Phenol adsorption on high microporous activated carbons prepared from oily sludge: equilibrium, kinetic and thermodynamic studies

The purpose of this study was the preparation, characterization and application of high-performance activated carbons (ACs) derived from oily sludge through chemical activation by KOH. The produced ACs were characterized using iodine number, N2 adsorption-desorption, Fourier-transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM). The activated carbon prepared under optimum conditions showed a predominantly microporous structure with a BET surface area of 2263 m2 g−1, a total pore volume of 1.37 cm3 g−1 and a micro pore volume of 1.004 cm3 g−1. The kinetics and equilibrium adsorption data of phenol fitted well to the pseudo second order model (R2 = 0.99) and Freundlich isotherm (R2 = 0.99), respectively. The maximum adsorption capacity based on the Langmuir model (434 mg g−1) with a relatively fast adsorption rate (equilibrium time of 30 min) was achieved under an optimum pH value of 6.0. Thermodynamic parameters were negative and showed that adsorption of phenol onto the activated carbon was feasible, spontaneous and exothermic. Desorption of phenol from the adsorbent using 0.1 M NaOH was about 87.8% in the first adsorption/desorption cycle and did not decrease significantly after three cycles. Overall, the synthesized activated carbon from oily sludge could be a promising adsorbent for the removal of phenol from polluted water.

characteristics of activated carbons. It can be clearly observed from Table 1 that the iodine number of ACs increases with increasing the activation temperature from 600 to 800 °C and the impregnation ratio from 1:1 to 2:1. According to the obtained results, the highest porosity was obtained at an activation temperature of 800 °C and an impregnation ratio of 2:1. Activation temperature plays an important role in the formation of micro-or mesopores in the ACs 14 . Generally, an increase of activation temperature from 600 °C to 800 °C, resulted the development of higher porous structure. This could be due to the reaction of KOH and carbon surface, vaporization of volatile material and development of new pores in the structure of raw materials 15 . However, a further increase in the activation temperature up to 900 °C led to a contraction of the porous structure. Small amounts of metallic potassium are produced at temperatures higher than 700 °C that intercalate between the carbon layers. Excessive temperature may cause the destruction of pore walls and transform micropores into meso-or macropores 16 , resulting in a porosity loss in agreement with the results of this work. Furthermore, it is also reported that very high temperatures increase the rate of gasification reaction resulting in destruction of micropores and reducing the surface area of AC 17 . The increase of the impregnation ratio from 1:1 to 2:1 resulted in a porosity development due to the presence of more KOH that could react with the carbon surface causing more release of gases and formation of more volume of pores on surface of adsorbent. However at higher impregnation ratios, the micropores might coalescence as a result of the deeper carbon decomposition 18 . At the highest impregnation ratio of 3:1, pore volume decreased as a consequence of destructive effect of chemicals on the micropore structure as reported by Yahya et al. 19 .

Sample
S BET (m 2 g −1 ) S mic (m 2 g −1 ) S ext (m 2 g −1 ) V total (m 3 g −1 ) V mic (m 3 g −1 ) V mic (%) Iodine number Yield (%) www.nature.com/scientificreports www.nature.com/scientificreports/ The N 2 adsorption/desorption isotherms of ACs prepared at different activation temperatures and an impregnation ratio of 1:1 are shown in Fig. 1(A). The adsorption isotherms for all those ACs belong to the combination of type І and type ІV isotherms according to the IUPAC classification 20 , which indicates the existence of both microand mesopore structures. According to the plots, with increasing the temperature from 600 to 800 °C, the amount of N 2 adsorbed at low relative pressures increases demonstrating the higher microporosity of the produced activated carbon 20 . Type І isotherm is typically attributed to the microporous samples with a relatively small external surface, while some mesopores exist in the AC due to slight increase of N 2 adsorption at higher relative pressure after filling the micropores 21 . After progressive increasing of P/P 0 in the isotherm of type I, a roughly horizontal line extended up. This isotherm describes the monolayer adsorption, which is consistent with the Langmuir equation 21 . Figure 1(B) shows N 2 adsorption/desorption isotherms of the obtained ACs with different impregnation ratios and at activation temperature of 800 °C. All the ACs showed again a combination of type I and type ІV isotherms of the IUPAC classification demonstrating the co-existence of micro and mesopores in the structure of ACs. The hysteresis loops in all desorption isotherms are characteristics of the existence of slit-shaped micropores and/or mesopores in the structure of these samples 22 . The maximum percentage of micropore volume (73%) was obtained at an impregnation ratio of 2:1 (mass ratio of activating agent to carbonized oily sludge) and an activation temperature of 800 °C. This sample (CK2T800), which shows the maximum BET value of 2263 m 2 g −1 and iodine number of 2003 mg g −1 , was selected for further characterization.
