Abstract
This research aimed to transform tobacco that originated from smuggled cigarettes into activated carbons, and to study its application in the remediation of water contaminated by cadmium and lead. For this, a triple activation was performed (thermal – chemical H3PO4 – physical CO2), resulting in the studied activated carbon (CT H3PO4 + CO2). The carbon was characterized by its chemical composition, pHPZC, Scanning Electron Microscopy (SEM), FT-IR, Brunauer-Emmett-Teller (BET), and BJH. The influence of the pH of Cd2+ and Pb2+ solutions, and the influence of the adsorbent dose, were studied. Isotherms were constructed by linear and nonlinear models of Langmuir, Freundlich, Sips, Dubinin-Radushkevich, and Temkin. CT H3PO4 + CO2 has carbonates, hydroxyls, carboxylic, and carbonic acids as surface functional groups. The triple activation caused changes in pHPZC (from 5.40 to 9.59). The SEM revealed a carbon surface full of irregular and heterogeneous structures, spongy aspect, with an estimated Specific Surface Area (SSA) increase of 395x and pore volume of 37x. The produced activated carbon has applicability over a wide pH range (3.00–7.00), with an ideal dose estimated at 4 g L−1 for greater efficiency in removing Cd2+ and Pb2. The linear and nonlinear models reveal that metal adsorption is predominantly physical and multi-layered, with a possible reuse in new sorption cycles.
Highlights
Triple activation was performed (thermal – chemical H3PO4 – physical CO2), resulting in the CT H3PO4 + CO2 activated carbon;
CT H3PO4 + CO2 activated carbon has on its surface carbonates, hydroxyls, carboxylic, and carbonic acids, which contribute to the adsorption process;
Activation increased the Specific Surface Area 395 times and pore volume 37 times;
CT H3PO4 + CO2 activated carbon has applicability over a wide pH range (3.00–7.00), with an ideal dose of 4 g L−1 for greater efficiency in removing Cd2+ and Pb2+.
INTRODUCTION
The quality of water matrix is a notorious subject, which has been discussed internationally in recent years, especially due to the dependence of different human activities on this precious resource (Schiller et al. 2018). Sustainable Development has been discussed since 1992, in the Rio' 92 UN Conference on Environment and Development. From that conference, many others have been conducted at different periods and locations to discuss problems and try to find solutions, including the water quality concerns (De Santis & Bortone 2018). In this sense, in 2015, the United Nations (UN) launched the 2030 Agenda, which established, among the 17 Sustainable Development Goals (SDGs), the guarantee of safe drinking water and sanitation (UN 2015).
Among the main anthropogenic activities that impact on water resources, we highlight those related to the industrial and agricultural sector (Schiller et al. 2019). Among the substances released by these activities, which negatively impact water resources, the most important are toxic metals. These metals can negatively affect the integrity of human health, and cause damage to the environment as a whole (Mahmood & Malik 2014). Moreover, these substances can be bioaccumulated by organisms; that is, they have the ability to persist in different environmental compartments (Tural et al. 2016).
Due to their high toxicity, cadmium (Cd2+) and lead (Pb2+) are among the metals that cause serious damage to humans and the environment (Bassegio et al. 2019; Conradi et al. 2019). The human toxicity for Cd and Pb occurs mostly by contact with inorganic compounds released from sources such as gasoline, paints and ceramic products, batteries, pigments, plastics and agrochemicals, and many others (May et al. 2018; Kubier et al. 2019).
Cadmium (Cd) is a heavy metal and one of the most toxic metals, classified as a category I carcinogen by the International Cancer Research Agency (IARC 2018). It is widely distributed in the environment as Cd acetate, sulfate, chloride, oxide, and carbonate as a result of various anthropogenic activities related to the industrial processes of Zn, the nickel-cadmium battery, as well as the combustion of fuels that produce gases and particulate matter, tobacco smoke, use of fertilizers and paint additives (Reyes-Hinojosa et al. 2019).
More than 80% of the ingested Cd is estimated to come from cereals, mainly rice and wheat (Gonçalves Jr. et al. 2020). The absorption of Cd by the lungs through tobacco smoke is estimated to be between 10 and 50%, compared to 8% obtained from food at the gastrointestinal level (Ganguly et al. 2018). Cd can affect health in many different ways, from poisoning and anemia to bone, liver, kidney, and lung disease. Also, it can increase the risk of heart failure due to coronary artery damage, diabetes, and oxidative stress (Huang et al. 2019; Reyes-Hinojosa et al. 2019).
Lead (Pb) is a highly harmful heavy metal comprising 0.002% of the earth's crust. In addition, Pb is the second most toxic metal after As (Zulfiqar et al. 2019). The manufacture of acid batteries, insecticides containing Pb, mining, use of fuel containing Pb, printing, and so on, are the main contaminant sources of Pb (Bassegio et al. 2019).
Pb does not degrade once mobilized. As a result, significant health risks persist due to residual lead particles released into the environment decades ago (Gonçalves et al. 2014). Pb paint remains a problem due to the deterioration of old buildings and the depletion of lead-based vehicles from decades ago. Approximately 24 million households are estimated to have hazardous levels of Pb-based paint or dust, and low-income households are eight times more likely to be at risk. Pb particles from these sources and others deposit in the soil, resurfacing with rain and construction (Rocha & Trujillo 2019), often contaminating water sources through erosion or incorrect discharge of effluents.
Considering the toxicity of Cd and Pb, the World Health Organization established guideline values for Cd and Pb presence in drinking water, which are 0.003 mg L−1 for Cd and 0.01 L−1 for Pb. However, the guideline value for Pb is provisional, it is based on the treatment performance and analytical achievability (WHO 2011a, 2011b). In addition, different countries in the world have assumed maximum allowed values in drinking water for these contaminants; in Brazil, for example, these values are 0.005 mg L−1 for Cd and 0.01 L−1 for Pb (Brasil 2011a).
Aiming at the decontamination of waters containing these toxic metals, the adsorption process stands out as an alternative with great potential. The adsorption process can be used to remove high or low concentrations of pollutants, being an attractive technology for treatment and post-treatment systems (Gonçalves et al. 2016; Gonçalves Jr. et al. 2016).
In this sense, the use of commercial activated carbon demands a complex and costly production process (Schiller et al. 2020). Given this, the development of activated carbons using alternative materials and processes that result in a lower cost and high performance product becomes attractive to industry as well as to scientific research (Barbosa et al. 2014). Among the alternative materials that can be used is tobacco from the seizure of smuggled cigarettes in the western region of the state of Paraná – Brazil (Conradi et al. 2019).
The number of cigarettes seized is large and continues to grow each year (Nobrega 2019; Agência Brasil 2018), and its usual disposal is the incineration of the material (Conradi et al. 2019). However, if this disposal does not follow the appropriate and environmentally friendly parameters, it may cause environmental damage, such as soil contamination and air pollution. The disposal of this waste is not environmentally appropriate, creating environmental problems on both sides of the Brazil-Paraguay border (Observatory of the National Policy on Tobacco Control 2020). It is important to mention that in cigarettes are found more than 16 PAHs (polycyclic aromatic hydrocarbons), such as naphthalene, acenaphthene, acenaphthene, fluorene, pyrene, chrysene, and so on, in addition to pesticides and toxic metal residues from the agricultural tobacco period (Conradi et al. 2019). When these materials are disposed of in the environment, these contaminants quickly leak into the environment (Torkashvand et al. 2020). As described above, innovative technological attempts to transform cigarettes and their waste, or to give them a new and correct destination, are sorely needed, especially because these wastes today have improper disposal, causing environmental problems.
The production of an activated carbon can occurs from different activation techniques, including thermal (incomplete pyrolysis), physical (with physical agents such as CO2 and steam) or chemical (using modifying agents such as ZnCl2, NaOH, H3PO4, etc.) (Hokkanen et al. 2016). Traditionally, controlled pyrolysis of wood, carbon, and coconut shells produces commercially active carbons (Schiller et al. 2020). However, this production usually requires some polymers, all of a certain purity, which may increase the cost of production.
In the production of activated carbon, it is important to note that some characteristics are very desirable, for example, a large surface area and a high mesoporous volume generally determine a good adsorbent product (Conradi et al. 2019). These characteristics guarantee to the product an application in different processes, such as separation, discoloration, deodorization, purification, and filtration operations (Xing et al. 2019). Also, chemical composition, methods, and process conditions used during the activation, play an important role in the characteristics mentioned before. Consequently, the adsorption properties of activated carbon vary from one source of plant species to another (Reyes-Hinojosa et al. 2019).
