been investigated by this author.
Another case of the synergistic effect of titania with AC has been investigated by
Khan (2003). Their study was performed by coating TiO2 (Degussa P25) on various
commercially available ACs. The coated ACs were then used in the adsorption/
degradation of dyes (i.e., methylene blue, reactive red). The results seen by Khan (2003)
showed that the activated carbons F400 (Calgon), HD4000 (NORIT) and a wood-based
activated carbon synthesized in their laboratory showed an increased removal of the dye
in the presence of UV as opposed to in the absence of it. On the other hand, Bionuchar
(Westvaco) showed the opposite trend. It was concluded that the presence of ionic metals
such as iron (Fe+) in the coal based carbons (F400, HD4000) resulted in the synergy.
This was based on the hypothesis that the ionic metals present in the coal based carbons
act as sinks for electrons, thereby hindering electron hole recombination and hence
enhancing photocatalysis. Iron in the form of Fe+ is commonly used in the doping of
TiO2 to enhance photocatalysis. Iron is also a good conductor and would hence contribute
to the electrical conductivity of the material in which it is present. Hence, from the theory
of Fermi levels and Schottky barrier explained earlier, it is believed that this hypothesis
by Khan (2003) could hold true in the case of dyes. It is a topic of interest to see whether
the synergy noticed in the case of dyes also holds true in the case of organic compounds
such as phenol.
The work by Khan (2003) is the basis for the hypothesis formed by this author
which attributes the conductivity of activated carbons (due primarily to metals) as an
influence over the photocatalysis observed with the TiO2-AC composite.
MATERIALS AND METHODS
3.1. Target Pollutant
The target pollutant, phenol (C6H60), was selected because it is representative of
pollutants experienced in water treatment. It was obtained in the liquid form (90 % w/w)
from Fisher Scientific. The stock solution was diluted to obtain the required
concentration (average concentration 55 mg/L 2 mg/L). The molecular weight of
phenol is 94.11. The pH of phenol solution is 5-6.
3.2 Carbon Materials
3.2.1 TiO2-AC Composites and Coating Procedures
Each AC differed in terms of physical and chemical properties due to differences in
precursor and activation technique. A summary of the properties of the AC's is provided
in Table 3-1.
Table 3-1. Summary of AC properties
Activated Precursor Activation BET surface area Micro Meso
carbon type (m2/gm) pores pores
___ __(%) (%)
Bionuchar Wood Chemical 1509 (+191, 270) 32 58
F400 Bituminous Thermal 1000 (+134, -200) 60 32
HD4000 Lignite coal Thermal 539 (+80, 114) 25 66
The Bionuchar which was chemically activated is an AC supersaturated with
carboxylic and phosphoric groups and hence has a very acidic surface. The F400 is a
more basic AC whereas the HD4000, although it has been thermally activated has an
acidic surface due to an acid wash after its activation procedure.
The activated carbons were standardized to a 20 x 35 mesh using American
Standard Sieves (Fisher Scientific) and a Ro-Tap sieve shaker (Fisher Scientific). They
were then subjected to a number of coating procedures which were assessed based on
simplicity and efficiency.
Particle synthesis. In order to synthesize the TiO2-carbon particles, four coating
techniques were considered. Some of these coating techniques were only applicable to
activated carbons, versus the low surface area raw materials, due to its porous nature. The
coating techniques that were considered are as follows:
* Boil deposition
* Pore Volume Impregnation (PVI)
* Chemical Attachment (Modified Sol-Gel technique)
* Mechanical Attachment (Theta Composer)
It was found during the course of experiments that a 3-5% (m/m) coating provided
efficient and optimum activity for photocatalysis. The techniques were assessed based on
simplicity and efficiency as the ultimate engineering goal is the development of a
portable RAC filter.
Boil deposition. The boil deposition procedure is the most convenient method of
coating a substrate with TiO2. It is based on the mechanism of agitation of a mixture
during boiling. The theory implies that the TiO2 would be deposited on the carbon surface
due to the agitation of water, TiO2 and carbon while boiling, resulting in the impaction
and hence deposition of TiO2 on the carbon/AC surface in the presence of heat. Although
this procedure is not very consistent, it results in convenient coating of the particles and is
efficient in terms of performance of carbons or activated carbons.
The procedure was adapted from the technique used by Khan (2003). In this
procedure, the carbon/activated carbon sample to be coated was first washed with
deionized water. The sample was added to a 250 ml Erlenmeyer flask and a mass of TiO2
was also added to the flask corresponding to the required extent of coating. Deionized
water (200-220 ml) was added to the mixture. The flask was then placed on a hot plate
and heated to 1200 C and allowed to boil. Once all the water had evaporated, the sample
was once again rinsed with DI water and was left to dry in a furnace maintained at 1200C
to remove any residual moisture.
Pore Volume Impregnation (PVI). Impregnation is a commonly used method in
supported catalyst preparation. The impregnation method involves three steps: 1)
Contacting the support with the impregnating solution for a certain period of time, 2)
Drying the support to remove the impregnated/absorbed liquid and 3) Activating the
catalyst by calcinations or reduction (Tao, 2003). The principle of impregnation involves
using a volume of the precursor solution such that it is equivalent to the pore volume of
the support. Hence, it is evident that this technique can only be successful for activated
carbons or substrates with a well established pore structure.
A weighed quantity of AC was taken in a 40 ml vial. The total pore volume of that
quantity of AC was determined. The precursor (titanium isopropoxide (TTIP), ACROS)
and isopropanol (Fisher Scientific) were then mixed with the AC sample such that the
volume of the solution was equal to the total pore volume of the AC sample. Hence, for a
3% mass loading of titanium dioxide on the AC, the solution of TTIP and isopropanol
would be prepared such that the TTIP was 3% of the solution (v/v). The solution of TTIP
and isopropanol was introduced into the 40 ml vial containing the AC. The contents of
the vial were agitated for a few minutes. The vial was then left open under the fume hood
for 24 hours to allow the TTIP-isopropanol solution to be impregnated in the AC. After
24 hours of drying, the activated carbon was heat treated in a furnace at 3000C to convert
the titanium dioxide to the anatase phase. The performance of this coating procedure was
questioned due to the fact that the titanium dioxide is impregnated predominantly in the
pores of the AC. This was believed to hinder the incidence of UV light on the titanium
dioxide given the random structure of the pores as well as the opacity of activated carbon.
Also, as stated by Matos et al. (1998), the presence of TiO2 in the pores as opposed to at
the surface of AC would mean the absence of a concentration gradient forming, resulting
in hindered photocatalytic activity. The results from the tests performed with AC's coated
with TiO2 by the PVI technique seemed to agree with this argument.
Chemical attachment (modified sol-gel technique). The chemical attachment
coating procedure makes use of a sol gel to coat the activated carbon surface. This
technique was adapted from the technique employed to coat barium ferrite particles.
(Drwiega, 2004). Four grams of AC (HD4000, 20 x 35) and some volume of
polyethyleneimine (PEI (Alfa Aesar)) and 80 mL of nanopure were combined and mixed
for 10 minutes until the entire PEI was dissolved. 250 mL of nanopure was added to a 3-
mouth flask. The carbon particles were then introduced into the 3-mouth flask. The
glassware consisting of the reflux tube, 3-mouth flask, gas adapter, funnel and
thermometer were assembled. The nitrogen and water flow were then turned on for a few
minutes. Five milliliters of isopropanol was added to the closed funnel after which 100 [tl
of titanium isopropoxide (TTIP) was added to the closed funnel. The particles were
slowly stirred with a magnetic stirrer. The solution in the funnel was slowly introduced
into the water. The mixing was continued for 10 minutes and the solution with the
particles was heated to 950C. This was continued for 20-24 hours. The particles were
then removed and rinsed thoroughly with deionized water. The AC particles were then
heated in an oven at 1100 C for 1 hour to dry the particles. The particles then underwent
heat treatment at 350-4000C in presence of nitrogen flow for 3 hours to convert the
titanium dioxide to the anatase phase. This coating technique performed comparably well
to the boil deposition. However, the main disadvantage of this technique was that it
required a complex setup of equipment. Since one of the ultimate goals of the project is
the development of a household filter, it was not selected as the coating technique for the
Mechanical attachment. Mechanical attachment consists of forcing the carbon
and TiO2 through a 1 mm gap at the same instant, hence implanting the TiO2 onto the
carbon surface due to the high forces on the particles. The technique, also known as
mechanofusion consists of equipment known as a Theta-Composer (Tokuju Inc, Japan).
The Theta-Composer basically consists of a rotor and a vessel set in a way such that they
are concentric and rotate in opposite directions (rotor speed: 2500 rpm; vessel speed: 77
rpm). This is shown in Figure 3-1 (Adapted from Khan, 2003). The AC particles (Sieved
to 20 x 35 mesh) and 3% TiO2 (m/m) was added to the vessel and the theta-composer was
allowed to run for 15 minutes.
The disadvantage of this technique was the separation of TiO2 from the carbon
particles when introduced in a liquid medium as well as the powdering of the AC due to
the natural mechanical motion and abrasion of the theta-composer.
After selection of the boil deposition procedure as the preferred coating technique,
the carbons were coated on two bases:
* 3% TiO2 on the basis of mass of carbon.
* 3% TiO2 on the basis of moles of elemental carbon in the AC.
(a) (b) (c)
Nanosized TiO2 & micrometer-sized
magnetic substrate particles
Figure 3-1. Coating activated carbon with TiO2 via mechanical attachment (Theta
Composer, Tokuju Corp., Japan)
3.2.2 Mass Based Coating (m/m)
The activated carbons were coated with TiO2 in the ratio of 3:100 (weight of
TiO2/weight of AC). The coating percentage was decided upon using evidence from
previous experiments which showed an optimal coating percentage range of 3%-5%
(mass/mass). For coating the ACs on a mass basis, a certain mass of AC (1-2 gms) was
mixed with TiO2 (3% of that mass). The boil deposition procedure was then used to coat
the TiO2 onto the ACs. After coating, the carbons were loaded to their reactors on the
basis of mass and in other experiments on the basis of volume. They may be denoted in
the future as 'mass-mass' and 'mass-volume' respectively.
3.2.3 Molar Based Coating (mol/mol)
For coating the carbons with TiO2 on a molar basis, the amount of elemental
carbon (C) in each sample was found by performing an ash analysis in which the AC's
were heated to approximately 5750 C. The difference between the initial masses of the
ACs and their final masses (of the ash) provided the amount of elemental carbon in each
sample. The following table (Table 3-2) provides a summary of the elemental carbon
contents of the ACs followed by a sample calculation of the TiO2-elemental Carbon
ratio's used in the composites.