FTIR analysis was used for identifying surface functional groups of the activated carbon. Figure 2 represents the FT-IR spectrum of CK2T800 sample. The broad peak around 3422 cm −1 is associated to the bands of O-H group due to the vibration of water molecules. The noticeable peak at 2924 cm −1 is attributed to presence of aliphatic C-H stretch of CH, CH 2 and CH 3 groups 23 . The peak at 2854 cm −1 is assigned to the CH 2 symmetric stretching. The peak observed at 2345 cm −1 arises due to the presence of C = C groups. The peak at 2085 cm −1 is related to the C = N stretching vibrations 24 . The peak which is presented at 1622 cm −1 can also be corresponded to C = O stretching of carboxylic acids 25 . The band 1444 cm −1 assigns to C-H asymmetric and symmetric bending vibrations 26 . The weak band in the range between 900 to 1300 cm −1 may be due to the presence of C-O group in the sample. The band found at 874 cm −1 is related to the stretching vibrations of C−H out-of-plane band 27 .
The morphological structures of the oily residue sludge char and CK2T800 activated carbon were analyzed by SEM. SEM images revealed that the external surface of activated carbon was very irregular and full of cavities with different shapes and sizes comparing to that of the non-activated char. This can be a consequence of the gasification and release of volatile matter produced during the activation process (Fig. 3). EDX analysis indicated the presence of C and O in the prepared AC. The atomic percentage of C and O calculated from the quantification of the peaks, gave the value of about 81 and 6, respectively. This relatively high carbon content and the absence of alkali and alkaline earth metal indicated that the carbonization process and the washing steps were successful. phenol adsorption. Initial adsorption tests were performed with all the synthesized ACs in order to select the one with the most suitable properties for the phenol removal from aqueous phase. Figure 4 shows the amounts of phenol adsorbed (mg g −1 ) by the different ACs and their BET surface areas. It can be observed a relation between the BET values and the adsorption capacities of the activated carbons 28 . As shown in Fig. 4, the highest phenol adsorption was obtained with CK2T800 activated carbon, which is the AC with the highest surface area and more developed porous structure. Therefore, this carbon was selected for the subsequent adsorption experiments. Furthermore, it should be mentioned that the adsorption capacity of phenol on the non-activated char was only 14.28 mg g −1 , which was significantly lower than of activated carbons in agreement with its much lower porosity development. This confirms that the activation process by KOH improves the adsorption capacity of the carbon through increasing the surface area and porous structure.
Adsorption kinetic models. To understand the dynamics of adsorption process of phenol on AC and to describe the adsorption rate, the adsorption kinetic data were fitted to the pseudo-first order and pseudo-second order models. The Lagergren equation (or pseudo first order equation, Eq. 1) and pseudo second order equation (Eq. 2) are represented as follows, respectively: Figure 3. SEM micrographs and EDX analysis of (a) the prepared activated carbon at temperature of 800 °C and impregnation ratio of 2:1 (KOH to carbon) and (b) char produced from oily sludge. www.nature.com/scientificreports www.nature.com/scientificreports/ where q t and q e are the adsorbed amount (mg·g −1 ) of phenol at any time t and at equilibrium, respectively. K 1 is the rate constant for pseudo-first order model (min −1 ) and h (mg·g −1 min −1 ) is the initial adsorption rate equal to K q 2 e 2 , K 2 is the rate constant for the pseudo second order model (g mg −1 min −1 ). The values of K 1 and R 2 at different initial concentrations of phenol were obtained by plotting a graph between log (q e − q t ) vs t (Table 2). Besides, q e and k 2 were determined from the slope (1/q e ) and intercept (1/h) of the straight line of t q t versus t, passing through origin for all three different initial concentrations of phenol ( Table 2).
