CET Volume 86
DOI: 10.3303/CET2186107
Paper Received: 18 September 2020; Revised: 5 March 2021; Accepted: 7 April 2021
Please cite this article as: Andreoli S., Eser S., 2021, Composition of Surface Groups and Adsorption Properties of Activated Carbons Derived
from Different Bio-precursors, Chemical Engineering Transactions, 86, 637-642 DOI:10.3303/CET2186107
CHEMICAL ENGINEERING TRANSACTIONS
VOL. 86, 2021
A publication of
The Italian Association
of Chemical Engineering
Online at www.cetjournal.it
Guest Editors: Sauro Pierucci, Jiří Jaromír Klemeš
Copyright © 2021, AIDIC Servizi S.r.l.
ISBN 978-88-95608-84-6; ISSN 2283-9216
Composition of Surface Groups and Adsorption Properties of
Activated Carbons Derived from Different Bio-Precursors
Sara Andreoli*, Semih Eser*
EMS Energy Institute, Penn State University, University Park, USA
sua437@psu.edu, seser@psu.edu
Activated carbons (AC) are widely used in a variety of applications because of their controllable porosity and
surface functionalities. In this work, AC were prepared from different hard woods through one- and two-step
pyrolysis/activation for the adsorption of organic pollutants in water. Water vapor was used as the activating
agent. The influence of the precursor and preparation methods on the properties of the resulting AC was
evaluated through multiple techniques. Temperature-programmed oxidation (TPO) measured the reactivity
toward oxygen of chars and AC, while temperature-programmed desorption coupled with mass spectrometry
(TPD/MS) revealed functional groups on AC surface. Methylene blue adsorption tests evaluated the
adsorption capacity of the prepared AC and the presence of mesopores. Depending on the oxidation reactivity
of the char produced by pyrolysis, the resulting AC show different surface composition and adsorption
performance. With the increasing char oxidation reactivity, the AC has higher degree of surface
functionalization and enhanced adsorption toward methylene blue. In particular, Red Maple char is the more
reactive one, followed by Birch and Ironwood. Thus, Red Maple AC prepared through the two-step process,
exhibits lower activation yield, higher concentration of surface functionalities and improved methylene blue
adsorption. TPO technique could predict the degree of surface functionalization and adsorption properties of
the final AC.
1. Introduction
Activated carbon (AC) is commonly used in purification and separation processes on a large scale due to its
large specific surface area and porous structure (Ma et al., 2017). Beside the low operating cost and simple
system design, the environmental and economic advantages of using AC increase when this material is
obtained from renewable biomass and available in large quantities (Ruiz-Rosas et al., 2019). Surface area
and porosity together with the surface chemistry affect the AC performance in many different applications.
Still, controlling and tuning these properties for specific applications remains a challenge for researchers. AC
can be prepared from a wide variety of raw materials using a chemical or physical activation process to create
the desired porosity and surface composition. Chemical activation requires an activating agent and relatively
low temperatures compared to physical activation which is conventionally performed with steam or carbon
dioxide following the carbonization of the raw materials. A one-step pyrolysis/activation method was also
reported in the literature where water vapor is introduced during the pyrolysis of biomass or other raw
materials (Gergova, 1996). Oxygen-containing groups are the most abundant on the AC surface. The nature
and concentration of surface groups is modified through different treatments, such as oxidative treatments
which can increase their concentration (Figueiredo et al., 1999) or thermal treatments in inert atmosphere
which remove some of them (Szymanski et al., 2002). Multiple methods characterize the surface of
carbonaceous materials, such as potentiometric titrations, electrochemical methods, temperature-programmed
desorption (TPD) and spectroscopic methods (Bandosz, 2008). Moreover, temperature-programmed
desorption can be coupled with mass spectrometry (TPD/MS) to determine the composition of gases evolved
upon heating the samples in an inert atmosphere. Since oxidation reactions are involved in the activation and
surface functionalization of AC, the oxidation reactivity of carbons becomes an important property which can
be determined through temperature-programmed oxidation (TPO) analysis (Andreoli, 2020).
637
The principal focus of this study was to prepare mesoporous AC to remove methylene blue used as model
pollutant. AC were prepared from different types of wood through one- and two-step activation with water
vapor. The surface composition, oxidation reactivity, and adsorption capacity of the prepared materials were
investigated and related to the AC precursors properties and preparation methods.