Table 3-2. Carbon contents of virgin and titania coated ((m/m) and (mol/mol)) activated
carbons and their surface areas
Activated Carbon content Surface area (m/m) Surface area (mol/mol)
carbon(%) (m2/gm) (m2/gmm3)
Bionuchar 92 98 1483 1380
F400 86 94 1034 1009
HD4000 82 88 602 624
A sample calculation is provided.
Activated carbon A contains X mg Carbon/ gm AC
For 3% (mol/mol) loading:
MolesTiO2 / MoleScarbon = 3%
> (m/M)Tio2 / (m/M)c = 3/100
S (m/79.86)Tio2 / (X/12)c = 0.03
S mTi2 / Xcarbon = 0.19965
mTiO2 = Mass of TiO2 to be added
mc = Mass of Carbon (Elemental Carbon)
MTiO2 = Molecular weight of TiO2 = 79.86
Mc = Molecular weight of Carbon = 12
The TiO2 and AC mixture then underwent boil deposition, standardizing the
Carbon-TiO2 ratio in all the carbons. The carbons after coating were loaded to their
reactors on the basis of volume. They may be denoted in the future as 'mole-volume'.
It must be noted that the tests performed with the coated ACs were performed using
TiO2-AC composites prepared in different batches. Hence the replicate sets of
experiments are indicative of the performances of different batches of composite as
opposed to composites prepared in the same batch. In this regard, the consistency of the
batches with respect to performance of the composites will also be observed. The TiO2-
low surface area carbon composites (explained in section 3.2.4) were also prepared and
3.2.4 Low Surface Area Carbon-TiO2 Composite
The low surface area carbons used herein consisted of Pitch Coke (Asbury Graphite
Mills, Asbury, NJ), Graphite (Asbury Graphite Mills, Asbury, NJ), Anthracite Coal
(Reading Anthracite Coal Company, Mahantongo St., Pottsville, PA), and Bituminous
Coal (International Industries, Logan, WV). The titanium dioxide was commercially
available Degussa P25 (70% Anatase, 30% Rutile). The virgin carbons used in these
studies had extremely low surface areas as shown in Table 3-3. In fact, the surface areas
of some of the samples were below the minimum value that was detectable. The values
are still provided although they may contain some error.
Table 3-3. Surface areas of virgin and coated low surface area carbons
Carbon Surface area (m /gm)
Virgin 3% TiO2 3% TiO2
carbons (mass/mass coating) (mole/mole of C)
Pitch coke 0.62 4.21 8.15
Anthracite 0.92 4.1 7.51
Graphite 1.35 3.63 7.99
Bituminous 0.5 1.00 3.34
Although in both cases (i.e., (m/m) and (mol/mol)) there was a considerable
percentage increase in the surface areas of the coated carbons as compared to the virgin
carbons, it was still extremely low for any significant sorption phenomena to occur.
We also analyzed the amount of elemental carbon in each sample (as for the AC's) in
order to coat the carbons on the basis of mole/mole which is provided in Table 3-4. The
coating technique was exactly the same as that used to coat the ACs.
Table 3-4. Carbon contents of low surface area carbons
Low Surface Area Carbon Carbon Content (%)
Pitch Coke 93
3.3 Experimental Setup
3.3.1 Batch Reactor Configuration
Batch testing was performed initially to investigate how the carbons with and
without TiO2 performed for phenol removal. The setup for the batch tests consisted of a
rotator (Analytical Research Systems, Gainesville, FL) onto which were mounted 4
reactors. The reactors consisted of 100 mL glass syringes (SGE; Obtained from Fischer
Scientific). The syringes were fitted with gas-tight plungers, the ends of which were
made of Teflon. The syringes were also mounted horizontally on the rotator such that
they were parallel to the axis of the rotator. The entire system rotated at a rate of 3 rpm.
Three UV lamps (365 nm wavelength) were placed such that they were 1200 apart and at
a distance of 0.5 cm from the outer wall of the syringes at the nearest point. The entire
assembly was enclosed in a wooden box, the walls of which were covered with reflective
aluminum foil so that any stray radiation could be reflected back towards the reactors.
3.3.2 Analytical Equipment
The concentration of phenol solution was determined on a UV-VIS
spectrophotometer (HACH DR/4000U). The BET surface area analyses for the activated
and low surface area carbons as well as their composites were performed using a
Quantachrome NOVA 2200e Gas Sorption Analyzer (Boynton Beach, Florida). Samples
were outgassed for 24 hours prior to testing. Before outgassing, the samples were dried at
110 C. The SEM images and the EDS scans were performed using a scanning electron
microscope (SEM JEOL JSM 6400). The images and scans were performed at the major
analytical instrumentation center (MAIC, Department of Materials Science, University of
Florida, Gainesville). Multiple images and scans of the same sample as well as different
particles of the same sample were taken. Density measurements of the Carbon-TiO2 and
the AC-TiO2 composites were performed on an Ultrapychnometer (Quantachrome).
Replicate sets of measurements were taken. Multiple samples from within batches as well
as across different batched were taken. The value provided is the average of all the values
with the maximum and minimum values also provided.
3.4 Experimental Procedure
3.4.1 Concentration Measurements
The concentration of the phenol solution was measured at the peak absorbance
wavelength X = 270nm for phenol which was determined by performing an absorbance
scan of the phenol solution. The concentration vs. absorbance profile for a phenol
solution measured at 270 nm on the UV-VIS spectrophotometer is shown in Figure 3-1
and obeys Beer's law. The R2 for the profile was found to be 0.97.
Absorbance Profile for
Linear (Concentration vs
Absorbance Profile for
10 y= 62.445x
R2 = 0.97
0 0.2 0.4 0.6 0.8 1
Figure 3-2. Concentration vs. Absorbance correlation for phenol on UV
spectrophotometer measured at 270nm wavelength
3.4.2 Batch Adsorption/Oxidation Tests with Activated Carbon
The synergistic phenomena observed in certain TiO2-AC composites were
attempted to be determined for phenol by performing batch tests with the various
composites. The three activated carbons selected Bionuchar, F400 and HD4000 were
tested initially in a batch system. The tests were performed using granular activated
carbon (GAC) for the convenience of separation and to simulate a real world system.
Experiments were performed with ACs coated with TiO2 on the basis of mass as well as
on a molar basis, which has been explained earlier. The ACs coated by mass were loaded
into the reactors on two bases: 1) mass basis and 2) volume basis. The loading of reactors
on the basis of mass was done such that 50 mg of AC was loaded for 80 mL of phenol
solution whose concentration was 55 mg/L. The volume loading of the composites in the
reactors was done such that 0.025 mL of AC was used for 80 mL of phenol solution
whose concentration was 55 mg/L. The ACs coated on a molar basis were loaded in the
reactors only on the basis of volume. The volume based loading of reactors with these
composites was done such that 0.025 mL of AC was used for 80 mL of phenol solution
with a concentration of 55 mg/L. The procedure of coating the ACs by molar ratio's and
loading the composites on a volume basis was done to normalize the TiO2-carbon ratio
for each sample since the AC's were of varying densities, resulting in different quantities
of AC (and hence TiO2) in each reactor.
The parameters such as initial concentration of phenol solution, volume of phenol
solution used in each reactor and the time for each run remained constant for the
adsorption-oxidation studies for both types of AC loading. Both experiments were carried
out for a period of 24 hours. For adsorption studies, experiments were performed in the
absence of UV radiation. At the end of 24 hours, the GAC particles were separated from
the phenol solution by filtration with 0.45 [tm filters (Fisher Scientific) using vacuum
filtration. The final concentration of the phenol solution was measured using a UV-VIS
spectrophotometer. The difference between the initial concentration of the stock solution
and the final concentrations were attributed to adsorption. For the degradation studies, the
experiments took place in the presence of UV light. Again, the experiments were carried
out for 24 hrs at the end of which the GAC particles were separated from the phenol
solution by vacuum-filtration. The final concentration of the separated phenol solution
was measured using the spectrophotometer. The difference between the initial
concentration of the stock solution and the final concentration of the phenol sample in
this case was attributed to a combination of adsorption and photocatalytic degradation.
In addition to adsorption and oxidation tests, tests were also performed with virgin
ACs in phenol solution in the presence of UV.
Replicate sets were run for each sample and the error between the samples has been
denoted as the maximum and minimum values that were recorded. The actual value
plotted is the average of all the values obtained for that result.
3.4.3 Batch Adsorption/Oxidation Tests with Low Surface Area Carbons
As explained earlier, the configuration for the batch tests involving the low surface
area carbons was identical to that employed for the AC-TiO2 tests. The TiO2-carbon
composites (35 x 200 sieve) were coated on the basis of mass as well as on the basis of
mole. For the carbons coated on a mass basis, the composites were loaded into the
reactors on two bases: by mass and by volume. For the mass loading, 80 mg of the
carbons were loaded into the reactors which contained 80 mL of 55 mg/L phenol
solution. For the volume loading, 0.05 mL of the carbons were loaded into the reactors
containing 80 mL of 55 mg/L solution. The volume loading was done in order to
normalize to some extent the amount of carbon (and hence titania) in each reactor.
For the carbons coated on a molar basis, the composites were loaded into the
reactors only on a volume basis. Here again, 0.05 mL of sample were loaded into reactors
containing 80 mL of 55 mg/L phenol solution. This would in theory completely
normalize the ratio of titania to carbon (elemental) in each reactor.
The testing procedure for both cases was identical to the one employed for the AC
TiO2 composites as far as runtime and sample analysis procedures are concerned.
Replicate sets were run for each sample and the error between the samples has been
denoted as the maximum and minimum values that were recorded. The actual value
plotted is the average of all the values obtained for that result.
RESULTS AND DISCUSSION
4.1 Evaluation of Carbon Coating Strategies
Titanium dioxide was coated on the activated carbons using 4 different coating
techniques: boil deposition, pore volume impregnation (PVI), chemical attachment, and
Boil deposition was by far the easiest and most convenient coating technique as it
required using a commercially available TiO2 (Degussa P25), water and heat. It was
nevertheless a very inconsistent and unpredictable of the coating methods as the TiO2
deposited not only on the AC but also on the beaker in which it was being carried out.
Therefore, although a known mass of titania was used in the coating process, the actual
amount of TiO2 deposited on the carbon was likely less than the targeted value. The
average phenol removal observed for HD4000 coated with titania was 78%.