The experimental data are well described by both kinetic models. However, the correlation of coefficient (R 2 = 0.99) for three different initial concentrations of phenol is higher for the second-order kinetic model. The values of K 1 and K 2 decreased with increasing the initial phenol concentrations (50-200 mg L −1 ) suggesting that the chemisorption was not the only rate limiting factor, and that both physisorption and chemisorption controled the adsorption process 29,30 . Adsorption isotherm models. The empirical Freundlich isotherm model supposes that the adsorbent involves no uniform sorption and the sorption of adsorbate occurs under a heterogeneous adsorption system. In contrast, the Langmuir isotherm model describes a monolayer adsorbent containing a finite number of adsorption sites with a homogeneous surface. Temkin isotherm is more useful in describing the adsorption energy distribution on heterogeneous surface, while Dubinin-Radushkevich isotherm is an empirical adsorption model indicating that the adsorption process occurred onto both homogeneous and heterogeneous surfaces 29 . The logarithmic form of the equation of the Freundlich model is expressed as follows: where C e (mg L −1 ) is the equilibrium concentration of phenol and q e (mg g −1 ) is the amount of phenol adsorbed per gram of AC at equilibrium. K f and n are the Freundlich model constants being indicative of adsorption capacity and adsorption intensity, respectively. These values can be determined from the slope and intercept of the produced straight line of the linear plot of log C e against log (q e ), respectively. The magnitudes of K f and n present desirable sorption and easy separation of adsorbate from the aqueous solution. Also, the linearized form of the Langmuir equation is described as follows: e e m e L m where q m (mg g −1 ) and K L (L mg −1 ) are the Langmuir isotherm constants related to the maximum adsorption capacity of the sorbent (mg g −1 ) and the free energy of adsorption (L mg −1 ), respectively. These two values were calculated from the slope (1/q m ) and intercept (1/K L q m ) of the line obtained by plotting C e /q e against C e . Temkin isotherm has the linear forms as following equation:  In above equation = RT b B / t , R is the gas constant (8.314 J mol −1 K −1 ), T (°K) is the absolute temperature, b is the Temkin energy constant related to the heat of adsorption (J mol −1 ), K t is the Temkin isotherm constant (L g −1 ). The values of b, B and K t are determined from the slope and intercept of linear plot of q e versus ln C e . Dubinin-Radushkevich model were calculated from the following equation: where K DR is a constant related to the mean free energy of adsorption (mol 2 K −1 J −2 ). By plotting ln q e against ε 2 , the value of Dubinin-Radushkevich constant can be determined. All the model parameters were calculated using Origin 8.0 software. The standard errors (SE) for each parameter and the correlation coefficient (R 2 ) were used to measure the goodness-of-fit and determined the best-fitting isotherm to the experimental data, respectively (Table 3). Based on the high values of correlation coefficient (greater than 0.99), both Langmuir and Freundlich models described better phenol adsorption onto AC than Temkin and Dubinin-Radushkevich isotherms models. In this regard, Issabayeva et al 31 . described that many studies on the phenol adsorption reported good fitness of the phenol adsorption data into both isotherm models, indicating a mixed adsorption behavior. They also stated that the majority of phenol adsorption studies were in good agreement with better description of adsorption process of phenol by Freundlich isotherm. In this study, although the experimental equilibrium data showed a satisfactory fit to both models, the values of SE for each parameter obtained from Langmuir isotherm model were found to be higher than that of Freundlich isotherm model, as shown in Table 3. This suggests that the phenol adsorption process on the produced AC can be relatively heterogeneous 30 .
It is also worth mentioning that the the selected AC showed a very high phenol adsorption capacity of 434 mg g −1 . This value of q m demonstrates that the synthesized AC can be used as an effective adsorbent for the removal of phenol from aqueous solution and also probably of other water pollutants. The well-developed porous structure, microporosity and the surface chemistry of activated carbons are the most important factors affecting the adsorption of phenol onto the adsorbents 32,33 . Table 4 summarizes phenol maximum adsorption capacities and BET surface areas for different ACs reported in the literature. It reveals the high potential of the produced AC in the current study for phenol removal from aqueous solutions in comparison to the other sorbents.
Adsorption mechanism. The affinities of the adsorbent towards the components of interest can be determined from the content of surface functional groups and pH pzc . The pH pzc indicates zero net surface charge of the adsorbents that implies their electronic surface charges as well as surface oxygen complexes. At pH values less than the pH pzc (pH < pH pzc ), water donate more H + than OH − groups, therefore the adsorbent surface has a net positive charge and attracts anions.