2. Materials and Methods
Lignocellulosic materials, like wood, account for around 45 % of the total raw materials used for the
manufacture of activated carbon (Dobele et al., 2012). In this study, the starting materials for AC (Red Maple,
Birch and Ironwood) were selected among hard wood species which have a higher content of cellulose and
hemicellulose with respect to the soft woods which are rich in lignin (Bhaskar et al. 2008). Cellulose
and lignin contents are 45 % and 23 % for Red Maple, 38 % and 22 % for Birch (Fengel, 1989), 44-45 % and
22 % for Ironwood (Kiaei et al., 2015), respectively.
2.1 AC preparation
AC were prepared form three different hard woods by a one-step pyrolysis/activation and a more conventional
two-step procedure. A diagram of the experimental set up used for the pyrolysis/activation experiments is
reported in Figure 1.
Figure 1: Diagram of the experimental set up used for pyrolysis/activation experiments.
Pyrolysis and activation take place in a vertical furnace while a horizontal furnace (set at 120 °C) and a HPLC
pump are used for water vapor generation. For the two-step procedure, 10 g samples of each wood (18-40
mesh) were pyrolyzed in a stainless-steel tube reactor at 550 °C for 30 minutes under nitrogen flow. After
cooling down, the char was heated and activated at 750 °C for 30 minutes with water vapor introduced in the
reactor with a liquid water flow rate of 1 mL/min. For the one-step procedure, 10 g samples of each wood were
heated at 20 °C/min to 750 °C and maintained at that temperature for 30 minutes. In this case water vapor
was introduced in the reactor at the beginning of the pyrolysis/activation at the same rate of two-steps
procedure.
2.2 AC characterization
For the composition of the surface functional groups, temperature-programmed desorption and mass-
spectrometry (TPD/MS) experiments were carried out on 60-80 mg of sample with a Micromeritics TPD/TPR
2910 equipment. After degassing at 120 °C for 30 minutes, samples were heated at 10 °C/min to 900 °C and
maintained at that temperature for 10 minutes under argon flow (20 mL/min). An on-line quadrupole mass
spectrometer (Ametek Dycor, USA) analysed the exit gas stream and the amounts of CO and CO2 evolved
were obtained by integration of the area under the corresponding peaks. Temperature-programmed oxidation
(TPO) analyses were carried out in a LECO RC612 Multi-phase Carbon Analyzer with 50 mg of ground
samples under a continuous flow of oxygen (1 L/min) from 100 to 550 °C. All carbon species from carbon
oxidation were converted to CO2 on a downstream CuO catalyst bed and the amount of evolving CO2 was
quantified with a calibrated IR detector. Methylene blue adsorption tests were performed on 20 mg of AC
638
placed in 100 mL of dye solution (0.1 g/L) and stirred at room temperature. After 24 hours, the amount of
methylene blue left in the solution was measured with a Milton Roy Spectronic 401 at 663 nm after dilution.
3. Results and discussion
The AC samples, prepared as described above, were characterized with respect of their oxidation reactivity,
surface composition and adsorption capacity. Results from the different characterization techniques will be
discussed in the following sections. Moreover, yields for the pyrolysis/activation were examined based on the
different precursors and number of steps used for AC preparation.
3.1 Char and activated carbon yields
Type of precursors and number of steps for pyrolysis/activation are reported in Table 1. For the one-step
procedure, only the final yield was calculated since pyrolysis and activation were performed at the same time.
The activation yield for the two-step process was calculated using the weight of the char (after pyrolysis) as
the starting weight.
Table 1: AC precursors, number of steps for pyrolysis/activation and corresponding yields.
AC
precursors
Number
of steps
Pyrolysis
Yield
(%)
Activation
Yield
(%)
Final
Yield
(%)
Ironwood 1 - - 18
Ironwood 2 22 73 16
Red Maple 1 - - 4
Red Maple 2 27 48 14
Birch 1 - - 4
Birch 2 26 65 17
Ironwood yields from one- and two-step procedures are comparable, 18 and 16 %, respectively. However, for
the other two wood samples, the yield for the one-step preparation was very low (4 %), compared to the two-
step method, which gave 14 % for Red Maple and 17 % for Birch. After one-step pyrolysis/activation, Red
Maple and Birch gave very low yields compared to Ironwood. The difference in the yield values for different
biomasses treated under the same conditions is often related to their lignin content. However, in this study,
different yield values were observed despite the similar content of lignin of the considered woods (22-23 %).
This can be explained by the lower density of Red Maple and Birch compared to Ironwood. The high density of
Ironwood could prevent the migration of volatile matters to the surface of the wood and consequently favour
the formation of secondary charcoal, as suggested by Dofourny et al. (2019). For the two-step process, the
pyrolysis yields were found to be comparable for the different precursors (22-27 %), while the activation yields
of Ironwood was higher (73 %) than Birch (65 %) and Red Maple (48 %). Beside the density of the starting
woods, the differences in the activation yields could be explained by TPO results shown in the next section.