Pore volume impregnation resulted in the immobilization of the TiO2 within the
pores of the AC. The titania depositing in the pores of the ACs was considered a
disadvantage for this coating technique, particularly due to the opacity of activated
carbon and the random orientation of its pores. It was hypothesized that due to titania
deposition within the pores it would result in a low probability of the UV light reaching
the titanium dioxide, resulting in a system with reduced photocatalytic activity. The
results obtained seemed to agree with this hypothesis. The photocatalytic tests which
were performed with HD4000 showed a removal of only 42%.
Chemical attachment of the TiO2 to the carbon was also attempted to manufacture a
more rigid composite. Although this coating technique would provide a more consistent
coating on the surface of the ACs, the performance of the ACs coated by this technique
did not show any marked improvement over the boil deposition technique. The sol-gel
coated HD4000 resulted in 81% removal via adsorption and photocatalysis, which was
comparable to the removal seen with the boil deposition method.
Table 4-1. Summary of coating techniques
Coating Performance Advantages Disadvantages
Simple procedure 1) Inconsistent coating
Boil deposition 78% 2) Poor control of TiO2
Good performance loading
1) Requires additional
Pore volume hydrolysis and calcination
impregnation 62% Good coating control steps
( ) 2) Titania deposition in
macropores (Tao, 2005)
3) Poor performance
d c g c l 1) Requires additional
Chemical 81% hydrolysis step
attachmentGood performance 2) Experimental setup is
1) Attrition of AC
attachment 2) TiO2 separation from
Mechanical attachment was achieved through impaction using the Theta Composer
instrument (Tokuju Inc., Japan) which resulted in the attrition of the granular activated
carbon (GAC) particles, which originated as approximately 1 mm in size, to particles less
than 45 [tm (Powdered Activated Carbon or PAC). The technique also resulted in
separation of the TiO2 particles from the substrate upon introduction in water. The
advantages and the disadvantages of each technique are summarized in Table 4-1. It must
be noted that multiple batches of the TiO2-AC composites were tested. Based on these
results, the boil deposition technique was selected for coating the carbons with titania.
4.2 Activated Carbon Studies
Khan et al. (2003) showed that F400 and HD4000 showed a synergy when coated
with TiO2, whereas Bionuchar did not. The authors concluded that metals, naturally
occurring in activated carbon may have been responsible for the observed synergy (i.e.,
removal of target pollutant increased when titania was coated on AC as compared to
activated carbon alone). Nevertheless, they did not investigate this phenomenon in
further detail. It was hence a viable exercise to investigate whether metals, in particular
electrically conductive metals such as iron and aluminum, could be a plausible
mechanism for creating the described synergy.
4.2.1 Ash Analyses of Activated Carbons
To determine the metallic content of Bionuchar, F400 and HD4000, the ACs used
by Khan (2003), the uncoated (i.e., virgin) carbons were combusted in air at 5750C until
only an ash residue remained. The ash residue of each AC was then analyzed by a
scanning electron microscope (SEM) to perform an energy-dispersive spectroscopy
(EDS) study. EDS was also used to verify the presence of titania on the surface of the
coated ACs. Note that the ACs of which the ash residues were examined were not coated
with titania prior to ashing. Even though the EDS data cannot be used as a complete
quantitative comparison, it is still useful as a semi-quantitative as well as a qualitative
tool for characterizing elemental composition. The percentage of ash (by weight) in each
sample is provided in Table 4-2. Also provided in the table is the amount of titania (as a
percentage) by weight actually coated on the AC samples. This was quantified simply as
the difference between the weight of the ash residue of a virgin AC and that of a coated
AC sample. The theoretical amount of titania which was to be coated on the surface of
each of the ACs was 3% (i.e., weight of titania added equals 3% of the weight of the AC
sample). Table 4-2 shows the inconsistency of the boil deposition procedure in coating
the ACs. On an average it is seen from the data provided in Table 4-2 that Bionuchar has
the lowest TiO2 deposited on its surface. This is attributed to morphological differences
between the wood based Bionuchar and the coal based F400 and HD4000.
Table 4-2. Ash contents of ACs and mass % of titania deposited (m/m) on AC surface
Activated carbon Precursor Carbon Ash% (virgin ACs) TiO2 %
Bionuchar Wood 92-98% 2-8% 0.5-1.3%
F400 Bituminous 86-94% 6-14% 1.6-2.2%
HD4000 Lignite 82-88% 12-18% 1.4-2.3%
The ash analyses of the ACs provided an estimate as to the magnitude of the
composition of elements seen in the EDS scans. Figures 4-1(A-F) show the EDS scans of
the ash residues of the virgin ACs and the scans of the coated ACs.
The EDS scan of the Bionuchar ash residue (Figure 4-1A) shows that it consists of
primarily phosphorous, silicon and sodium. A very faint trace of aluminum as well as a
peak corresponding to iron is also observed. The phosphorous peak is likely remnant
from the chemical activation technique using phosphoric acid.
Figure 4-1B represents an EDS image of a Bionuchar sample coated with titania. It
can be seen from the scan that smaller peaks corresponding to phosphorous, silicon and
sodium are again observed. A titanium peak is also seen in addition to the peaks seen in
the ash residue of the same AC. It is important to note than the total ash content of the
Bionuchar samples is a maximum of only 10% of the weight of the sample which is
inclusive of the titania coated on the surface. Hence, coupled with the results from the
EDS scan, it may be concluded that a very small amount of each of the elements is
present in Bionuchar. Hence the low presence of conductive metals such as aluminum in
Bionuchar (also observed by Khan (2003)) may result in very few conductive sites on the
surface of Bionuchar.
The residual ash sample of F400 (Figure 4-1C) consisted of very large peaks of
aluminum and silicon. Smaller peaks pertaining to iron, sodium, magnesium and calcium
were also noticed. In comparison to the EDS scan performed on the ash of the Bionuchar
sample which showed an aluminum peak of about 200 counts (in replicate sets of scans),
the F400 ash residue showed aluminum peaks having magnitudes between 3800 and 4000
counts in the replicate sets of scans preformed. The presence of iron should also be noted.
The activated carbon sample of TiO2-F400 (Figure 4-1D) shows the presence of
aluminum, silicon, iron and titanium. What is interesting to note is the extremely high
titania peak of about 5000 counts. It may be recalled that the corresponding peak in the
Bionuchar sample did not have a magnitude of even 2000 counts. This suggests that the
morphological difference between the wood based Bionuchar and the coal based F400
resulted in a difference in TiO2 coating. For instance, variations in roughness may cause
the titania to adhere to AC surfaces differently. This would then affect the adsorption of
phenol as well as photocatalytic mechanisms. Another observation to be made is the
higher ash content of F400 (6-14%) as compared to Bionuchar. This indicates the
presence of larger quantities of the elements seen in the EDS scans of F400 relative to
2000 Na Si
C AAl S K Fe
0 2 4 8 8
2000 p Ti
Na Si Ti
0 2 4 0 8
Na S Ca
C e P K Ca Ti Ti Fe
0 1 1 I . I . .-. ,. . I . ,
0 2 4 8 8
Figure 4-1. EDS of A) ash residue of uncoated Bionuchar, B) Bionuchar coated with
TiO2, C) ash residue of uncoated F400, D) F400 coated with TiO2, E) ash
residue of uncoated HD4000, F) HD4000 coated with TiO2
Si Ti T
l S Ca Fe
0 2 4 6 8 10
C Na Ca Ti
0 2 4 6 8
C Si Ti Ti
0 Al S
0 2 4 6 8 10
Figure 4-1. Continued
The differences in the ash compositions as well as the compositions of the ACs
(Bionuchar and F400) are noticed, with the prominent peak of aluminum and the
presence of iron being the most pertinent to this study. The relatively high conductivity of
aluminum and iron may play a role in the distinction between the performances of
Bionuchar and F400 that were observed in the work of Khan (2003). This may be a
possible cause for F400 showing a synergistic phenomena with respect to dyes. The
results of the EDS scan of F400 also agree with Khan (2003) about the presence of iron
F400 would have different metals contents because it is a coal based activated
carbon that originated as bituminous coal. Therefore, metals that are present in the
subterranean environment would be inherent in the coal. However, Bionuchar is a wood
based activated carbon that not only would have different metals than coal based
activated carbons, but since it is chemically activated, during the chemical impregnation
step at low pH, some of the metals would become soluble and leach out of the wood.
In the case of HD4000 the EDS image (Figure 4-1E) of the ash residue prominently
consisted of silicon and aluminum. In addition, magnesium and sodium were also fairly
prominent. However, iron that was hypothesized to be important for the observed synergy
in the work of Khan (2003), was absent in the EDS scan.
The presence of aluminum and silicon were also detected in the TiO2 coated
HD4000 sample (Figure 4-1F). The EDS image also showed a very high titanium peak
(about 5000 counts). In replicate sets which provided reproducible results, counts as high
as 8000 were found for titanium. The EDS scans for the HD4000 samples also showed a
relatively low carbon peak indicating that a titanium dioxide agglomerate may have been
focused on. The fact that this was the case in replicate sets between different TiO2 coated
HD4000 samples may also indicate to some extent a uniform coating on the surface of
this AC. Noteworthy is the fact that HD4000, like F400 is also a coal based AC. This
strengthens the argument that morphology may play a role in the deposition of titania on
the AC. This can also be attributed to the enhanced performances seen in the coal based
carbons with respect to adsorption and photo-oxidation of dyes (Khan, 2003). The fact
that HD4000 also contains highly conductive metals such as aluminum strengthens the
argument that conductivity may play a role in enhanced photocatalysis. Also, the fact that
the ash content of HD4000, a lignite based AC, is relatively high (12-18%) increases the
probability of the coated titanium dioxide being in the vicinity of a site on the AC which
possesses a higher conductivity (due to the presence of conductive metal ions).
4.2.2 SEM Images of Titania Coated ACs
SEM images were used to further characterize the coating of titania on the ACs.
The images showed a variation in the deposition of titania on the activated carbons as
well as the degree to which each AC had been coated. Figure 4-2 shows an image of
titania coated Bionuchar.
'- U "
Figure 4-2. Titania coated Bionuchar with the TiO2 blocked pores circled
It can be noticed from the image that the surface of Bionuchar seems very smooth.
The smooth nature of the AC surface may cause a lack of adherence of the titania to the
AC surface which may result in decreased photocatalytic activity. Another notable aspect
of this image is the fact that a number of pores (shown circled) appear to be partially or
completely blocked by the titanium dioxide. This may in fact be due to the poor
adherence of the TiO2 surface resulting in the titania being deposited inside the pores of
the AC. Deposition in the pores would of course reduce the diffusion of phenol to
adsorption sites, thereby decreasing the adsorption of phenol (as discussed in subsequent
sections). The wood based nature of this AC is evident in Figure 4-3.