Models Parameters
Freundlich K f mg g −1 ) (L mg   The pH pzc of the prepared AC was 7.9 that indicated its slightly basic character. The ACs with basic surface characteristics, such as the adsorbents of this study, are considered as the most appropriate and effective adsorbents for the removal of phenol due to the weak acid nature of this molecule. The adsorption of organic compounds onto the ACs mainly contribute to three types of surface-phenol interactions namely, (i) π-π dispersion interaction, (ii) the electron-donor-acceptor complex formation and (iii) the hydrogen-bonding formation 34 that may occur simultaneously. Surface oxygen complexes sites are of Bronsted type. The appearance of acidic functional groups such as carboxyl and phenolic groups, due to oxidation under aerobic conditions, causes the π electron to be removed from carbon matrix and decreases the amount of adsorbed phenol from the solutions. But the increase in the surface basicity of the activated carbon due to the fixation of surface oxygen groups of basic character during chemical activation favors the formation of electron donor acceptor complex. The existence of surface oxygen groups located at π-electron-rich regions provide the formation of mentioned complexes with aromatic rings of phenol. These Lewis basic centers predominate on the surface of adsorbents with low oxygen content. During the adsorption of phenol on activated carbon, these regions act as donor and the aromatic rings of phenol as acceptor. Phenol adsorption onto the carbons is controlled by dispersive force between π electrons 35 .
Beside the surface chemistry of activated carbon, pore structure also affects the adsorption process. The porosity of carbonaceous material has been considered as an important factor in adsorption processes of organic compounds from aqueous solutions. It has been reported that the adsorption capacity of small molecules such as phenol to the inner surface of carbon correlates with the content of micropores and BET surface area, while for mesoporous ACs, substituent group in the phenol and nature of the carbon controlled the phenol adsorption as well 35 . thermodynamics. The influence of adsorption temperature (25 to 40 °C) on the adsorption of phenol was investigated. It can be observed that the adsorption of phenol decreases with increasing temperature. Thus, the maximum adsorption was achieved at the lowest adsorption temperature (25 °C). This probably indicates a poor chemical reaction interplay between adsorbate and surface functionalities of AC 36 .
Thermodynamic parameters such as free energy change (∆G°), enthalpy (∆H°), and entropy (∆S°) were estimated according to the following equations: 37 where R is the gas constant (8.314 J·mol −1 K −1 ), T is the temperature (K), K d is the distribution coefficient. q e and C e are the equilibrium concentrations of phenol on the adsorbent (mg L −1 ) and in the solution (mg L −1 ), respectively. The values of ∆H° and ∆S° were obtained from the slope and the intercept of plot of ln K o versus 1/T for initial phenol concentrations of 50, 100 and 200 mg·L −1 . The thermodynamic parameters are given in Table 5. The negative ∆G° values were observed in all cases, indicating the spontaneous nature of phenol adsorption onto AC. In addition, the decrease in the magnitude of ∆G° at the higher temperatures showed the diminishing of the spontaneous of the process so the adsorption was not favorable at higher temperatures. Moreover, it was observed that all the values of ∆H° were negative demonstrating the exothermic nature of the adsorption interaction which was also supported by the experimental observations. The decreased randomness at the solid-solution interface during the sorption of phenol on AC is well explained by the negative value of the entropy 38 .

Removal of phenol from real industrial wastewater. Phenol adsorption capacity from industrial
wastewater was obtained 90.11 mg·g −1 . The result indicated that, as expected, the phenol removal efficiency of this AC in synthetic wastewater was greater than in real effluents. This is probably due to the pH of real effluents Temperature (°k) C 0 (mg L −1 ) q e (mg L −1 ) ln K 0 ∆G (kJ mol −1 ) ∆H (kJ mol −1 ) ∆S (J kmol −1 ) www.nature.com/scientificreports www.nature.com/scientificreports/ as well as the presence of other chemical species and impurities in textile effluent that could interfere and/or compete in the process of adsorption.
Regeneration studies. The reversibility of the adsorption process and the feasibility of adsorption and desorption behavior of the AC for the phenol removal were examined using hot water, HCl and NaOH solutions. The better regeneration performance of the NaOH solution compared to the HCl solution (Fig. 5(A)) can be explained by the fact that phenol react with NaOH, resulting in the formation of soluble salt C 6 H 5 O − Na + which facilitates the desorption of phenol from the adsorbent 39 .
The higher desorption amount obtained when using NaOH solution compared to hot water could be due to 3 simultaneous mechanisms: (1) hydrolysis of some chemical bonds between phenol hydroxyl groups (−OH) and surface oxygen groups of the adsorbent, (2) higher solubility of phenol (reactive species under alkaline conditions) than that of neutral molecules, and (3) favorable desorption of phenol in NaOH solution due to deprotonation of acidic groups of the adsorbent surface and adsorbed phenol, as well as increasing electrostatic repulsion between these molecules 40,41 . As shown in Fig. 5(B), 0.1 M NaOH was found to be the best desorbing agent for regenerating the AC. Ozkaya 42 also reported that more than 60% phenol was desorbed from the AC using 0.15 M NaOH solution.