3.2 Oxidation reactivity of chars and activated carbons
TPO technique was used to characterize the oxidation reactivity of chars and activated carbons (Figure 2).
Figure 2: a) TPO profiles for two-step chars (dashed lines) and AC (solid lines); b) TPO profiles for AC from
one-step procedure. Number of steps is indicated with 1 for one-step and 2 for two-step.
639
During TPO, the reaction of solid carbon with molecular oxygen takes place on specific active sites, including
structural defects. Thus, there is a relationship between the intrinsic oxygen reactivity of solid carbons and
their structure (Alonso-Morales et al., 2013). In the TPO profiles of carbonaceous substances, the more
reactive species shows oxidation peaks at lower temperatures, while the less reactive species is oxidized at
higher temperatures.
In Figure 2a, the chars and corresponding AC, obtained from the three wood samples through two-step
process, show oxidation activity in a wide temperature range between 300 and 500 °C and the temperatures
of maximum rate of CO2 evolution vary in the range from 360 to 445 °C, suggesting significant differences in
the reactivities of the chars and AC samples. This is an interesting observation considering that the pyrolysis
of the wood samples and the subsequent activation of pyrolysis chars were conducted under the same
conditions, respectively. For all samples, the AC profile peaks are shifted to higher temperatures compared to
the parent chars, as expected, because the more reactive part of the chars is removed by oxidation.
Moreover, the char and AC derived from Red Maple are the most reactive ones, followed by the Birch and the
Ironwood products, respectively. To summarize, the nature of the starting wood has a strong effect on the
oxidation reactivity of the pyrolysis chars and the corresponding activated carbons.
The TPO profiles of the two-step products (Figured 2a) can be related to the differences observed among the
corresponding AC yields. The Red Maple sample that has the highest oxidation reactivity gave the lowest yield
(48%) after activation because this material is more prone to oxidation. The Birch sample shows an
intermediate reactivity followed by the Ironwood sample which is the least reactive among the three. Hence,
the AC yield is intermediate for Birch (65%) and the highest for Ironwood (73%). At the same time Ironwood
has the highest density among the three woods. Thus, the density of the starting woods appears to be an
important factor that could affect the oxidation reactivity. Referring back to Table 1, the overall AC yields
based on the starting wood sample are not very different, 13, 16, and 17 %, for Red Maple, Ironwood, and
Birch, respectively, while the AC yields from the corresponding chars vary widely, 48, 73, and 65 %, in the
same order. Thus, the principal influence of the nature of different woods is manifested in the reactivity of the
pyrolysis chars, and, consequently, in the AC yields obtained from the pyrolysis chars.
In Figure 2b, the AC produced through one-step pyrolysis/activation show a reactivity trend similar to the AC
derived from the two-step process. Indeed, Red Maple AC has the higher reactivity with a maximum of CO2
evolution around 400 °C, followed by Birch AC at 450 °C and Ironwood AC at 466 °C. Comparing the
temperatures at which the maximum evolution of CO2 is observed, there is a shift toward higher temperature
for one-step derived AC, suggesting a lower reactivity for these samples with respect to the two-step AC. This
observation is consistent with an earlier report proving that the presence of water vapor introduced at the
beginning of the pyrolysis could lead to a considerable activation of the carbon (Gergova, 1996). Thus, a
considerable portion of the more reactive part of carbonaceous material was removed during the
pyrolysis/activation. This could explain also the very low yields of AC (4 %) obtained from the Red Maple and
Birch samples, the woods with the lower density compared to Ironwood which gave a rather high AC yield (18
%) comparable to that obtained by the two-step run (16 %).
3.3 Surface functionalities
TPD/MS analysis provides quantitative information on the concentration of functional groups on carbon
surfaces TPD/MS data may help explain the relationships between the oxidation reactivity of pyrolysis chars
and the composition of the surface functional groups associated with the different AC. Table 2 shows the
amounts of CO2 and CO evolved from the AC samples during the TPD/MS analyses.
Table 2: Amount of CO2 and CO evolved during the TPD/MS analyses from the activated carbons obtained by
one- and two-step runs from the three samples, as indicated with 1 for one-step, and 2 for two-step.