Thus, the objective of this research was to develop tobacco activated carbons from smuggled cigarettes on the Brazil-Paraguay border using a triple activation process (thermal, physical, and chemical). In addition, this research aims to produce a new and correct destination for these hazardous wastes by generating an environmentally friendly solution by testing these new activated carbons to remove Cd2+ and Pb2+ from water.
Next, by treating a hazardous toxic waste (cigarette) and turning it into a material to be used in remediation processes such as water treatment facilities and industries, this research highlights a promising and innovative solution to two distinct problems, to mention: the possible cheapening of activated carbon production and environmentally sound disposal of smuggled cigarettes on the Brazil-Paraguay border.
MATERIAL AND METHODS
Obtaining of T in natura and development of CT H3PO4 + CO2
The tobacco used in this study came from the Brazilian Internal Revenue Service (IRS) through seizures in the western region of Paraná, South of Brazil. The tobacco in its fresh form was sent to the Laboratory of Environmental and Instrumental Chemistry of the State University of Western Paraná – UNIOESTE, campus of Marechal Cândido Rondon, for further transformation into activated carbon.
The tobacco was dried in a forced-air oven at 65 °C for a period of 24 h, ground and sieved to standardize particle size on 14 and 65 mesh sieves, obtaining particles between 0.212 and 1.40 mm. For thermal activation, the pyrolysis of the material was performed in a tubular oven (FT 1200 1Z, with internal dimension of 120 × 300 mm, model FE50RPN digital controller). The carbonization occurred only under continuous flow of inert gas N2 and absence of O2, until reaching a predetermined temperature of 500 °C for 60 min.
Subsequently, the decanted material was washed with ultrapure water (up to neutral pH) and, subsequently, taken to a lab oven for drying for 4 h at 110 °C (Conradi et al. 2019). For chemical activation, 1,000 mL of a 1 mol L−1 H3PO4 solution was added to the material in the mixing ratio of 10 g of tobacco per 100 mL of solution.
The mixture containing the chemical solution and the material to be activated was kept for 6 h in direct contact and under constant stirring at 45 °C. The resulting material was filtered, washed with ultrapure water and then oven dried for 24 h at 65 °C.
It should be noted that throughout this analytical procedure, the degree of carbonization (burn-off) was determined (Gonçalves et al. 2016). As the last step of the activation process, physical activation was conducted. For this, the carbonization occurred only under continuous flow of inert gas N2 and absence of O2, until reaching a predetermined temperature of 750 °C. After reaching 750 °C, activation occurred only under continuous flow of CO2 gas for 60 minutes. All this procedure resulted in the activated carbon named CT H3PO4 + CO2.
Characterization of CT H3PO4 + CO2 and T in natura
Activated carbon CT H3PO4 + CO2 (carbon from tobacco chemically activated by H3PO4, physically activated by CO2, and thermal activated at 500 °C) and tobacco sample (T in natura) were subjected to chemical characterization by nitroperchloric digestion (AOAC 2012), with subsequent determination of metal concentrations in its composition by FAAS (Welz & Sperling 2008).
In order to study the morphological, chemical, and physical characteristics of CT H3PO4 + CO2 and its source material (T in natura), both materials were subjected to the following analytical determinations: pH of the point of zero charge (pHPZC) as adapted from Mimura et al. (2010), Scanning Electron Microscopy – SEM, Infrared spectrum – FT-IR, BET - Brunauer et al. (1938), and BJH - Barrett et al. (1951).
Contaminated water samples
To assess the adsorption capacity of the activated carbon in contaminated water, synthetic contamination of the water samples with toxic metals was performed, which were prepared from certified standards diluted in ultrapure water. The solutions were prepared from certified Cd2+ standards with cadmium nitrate salts (Cd(NO3)2. 4H2O P.A. ≥99.0% Sigma-Aldrich), and Pb2+ with lead nitrate salts (Pb(NO3)2 P.A. ≥99.0%).
The concentrations studied were defined based on the National Council of Environment – CONAMA resolutions N° 357 of 2005, N°. 430 of 2011 (Brasil 2005,, 2011b), and N°. 2,914 of 2011 of the Ministry of Health (Brasil 2004 , 2011a), which establish the maximum permitted values of toxic metals and other contaminants in water.
Adsorption studies
In order to investigate the adsorption process, studies were carried out to evaluate the influence of adsorbent dose (CT H3PO4 + CO2 and T in natura), as well as the influence of medium pH on the removal process of Cd2+ and Pb2+.
Thus, from the 1,000 mg L−1 mono-elemental solutions, the solutions with the desired concentrations for each study were prepared, buffered to the predetermined pH values, as presented in Table 1, by addition of NaOH and HCl, both at 0.1 mol L−1.
Central matrix of the DCCR planning
Tests . | X1 . | Mass (mg) . | X2 . | pH . |
---|---|---|---|---|
1 | − 1.00 | 396.39 | − 1.00 | 3.60 |
2 | 1.00 | 1,103.61 | − 1.00 | 3.60 |
3 | − 1.00 | 396.39 | 1.00 | 6.40 |
4 | 1.00 | 1,103.61 | 1.00 | 6.40 |
5 | 0.00 | 750.00 | 0.00 | 5.00 |
6 | − 1.41 | 250.00 | 0.00 | 5.00 |
7 | 0.00 | 750.00 | 1.41 | 7.00 |
8 | 1.41 | 1,250.00 | 0.00 | 5.00 |
9 | 0.00 | 750.00 | − 1.41 | 3.00 |
10 | 0.00 | 750.00 | 0.00 | 5.00 |
11 | 0.00 | 750.00 | 0.00 | 5.00 |
12 | 0.00 | 750.00 | 0.00 | 5.00 |
Tests . | X1 . | Mass (mg) . | X2 . | pH . |
---|---|---|---|---|
1 | − 1.00 | 396.39 | − 1.00 | 3.60 |
2 | 1.00 | 1,103.61 | − 1.00 | 3.60 |
3 | − 1.00 | 396.39 | 1.00 | 6.40 |
4 | 1.00 | 1,103.61 | 1.00 | 6.40 |
5 | 0.00 | 750.00 | 0.00 | 5.00 |
6 | − 1.41 | 250.00 | 0.00 | 5.00 |
7 | 0.00 | 750.00 | 1.41 | 7.00 |
8 | 1.41 | 1,250.00 | 0.00 | 5.00 |
9 | 0.00 | 750.00 | − 1.41 | 3.00 |
10 | 0.00 | 750.00 | 0.00 | 5.00 |
11 | 0.00 | 750.00 | 0.00 | 5.00 |
12 | 0.00 | 750.00 | 0.00 | 5.00 |
Note: X1: codification for the variable mass. X2: codifications for the variable pH. Based on the study of Conradi et al. (2019).
Using a Central Rotational Compound Design (CRCD) (Barros Bruns & Scarminio 2010), the doses of adsorbents in the range of 5–25 g L−1 (250–1,250 mg in each 125 mL reactor) were evaluated, with pH variations within the range from 3.00 to 7.00 (Table 1).
To this end, 50 mL of the mono-elemental solution containing Cd2+ or Pb2+ in 125 mL conical flasks was added, with the modified adsorbent masses, which were then placed in a thermostated Dubnoff system with constant stirring at 200 rpm for 1.5 hours at 25 °C.
In equation (1), qe is the amount of adsorbed ion per g of adsorbent in equilibrium (mg g−1), m is the used mass of adsorbent (g), C0 is the initial ion concentration (mg L−1), Ce is the concentration of ion in equilibrium (mg L−1) and V is the used volume of solution (L).
In which % R is the percentage of metal removal by the adsorbent, Cf is the metal final concentration (mg L−1) and C0 is the initial concentration of metal in the solution (mg L−1).
Adsorption isotherms were also constructed. For this purpose, samples of CT H3PO4 + CO2 and T in natura adsorbents were placed in 125 mL conical flasks, and set in contact with 50 mL of Cd2+ and Pb2+ solutions. The experimental conditions for this study were: adsorbent dose 4 g L−1, pH of the contaminant solution of 5.00, constant system temperature of 25 °C, and contact time between adsorbent and adsorbate of 90 min. The concentrations of Cd2+ and Pb2+ evaluated were: 5, 30, 60, 90, 120, 150, 180, 210, 240, 270, and 300 mg L−1. The obtained results were evaluated by the linear and nonlinear models of Langmuir, Freundlich, Dubinin-Radushkevich, Sips, and Temkin.