Figure 4-3. Bionuchar showing cellulose structure of the wood based AC.
The Bionuchar AC particle seen in this image shows what seems to be a cellulose-
like structure due to the wood precursor. It may also be noticed that the particle appears
to have a smooth honeycomb structure which as mentioned earlier may not allow TiO2 to
adhere well to its surface. The author considers the morphology of the Bionuchar surface
to be one of the reasons why a relatively small titanium peak is detected in the EDS of
Bionuchar and its ash residue. Indeed, Table 4-2 showed that the percentage of titania on
the Bionuchar samples was less than the other two samples.
In contrast to the morphology of Bionuchar, the titania seems to be uniformly
coated on the surface of F400 (Figure 4-4).
Figure 4-4. Titania coated on the surface of F400
The uniformity in coating is thought to be an explanation for the well defined peaks
of titania seen in the EDS scans of F400 and could also be an explanation for the synergy
seen in the study by Khan (2003). A uniform coating would increase the probability of
titania contacting metals present on the AC surface which could hypothetically improve
An SEM image of HD4000 coated with titania is shown in Figure 4-5. It may be
noticed that the TiO2 agglomerates on the surface of HD4000; here the TiO2 agglomerate
is near the opening of a macropore (see circled region). The size of the TiO2
agglomerates (about 15 [tm) is observed to be much larger than the individual TiO2
particles (i.e., 27 nm) (Li et al., 2004).
Figure 4-5. TiO2 agglomerate (shown circled) near a macropore on the surface of
However, it may also be noticed that similar to the case of F400, the titania is
coated more uniformly on the surface than in the case of Bionuchar; a phenomenon
believed to be dictated by the morphology of HD4000. Figure 4-6 shows the uniform
coating and the agglomeration of titania particles along with the rough surface of
HD4000. As in the case of F400, the uniformity in coating may be an explanation for the
large titania presence seen in the EDS scans for HD4000. This may also be a possible
explanation for the synergy noticed by Khan (2003) for the same reasons as mentioned
Figure 4-6. Surface of HD4000 showing uniform coating of TiO2 which tends to form
Hence, it may be summarized that the wood based carbon and the coal based
carbons differ morphologically. Titania coated on Bionuchar seems to concentrate around
the pores. It is also believed that due to the smoothness of the Bionuchar surface, there is
less adherence of the TiO2 particles to the Bionuchar surface. However in the case of both
F400 and HD4000, a more uniform coating of titania is observed with the titania forming
agglomerates on the surface of HD4000.
The ash analysis, EDS scans and the SEM images together provide us with some
possible explanations for the synergistic phenomenon of certain TiO2-AC particles
observed by Khan (2003). Both F400 and HD4000 contain relatively high ash contents
as compared to Bionuchar. Their EDS scans also show a very high titanium presence on
their surface in addition to the higher presence of electrically conductive metals (i.e.,
It may be concluded that there is a greater possibility of titania being in contact
with a conductive site in the case of F400 and HD4000 relative to Bionuchar.
4.2.3 Activated Carbon Batch Studies
The results of the batch tests conducted on the activated carbons which compared
the performance of the ACs (or the TiO2-AC composites) as a function of the removal of
phenol are presented. It must be noted that differences in phenol removal of less than 4-
5% were not considered significant.
Virgin activated carbons. In the work of Khan (2003), the focus was on the
removal of dyes. Herein, the focus was on a model aromatic adsorbate (phenol) and the
intent was to learn if the conclusions made by Khan (2003) and as previously discussed,
were applicable to the removal of phenol.
To determine the adsorption capacity of the ACs, batch tests were performed in the
dark for a period of 24 hours. The studies were first performed with virgin ACs. Three
activated carbons were used: Bionuchar, F400 and HD4000. The results of the tests
performed with the virgin ACs are presented in Figure 4-7.
It can be seen from the results that the virgin F400 showed the most adsorption
capacity with 87% removal whereas Bionuchar showed the least adsorption capacity with
only 52% removal. HD4000 performed relatively well with about 76% removal. The high
adsorption capacity of F400 may be explained by the fact that its pores are predominantly
(60%) microporous. The molecular size of phenol is about 6 A (0.6 nm). Therefore, it is
expected to adsorb mainly in pores having a microporous size range. The low adsorption
capacity observed for Bionuchar is attributed to the acidity of its surface. The surface of
Bionuchar is highly acidic. It is well documented in the literature that 'L-type carbons' or
'acidic' activated carbons show reduced adsorption of organic such as phenol (Matos et
al., 1998; Salame et al., 2003; Coughlin et al., 1968, Mattson et al., 1969).
40 m Initial Concentration
re Virgin Bionuchar
0 E0 Virgin F400
o Virgin HD4000
Figure 4-7. Batch adsorption studies of virgin ACs in the absence of UV
It is thought that carboxyl and other such electron withdrawing functional groups
result in a lowering of phenol adsorption capacity by removal of the 7t-electron from the
AC aromatic ring. This decrease in the density of the electron cloud in the carbon basal
planes causes a decrease in the strength of interactions between the benzene ring of
phenol and the basal planes of AC. Phenol also reacts with carboxylic groups on the
carbon surface forming ester bonds which contributes to this phenomenon (Salame et al.,
2003; Nevskaia et al., 1998). As a result, there is decreased adsorption of phenol on the
basal planes of carbon. The case of HD4000 however is peculiar at first glance. HD4000
is acidic due to its activation procedure, but this acidity is a function of residual acid used
for acid-washing versus the presence of oxygenated functional groups located on the edge
sites. Therefore, although it displays characteristics of an L-type carbon, the manner in
which this is manifested is different than "traditional" L-type carbons. At the nano-scale,
HD4000 behaves more like an H-type carbon, which suggests that 7t-7t bonding is the
dominant adsorption mechanism.
Adsorption capacity for titania-coated activated carbons. Similar trends were
seen when batch adsorption studies were performed in the dark with 3% (m/m) (i.e., mass
of titania per mass of activated carbon) titania-coated AC samples (Figure 4-8). Note that
the targeted titania loading was 3%, but as was shown earlier the actual amount of titania
on the surface of each AC was less than 3 %.
50 50 mg 3% Bionuchar
2" 50 mg 3% F400
I[3 50 mg 3% HD4000
40 1.5 mg 1T02 Slurry
0 2.4 mg 1T02 Slurry
Figure 4-8. Batch adsorption studies of 3% (m/m) coated ACs in the dark compared to
A slight decrease in adsorption capacity as compared to the virgin ACs was
observed in all the ACs. This was attributed to the partial blockage of pores and hence
reduced surface area and active sites available for adsorption. In addition to the ACs,
Figure 8 presents the adsorption capacities of two titania slurries (no AC). These masses
of titania were chosen because they are very near the masses of titania that was present on
the ACs. Clearly the titania slurries show very low adsorption capacities (approximately
8% removal). In fact, it is noticed that even when a higher amount of titania is loaded to
the reactor, the decrease in phenol concentration does not change.
Adsorption and photocatalysis for titania-coated activated carbons. The next
step was to determine the performance of the coated ACs in the presence of UV light. As
in the earlier experiment, performances of two titania slurries were also included. It was
expected that UV would initiate photocatalysis and therefore considerably enhance the
removal of phenol through the combined functions of adsorption and photocatalysis.
The tests with the titania coated ACs were conducted in a sequence which would
eventually normalize the amount of carbon and titanium dioxide for each sample. The
initial batch tests consisted of ACs coated with titania on the basis of mass. The AC's
were loaded to their respective reactors such that each reactor contained 50 mg of the
titania-coated AC. These samples have been denoted in places as 'mass-mass' or 'm/m'.
The amounts of titanium dioxide for the slurry systems were 1.5 (3% of 50 mg) and 2.4
mg (3% of 80 mg as a comparison) respectively. It must be noted that the ACs do not
have the same density which resulted in each reactor having different volumes of ACs.
The results of the tests are shown in Figure 4-9.
The irradiated Bionuchar composite with only 39% removal performed poorly
compared to HD4000 (78% phenol removal) and F400 (88% phenol removal). The
performance of both titania-HD4000 and titania-F400 composites slightly improved in
the presence of UV; a fact that is attributed to photocatalysis by UV. However, this
refutes Matos et al. (1998, 2001) who claim that an L-type AC inhibits photocatalysis
whereas an H-type AC enhances photocatalysis. The surface of HD4000 is acidic and can
therefore be categorized as an L-type carbon, but as was previously mentioned, the
manner in which HD4000 exhibits a low pH is mechanistically different than
"traditional" L-type carbons. Therefore, further investigation into this phenomenon is
,50 -9 50 mg 3% Bionuchar
[ 50 mg 3% F400
E [0 50 mg 3% HD 4000
40 1.5 ng Ti02 Slurry
0 2.4 mng Ti02 Slurry
Figure 4-9. Batch adsorption-photocatalysis studies of 3% (m/m) coated ACs in the
presence of UV
Matos et al. (1998, 2001) clearly erred in their oversight that L-type carbons
exhibit less adsorption than H-type carbons. Therefore, decreased adsorption was
actually responsible for their results as opposed to decreased photocatalysis.
As was previously discussed in detail, a possible explanation as to why the coal-
based carbons behaved synergistically is because both contained conductive metals. The
enhanced performance of the ACs may point to a favorable difference in Fermi energies
between the titania coated on their surface and a conductive metal in their vicinity,
resulting in longer recombination times and improved photocatalysis. As explained
earlier, due to the relatively large ash content and metals composition, there is a greater
probability of the titania on the surface of AC being in the vicinity of a conductive metal.
Contrary to the performance of the coal based carbons, the performance of the
irradiated Bionuchar-titania composite, which is acidic, was considerably worse than the
performance of the virgin Bionuchar (13% difference). Whereas this phenomenon may be
attributed to pore blockages in the coated Bionuchar, the observation that the irradiated
Bionuchar composite performed worse than the Bionuchar composite (6% difference) in
the absence of UV makes for an interesting discussion (Figure 4-9).