In order to investigate the adsorbent regeneration ability, consecutive adsorption-desorption cycles were repeated four times using recycled AC as the regenerated sorbent and 0.1 M NaOH as the effective eluting agent (Fig. 5(C,D)). As can be seen, the synthesized AC maintained most of its adsorption capacity through successive adsorption-desorption cycles.

conclusion
Highly microporous activated carbons were prepared from oily sludge using two-step method under different synthesis conditions. The optimal activated carbon that obtained at activation temperature of 800 °C and impregnation ratio of 2:1 (KOH: precursor) had unique texture structure such as, high BET surface area, well-developed pore structure and high volume of micropores and total pore volume. Phenol adsorption experiments showed a maximum adsorption capacity of 434 mg g −1 . The Freundlich isotherm showed the best fit to the equilibrium experimental data. The kinetic data were fitted well to pseudo-second-order kinetic model. Since removal of phenol as a weak acid was pH dependent and better in basic medium, the oily sludge derived activated carbon with basic surface characteristic and porous structure showed effective adsorbent for phenol treatment from aqueous solution. The negative values of both free energy change and enthalpy change implied the spontaneous and exothermic nature of the adsorption process. The 0.1 M NaOH solution was the most effective desorbing agent for adsorbed phenol from loaded AC. Regenerating of AC demonstrated that the produced sorbent could be reused for phenol removal repeatedly from aqueous solution without significant loss of adsorption capacity. The results of this work demonstrated the potential use of oily sludge as a low cost precursor for production of AC with a high . To investigate the specific surface area, pore volume, and pore diameter of the obtained activated carbons, N 2 adsorption-desorption at −196 °C was performed in a Micromeritics TriStar II 3020 apparatus at sub atmospheric pressures. The micropore volume was calculated by using t-Plot micropore volume. The samples were previously degassed under vacuum at 150 °C for at least 8 h. Brunauer-Emmett-Teller (BET) method was applied to quantify the total surface areas (S BET ). The t-plot method was used to determine the micropore volume, V mic , micropore surface area (S mic ), external or non-microporous area (S ext ), and the mean micropore diameter (D mic ). The pore structure and surface morphology of the obtained AC with the highest surface area was studied by scanning electron microscopy (SEM, Tescan, VEGA III) after gold coating of the sample. www.nature.com/scientificreports www.nature.com/scientificreports/ The adsorption kinetics were done following a similar procedure at temperature of 25 °C and natural pH in three initial phenol concentrations of C 0 = 50, 100 and 200 mg L −1 . The samples were detached at predefined time spacing and phenol uptake at time t was determined using Eq. 13.
t t 0 where C t (mg L −1 ) is liquid concentration of phenol at any time.
In this study, the adsorption equilibrium data were analyzed using four types of sorption isotherm models including Langmuir, Freundlich, Temkin and Dubinin-Radushkevich. To investigate the adsorption isotherms, the experiments were conducted by adding a known amount of adsorbent (0.8 g L −1 ) to the phenol solutions with different initial concentrations of C 0 = 20-100 mg L −1 at 25 °C and equilibrium time.
Furthermore, to elucidate the effect of temperature on the adsorption, 0.8 g L −1 adsorbent dose was added to each sample contains 50 mL of phenol solution with different initial concentrations of 50, 100 and 200 mg L −1 . The suspensions were agitated at 200 rpm for 1 h in a water bath shaker at different temperatures in the range of 25 to 40 °C. Adsorbent regeneration. For the purpose of regenerating the adsorbent, various solvents including HCl, NaOH and hot water were used for the elution of phenol from the AC. The AC loaded with phenol, obtained after the adsorption process under optimized conditions, was added into 50 mL of eluent solutions. The samples were shaken (200 rpm at 25 ± 2 °C) for 3 hours. Then the effluent was analyzed to determine the concentration of desorbed phenol. The adsorption/desorption procedure of AC with the best desorbing agent were repeated in the four cycles on the same matrix to study the reusability of the AC. The percentage of desorbed phenol was calculated by the following equations: where, q des is the content of desorbed phenol (mg g −1 ) and C des (mg L −1 ) is the concentration of phenol in the solution with volume V (L) and W is AC weight (g).

Data availability
All data generated or analyzed during this study are included in this manuscript.