All the aforementioned experimental routine was performed in triplicate, with certified standards and ultrapure water. Graphs and results of linear and nonlinear models were generated using Origin 8.0 software. The equations employed in this study are exhibited in Table 2.
Mathematical models used for equilibrium studies
Used adsorption models . | ||
---|---|---|
Langmuir (Langmuir 1918) (linear) | ![]() | Ce (mg L−1): concentration of the metal in the equilibrium; qe (mg g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; qmax: maximum capacity of adsorption (mg g−1); kL or qmb: adsorbent-adsorbate interaction forces (L mg−1). |
(nonlinear) | ![]() | |
Freundlich (Freundlich 1906) (linear) | ![]() | Ce (mg L1): concentration of the metal in the equilibrium; qe (mg g1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; kf [mg g−1 (mg L−1)−1/n] and n (dimensionless): Freundlich parameters. |
(nonlinear) | ![]() | |
Dubinin-Radushkevich (Dubinin e Radushkevich 1947) (linear) | ![]() | Ce (mol L−1): concentration of the metal in the equilibrium; qe (mol g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; qd (mol L−1): adsorption capacity; qmax (mol g−1): maximum capacity of adsorption; Bd (mol2J−2): coefficient related to sorption energy; ‘ε': Polanyi potential; R: universal gas constant (8.314 J mol−1 K−1); T: temperature (K). |
![]() | ||
(nonlinear) | ![]() | |
Sips (Sips 1948) (linear) | ![]() | Ce (mg L−1): concentration of the metal in the equilibrium; qe (mg g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; qmax (mg g−1) maximum capacity of adsorption; n (dimensionless) relates to the heterogeneity of adsorbent material; kL is the adsorbent-adsorbate interaction forces. |
(nonlinear) | ![]() | |
Temkin (Temkin & Pyzhev 1940) (linear) | ![]() | Ce (mg L−1): concentration of the metal in the equilibrium; qe (mg g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; At (L g−1) is the equilibrium binding constant; bt is the Temkin isotherm constant (dimensionless); R: universal gas constant (8.314 J mol−1 K−1); T: temperature at 298 K. B (J mol−1): constant related to the heat of sorption, obtained by the expression B = RT/bt. |
(nonlinear) | ![]() |
Used adsorption models . | ||
---|---|---|
Langmuir (Langmuir 1918) (linear) | ![]() | Ce (mg L−1): concentration of the metal in the equilibrium; qe (mg g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; qmax: maximum capacity of adsorption (mg g−1); kL or qmb: adsorbent-adsorbate interaction forces (L mg−1). |
(nonlinear) | ![]() | |
Freundlich (Freundlich 1906) (linear) | ![]() | Ce (mg L1): concentration of the metal in the equilibrium; qe (mg g1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; kf [mg g−1 (mg L−1)−1/n] and n (dimensionless): Freundlich parameters. |
(nonlinear) | ![]() | |
Dubinin-Radushkevich (Dubinin e Radushkevich 1947) (linear) | ![]() | Ce (mol L−1): concentration of the metal in the equilibrium; qe (mol g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; qd (mol L−1): adsorption capacity; qmax (mol g−1): maximum capacity of adsorption; Bd (mol2J−2): coefficient related to sorption energy; ‘ε': Polanyi potential; R: universal gas constant (8.314 J mol−1 K−1); T: temperature (K). |
![]() | ||
(nonlinear) | ![]() | |
Sips (Sips 1948) (linear) | ![]() | Ce (mg L−1): concentration of the metal in the equilibrium; qe (mg g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; qmax (mg g−1) maximum capacity of adsorption; n (dimensionless) relates to the heterogeneity of adsorbent material; kL is the adsorbent-adsorbate interaction forces. |
(nonlinear) | ![]() | |
Temkin (Temkin & Pyzhev 1940) (linear) | ![]() | Ce (mg L−1): concentration of the metal in the equilibrium; qe (mg g−1): amount of metal adsorbed in the equilibrium per unit of adsorbent mass; At (L g−1) is the equilibrium binding constant; bt is the Temkin isotherm constant (dimensionless); R: universal gas constant (8.314 J mol−1 K−1); T: temperature at 298 K. B (J mol−1): constant related to the heat of sorption, obtained by the expression B = RT/bt. |
(nonlinear) | ![]() |
RESULTS AND DISCUSSION
Degree of materials carbonization and characterizations of the materials
The carbonization degree of CT H3PO4 + CO2 obtained was 37.38%. According to Gonçalves et al. (2016), the precursor material as well as the activations employed directly influence the material structure and characteristics at the end of the process. Similarly, material profitability, expressed by the burn-off index, is directly related to the precursor material and the established activation method. The same author also mentions that the profitability of activated carbon after carbonization varies on average between 40 and 80%.
Characterization of CT H3PO4 + CO2 and T in natura
The results in Table 3 show a change in the concentration of chemical elements in CT H3PO4 + CO2 when compared to T in natura (Conradi et al. 2019), a decrease of P (0.84x), K (0.09x), Ca (0.56x), Mg (0.56x), Mn (0.77x) and Pb (0.07x) being observed; whereas Cu (4.70x), Zn (21.42x), and Fe (2.84x) concentrations increase.
Chemical composition of the materials T in natura and CT H3PO4 + CO2
Adsorbents . | P . | K . | Ca . | Mg . | Cu . | Zn . | Mn . | Fe . | Cd . | Pb . | Cr . |
---|---|---|---|---|---|---|---|---|---|---|---|
g Kg−1 . | mg kg−1 . | ||||||||||
T in natura | 5.92 | 44.80 | 26.10 | 5.88 | 10.00 | 19.00 | 175.67 | 283.00 | < LQ | 26.67 | < LQ |
CT H3PO4 + CO2 | 4.96 | 3.85 | 14.70 | 3.30 | 47.00 | 407.00 | 136.00 | 805.00 | 1.00 | 2.00 | 75.00 |
T in natura/activated carbon | 0.84 | 0.09 | 0.56 | 0.56 | 4.70 | 21.42 | 0.77 | 2.84 | – | 0.07 | – |
Adsorbents . | P . | K . | Ca . | Mg . | Cu . | Zn . | Mn . | Fe . | Cd . | Pb . | Cr . |
---|---|---|---|---|---|---|---|---|---|---|---|
g Kg−1 . | mg kg−1 . | ||||||||||
T in natura | 5.92 | 44.80 | 26.10 | 5.88 | 10.00 | 19.00 | 175.67 | 283.00 | < LQ | 26.67 | < LQ |
CT H3PO4 + CO2 | 4.96 | 3.85 | 14.70 | 3.30 | 47.00 | 407.00 | 136.00 | 805.00 | 1.00 | 2.00 | 75.00 |
T in natura/activated carbon | 0.84 | 0.09 | 0.56 | 0.56 | 4.70 | 21.42 | 0.77 | 2.84 | – | 0.07 | – |
Note: LQ (limits of quantification in mg L−1): K = 0.01, Ca = 0.005, Mg = 0.005, Cu = 0.005, Fe = 0.01, Mn = 0.01, Zn = 0.005, Cd = 0.005, Pb = 0.01, Cr = 0.01 (mg kg−1). Red values indicate the loss of the element concentration comparing to the precursor T in natura (by T in natura/activated carbon). Blue values indicate the increase of the element concentration comparing to the precursor T in natura (by T in natura/activated carbon). T in natura values described by Conradi et al. (2019).
Phosphoric acid acts as a cleansing agent and cleans the tars deposited in the biochar pores (Oginni et al. 2019), this satisfactorily explains the decrease of P, K, Ca, Mg, Mn and Pb concentration in the CT H3PO4 + CO2.
It is noteworthy that among the activating agents, H3PO4 is one of the most widely used for the activation of lignocellulosic biomass, because it is less dangerous for the environment, is easily recovered by washing with water, is reusable, produces less corrosion than alkaline hydroxides, and produces high yields (Negara et al. 2020). In addition, it is possible to adjust the textural properties and surface chemistry of activated carbon by precisely selecting the conditions used during the activation process with H3PO4 (Quesada-Plata et al. 2016).
Similar changes in adsorbent composition were also observed by Schwantes et al. (2018a), when performing chemical modifications with H2SO4, NaOH, and H2O2 in different vegetable biomass. According to Kan et al. (2018) the volatilization of components during material carbonization (thermal activation) is responsible for the loss of volatiles and the increase in the content of various nonvolatile elements such as metals. Then, it can explain the increase in Cu, Zn, and Fe.