A trend similar to the one seen in Figures 4-8 and 4-9 was noticed in the
experiments performed by Khan (2003). It was found that irradiated dyes (denoted as
Dye*) showed tendencies to adsorb differently than dyes which had not been irradiated
(denoted as Dye). According to Khan (2003) and Khan et al. (2006), the irradiation of the
dye caused the excitation of the delocalized electrons in the aromatic structure which
would affect the adsorption of the compound (Dye*). These delocalized electrons in the
aromatic structure would then interact with the carbon basal planes of Bionuchar which
due to acidic functional groups would contain a less dense electron cloud resulting in
decreased uptake (due to decreased 7t-7t interactions). The decreased uptake would also
result in decreased photocatalysis. Hence, ACs which preferentially adsorbed the
unadulterated dye compounds and had reduced uptake for 'Dye*' would have a reduced
overall performance. This was due to the fact that their affinity for the unadulterated dye
compound resulted in the overall hindrance to adsorption of the dye compound. Hence, in
the case of dyes, the two theories cumulatively may explain the reduced performance of
titania-Bionuchar composites in the presence of UV. Continuing in the same vein, the
synergy observed with F400 and HD4000 in the presence of UV was attributed to their
relative affinity for 'Dye*'. In the case of ACs that showed good removal, a situation of
augmented adsorption was theorized wherein activated carbons having an affinity to
adsorb both 'Dye' and 'Dye*' at comparable levels showed increased decolorization of
the dye solution (Dye + Dye*). Whether the same theory may apply to phenol (hence
forming phenol*) was thought to be an idea worth investigating and is discussed later.
Regarding the performance of the titania coated ACs and titania alone, each of the
ACs showed more removal of phenol than the titania slurries. In fact, the titania slurry
containing a lower concentration of titania slightly outperformed the slurry with a higher
concentration. Although this may be attributed to decreased penetration of UV with
increasing TiO2 concentration, the fact still remains that the removal was considerably
lower than the AC-TiO2 composites. Therefore, it must be noted that although titanium
dioxide may result in a decrease in adsorption capacity when coated on ACs, the ACs
coated with titania provide for better overall removal of phenol in the presence of UV as
compared to the titania slurry.
Volumetric loading of the reactors and mol/mol coating. Due to the variation in
the densities of the ACs, it was deemed necessary to normalize the differences in the
volumes of ACs loaded to the reactors. In other words, since the wood-based carbon was
less dense than the coal based carbons, the mass of carbon added for each experiment was
less. Since the mass of carbon was less, then mass of titania would be less. Therefore,
the system could be normalized by adding the carbons to the reactor based on their
volumes versus mass. Hence, 0.025 mL of AC was loaded to each of the respective
The results of the experiment carried out under these conditions are shown in
=E U Initial Concentration
S40 0 0.025ml 3% Bionuchar
0 0.025ml 3% F400
IM c3 0.025ml 3% HD4000
Figure 4-10. Batch adsorption-photocatalysis studies of 3% (m/m) coated ACs loaded to
the reactors on a volume basis in the presence of UV
The tests did not result in any significant difference in performance of the ACs. A
trend similar to the one seen in Figure 4-9 was seen for this experiment as well. With this
said, the change in basis of loading from mass to volume only succeeded in normalizing
the amount of AC in each reactor. It was also of interest to learn if normalizing the
amount of titania on each AC based on titania to carbon would impact the results. This
ratio would be different for each carbon because of the different ash contents. To
normalize the titania to carbon loading, the ACs were coated such that 3 moles of titania
were present for every 100 moles of elemental carbon. In doing so, the titania to
elemental carbon ratio in each AC remained constant. In addition to the change in
coating basis, the titania-coated ACs were continued to be loaded by volume into their
respective reactors. Using these bases for coating the titania on the ACs and loading them
to the reactors, tests were performed in the absence as well as in the presence of UV light.
* Initial Concentration
* 3% Bionuchar (mol/mol) loaded by volume
O 3% F400 (mol/mol) loaded by volume
O 3% HD4000 (mol/mol) loaded by volume
* Initial Concentration
m 3% Bionuchar (mol/mol) loaded by volume
O 3% F400 (mol/mol) loaded by volume
O 3% HD4000 (mol/mol) loaded by volume
Figure 4-11. Batch adsorption studies: A) 3% (mol/mol) coated ACs in the absence of
UV loaded to the reactors by volume, B) Photocatalysis studies of 3%
(mol/mol) coated ACs in the presence of UV loaded to the reactors on a
0 20 -
Tests in the absence of UV were performed to assess as to how much the change in
the basis of coating affected adsorption of phenol since in each case the amount of coated
titania increased. Similarly, tests in the presence of UV were performed to assess the
effect that the increased titania coating on the ACs had on simultaneous adsorption-
photocatalysis. Figures 4-1 la and 4-1 lb show the results of the experiments in the
absence and presence of UV light respectively.
As in the earlier tests, an analysis of the data revealed no significant differences in
the trends of their performances. Although there are signs of photocatalysis having some
effect on phenol removal, the differences noticed in the trends are not of sufficient
magnitude to arrive at any definite conclusion with respect to how photocatalysis was
influenced by the variations.
The results seemed to show that adsorption dominated over photocatalysis as a
mechanism. Hence, the extent to which photocatalysis plays a role in the removal of
phenol in a co-adsorbent with high surface area was questioned. Unlike results shown by
Matos et al. (1998) in the case of organic compounds and Khan (2003) in the case of
dyes, the results seen herein failed to show any clear evidence of a synergy of any kind
existing between ACs and TiO2 with respect to performance. The premise that a co-
adsorbent with high surface area would result in an increase in photocatalytic efficiency
of titania (Matos et al., 1998; Takeda et al., 1995) can hence be questioned. The theory by
Matos et al. (2001) that L-type ACs inhibit photocatalysis whereas H-type ACs enhance
photocatalysis has also gone unfounded based on the data that has been presented. The
failure by Matos et al. (2001) to perform adsorption control studies seems to be the cause
of the unverifiable hypothesis.
4.2.4. Determination of Existence of Phenol* (Phenol with Delocalized Electron) in
the Presence of UV
The trends observed in the titania-AC composites, particularly when compared to
the performances of the virgin ACs resulted in the author investigating the possibility of a
phenol* phenomenon, similar to the dye* phenomenon noticed by Khan et al (2006). The
virgin ACs (50 mg) were exposed to UV for 24 hours without the presence of any
photocatalyst. As a control study, phenol solution (no AC, no titania) was also exposed to
UV light for 24 hours. The resulting data are provided in Figure 4-12.
o) 1 Virgin AC (NO UV)
-40 Coated AC (No UV)
o O Coated AC (Wth UV)
SVirgin AC (Wcth UV)
Initial Concentration Bionuchar F400 HD4000
Figure 4-12. Comparison of irradiated virgin ACs with virgin and coated ACs in the
presence and absence of UV
The results of the batch tests performed in section 4.2.3 are provided in Table 4-3.
The table quantifies to an extent the domination of the adsorption mechanism over
Table 4-3. Phenol removal as a function of AC coating strategy with and without
Activated Precursor Virgin AC 3% (m/m) AC 3% (m/v) 3% (mol/v)
carbons performance performance AC AC
(% removal) (% removal) performance performance
(% removal) (% removal)
No With No With With UV No With
UV UV UV UV UV UV
Bionuchar Wood 51.5 36.9 44.7 38.7 37.7 37.1 37.9
F400 Bituminous 87.1 84.7 82.3 88.6 91.3 91.3 92.5
HD4000 Lignite 75.9 61.3 73.4 78.2 74.2 74.5 75.5
Comparison of Bionuchar systems. In the case of Bionuchar, the virgin AC in the
presence of UV (36.9% removal) performed poorly compared to its performance in the
dark (51.5% removal). This can be accounted for by the theory put forth by Khan (2006)
whereby the excitation of delocalized electrons in the aromatic structure of phenol results
in reduced uptake by Bionuchar whose surface is supersaturated with acidic functional
groups. The decrease in the electron density in the basal planes of Bionuchar due to the
acidic functional groups coupled with the excitation of delocalized electrons in phenol
(resulting in phenol*) contributes to decreased xt-7t interactions and hence decreased
adsorption. Bionuchar's low affinity for phenol* results in the overall low adsorption of
phenol. The coated Bionuchar exposed to UV may show a similarly low performance due
to a combined effect of pore blockage and the effect of phenol* on its adsorption.
Comparison of F400 systems. F400, unlike Bionuchar and HD4000 does not show
much variation regardless of whether the carbon was coated with titania and exposed to
UV or if the virgin carbon was exposed to UV. The virgin F400 in the absence of UV
(87% removal) is observed to perform as well as the coated F400 which was irradiated
The similarities in the performances of the two systems could be a result of either
of two processes.
* The coated F400 compensates for the blockage of pores through photocatalytic
degradation of phenol or
* The coated F400 in the presence of UV behaves the same as the virgin F400. Hence
pore blockage is not a factor. However, this would also put to question the presence
of any photocatalytic effect.
It is also noticed that virgin F400 in the presence of UV performs slightly worse
(ca. 85%) than the F400 experiment in the dark (ca. 87%). This difference is however not
considered to be significant.
Comparison of HD4000 systems. Finally, for HD4000, the irradiated titania
coated sample displays maximum performance (78% removal) which is only slightly
greater than the performances of the virgin HD4000 (no UV) with 76% removal and the
coated HD4000 (no UV). It is however interesting to note that the irradiated samples of
virgin HD4000 perform considerably worse (61% removal) than the other samples. This
may point to phenol* having a large influence over the adsorption capacity of HD4000.
Noticing that the coated HD4000 in the presence of UV performs well, it may be
theorized that the performance of HD4000 observed in irradiated samples may be
attributed more to enhanced photocatalysis. It may be this enhanced photocatalysis which
compensates for the poor adsorption of phenol* attributed to its surface acidity.
The performances of the ACs in the remaining systems do not show any significant
deviation. The slight changes in performance are attributed to a combination of
adsorption and photocatalysis.
Hence, in summary, from the explanation provided above, it is clear that in terms of
photocatalytic activity, HD4000 seems to be the most active AC. The acidic nature of its
surface is thought to slightly inhibit adsorption. However, as seen from the SEM and
EDS images, it seems to have the most uniform coating and the most adhesive surface
towards TiO2. This could possibly be an explanation to its superior photocatalytic
activity. The poor performance of Bionuchar has been extensively discussed and
attributed to its characteristic acidic surface. A point has also been made regarding the
morphology of the Bionuchar resulting in deposition of TiO2 in its pores, hence resulting
in decreased adsorption. The poor adherence of TiO2 to the Bionuchar surface is the
cause of the low photocatalytic degradation of phenol observed with the TiO2-Bionuchar
composite. F400 largely displayed a consistent performance over each of its systems. Its
more basic surface, morphology resulting in good affinity towards TiO2 and high
adsorption capacity seem to be important factors.