Sample CO2 amount
(µmol/g)
CO amount
(µmol/g)
Ironwood 1 687 429
Ironwood 2 307 359
Red Maple 1 985 664
Red Maple 2 871 928
Birch 1 2171 1057
Birch 2 602 739
During TPD analysis, carbonaceous materials are heated in inert atmosphere and their surface functionalities
decompose releasing primarily CO2, CO and H2O. Carboxylic acids and lactones generate CO2. Carboxylic
anhydrides produce simultaneously CO and CO2. Phenols, ethers and carbonyl groups release CO (Aksoylu
640
et al., 2001). Therefore, the higher the amount of CO2 and CO released, the larger is the extent of
functionalization on sample surfaces. Amounts of CO2 and CO released are of the same order of magnitude
reported for commercial activated carbons (Pereira et al., 2003). In general, AC produced through one-step
activation released larger amounts of CO2 compared to those derived from the two-step runs. As reported in
the section regarding the oxidation reactivity of chars and activated carbons, the presence of water vapor
introduced at the beginning of the pyrolysis could lead to a considerable activation of the carbon. Thus, a
higher amount of surface functionalities could be expected for the one-step AC. In particular, the difference is
more marked for the amount of CO2 released from Birch AC after one- and two-step procedures. At the same
time, a larger shift toward higher temperature was observed for Birch AC from one- and two-step procedure in
the TPO profiles. Indeed, the temperature for the maximum of CO2 evolution was 43 °C higher for the one-
step birch AC with respect to the corresponding two-step sample. For the one-step AC derived from Red
Maple and Ironwood, a smaller shift was observed with respect to the corresponding two-step AC, 25 and 21
°C, respectively. Same trend is observed for the CO amounts, with the exception of Red Maple AC (Red
Maple 2) sample which shows a higher amount of CO although it was produced through the two-step process.
This may be explained by the higher reactivity of Red Maple char. AC derived from different sources through
the two-step procedure show difference in surface composition and the reason of this can be find again in the
reactivity of the parent chars. From TPO analysis, Red Maple was found to be the more reactive char and the
corresponding AC (Red Maple 2) shows the higher amount of surface functionalities (higher amount of CO2
and CO). Chars from Birch and Ironwood showed respectively an intermediate and lower reactivity. Indeed,
CO2 and CO amounts from TPD/MS analysis are intermediate for Birch (Birch 2) and lower for Ironwood
(Ironwood 2) derived AC. Thus, TPO technique could predict the extent of surface functionalization based on
the tendency of the char to oxidized during the activation process. The higher the oxidation tendency of the
char, the larger the amount of surface functionalities on AC is.
3.4 Methylene blue adsorption
Methylene blue adsorption tests indicate the capacity of AC to adsorb molecules having similar dimensions to
methylene blue and to determine the presence of pores larger than 1.5 nm (Hui, 2015).The presence of
surface functionalities, as well as the porous structure, are known to affect the adsorption performance of AC
(Spagnoli et al., 2017). Methylene blue can interact with carbonaceous adsorbent through electrostatic
interactions, like dipole-dipole interaction between methylene blue and surface phenols or hydrogen bonding
between hydroxyl group of phenols and aromatic ring of methylene blue (Spagnoli et al., 2017). Moreover, an
increase in adsorption of methylene blue was observed for materials having carboxylic acid groups on the
surface (Laksaci et al., 2017). Figure 3 shows the trend of methylene blue adsorption for two steps AC with
the amount of CO and CO2 evolved from the decomposition of surface functionalities.
Figure 3: CO and CO2 amounts and methylene blue uptake for two-step AC.
The Red Maple AC shows the higher amount of surface groups and the higher methylene blue uptake,
followed by Birch and Ironwood AC. This trend suggests that the presence of surface groups improved the
adsorption performance. However, in the adsorption process, the porosity could play an important role as well.
The adsorption capacity toward methylene blue of the prepared AC is comparable to the one reported by
Zubair et al. (2020) for biomass-derived AC.
4. Conclusions
Significant differences in the yields, oxidation reactivities, surface compositions and adsorption properties
were observed for wood samples pyrolyzed and activated under the same conditions. Char reactivity toward
oxygen and density of starting woods affect the pyrolysis/activation yields and the properties of the resulting
materials. Ironwood has the higher density and provides higher yields. Red Maple char shows a higher
641
tendency to oxidation and formed AC with lower yield, higher content of surface functionalities and improved
adsorption toward methylene blue. Determining the oxidation reactivity of chars through TPO may help in
predicting the properties of the resulting AC as extent of surface functionalization which affects the adsorption
performance.
Acknowledgments
We thank Nate A. Rachor for the pyrolysis/activation experiments.