In addition, activating chemical solutions used in the chemical activation step can extract or modify the composition of the final carbon. Another factor to consider, which explains the decrease in element concentration, is the washing of carbon to neutral pH, which possibly removed much of what was once solubilized in the adsorbent (Nacke et al. 2017; Schwantes et al. 2018b).
The following pHPZC values were obtained: 5.40 for T in natura (Conradi et al. 2019) and 9.59 for CT H3PO4 + CO2 (Figure 1). According to Schwantes et al. (2018b), the adsorbent, upon contact with the liquid solution with pH lower than pHPZC, will have its surface positively charged, suggesting the preference for adsorption of anions by the active sites formed, due to the protonation and deprotonation of groups on the surface materials.
Graph expressing the pHPZC for the studied adsorbents, exhibiting the predominance of surface + or – charges, depending on the pH of the solution and the pHPZC of the material.
Graph expressing the pHPZC for the studied adsorbents, exhibiting the predominance of surface + or – charges, depending on the pH of the solution and the pHPZC of the material.
Similarly, in aqueous solutions with a pH higher than pHPZC, the adsorbent surface is negatively charged, preferentially absorbing more cations than anions, a process that can be explained by the electrostatic attraction between the electrical charge generated on the adsorbent surface and the anionic group or cationic solution (Mimura et al. 2010; Tagliaferro et al. 2011).
Nacke et al. (2017) obtained different pHPZC in biosorbents from different structures of Jatropha curcas L., Bohli et al. (2015); when producing activated carbon from olive stones, obtained pHPZC much lower than those found in the present study. This shows that the great variation of the adsorbent pHPZC may depend on the material source, the process used to obtain the adsorbent, and moreover the types of adopted activations which will characterize the material and may provoke changes in its structure (Conradi et al. 2019).
The visualization of the samples obtained by SEM (Figure 2) shows variations in the surface morphology of the adsorbents. T in natura has some homogeneity in its surface, whereas the activated carbon CT H3PO4 + CO2 exhibited irregular and heterogeneous structures, spongy aspect, reminding to a lesser extent its precursor.
Scanning Electron Microscopy (SEM) for the materials T in natura (Conradi et al. 2019) in approximations of 40, 600, and 2.400 times. For CT H3PO4 + CO2 in approximations of 100, 1.000, and 5.000 times.
Scanning Electron Microscopy (SEM) for the materials T in natura (Conradi et al. 2019) in approximations of 40, 600, and 2.400 times. For CT H3PO4 + CO2 in approximations of 100, 1.000, and 5.000 times.
The characteristics observed in tobacco adsorbents are very similar to those observed by Nacke et al. (2017) and Schwantes et al. (2018a) in biosorbents from Jatropha curcas L. and cassava and pine bark. According to Nacke et al. (2017) and Conradi et al. (2019), characteristics such heterogeneity and spongy aspect may indicate good adsorption capacity.
It can be inferred from Table 4 that the FT-IR spectra observed in Figure 3 show that the main functional groups capable of acting on the adsorption of Cd2+ and Pb2+ are hydroxyl, carbonyl, carboxylic, amine, and carbonate.
Summary of selected groups in the infrared spectrum in the region between 722 and 3,449 cm
Wavelength (cm−1) . | Vibration . | Functional group . |
---|---|---|
3,449–3,382 | O-H | Bonded and unbound hydroxyl groups and water a,b |
2,361–2,356 | C = O | Different isotopic CO2 molecules formed in reactionsa,c |
1,623 | C-C | Nucleic acids (carbonyl and carboxyl groups)d |
1,445–1,400 | C-H | Lipids (fatty acids), polysaccharides (pectin), and preoteinse |
1,057–1,038 | C-O | Functional groups present in lignin (amines and hydroxyls)a,f |
874–722 | C-O | Carbonatef,g |
Wavelength (cm−1) . | Vibration . | Functional group . |
---|---|---|
3,449–3,382 | O-H | Bonded and unbound hydroxyl groups and water a,b |
2,361–2,356 | C = O | Different isotopic CO2 molecules formed in reactionsa,c |
1,623 | C-C | Nucleic acids (carbonyl and carboxyl groups)d |
1,445–1,400 | C-H | Lipids (fatty acids), polysaccharides (pectin), and preoteinse |
1,057–1,038 | C-O | Functional groups present in lignin (amines and hydroxyls)a,f |
874–722 | C-O | Carbonatef,g |
Infrared spectrum analysis from 4,000 a 400 nm−1 for T in natura and CT H3PO4 + CO2.
Infrared spectrum analysis from 4,000 a 400 nm−1 for T in natura and CT H3PO4 + CO2.
The wave lengths found near 3,400 cm−1 suggest the presence of O-H stretch vibrations due to the presence of water (Feng et al. 2011; Silverstein et al. 2014). The bands with values between 2,361 and 2,356, found for adsorbents T in natura and CT H3PO4 + CO2, indicate vibrational stretching C = O; suggesting the presence of different isotopic molecules of CO2 formed in reactions (Silverstein et al. 2014; Conradi et al. 2019).
In T in natura it is possible to observe C-C vibrational stretching, with the presence of bands at 1,623 cm−1, which can be attributed to the presence of carbonyl and carboxyl functional groups (Onorevoli et al. 2018). For CT H3PO4 + CO2 bands from 1,445 to 1,400 cm−1 were found, suggesting the presence of C-H vibrational strains, possibly due to the presence of lipids, polysaccharides, and proteins (Peng & Wu 2010).
The bands between 1,038 and 1,057 cm−1, referring to the C-O vibrational stretches, can be attributed to the presence of hydroxyls, and amines in the source material. Such a characteristic is of great importance as it may favor the interaction between the metal ion and the adsorbent surface (Silverstein et al. 2014; Chen et al. 2017).
The wavelengths between 874 cm−1 and 722 cm−1 show the presence of CO, suggesting carbonates in CT H3PO4 + CO2. It is a very positive aspect, since this group can act on retention of metal ions (He et al. 2015; Chen et al. 2017).
When comparing the infrared spectra of T in natura and CT H3PO4 + CO2, many of the groups are repeated, demonstrating that the modified material retains characteristics of its precursor, except for the formation of carbonates, exclusively in CT H3PO4 + CO2.
Table 5 shows that the surface area found for T in natura (0.27 m2 g−1) is similar to other in natura adsorbents present in the literature, as found by Penha et al. (2016) 1.13 m2 g−1 in rice husk and by Oliveira & Franca (2011) 0.46 m2 g−1 in rice bran. However, a 395-fold increase in surface area for CT H3PO4 + CO2 was observed, as well as a 37-fold increase in pore volume.
Specific surface area (SSA), pore volume, and pore diameter for tobacco adsorbents
Materials . | Parameters . | ||
---|---|---|---|
SSA (m2 g−1) . | Pore volume (cm3 g−1) . | Pore diameter (nm) . | |
T in natura | 0.27 | 0.0008 | 2.18 |
CT H3PO4 + CO2 | 106.80 | 0.0296 | 1.68 |
Relation [activated carbon/T in natura] | 395.56 | 37.00 | 0.77 |
Materials . | Parameters . | ||
---|---|---|---|
SSA (m2 g−1) . | Pore volume (cm3 g−1) . | Pore diameter (nm) . | |
T in natura | 0.27 | 0.0008 | 2.18 |
CT H3PO4 + CO2 | 106.80 | 0.0296 | 1.68 |
Relation [activated carbon/T in natura] | 395.56 | 37.00 | 0.77 |
Note: Red values indicate the decrease of the parameters comparing to the precursor T in natura (by activated carbon/T in natura); Blue values indicate the increase of the parameter comparing to the precursor T in natura (by activated carbon/T in natura). T in natura values described by Conradi et al. (2019).
Ibrahim et al. (2016) obtained similar values to those found for CT ZnCl2 when producing activated carbon produced from the aquatic plant Ulva lactuca (345.40 m2 g−1).
It is worth noting that these values are lower when compared to the values obtained for commercial activated carbon (894.00 m2 g−1) (Merk®). Bohli et al. (2015) found SSA in activated carbon from olive pits higher than commercial carbon, reaching 1,194 m2 g−1, demonstrating that the use of alternative materials in the production of activated carbon can be extremely promising in the development of new adsorbent materials.
Adsorbent T in natura (0.0008 cm3 g−1) presented pore volume similar to that observed by Penha et al. (2016) in rice husks (0.0019 cm3 g−1) and Schwantes et al. (2018a) in NaOH modified cassava bark (0.0015 cm3 g−1).