The effect of conductivity, if any, seems to be overwhelmed by other factors such
as adsorption and morphology of the surface. In fact, in the case of activated carbons, the
phenomenon of irradiated phenol resulting in a change in adsorption characteristics
proves that a good photocatalytic system would require adsorption as a primary
Although it has not been investigated, the author also proposes the possibility of
UV irradiation having an effect on the AC surface. Hence, it is possible that the UV
irradiation affects the AC such that the carbon 7n system changes. This would also result
in noticeable changes in adsorption of phenol by each of the ACs. For instance, excitation
of the 7t electrons in the carbon basal planes may result in a phenomenon similar to the
one proposed with the phenol*. The author has however not investigated the possibility
of this phenomenon in any further detail.
The replicate sets for the experiments involving the virgin ACs in the absence of
UV radiation is provided in appendix A.
4.3 Low Surface Area Carbon Studies
The multitude of variables observed in the performance of ACs with respect to
adsorption and photocatalysis resulted in the investigation of carbon substrates having
low surface areas. To further test whether conductivity and metals play a role in
enhancing photocatalysis, low surface area carbons (< 8 m2/g) with varying
conductivities were used in further tests. By selecting a variety of carbons such that they
differed in their properties that were conducive or inhibitive to photocatalysis, it would be
possible to relate specific properties of the carbons to the observed synergy. As
hypothesized by Khan (2003), the reason for the superior performance of certain ACs is
the presence of ionic metals. These metals act as electron sinks, causing the hindrance of
electron-hole recombinations, thus enhancing photocatalysis. As explained earlier, this is
interpreted by this author as enhanced photocatalytic activity being a function of higher
electrical conductivity primarily due to specific sites on the carbons where metals are
present. In theory, the presence of titania in the vicinity of a conductive site (metal) will
result in a favorable difference in Fermi energies between the TiO2 and the metal. The
resulting effect is the change in the distribution of electrons of the semiconductor (TiO2).
The formation of the Schottky barrier will cause the migration of electrons from the
titania to the metal increasing the electron-hole recombination times and therefore
It should be noted that in general, the addition of conductive materials to
semiconductors or insulators has a positive impact on their conductivity. This property
may be approximated as a summation of the conductivities of the materials involved.
However, at the same time it is important to note that metals content is by no means the
sole factor influencing electrical conductivity. In the case of graphite, a major influence
of its electrical conductivity is its crystalline structure. Crystalline graphite consists of
parallel sheets of carbon atoms, each sheet containing hexagonal arrays of carbon atoms.
Each carbon atom exhibits sp2 hybridization and forms sigma bonds with its three nearest
neighbors. There is also distributed pi bonding between the carbon atoms in the sheet.
This delocalized pi system is responsible for the electrical conductivity of graphite. The
disruption of this graphitic structure is the cause for a loss in electrical conductivity in
With this in mind, carbon materials having varying electrical conductivites were
selected. The carbons selected for the studies were graphite, pitch coke, bituminous coal
and anthracite coal. The gradient in electrical conductivity between the selected carbons,
graphite being the highest and bituminous the lowest is shown in Table 4-4. These
conductivities were measured as resistivities by compacting the carbons in a column at
pressures between 125-150 psi and measuring their resistivity to charge (Data provided
by and experiments performed by Asbury Carbons, NJ). Hence the resistance of the
compacted powdered carbons was measured and obtained which was then easily
translated to electrical conductivity by the reciprocal of the resistivity.
Table 4-4. Electrical conductivity values for low surface area carbons
Carbon Conductivity (mQ' cm' ) (mho)
Pitch Coke 10
Bituminous < 0.001
4.3.1 Metals Content of Low Surface Area Carbons
Before discussing how the titania coated low surface area carbons performed for
phenol adsorption, it was of interest to also investigate the ash contents of these samples.
To determine if conductive metals were indeed present in any of the carbons, an EDS
study similar to the one performed on the ACs was performed. Prior to that, as in the case
of ACs, the quantity of ash for each carbon was determined. This would enable us to
analyze the EDS scans with more perspective. The ash contents of the low surface area
carbons are provided in Table 4-5.
Table 4-5. Elemental carbon and ash contents of low surface area carbons
Carbon Elemental carbon (%) Ash (%)
Graphite 99% 1%
Pitch coke 93% 7%
Anthracite 90% 10%
Bituminous 65-72% 28-35%
EDS scans of all the carbons were performed with the exception of graphite. The
electrical conductivity of graphite is extremely high and its ash content is negligible.
Hence, the amount of impurities found in graphite (metals) would be negligible and
would not make any significant difference. However, it is interesting to see whether the
conductivity due to its structure plays a role in photocatalysis.
The EDS scan of the ash residue of the anthracite coal is shown in Figure 4-13. It
shows prominent peaks corresponding to aluminum and iron. These metals have very
high electrical conductivities (376/mohm-cm and 102 /mohm-cm respectively) as
compared to titanium (23 /mohm-cm). A smaller peak of sodium is also seen. Another
peak present is that of silicon (electrical conductivity of silicon = 0.012 /mohm-cm). A
number of such EDS scans were taken, all of which showed similar results.
K eTi Fe
0 . .. . I.-' . ..
0 2 4 6 8
Figure 4-13. EDS scan of ash residue of uncoated anthracite coal
Again, it must be noted that EDS is by no means a quantitative analytical tool and
is at best semi-quantitative. However, the images do show a greater prominence of some
elements over others.
Figure 4-14 shows an EDS scan of the ash residue of bituminous coal. The image shows
a prominent aluminum peak. In addition to this, smaller peaks of iron, calcium and
magnesium and sodium are also seen. Again, as in the case of anthracite, a silicon peak is
0 2 4
Figure 4-14. EDS scan of ash residue of uncoated bituminous coal
Although anthracite and bituminous are low conductivity carbons, the presence of
metals such as aluminum and iron means that there is a high probability of a number of
conductive sites on the surface of these carbons. The amount of ash contained in these
carbons (10% and 28-35% respectively) also points towards there being a relatively large
amount of these metals in these carbons.
A similar scan was performed on the ash of pitch coke (Figure 4-15). Small peaks
of aluminum and silicon in addition to small peaks of sodium and sulfur are seen.
Na Al S
0 I A . . i ,-,-,--
0 2 4 0 8
Figure 4-15. EDS scan of ash residue of pitch coke
From the EDS scans of the low surface area carbons, it is evident that aluminum is
commonly found in each of the carbons. A very prominent iron peak is seen in anthracite
coal with the bituminous sample also showing some presence of iron. Recalling the study
by Khan (2003) which hypothesized metals (specifically iron) to be the cause of
increased photocatalysis, it may be expected that anthracite and bituminous should show
the greatest removal when the carbons are subjected to batch adsorption-oxidation
4.3.2 Batch Studies with Low Surface Area Carbons
Batch studies were performed with the low surface area carbons. These tests were
performed in a procedure identical to that used in the activated carbon studies. The
carbons used were sized to 35 x 200 sieve in order to normalize as far as possible the
external surface area of the carbons. Titanium dioxide was coated on the ACs via boil
deposition on a mass basis. A mass of 80 mg of each sample was used with 80 mL of 55
ppm phenol solution.
Adsorption studies performed on coated low surface area carbons in the dark.
In order to verify the negligible adsorption capacity of the carbons, they were put through
adsorption studies in the absence of UV light. In addition to the carbons, a titanium
dioxide (Degussa P25) slurry was also tested as a comparison to the coated substrates.
The titanium dioxide was weighed such that its mass was equal to the theoretical mass of
the titania on the surface of the carbons (3% of 80 mg, i.e., 2.4mg). It was expected that
all the samples including the titanium dioxide slurry would show negligible adsorption.
Figure 4-16 shows the results of the adsorption studies.
3% Pitch Coke
o 3% Graphite
60 0 3% Anthracite
0 3% Bituminous
0 2.4 mg TiO2 Slurry
Figure 4-16. Adsorption studies of coated low surface area carbons and titanium dioxide
It can be seen that the adsorption seen in each of the carbon samples including the
slurry is negligible. In fact the maximum removal seen is only 4.6%. This emphatically
proves that the carbon composites are indeed of low surface area, incapable of any
significant adsorption. It was thought unnecessary to perform adsorption studies for the
virgin carbons due to their low surface areas.
Photocatalysis studies performed with coated low surface area carbons. The
carbons and the titania slurry were then made to undergo photocatalysis studies. Using an
identical configuration, the samples were exposed to 365 nm UV light. It was expected
that carbons containing a larger concentration of metals in their matrices would perform
better than those with a lower content of metals (Khan, 2003). Although all the carbons
with the exception of graphite contained a considerable amount of metals, due to the
presence of metals and the relatively high natural conductivity of pitch coke, it was
expected to perform better than the other carbon composites. Due to better mass transfer,
the titanium dioxide was also expected to perform relatively well. The results of the
experiment are shown in Figure 4-17.
Although there was little distinction in the performance of the carbons, it was
noticed that the titania coated bituminous coal showed the greatest removal of 20%.
Anthracite coal with 18% removal also performed relatively well. Pitch coke which was
expected to perform well only removed 6% of the phenol. The graphite which is the most
electrically conductive, although lacking in ash content showed a removal of 13%.
However, a surprising result was that of the titania slurry which only accounted for 11%
removal. This was attributed to a reduction of the surface area of TiO2 available for
irradiation. It was thought that the amount of titania loaded was too high.
3% Ptch Coke
60 3% Graphite
O[ 3% Anthracite
su- w 3% Bituminous
S50 2.4 mg Ti02 Slurry
Figure 4-17. Batch photocatalysis studies of carbons and titanium dioxide slurry (m/m)
Hence for the specific configuration of the reactor, the titania concentration in the
slurry was greater than the optimal loading, causing low penetration of UV through the
slurry. It has been previously observed in Figure 4-9 that a slurry with a lower loading of
titania has performed better.
Comparing the oxidation studies to the adsorption studies in the absence of UV, it is
evident that there is photocatalysis taking place. A comparison between the carbon-titania
particles shows no trend with respect to conductivity or metals content in their ash
residues. However, a trend was noticed in the difference in performances being similar to
the carbons' differences in densities. Due to this, the volume of carbons and hence the
amount of titania in each reactor would be different. Higher density would result in a
smaller volume of that particular carbon resulting in a lower TiO2 dose in the respective
reactor. Table 4-6 summarizes the densities of the coated carbons.
Table 4-6. Average densities of 3% (m/m) titanium dioxide coated carbons
Pitch coke 2.25
From the table it can be seen that bituminous coal and anthracite coal are the two
carbons with the least density. This difference in density is reflected in the difference
seen between the performances of the carbons Since the adsorption capacities for all the
carbons were negligible, the coated carbons were loaded on the basis of volume in order
to normalize the volumes and titanium dioxide loaded to each reactor. This was expected
to magnify differences in phenol decomposition between the carbons. The following
graph shows the results of the batch studies in which the carbons were coated with TiO2
on the basis of mass and loaded to the reactors such that their volumes equaled 0.05 mL.