References
Aksoylu A. E., Madalena M., Freitas A., Pereira M. F. R., Figueiredo J. L., 2001, The effects of different
activated carbon supports and support modifications on the properties of Pt/AC catalysts, Carbon, 39 175–
185.
Alonso-Morales N., Gilarranz M. A., Heras F., Rodriguez J. J., Eser S., 2013, Oxidation Reactivity and
Structure of LDPE-Derived Solid Carbons: A Temperature-Programmed Oxidation Study, Energy Fuels,
27, 1151−1161.
Andreoli S., Eser S., 2020, Relating reactivity to structure in cokes and carbon materials: Temperature-
programmed oxidation and microscopy techniques, Carbon, 168, 362-371.
Bandosz, T. J., 2008, Surface Chemistry of Carbon Materials, Chapter 2. In: Serp P., Figueiredo J. L., (Ed),
Carbon Materials for Catalysis, John Wiley and Sons INC, Hoboken, New Jersey, USA.
Bhaskar T., Sera A., Muto A., Sakata Y., 2008, Hydrothermal upgrading of wood biomass: Influence of the
addition of K2CO3 and cellulose/lignin ratio, Fuel, 87, 2236–2242.
Dobele G., Dizhbite T., Gil M. V., Volperts A., Centeno T.A., 2012, Production of nanoporous carbons from
wood processing wastes and their use in supercapacitors and CO2 capture, Biomass and Bioenergy, 46,
145-154.
Dufourny A., Van De Steene L., Humbert G., Guibal D., Martin L., Blin J., 2019, Influence of pyrolysis
conditions and the nature of the wood on the quality of charcoal as a reducing agent, Journal of Analytical
and Applied Pyrolysis, 137, 1–13.
Kiaei M., Kord K., Chehalmardian A., Moya R., Farsi M., 2015, Mineral content in relation to radial position,
altitude, chemical properties and density of Persian ironwood, Maderas. Ciencia y tecnología, 17 (3), 657-
672.
Ma Q., Yu Y., Sindoro M., Fane A. G., Wang R., Zhang H., 2017, Carbon-Based Functional Materials Derived
from Waste for Water Remediation and Energy Storage, Advanced Materials, 29, 1605361.
Fengel D. (Ed), 1989, Wood: Chemistry, Ultrastructure, Reactions, Walter de Gruyter & Co., Berlin, New York,
USA.
Figueiredo J. L., Pereira M. F. R., Freitas M. M. A., Orfao J.J.M., 1999, Modification of the surface chemistry of
activated carbons, Carbon, 37, 1379–1389.
Gergova, K., Eser S., 1996, Effects of activation method on the pore structure of activated carbons from
apricot stones, Carbon, 34 (7), 879-888.
Hui T. S., Zaini M. A. A., 2015, Potassium hydroxide activation of activated carbon: a commentary, Carbon
Letters, 16 (4), 275-280.
Laksaci H., Khelifi A., Belhamdi B., Trari M., 2017, Valorization of coffee grounds into activated carbon using
physic—chemical activation by KOH/CO2, Journal of Environmental Chemical Engineering, 5, 5061–5066.
Pereira M. F. R., Soares S. F., Orfao J. M. J, Figueiredo J. L., 2003, Adsorption of dyes on activated carbons:
influence of surface chemical groups, Carbon, 41, 811-821.
Ruiz-Rosas R., García-Mateos F. J., Gutiérrez M. C., Rodríguez-Mirasol J., Cordero T., 2019, About the Role
of Porosity and Surface Chemistry of Phosphorus-Containing Activated Carbons in the Removal of
Micropollutants, Frontiers in Materials, 6, 1-14.
Spagnoli A., Giannakoudakis D. A., Bashkova S., 2017, Adsorption of methylene blue on cashew nut shell
based carbons activated with zinc chloride: The role of surface and structural parameters, Journal of
Molecular Liquids, 229, 465–471.
Szymanski G. S., Karpinski Z., Biniaka S., Swiatkowski A., 2002, The effect of the gradual thermal
decomposition of surface oxygen species on the chemical and catalytic properties of oxidized activated
carbon, Carbon, 40, 2627–2639.
Zubair M., Mu’azu N. D., Jarrah N., Blaisi N. I., Aziz H. A., Al-Harthi M. A., 2020, Adsorption Behavior and
Mechanism of Methylene Blue, Crystal Violet, Eriochrome Black T, and Methyl Orange Dyes onto Biochar-
Derived Date Palm Fronds Waste Produced at Different Pyrolysis Conditions, Water, Air & Soil Pollution,
231, 240.
642