CT H3PO4 + CO2 has lower pore volume than those found by Ibrahim et al. (2016), which obtained values of 0.320 cm3 g−1 for Ulva lactuva activated carbon and 0.108 cm3 g−1 for Ulva lactuca biosorbent.
It is important to mention that the pore diameter of precursor material and CT H3PO4 + CO2 is larger than several other adsorbents of plant origin. Schwantes et al. (2018a) obtained, in a study with biosorbent of cassava peels modified with H2O2, a diameter of 1.73 nm and for NaOH modified shells of diameter 1.92 nm. Moreover, Ibrahim et al. (2016) found a pore volume of 1.11 nm and 1.85 nm for Ulva lactuva as biosorbent and activated carbon, respectively.
Thus, the results of porous structure analysis of the samples presented in Table 5 indicate that the activation process caused a decrease in pore diameter and, in general, an increase in SSA and pore volume. However, the intraparticle adsorption process of Cd2+ and Pb2+ may be hampered by reducing the pore diameter too much, since the estimated ionic radius of Cd2+ is 0.95 nm and Pb2+ is 0.118 nm (Marcus 1993).
The results show that both the precursor material and activated carbon present predominantly microporous texture, since the pore diameter found is smaller or very close to 2.00 nm (IUPAC 1985), being one of the main characteristics that favor the adsorption process (Kawahigashi 2012).
In addition, CT H3PO4 + CO2 exhibits Type I isotherm (Figure 4), whose pores are typically micropores, with the exposed surface residing almost exclusively within the micropores, which once occupied, leave little or no external surface for later adsorption (Shamsudin et al. 2019).
BET adsorption and desorption isotherms for T in natura and CT H3PO4 + CO2. T in natura values described by Conradi et al. (2019).
BET adsorption and desorption isotherms for T in natura and CT H3PO4 + CO2. T in natura values described by Conradi et al. (2019).
T in natura exhibits a type II isotherm (sigmoid isotherm) characteristic of low porosity systems. These are most often found when adsorption occurs on non-porous powders or diameters larger than micropores. The tipping point occurs near the completion of the first adsorbed monolayer (Thommes et al. 2015).
As noted in the above results, changes in adsorbent materials are evident, including increased porosity and specific surface area (Table 5), loss or gain of functional groups originally present in tobacco biomass (Figure 3 and Table 4), besides modifications in the structural morphology of CT H3PO4 + CO2 (Figure 2), which may influence the removal rate of organic and inorganic contaminants in aqueous solutions.
Adsorption studies
As can be observed in Table 6, a statistical difference at a 5% level of significance for the adsorbent doses was found, showing that the metal adsorption (Pb or Cd) depends closely on the adsorbent dose, not depending on the studied pH range (3.0–7; 0).
ANOVA of the predicted model for the influence of adsorbent doses and solution pH on the removal of Cd2+ and Pb2+ from contaminated waters
CV . | DF . | Cd2+ . | Pb2+ . | ||
---|---|---|---|---|---|
T in natura . | CT H3PO4 + CO2 . | T in natura . | CT H3PO4 + CO2 . | ||
Dose (L) | 1 | 10,658.71** | 11,500.39** | 9,041.788** | 3,160.528** |
Dose (Q) | 1 | 2,141.89** | 2,442.39** | 1,491.599** | 514.707** |
pH (L) | 1 | 0.03 | 49.43 | 1.910 | 169.297 |
pH (Q) | 1 | 5.31 | 17.55 | 0.137 | 91.999 |
Dose x pH | 1 | 0.00 | 40.75 | 6.625 | 4.803 |
Error | 6 | 51.61 | 56.72 | 21.352 | 28.511 |
Total | 11 |
CV . | DF . | Cd2+ . | Pb2+ . | ||
---|---|---|---|---|---|
T in natura . | CT H3PO4 + CO2 . | T in natura . | CT H3PO4 + CO2 . | ||
Dose (L) | 1 | 10,658.71** | 11,500.39** | 9,041.788** | 3,160.528** |
Dose (Q) | 1 | 2,141.89** | 2,442.39** | 1,491.599** | 514.707** |
pH (L) | 1 | 0.03 | 49.43 | 1.910 | 169.297 |
pH (Q) | 1 | 5.31 | 17.55 | 0.137 | 91.999 |
Dose x pH | 1 | 0.00 | 40.75 | 6.625 | 4.803 |
Error | 6 | 51.61 | 56.72 | 21.352 | 28.511 |
Total | 11 |
Note: CV, coefficient of variation. DF, degrees of freedom. ** Statistically significant at 1% of probability.
Figure 5 shows the response surfaces for the removal of Cd2+ and Pb2+ by increasing adsorbent doses versus the influence of the solution pH. The highest removal capacity of Cd2+ and Pb2+ was observed with the smallest amount of available material (4 g L−1). This is a very positive result, since it is evident that even in dosages considered low, the removal is significantly high.
Response surfaces for Cd2+ and Pb2+ removal as a function of adsorbing dose and pH effect of contaminant solution. Experimental conditions: 1.5 h stirring; 200 rpm at 25 °C; Masses of 200–1,400 mg in 50 mL, i.e. 4 g L−1–28 g−1. [T in natura; Cd2+] z = 201.36116 − 16.15196x + 0.36622x2 + 4.52439y − 0.45775y2 + 0.00071xy; Adj-R2 = 0.976 [T in natura; Pb2+] z = 259.81646 − 21.93129x + 0.49680x2 + 8.25111y − 0.86341y2 + 0.04926xy; Adj-R2 = 0.949 [CT H3PO4 + CO2; Cd2+] z = 170.79405 − 15.48180x + 0.39107x2 + 14.92367y − 0.83215y2 − 0.32237xy; Adj-R2 = 0.976 [CT H3PO4 + CO2; Pb2+] z = 64.80010 − 7.64316x + 0.17952x2 + 23.98017y − 1.90509y2 − 0.11066xy; Adj-R2 = 0.959.
Response surfaces for Cd2+ and Pb2+ removal as a function of adsorbing dose and pH effect of contaminant solution. Experimental conditions: 1.5 h stirring; 200 rpm at 25 °C; Masses of 200–1,400 mg in 50 mL, i.e. 4 g L−1–28 g−1. [T in natura; Cd2+] z = 201.36116 − 16.15196x + 0.36622x2 + 4.52439y − 0.45775y2 + 0.00071xy; Adj-R2 = 0.976 [T in natura; Pb2+] z = 259.81646 − 21.93129x + 0.49680x2 + 8.25111y − 0.86341y2 + 0.04926xy; Adj-R2 = 0.949 [CT H3PO4 + CO2; Cd2+] z = 170.79405 − 15.48180x + 0.39107x2 + 14.92367y − 0.83215y2 − 0.32237xy; Adj-R2 = 0.976 [CT H3PO4 + CO2; Pb2+] z = 64.80010 − 7.64316x + 0.17952x2 + 23.98017y − 1.90509y2 − 0.11066xy; Adj-R2 = 0.959.
It is noteworthy that, in certain cases, it is possible to observe the decrease of the contaminant removal capacity, due to the increase of the adsorbent availability, since the formation of agglomerates in the material can reduce the total surface area and therefore the number of active sites available for the process (El-Sadaawy & Abdelwahab 2014).
Another important aspect is the variation of the maximum adsorption capacity (qe) observed for each adsorbent, a fact that enhances the diversity of interactions between adsorbent and adsorbate, and its influence on contaminant removal potential (Rodriguez et al. 2018).
Excellent mathematical adjustments were obtained for the Langmuir linear model for CT H3PO4 + CO2; that is, this model represents the experimental data with high reliability due to the low values of σ (standard deviation) and high values for R2. This suggests the occurrence of adsorption on monolayers of Cd2+ and Pb2+ by CT H3PO4 + CO2.
Good mathematical adjustments were also obtained for the nonlinear Langmuir model, showing once again the occurrence of Cd2+ monolayers for T in natura, and CT H3PO4 + CO2, as well as Pb2+ monolayers observed for T in natura. When comparing these two cases, there is a great similarity of the linear and nonlinear estimation results, especially for CT H3PO4 + CO2 in the removal of Cd2+, and T in natura in the removal of Pb2+.
Estimated qm values for T in natura are higher when removing Cd2+ (qm T in natura = 135.1 mg g−1) and Pb2+ (qm T in natura = 56.8 mg g−1) when compared to the activated carbon, suggesting that for T in natura, adsorption preference occurs in monolayers. According to Araújo et al. (2018), Langmuir's model states that all sites are equivalent and each site can contain at most one adsorbate molecule, the sorting process being dependent on chemical affinity; that is, through chemical bonds between the adsorbent and adsorbate. This may be indicative of the low reuse capacity of T in natura biosorbents in the removal of Cd2+ and Pb2+.