This set, shown in Figure 4-18, is referred to as 'mass-volume' or 'm/v'.
The results reiterated the fact that there was no real difference in performance in the
carbons with respect to adsorption or photocatalysis. The lack of distinction in
photocatalytic performance as well as adsorption between the carbons prompted the
author to discard the theory regarding phenol* having an influence on photocatalysis
when adsorption was absent. It was postulated that the excitation of the electrons in
phenol only has an effect on its adsorption characteristics and does not affect
The influence of metals on photocatalysis was also questioned as clearly, there was
no distinction seen in photocatalysis between the carbons even though they differed in the
content and composition of metals.
3% Ptch Coke
60 0 3% Graphite
Fi 4 O 3% Anthracite
s a 3% Bituminous
Figure 4-18. Photocatalysis studies of coated low surface area carbons in the presence of
To further verify these results and normalize the amount of carbon and titania in
each reactor, tests were performed using a molar basis for coating the carbons with TiO2.
This basis, explained in detail in the materials and methods section consists of coating the
carbons such that the mole ratio of titanium dioxide to carbon is 3%. The carbons were
then loaded to their reactors on the basis of volume as in the earlier case (0.05 mL). This
set of samples has been referred to as 'mole-volume' or 'mol/v'. The results from these
tests which were otherwise performed in an identical manner to the previous batch tests
are shown in Figure 4-19. For the mass-volume and the mole-volume sets, it was deemed
unnecessary to perform control studies in the absence of UV light as the carbons
contained no internal pore structure. Hence, a higher concentration of TiO2 coated on the
carbon surface would not make any noticeable difference to adsorption. The results from
these tests showed a slight increase (less than 2%) in the photocatalytic degradation of
phenol. However, this was seen in all the samples, resulting again in no significant
distinction between the samples.
m Initial Concentration
3% Pitch Coke
60 O 3% Graphite
O 3% Anthracite
50 3% Bituminous
Figure 4-19. Photocatalysis studies of low surface area carbons in the presence of UV
It may be noticed that the performances of the ACs with the exception of the 'mole-
volume' samples which showed a slight increase in performance, remains relatively
constant. The only considerable distinction can be made in the case of the 'mass-mass'
carbons. However, this has been explained by the differences in their densities. It is
evident from these studies that when titanium dioxide is coated on a substrate such as
carbon, adsorption plays a very important role. More importantly, what is shown by the
low surface area carbon tests is that as previously hypothesized by this author (Section
2.4), conductivity may not be playing any significant role in the enhancement of
photocatalysis. Even though carbons were selected whose conductivities varied by orders
of magnitude, there was no significant effect or even any trend that was noticed in the
performance of the titania coated carbons. Hence, the conclusion by Khan (2003) about
metals such as iron enhancing photocatalysis due to their role as electron traps is
uncorroborated. At the same time, the theory by Matos et al. (2001) regarding the
synergy seen in the case of titania systems containing H-type ACs is also unsubstantiated.
Furthermore, the experiments conducted with the AC composites and the low
surface area carbonaceous composites show that a slight synergy may be noticed in TiO2-
AC composites, primarily driven by adsorption. The difference in the electrical
conductivities of the low surface area substrates did not show any considerable synergy.
SUMMARY AND CONCLUSIONS
In this study, activated carbons which differed in physical and chemical properties
were coated with titanium dioxide and subjected to a variety of experiments. The AC
composites were primarily tested with respect to the impact of electrical conductivity of
the substrate on photocatalysis in addition to photocatalysis as a function of morphology
and surface chemistry.
The results did not definitively support my initial hypothesis that electrical
conductivity of the ACs influences the photocatalytic properties of the AC-TiO2
composite. A slight improvement in photocatalysis was observed in coal based ACs
which were believed to possess higher conductivities. The improvement in performance
did not significantly prove that a synergy was present between TiO2 and the conductive
ACs (F400 and HD4000); however, the definite trend seen in the tests leads the author to
believe that conductive sites on the surface of AC may result in a localized enhancement
of photocatalysis. Moreover, this localized enhanced photocatalysis mechanism is
insignificant in the perspective of the adsorption mechanism at the scale at which it was
EDS scans of the ACs showed the absence of key metals (i.e., absence of iron in
HD4000) which had been attributed to the synergy observed in the tests performed by
Khan (2003). The premise that trace metals in the AC serve as electron traps, prolonging
electron-hole recombination times leading to enhanced photocatalysis was thus
questioned. In effect, the presence of metals and hence the influence of conductivity of
ACs on enhanced photocatalysis is a minor factor for photocatalysis due to their low
distribution and the low probability that a TiO2 particle would be deposited in their
vicinity. The argument was further strengthened after reviewing the performances of the
low surface area carbons which contained varying quantities of metal composition but
showed no significant differences in performances.
There is evidence that the morphology of the material (AC) to an extent governs
the amount of photocatalyst that adheres to it. The smoothness of surfaces such as in the
case of the wood based Bionuchar resulted in decreased amounts of titania being
deposited. In fact, it was noticed that of the titania that was deposited, a large quantity
was driven into the pores, pointing towards adsorption of titania in the pores during the
coating process as opposed to its deposition on the exterior surface of Bionuchar. The
pore blockages as a result of this are one of the reasons for Bionuchar's lower phenol
adsorption. In contrast, rough surfaces such as the coal based F400 and HD4000
displayed greater quantities of TiO2 deposited on their surfaces. SEM images of the
coated ACs also showed a more uniform coating of titania. It was noticed that the same
ACs also displayed greater adsorption capacity
Presence of surface functional groups on the AC is also believed to be a major
influence on adsorption and hence indirectly on photocatalysis. Surface functional groups
are well known to have an influence on the adsorption of organic compounds by activated
carbons. This fact was related to another phenomenon in explaining the variation in
trends seen in the ACs when they were subjected to UV. The phenomenon involved a
similar observation to the one by Khan (2003) in the case of dyes. Irradiated virgin ACs
showed a decrease in performance when compared to the adsorption demonstrated by
virgin ACs in the dark. Phenol* which is the excited state of phenol in which delocalized
electrons in phenol are excited is the suggested theory. The adulteration of the initial
compound is believed to be a reaction which occurs in the picosecond time scale,
explaining why irradiated phenol solution does not show any change in spectroscopic
Another possibility which has not been investigated by this author is that the UV
radiation affects the activated carbon itself. Hence, it is possible that there is a change in
the system of the 7n electrons in the carbon basal planes of the activated carbon resulting
in a decrease in the dispersive adsorption forces. Due to the decreased 7n-7n interactions; in
this case due to the carbon 7n electrons, a decrease in adsorption of phenol by the
irradiated ACs may be noticed. Another possibility is that both the theories may take
place simultaneously causing the observed results.
The superior performance of F400 across all the systems is attributed to a
combination of factors. The basic surface of F400 does not inhibit adsorption of the
organic contaminant as its 7n-7n interactions are strong. The pore blockages of F400 are
overcome by the superior photocatalysis that it shows. Although the presence of metals
such as aluminum might point towards a Fermi energy difference influencing the results,
it is highly unlikely given the large dominance of the adsorption mechanism.
On the basis of photocatalysis alone, HD4000 performed the best. The poor
adsorption capacity of HD4000 for irradiated phenol was compensated by photocatalysis.
The process of photocatalysis, especially when combined with the mechanism of
adsorption is one that is extremely complex and not very well understood. When the
mechanism of photocatalysis is combined with an adsorption mechanism, it is often
difficult to distinguish between and isolate the two mechanisms.
I propose that a synergy in activated carbon-TiO2 composites may indeed show no
significant improvements. However, even small improvements in performances may
actually be a synergistic phenomenon in which the AC is being photocatalytically
regenerated during its adsorption cycle. When the ACs are compared, it is evident that
there is a difference in their overall performance with the use of the titania-carbon
* In the case of Bionuchar, the decrease in the electron density in its basal planes due
to its acidic surface functional groups coupled with the excitation of delocalized
electrons in phenol explains the decrease in performance of its irradiated coated
composite as compared to its non-irradiated composite. It is due to a combination
of these factors that Bionuchar demonstrates poor performance.
* The synergy demonstrated by F400 is attributed to a combination of high
adsorption capacity of phenol as well as a balanced affinity for phenol which
would provide more pollutant to be photocatalytically degraded on the surface of
F400 which is uniformly coated with TiO2.
* Even though HD4000 may be considered an L-type AC, it actually shows a
synergistic effect with respect to adsorption and photocatalysis which refutes Matos
et al. (2001).
* Although electrical conductivity of the low surface area carbons varied by orders of
magnitude, they demonstrated the same performance. The absence of any
adsorption in fact decreased the distinction between carbons. Hence, adsorption is
clearly the dominant mechanism over photocatalysis. The key factors of adsorption
are the morphology of the AC, surface area and pore size of the AC and the surface
functional groups on the AC.
CONTRIBUTIONS TO SCIENCE AND ENGINEERING
6.1 Contributions to Science
The experiments carried out have yielded a very interesting set of results. An
analysis of the results has shown that combining two mechanisms (i.e., adsorption and
photocatalysis) creates a complex system which needs to be further understood. This
complex system consists of various mechanisms which may be dependent on each other.
The systems behave in such a manner that it is extremely difficult to distinguish between
the two mechanisms. In lieu of this, the contributions of this thesis to science in the
opinion of the author are as follows:
* It has been adequately demonstrated that surface functional groups primarily affect
adsorption and hence only indirectly affect photocatalysis. This is not in
accordance to what Matos et al. (1998, 2001) proposed. It is only due to the fact
that adsorption is a dominant mechanism that photocatalysis is observed to be
influenced by the surface functionality of the activated carbons.
* The phenomenon of the excitation of the delocalized electron in dyes (dye*) which
was observed by Khan et al. (2006) is also observed in the case of phenol (i.e.,
phenol*). However, whereas in the case of dyes it was observed to enhance or
inhibit adsorption depending upon the type of AC, in the case of phenol it was
primarily noticed to inhibit adsorption of both L-type and H-type ACs. It is also
possible that the UV radiation has an affect on the carbon 7t electron system,
thereby affecting adsorption.