Moreover, the more regular surface of T in natura, as observed in Figure 2, corroborates the hypothesis that the adsorption sites are more homogeneous for this natural adsorbent. This also agrees with the results obtained by Langmuir, which show the occurrence of chemical interactions between adsorbent-adsorbate and sorption in monolayers.
The kf values obtained by the linear modeling show that CT H3PO4 + CO2 has high Cd2+ adsorption capacity in multilayers (kf CT H3PO4 + CO2 = 3.2 mg g−1), which also suggests a possible predominance of physical interactions between adsorbent- adsorbate at many sites, which may be a good indication of reuse of carbons activated in more than one adsorption cycle.
This result may characterize that carbon activated by H3PO4 and CO2, interact with Cd2+ and Pb2+ in solution through physical forces, such as Van der Waals, enabling the process of material desorption and reuse (Conradi et al. 2019).
For the estimation of kf by nonlinear models, good adjustments were found only for Cd2+ adsorption by CT H3PO4 + CO2. In this case, the values of 1/n are between 0 and 0.7, which is evident in highly curved isotherms, as observed for the case cited, in Figure 6.
Nonlinear isotherms of adsorption for T in natura and CT H3PO4 + CO2 in the removal of Cd2+ and Pb2+.
Nonlinear isotherms of adsorption for T in natura and CT H3PO4 + CO2 in the removal of Cd2+ and Pb2+.
For linear modeling of D-R, the E obtained values (sorption energy) for CT H3PO4 + CO2 in Cd2+ adsorption suggest the predominance of a chemical process that governs sorption, since in this case E > 8 KJ mol−1. Similarly, adsorption of Cd2+ by T in natura and for biosorbent, and carbon in Pb2+ sorption, E < 8 KJ mol−1 is observed, suggesting physical sorption.
In contrast, when using the nonlinear Dubinin-Radushkevich model, E values slightly higher than 8 KJ mol−1 are observed, suggesting that the adsorption of Cd2+ and Pb2+ by natural adsorbents and activated carbon of tobacco is of a chemical nature.
As described by Králik (2014), chemical adsorption has the following characteristics: it is similar to a chemical bond, with high heat of adsorption (20–400 kJ mol−1), with the formation of a surface compound (monolayer). Thus, this process is irreversible because ‘desorbed compounds’ are different from adsorbed compounds; it has high activation energy (similar to a chemical reaction) and high specificity of adsorbent-adsorbate interactions.
According to Nethaji et al. (2013), when Sips parameter ns assumes values <1, the isotherm assumes a rectangular shape; if ns >2, the isotherm reaches equilibrium (qmax) at a lower metal concentration; when ns = 1, the Sips isotherm takes the form of the Langmuir isotherm, showing predominant adsorption in monolayers.
According to Table 7, regarding the use of linear model of Sips, ns = 1 values are observed for Pb2+ adsorption by T in natura, suggesting the occurrence of a surface with more homogeneous active sites, which corroborates the homogeneity observed in Figure 2 for this biosorbent, as well as results already evidenced by the Langmuir model.
Linear and nonlinear parameters of adsorption for Langmuir, Freundlich, D-R, Sips, and Temkin models for the removal of Cd2+ and Pb2+
Models . | Cd2+ . | Pb2+ . | ||||||
---|---|---|---|---|---|---|---|---|
T in natura . | σ . | CT H3PO4 + CO2 . | σ . | T in natura . | σ . | CT H3PO4 + CO2 . | σ . | |
Langmuir (linear)![]() ![]() | ||||||||
qm (mg g−1) | 135.135 | 6.21e−4 | 44.444 | 6.72e−4 | 56.818 | 0.001 | 44.053 | 0.152 |
kL (L mg−1) | 0.014 | 0.018 | 0.084 | 3.99e−2 | 0.018 | 0.043 | 0.110 | 0.002 |
RL | 0.311–0.964 | 0.129–0.899 | 0.159–0.919 | 0.128–0.898 | ||||
Adj-R2 | 0.973 | 0.995 | 0.923 | 0.984 | ||||
Freundlich (linear)![]() | ||||||||
kf[mg g−1 (mgL−1)−1/n] | 1.834 | 0.038 | 3.266 | 0.046 | 1.480 | 0.046 | 0.002 | 0.541 |
n | 1.089 | 0.031 | 1,460 | 0.035 | 1.303 | 0.030 | 0.375 | 0.413 |
Adj-R2 | 0.993 | 0.995 | 0.991 | 0.913 | ||||
D-R (linear)![]() ![]() ![]() | ||||||||
qd (mol L−1) | 0.010 | 0.127 | 0.001 | 0.049 | 0.009 | 0.175 | 1.00e−4 | 0.344 |
Bd | 0.008 | 2.74e−4 | 0.004 | 9.90e−5 | 0.009 | 3.66e−4 | 0.009 | 6.077 |
E (KJ mol−1) | 7.857 | 11.785 | 7.180 | 7.454 | ||||
Adj-R2 | 0.998 | 0.996 | 0.990 | 0.314 | ||||
Sips (linear)![]() | ||||||||
ns | 0.582 | 0.071 | 1.238 | 0.034 | 1.033 | 0.042 | 0.338 | 0.955 |
ks (L mg−1) | 0.017 | 0.21 | 0.048 | 0.108 | 0.018 | 0.139 | 0.026 | 0.317 |
Adj-R2 | 0.974 | 0.990 | 0.961 | 0.919 | ||||
Temkin (linear)![]() | ||||||||
At (L mg−1) | 0.488 | 0.815 | 0.225 | 0.815 | ||||
bt | 46.960 | 3.627 | 113.400 | 1.057 | 22.063 | 10.613 | 113.400 | 2.693 |
B (J mol−1) | 52.759 | 4.906 | 21.848 | 1.057 | 112.294 | 19.258 | 21.848 | 1.656 |
Adj-R2 | 0.977 | 0.991 | 0.974 | 0.991 | ||||
Langmuir (nonlinear)![]() ![]() | ||||||||
qm (mg g−1) | 200.278 | 101.367 | 45.861 | 2.346 | 56.980 | 17.602 | 82.892 | 24.219 |
kL (L mg−1) | 0.008 | 0.005 | 0.057 | 0.010 | 0.017 | 0.007 | 0.010 | 0.004 |
RL | 0.294–0.962 | 0.05–0.78 | 0.160–0.919 | 0.26–0.95 | ||||
Adj-R2 | 0.974 | 0.980 | 0.974 | 0.884 | ||||
Freundlich (nonlinear)![]() | ||||||||
kf[mg g−1 (mgL−1)−1/n] | 1.727 | 3.290 | 7.249 | 0.938 | 1.323 | 0.369 | 1.716 | 0.802 |
n | 1.081 | 0.078 | 2.707 | 0.227 | 1.259 | 0.147 | 1.454 | 0.241 |
Adj-R2 | 0.963 | 0.977 | 0.953 | 0.838 | ||||
D-R (nonlinear)![]() ![]() ![]() | ||||||||
qmax (mol L−1) | 0.007 | 0.001 | 0.001 | 7.51 e−5 | 0.002 | 4.91e−4 | 0.002 | 7.08 e−4 |
Bd | 0.007 | 3.63e−4 | 0.004 | 1.97 e−4 | 0.006 | 6.36e−4 | 0.006 | 9.61 e−4 |
E (KJ mol−1) | 8.265 | 11.801 | 9.098 | 9.098 | ||||
Adj-R2 | 0.991 | 0.991 | 0.977 | 0.977 | ||||
Sips (nonlinear)![]() | ||||||||
qmax | 77.806 | 22.607 | 62.275 | 8.506 | 35.914 | 12.843 | 45.512 | 3.479 |
ns | 1.275 | 0.145 | 0.653 | 0.075 | 1.240 | 0.264 | 2.459 | 0.533 |
ks (L mg−1) | 0.013 | 0.003 | 0.085 | 0.009 | 0.018 | 0.006 | 2.44 e−4 | 3.96 e-4 |
Adj-R2 | 0.983 | 0.994 | 0.974 | 0.956 | ||||
Temkin (nonlinear)![]() | ||||||||
At (L mg−1) | 0.529 | 0.066 | 2.317 | 0.877 | 0.453 | 0.066 | 0.101 | 0.010 |
bt | 231.384 | 26.271 | 368.077 | 36.141 | 373.068 | 38.168 | 136.992 | 11.685 |
B (J mol−1) | 10.708 | 6.731 | 6.641 | 18.085 | ||||
Adj-R2 | 0.901 | 0.937 | 0.950 | 0.952 |
Models . | Cd2+ . | Pb2+ . | ||||||
---|---|---|---|---|---|---|---|---|
T in natura . | σ . | CT H3PO4 + CO2 . | σ . | T in natura . | σ . | CT H3PO4 + CO2 . | σ . | |
Langmuir (linear)![]() ![]() | ||||||||
qm (mg g−1) | 135.135 | 6.21e−4 | 44.444 | 6.72e−4 | 56.818 | 0.001 | 44.053 | 0.152 |
kL (L mg−1) | 0.014 | 0.018 | 0.084 | 3.99e−2 | 0.018 | 0.043 | 0.110 | 0.002 |
RL | 0.311–0.964 | 0.129–0.899 | 0.159–0.919 | 0.128–0.898 | ||||
Adj-R2 | 0.973 | 0.995 | 0.923 | 0.984 | ||||
Freundlich (linear)![]() | ||||||||
kf[mg g−1 (mgL−1)−1/n] | 1.834 | 0.038 | 3.266 | 0.046 | 1.480 | 0.046 | 0.002 | 0.541 |
n | 1.089 | 0.031 | 1,460 | 0.035 | 1.303 | 0.030 | 0.375 | 0.413 |
Adj-R2 | 0.993 | 0.995 | 0.991 | 0.913 | ||||
D-R (linear)![]() ![]() ![]() | ||||||||
qd (mol L−1) | 0.010 | 0.127 | 0.001 | 0.049 | 0.009 | 0.175 | 1.00e−4 | 0.344 |
Bd | 0.008 | 2.74e−4 | 0.004 | 9.90e−5 | 0.009 | 3.66e−4 | 0.009 | 6.077 |
E (KJ mol−1) | 7.857 | 11.785 | 7.180 | 7.454 | ||||
Adj-R2 | 0.998 | 0.996 | 0.990 | 0.314 | ||||
Sips (linear)![]() | ||||||||
ns | 0.582 | 0.071 | 1.238 | 0.034 | 1.033 | 0.042 | 0.338 | 0.955 |
ks (L mg−1) | 0.017 | 0.21 | 0.048 | 0.108 | 0.018 | 0.139 | 0.026 | 0.317 |
Adj-R2 | 0.974 | 0.990 | 0.961 | 0.919 | ||||
Temkin (linear)![]() | ||||||||
At (L mg−1) | 0.488 | 0.815 | 0.225 | 0.815 | ||||
bt | 46.960 | 3.627 | 113.400 | 1.057 | 22.063 | 10.613 | 113.400 | 2.693 |
B (J mol−1) | 52.759 | 4.906 | 21.848 | 1.057 | 112.294 | 19.258 | 21.848 | 1.656 |
Adj-R2 | 0.977 | 0.991 | 0.974 | 0.991 | ||||
Langmuir (nonlinear)![]() ![]() | ||||||||
qm (mg g−1) | 200.278 | 101.367 | 45.861 | 2.346 | 56.980 | 17.602 | 82.892 | 24.219 |
kL (L mg−1) | 0.008 | 0.005 | 0.057 | 0.010 | 0.017 | 0.007 | 0.010 | 0.004 |
RL | 0.294–0.962 | 0.05–0.78 | 0.160–0.919 | 0.26–0.95 | ||||
Adj-R2 | 0.974 | 0.980 | 0.974 | 0.884 | ||||
Freundlich (nonlinear)![]() | ||||||||
kf[mg g−1 (mgL−1)−1/n] | 1.727 | 3.290 | 7.249 | 0.938 | 1.323 | 0.369 | 1.716 | 0.802 |
n | 1.081 | 0.078 | 2.707 | 0.227 | 1.259 | 0.147 | 1.454 | 0.241 |
Adj-R2 | 0.963 | 0.977 | 0.953 | 0.838 | ||||
D-R (nonlinear)![]() ![]() ![]() | ||||||||
qmax (mol L−1) | 0.007 | 0.001 | 0.001 | 7.51 e−5 | 0.002 | 4.91e−4 | 0.002 | 7.08 e−4 |
Bd | 0.007 | 3.63e−4 | 0.004 | 1.97 e−4 | 0.006 | 6.36e−4 | 0.006 | 9.61 e−4 |
E (KJ mol−1) | 8.265 | 11.801 | 9.098 | 9.098 | ||||
Adj-R2 | 0.991 | 0.991 | 0.977 | 0.977 | ||||
Sips (nonlinear)![]() | ||||||||
qmax | 77.806 | 22.607 | 62.275 | 8.506 | 35.914 | 12.843 | 45.512 | 3.479 |
ns | 1.275 | 0.145 | 0.653 | 0.075 | 1.240 | 0.264 | 2.459 | 0.533 |
ks (L mg−1) | 0.013 | 0.003 | 0.085 | 0.009 | 0.018 | 0.006 | 2.44 e−4 | 3.96 e-4 |
Adj-R2 | 0.983 | 0.994 | 0.974 | 0.956 | ||||
Temkin (nonlinear)![]() | ||||||||
At (L mg−1) | 0.529 | 0.066 | 2.317 | 0.877 | 0.453 | 0.066 | 0.101 | 0.010 |
bt | 231.384 | 26.271 | 368.077 | 36.141 | 373.068 | 38.168 | 136.992 | 11.685 |
B (J mol−1) | 10.708 | 6.731 | 6.641 | 18.085 | ||||
Adj-R2 | 0.901 | 0.937 | 0.950 | 0.952 |
Note: qmax: maximum adsorption capacity, kL: constant related to adsorbent/adsorbate interaction forces, adj-R2: adjusted determination coefficient, kf: related to adsorption capacity, n: related to solid heterogeneity, qd: maximum adsorption capacity, E: average sorption energy, ɛ: potential of Polanyi; Bd: coefficient related to sorption energy (Dubinin-Radushkevich); ks: is the Sips constant. At: Temkin isotherm equilibrium binding constant; bt: Temkin isotherm constant; B: Constant related to heat of sorption; σ: standard deviation. Equilibrium time = 45 minutes; pH of the contaminant solutions = 5.0.
For the non-linear Sips model, ns values ≈ 1 for T in natura are observed in Cd2+ and Pb2+ adsorption, again suggesting that in these cases the Sips model indicates the preference for adsorption in monolayers. Also noteworthy, the ns values >2 (for Pb2+ adsorption) by nonlinear Sips model for CT H3PO4 + CO2, suggesting rapid saturation of surface sites (Özer et al. 2016).
Overall, as observed in the micrographs of Figure 2, CT H3PO4 + CO2 has high heterogeneity. Due to these heterogeneous morphological and physiological characteristics, there is a possibility that there is more than one type of adsorption site interacting with the contaminant, and thus adsorption may occur in both mono and multilayers. This fact may be related to the morphological variation of the materials and the different characteristics of surface area, pore volume, and pore diameter, which can also be confirmed by the results obtained by Sips, which for CT H3PO4 + CO2 suggests the occurrence of adsorption of Cd2+ and Pb2+ in mono and multilayers.
The bt values <0 for linear and nonlinear Temkin corroborate that the process of adsorption of Cd2+ and Pb2+ by T in natura and CT H3PO4 + CO2 is exothermic (Pongener et al. 2018). In addition, B (sorption heat) values ranged from 6.73 to 112.29 J mol−1, demonstrating that the adsorbent-adsorbate interaction occurs by physical adsorption; that is, with values below 20 KJ mol−1.
According to Temkin, the heat of surface adsorption of ions decreases linearly and not logarithmically with the coating. Furthermore, according to this model, adsorption is characterized by a uniform distribution of binding energies on the adsorbent surface, encompassing interactions between adsorbent-adsorbent (Nascimento et al. 2014).
For chemical adsorption to occur, the heat of adsorption must be of the order of the heat of reaction; that is, above 20 Kcal mol−1 (Nascimento et al. 2014). In this sense, Temkin's B values (linear and nonlinear) are in the order of 1.60 cal mol−1–26.8 cal mol−1, which is far below, suggesting the predominance of physical forces in the sorption of Cd2+ and Pb2+, such as Van der Waals forces, and other intermolecular forces (Conradi et al. 2019).