* Finally, it has been shown that individual material properties do not have much of
an influence on the adsorption-photocatalysis system. It is more likely that a
synergy exists between the general mechanisms of photocatalysis and adsorption
with adsorption being the dominant mechanism. It is proposed that an enhancement
due to synergy may be noticed in systems in which the AC needs to be regenerated
in-situ. Prior to research being performed on individual material properties and
their effects on photocatalysis, further research must be performed to distinguish
between photocatalysis and adsorption and how they behave synergistically.
6.2 Contributions to Engineering
As mentioned in section 6.1, the research presented by the author demonstrates a
possibility to apply adsorption-photocatalysis systems to regenerate activated carbons in-
situ. Activated carbon in its contemporary use requires to be regenerated (usually off-site)
after its adsorption capacity has been reached. By the use of activated carbon coated with
titanium dioxide, it is possible to regenerate the ACs in-situ by exposing them to UV
radiation during their adsorption process.
Such a system would require a certain amount of fluidization to allow the radiation
to be incident on all of the TiO2-AC particles. Preliminary column studies were
performed by this author in which HD4000 and F400 were used in a column
configuration. Phenol was allowed to flow through the column and the particles were
slightly fluidized. The entire column was exposed to UV radiation.
Comparing the irradiated system to a system not exposed to UV, a small difference
was seen in the two sets of data with the irradiated system performing better. The data is
provided in Appendix B and Appendix C. It was found that although the two systems
were regenerated by flowing DI water to the same extent, the effluent showed lower
concentrations of phenol in the irradiated sample pointing towards photocatalytic
degradation of the phenol.
Hence, these bench scale column studies show the engineering application of
titanium dioxide coated activated carbon in the in-situ regeneration of activated carbon.
VIRGIN AC STUDIES IN PRESENCE OF UV
Figure A-i. Replicate data sets for virgin AC studies performed in the presence of UV
demonstrating the phenol* phenomenon observed in Section 4.2.4.
* With UV 1500 mL/hr
* Without UV 1500 mL/hr
x With UV Packed Bed
* Without UV Packed Bed
i i" n
Figure B-1. Recirculation column studies performed with F400 using different flowrates
(degrees of fluidization) in the presence and absence of UV.
COLUMN STUDIES II
1st Adsorption Run (Averaged)
S2nd Adsorption Run (Only DI)
1.00 2nd Adsorption Run (UV + DI)
0 100 200 300 400 500 600 700
Figure C-1. Adsorption runs of column tests performed with HD4000 in the presence and
absence of UV radiation measuring phenol concentration in the effluent
DI + UV Regeneration
1.60 m Only DI Regeneration
0 100 200 300 400 500 600 700 800
Figure C-2. Concentration of phenol
in effluent during regeneration runs of column
LIST OF REFERENCES
Bacsa RR and Kiwi J. Effect of rutile phase on the photocatalytic properties of
nanocrystalline titania during the degradation of p-coumaric acid. Applied Catalysis
B: Environmental. 1998; Vol.16; Issue 1: 19-29.
Byrne JA, Eggins BR, Brown NMD, McKinney B, Rouse M. Immobilization of TiO2
powder for the treatment of polluted water. Applied Catalysis B: Environmental.
1998; 17: 25-36.
Carp O, Huisman CL, Reller A. Photoinduced reactivity of titanium dioxide. Progress in
SolidState Chemistry. 2004; 32: 33-177.
Castilla CM, Rivera-Utrilla J, Lopez-Ramon MV, Carrasco-Marin F. Adsorption of some
substituted phenols on activated carbons from a bituminous coal. Carbon. 1995; 33:
Chhor K, Bocquet JF, Colbeau-Justin C. Comparitive studies of phenol and salicylic acid
photocatalytic degradation: Influence of adsorbed oxygen. Materials Chemistry and
Physics. 2004; 86: 123-131.
Coughlin RW and Ezra FS. Role of surface acidity in the adsorption of organic pollutants
on the surface of carbon. Environmental Science and Technology. 1968; vol.2; No.
Coughlin RW, Ezra FS, Tan RN. Influence of chemisorbed oxygen in adsorption onto
carbon from aqueous solution. Journal of Colloid andInterface Science. 1968; Vol.
28; No. 3: 386-396.
Dabrowski A. Adsorption from theory to practice. Advances in Colloid andInterface
Science. 2001; 93: 135-224.
Dijkstra MFJ, Buwalda H, De Jong AWF, Michorius A, Winkelman JGM, Beenackers
AACM. Experimental comparison of three reactors designs for photocatalytic
water purification. Chemical Engineering Science. 2001; 56: 547-555.
Dijkstra MFJ, Michorius A, Buwalda H, Panneman HJ, Winkelman JGM, Beenackers
AACM. Comparison of the efficiency of immobilized and suspended systems in
photocatalytic degradation. Catalysis Today. 2001; 66: 487-494.
Hobbs J. Adsorption of substituted aromatic compounds by activated carbon: A
mechanistic approach to quantitative structure activity relationships. 2005; Masters
Thesis, The University of Florida, Gainesville, FL-32611.
Hoffman MR, Martin ST, Choi W, Bahnemann DW. Environmental applications of
semiconductor photocatalysis. ChemicalReview. 1995; 95: 69-96.
Karthikeyan T, Raj gopal S, Miranda LR. Journal ofHazardous Materials. 2005; B 124:
Khan AY and Mazyck DW. The effect of UV irradiation on adsorption by activated
carbon/TiO2 composites. Letters to the Editor/Carbon. 2006; 44: 158-193.
Khan AY. Titanium dioxide coated activated carbon: A regenerative technology for water
recovery. 2003. Masters Thesis. The University of Florida, Gainesville, FL-32611.
Kinoshita K. Carbon. Electrochemical and Physicochemical Properties. Wiley, 1988.
Legrini O, Oliveros E, Braun AM. Photochemical processed for water treatment.
Chemical Review. 1993; 93: 671-698.
Leng CC and Pinto NG. Effects of surface properties of activated carbons on adsorption
behavior of selected aromatics. Carbon. 1997; Vol. 35; No.9: 1375-1385.
Leon y Leon C and Radovic L. Interfacial Chemistry and Electrochemistry of Carbon
Surfaces. In: Thrower P. editor: Chemistry and Physics or Carbon. Volume 25,
New York: Dekker, 1994.
Li W, Ni C, Lin H, Huang CP, Shah IS. Size dependence of thermal stability of TiO2
nanoparticles. Journal ofApplied Physics. 2004; Vol.96; No. 11.
Linsebigler AL, Guanquan L, Yates JT Jr. Photocatalysis on TiO2 surfaces: Principles,
mechanisms and selected results. Chemical Review. 1995; 95: 735-758.
Matos J, Laine J, Herrmann JM. Effect of the type of activated carbons on the
photocatalytic degradation of aqueous organic pollutants by UV-irradiated titania.
Journal of Catalysis. 2001; 200: 10-20.
Matos J, Laine J, Herrmann JM. Synergy effect in the photocatalytic degradation of
phenol on a suspended mixture of titania and activated carbon. Applied Catalysis B:
Environmental. 1998; 18: 281-291.
Mattson JS, Harry MB Jr., Malbin MD, Weber WJ Jr., Crittenden JC. Surface chemistry
of active carbon: specific adsorption of phenols. Journal of Colloid andInterfacial
Science. 1969; Vol. 31; No.1; 116-130.
Mattson JS, Mark HB. Active carbon surface chemistry and adsorption from solution.
Dekker, New York; 1971.
Mazyck DW. Steam-curing plus ramped temperature N2 treatment: A novel technology
for thermal reactivation of granular activated carbons (GAC). 2000; PhD
dissertation, The Pennsylvania State University, University Park, PA-16802.
Moreno-Castilla C, Carrasco-Marin F, Maldonado-Hodar F, Rivera-Utrilla J. Effect of
non-oxidant acid treatments on the surface properties of activated carbons with
very low ash content. Carbon. 1998; 36 (1-2): 145-151.
Nevskaia DM, Castillejos-Lopez E, Munoz V, Guererro-Ruiz A. Adsorption of aromatic
compounds from water by treated carbon materials. Environmental Science and
Technology. 2004; 38: 5786-5796.
Nevskaia DM, Santianes A, Munoz V, Guerrero-Ruiz A. Interaction of aqueous solutions
of phenol with commercial activated carbons: an adsorption and kinetic study.
Carbon. 1999; 37: 1065-1074.
Ohtani B, Kakimoto M, Nishimoto S, Kagiya T. Photocatalytic reaction of neat alcohols
by metals loaded titanium (IV) oxide particles. Journal of Photochemistry and
Photobiology A: Chemistry. 1993; Vol. 70; 70;3:265-272.
Ollis DF, Pelizzetti E, Serpone N. Photocatalyzed destruction of water contaminants.
Environmental Science and Technology. 1991; Vol. 25; 9: 1522-1529.
Radovic L, Sivla I, Ume J, Mendes J, Leon y Leon C, Scaroni A. An experimental and
theoretical study of the adsorption of aromatics possessing electron withdrawing
and electron donating functional groups by chemically modified activated carbons.
Carbon. 1997; 35(9): 1339-1348.
Ray AK and Beenackers AACM. Novel swirl flow reactor for kinetic studies of
semiconductor photocatalysis. A.I.Ch.E. Journal. 1997; 43(10): 2571-2578.
Salame II and Bandosz TJ. Role of surface chemistry in adsorption of phenol on activated
carbons. Journal of Colloid and Interfacial Sciences. 2003; 264: 307-312.
Serpone N. Brief introductory remarks on heterogeneous photocatalysis. Solar Energy
Materials and Solar Cells. 1995; 38: 369-379.
Tao Y, Schwartz S, Wu CY, Mazyck DW. Development of a TiO2/AC composite
photocatalyst by dry impregnation for the treatment of methanol in humid
airstreams. Industrial and Engineering Chemistry Research. Revision submitted:
Tessmer CH, Vidic RD, Uranowski L. Impact of oxygen containing surface functional
groups on activated carbon. Environmental Science and Technology. 1997; 31:
Trillas M, Peral J, Domenech X. Photocatalyzed degradation of phenol, 2,4-
Dichlorophenol, phenoxyacetic acid and 2,4-dichlorophenoxyacetic acid over
supported TiO2 in a flow system. Journal of chemical technology and
biotechnology. 1996; 67: 237-242.
Turchi CS and Ollis DF. Photocatalytic degradation of organic water contaminants.
Mechanisms involving hydroxyl radical attack. Journal of Catalysis. 1990; 122:
Yang P, Lu C, Hua N, Du Y. Titanium dioxide nanoparticles co-doped with Fe3+ and
Eu3 ions for photocatalysis. Materials Letters. 2002; 57: 794.