When comparing the developed adsorbents (T in natura and CT H3PO4 + CO2) with results from the literature (Table 8), superiority in the removal of Cd2+ and Pb2+ is observed. T in natura obtained 135.13 mg g−1 in Cd2+ removal, considerable value when compared to other biosorbents, such as Ulva lactuva (62.50 mg g−1) (Ibrahim et al. 2016) and Jatropha curcas (229.66 mg g−1) (Nacke et al. 2017).
Comparison of Cd2+ and Pb2+ removal capacity for T in natura and CT H3PO4 to other adsorbent materials cited in the literature
Adsorbent . | qm (mg g−1) Cd2+ . | Reference . | Adsorbent . | qm (mg g−1) Pb2+ . | Reference . |
---|---|---|---|---|---|
Olive stone (CA) | 11.72 | Alslaibi et al. (2014) | Ulva lactuca (CA) | 83.30 | Ibrahim et al. (2016) |
Bamboo (CA) | 0.19 | Lo et al. (2012) | Bamboo (CA) | 0.67 | Lo et al. (2012) |
Ulva Lactuca (CA) | 84.60 | Ibrahim et al. (2016) | Olive stone (CA) | 147.52 | Bohli et al. (2015) |
Grape residues (CA) | 58.20 | Sardella et al. (2015) | Cattle bones (CA) | 32.10 | Lo et al. (2012) |
Mangosteen stone (Biosorbent) | 3.15 | Zein et al. (2010) | Ulva lactuca (Biosorbent) | 68.90 | Ibrahim et al. (2016) |
Ulva lactuca (Biosorbent) | 62.50 | Ibrahim et al. (2016) | Jatropha Curcas L. (Biosorbent) | 22.90 | Nacke et al. (2017) |
Barks of Jatropha Curcas L. (Biosorbent) | 29.66 | Nacke et al. (2017) | Barks of cassava (Biosorbent) | 42.46 | Schwantes et al. (2018a) |
Barks of pine (Biosorbent) | 10.83 | Schwantes et al. (2018b) | Activated tobacco by ZnCl2 (CA) | 84.75 | Conradi et al. (2019) |
T in natura (Biosorbent) | 135.13 | This research | T in natura (Biosorbent) | 56.81 | This research |
CT H3PO4 + CO2 (CA) | 44.44 | This research | CT H3PO4 + CO2 (CA) | 44.05 | This research |
Adsorbent . | qm (mg g−1) Cd2+ . | Reference . | Adsorbent . | qm (mg g−1) Pb2+ . | Reference . |
---|---|---|---|---|---|
Olive stone (CA) | 11.72 | Alslaibi et al. (2014) | Ulva lactuca (CA) | 83.30 | Ibrahim et al. (2016) |
Bamboo (CA) | 0.19 | Lo et al. (2012) | Bamboo (CA) | 0.67 | Lo et al. (2012) |
Ulva Lactuca (CA) | 84.60 | Ibrahim et al. (2016) | Olive stone (CA) | 147.52 | Bohli et al. (2015) |
Grape residues (CA) | 58.20 | Sardella et al. (2015) | Cattle bones (CA) | 32.10 | Lo et al. (2012) |
Mangosteen stone (Biosorbent) | 3.15 | Zein et al. (2010) | Ulva lactuca (Biosorbent) | 68.90 | Ibrahim et al. (2016) |
Ulva lactuca (Biosorbent) | 62.50 | Ibrahim et al. (2016) | Jatropha Curcas L. (Biosorbent) | 22.90 | Nacke et al. (2017) |
Barks of Jatropha Curcas L. (Biosorbent) | 29.66 | Nacke et al. (2017) | Barks of cassava (Biosorbent) | 42.46 | Schwantes et al. (2018a) |
Barks of pine (Biosorbent) | 10.83 | Schwantes et al. (2018b) | Activated tobacco by ZnCl2 (CA) | 84.75 | Conradi et al. (2019) |
T in natura (Biosorbent) | 135.13 | This research | T in natura (Biosorbent) | 56.81 | This research |
CT H3PO4 + CO2 (CA) | 44.44 | This research | CT H3PO4 + CO2 (CA) | 44.05 | This research |
Note: CA means activated carbon.
However, the tobacco material in its fresh form (in natura) presents high toxicity and toxic elements in its composition. Torkashvand et al. (2020) mentions that there are more than 16 PAHs (polycyclic aromatic hydrocarbons) such as naphthalene, acenaphthylene, acenaphthene, fluorene, pyrene, chrysene, etc. In addition to pesticides and toxic metals, residues from the tobacco growing period are found in the cigarette after burning; when they are disposed of in the environment, there is a rapid leakage of these contaminants into the environment (Conradi et al. 2019).
Moreover, when this material is subjected to pyrolysis and activation processes resulting in activated carbon, most of the aforementioned toxic materials are volatilized during the process, ensuring that the final material has a very small amount of toxic compounds (CHEN et al. 2014), which even in residual amounts are unavailable and insolubilized in the carbonaceous matrix. This findings were also confirmed by Conradi et al. (2019), performing similar chemical activations on tobacco materials. The aforementioned authors state that this procedure (triple activation on tobacco wastes) is extremely attractive for the production of activated carbon, which can be used to treat water and effluents, contaminated especially with toxic metals such as Cd2+ and Pb2+.
Estimates and the Brazilian scenario
According to the estimates from the Brazilian Institute of Public Opinion and Statistics (Ibope), described by a report from Nobrega (2019) and exhibited in Figure 7, in 2019 about 76,080 metric tonnes of illegal tobacco in the Brazilian territory were commercialized, and according to our estimates, and if all that smuggled tobacco amount was seized, could be used to produce 30,432 metric tonnes of activated carbon (AC) that could be used in the decontamination of 7,608,000 m3 of water. Thus, considering a water consumption of 200 L hab−1day−1, this could mean that the produced amount of tobacco AC could be used for water treatment systems, during one year successfully removing pollutants and supplying water for a city of about 104,219 inhabitants.
Predicted legal and illegal tobacco sales in Brazil during 2017–2019 (Nobrega 2019, based on IBOPE data), and estimative of activated carbon (AC) production and cost. Note: considering 1 US$ = R$ 5.33; experimental oven of power = 4,000 W; 120 minutes for the production of 18 g of AC with an average yield of 40%; KWh price (based on São Paulo prices in 2020) = 0.16 US$; 10% of the cost is due to reagent and operational expenses; estimate cost to produce 1 Kg of AC = 78 US$. The estimates presented in Figure 7 were obtained by the experimental laboratory conditions of this research. Estimates that consider industrial conditions (oven, pyrolysis, reagents, etc.) could be 10 or 100 times cheaper.
Predicted legal and illegal tobacco sales in Brazil during 2017–2019 (Nobrega 2019, based on IBOPE data), and estimative of activated carbon (AC) production and cost. Note: considering 1 US$ = R$ 5.33; experimental oven of power = 4,000 W; 120 minutes for the production of 18 g of AC with an average yield of 40%; KWh price (based on São Paulo prices in 2020) = 0.16 US$; 10% of the cost is due to reagent and operational expenses; estimate cost to produce 1 Kg of AC = 78 US$. The estimates presented in Figure 7 were obtained by the experimental laboratory conditions of this research. Estimates that consider industrial conditions (oven, pyrolysis, reagents, etc.) could be 10 or 100 times cheaper.
CONCLUSION
The use of chemical activation with H3PO4, with subsequent carbonization at 750 °C, in addition to the physical activation with CO2 for activated carbon production, caused significant changes in tobacco, evidencing functional surface groups such as carbonates, hydroxyls, carboxylic, and carbonic acids (FT-IR). Also, this method of production showed a significant change in pHPZC (from 5.40 to 9.59), as well as a decrease in the concentration of most chemical elements (P, K, Ca, Mg, Mn, Pb) in the activated carbon chemical composition.
It is noteworthy that the carbon produced has irregular and heterogeneous structures, a spongy aspect (SEM), with an increase of 395× SSA and 37× pore volume, which makes it extremely attractive for use as a commercial adsorbent material.
Furthermore, activated carbon modified with H3PO4 and CO2 has applicability over a wide pH range (3.00–7.00), the ideal carbon dose being 4 g L−1, the recommended dose for an efficient removal of Cd2+ and Pb2+ from solutions.
The good experimental adjustments obtained for the linear and nonlinear models of Langmuir, Freundlich, Dubinin-Radushkevich, Sips, and Temkin demonstrate that, despite the heterogeneous structure, Cd2+ and Pb2+ adsorption by carbons is predominantly physical and multilayer, apart from some exceptions, where there is predominance of chemical affinity and adsorption in monolayers, such as is found in the precursor material T in natura.
ACKNOWLEDGEMENTS
This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.