Introduction

Today, burning fossil fuels as the main energy source causes increases in toxic gases emotion such as CO2, CO, and NOx. These toxic gases have negative effects on the environment such as air pollution, photochemical smog, acid rain, etc.1. Therefore scientists try to solve these environmental problems using practical and cost-effective technologies to capture materials like toxic gases2. Using solid adsorbents such as covalent organic frameworks (COFs), Mg(OH)23, DEA solution4, graphene oxide (GO), sodium hydroxide (NaOH), and Chitosan is more widely applied because of low equipment corrosion, high-performance good chemical and mechanical properties5.

Graphene oxide (GO) is a layer of sp2 carbon atoms that is obtained from natural graphite through oxidation6. GO contains various oxygenated functional groups7 on its basal planes and edges, such as carboxyl, carbonyl, hydroxyl, and epoxy groups8,9. These oxygenated functional groups play a significant role in creating either electrostatic interactions or chemical bonding, GO has the potential to serve as a nano-support for biologically active agents, which can be utilized in the creation of innovative catalysts, solid adsorbent, sensors, and drug delivery systems10. Furthermore, GO has been explored as a physical adsorbent for harmful greenhouse gases like CO2, whose elevated concentrations can pose serious risks to both the environment and human health11,12. So, it is need appropriate method for the declining distribution of CO2 into the biosphere13,14.

The oxygenated functional groups of GO caused alkali in nature. Since the CO2 molecules are acidic in nature, the strong interaction between GO and CO2 molecules is formed and helps CO2 adsorption process15,16. Furthermore, due to their hydrophilic nature, GO sheets can be easily integrated into a water-based solution and can be produced as single, double, or multiple layers with exceptional stability. This character makes the both GO surface and its active sites highly effective in improving the capacity of solid adsorption17. Although pure GO has limited ability to adsorb CO2 and N2, its capacity can be enhanced by forming diverse structures.

Bio-polymers like chitin, chitosan, and Persian gum have different advantages including economical, eco-friendly, safe, and suitable approaches. In addition, bio-polymer structures usually have amine groups which enhance the alkali properties of bio-polymer18. One of the bio-polymers is Persian gum (Pg). Persian gum is secreted from Amygdalus scoparia Spach and is a non-starch hydrocolloid with an acidic polysaccharide nature. The main part of its body consists of the sugar unit of galactose with a smaller amount of arabinose19. It is considered in the pharmaceutical, food, textile industries, etc. The resulting solutions cause the medium to become viscous and have emulsifying, texturing, stabilizing, and suspending properties. This gum consists of two insoluble and soluble parts. Iran is known as a good source of Persian gum20. The insoluble part of this gum is less used, therefore, in this work, this part is used for synthesis new adsorbent for CO2 capture by incorporating with GO.

In this study, Persian gum (Pg) as a biopolymer was used as a part of the solid adsorbent for the CO2 adsorption process. By mixing Pg with graphene oxide (GO), Pg@GO was synthesized and used as a novel adsorbent for the CO2 adsorption process. Characterization of the adsorbent was studied via X-ray diffraction (XRD), Field Emission Scanning Microscopy (FESEM), Energy Dispersive X-ray (EDX), Fourier transform infrared spectroscopy (FTIR), Thermogravimetric (TGA) analysis. The response surface methodology (RSM) was used to optimize the effect of various independent parameters including temperature, pressure, and adsorbent weight on CO2 adsorption capacity. Isotherm, kinetic and thermodynamic models were applied to evaluate the behavior and performance of the adsorption process. Finally, the regeneration of the adsorbent was evaluated at optimum conditions.

Materials and methods

Materials

Sulfuric acid (H2SO4, 98%), nitric acid (HNO3, 65%), hydrochloric acid (HCl, 37%), and potassium chloride (KClO3, 99%) were collected from Merck company and used without any purification. Carbon dioxide (CO2) gas was purchased from Hmtagas company.

Synthesis GO

The study involved creating GO using a modified version of the Hummers and Offeman method, which overcomes the drawbacks of previous methods and produces high-quality and pure GO21. To prepare the GO, two different acids including sulfuric acid and nitric acid were combined in a ratio of 2:1, and graphite powder was added. The mixture was then placed in an ice bath, and potassium chlorate as an oxidation agent was slowly added. After 7 days, the solution turned green. The prepared sample as the oxidized suspension was washed with hydrochloric acid and deionized water to remove impurities and neutralize the sample. Finally, the GO was dried at 60 °C in a vacuum oven.

Preparation of the insoluble part of Persian gum

For this purpose, a colloidal suspension of Persian gum was prepared. After an overnight, the insoluble part is separated and another solution is prepared, and this process is repeated three times. Finally, the separated insoluble part was dried in an oven and its particle size was reduced by grinding and meshing20.

Synthesis of Pg@GO adsorbent

To synthesis Pg@GO as a solid adsorbent, first, 0.5 g of Persian gum was dissolved in 50 mL deionized water for 2 h at 40 °C until a homogenous solution was achieved. Then, 0.5 g GO was added to the solution and stirred over the night. Next, the solution was dried at room temperature for two weeks. Finally, the prepared sample was used as a solid adsorbent for CO2 adsorption. Figure 1 shows each step of synthesis of Pg@GO as a solid adsorbent.

Figure 1
figure 1

The steps of the synthesis of Pg@GO as a solid adsorbent.

Full size image

CO2 adsorption setup

The laboratory-scale reactor depicted in Fig. 2 was utilized for the CO2 adsorption process and consisted of four main parts: gas injection, a reactor system, an instrument for controlling CO2 pressure variations during the adsorption process, and a thermocouple for monitoring test heat. During the adsorption process, CO2 gas flowed from a high-purity capsule into the chamber, with different weights of solid adsorbent used in each run. As the adsorption process commenced, the chamber's pressure decreased due to CO2 adsorption, as the volume of both the chamber and reactor remained constant. The amount of CO2 adsorbed was calculated based on the pressure reduction. In another word, when unadulterated CO2 was implanted into the system, the adsorption operation was as takes after. The CO2 gas stream was traded from the capsule to the chamber containing the specified adsorbent inside the reactor. At that point, the adsorption started, and CO2 was ingested by the adsorbent. Due to the steady volume of the reactor and the chamber, the CO2 adsorption decreased weight inside the chamber, and concurring to the entirety of weight diminish utilizing conditions related to adsorption, the amount of CO2 adsorption (%) and the value of CO2 adsorbed were measured by Eqs. (1) and (2):

$${\text{Adsorption}}(\% ) = \left( {{\raise0.7ex\hbox{${p_{i} - p_{f} }$} \!\mathord{\left/ {\vphantom {{p_{i} - p_{f} } {P_{i} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${P_{i} }$}}} \right) \times 100$$
(1)
$$n_{{{\text{CO}}_{2} }} = {\raise0.7ex\hbox{${\left( {P_{i} - P_{f} } \right)V}$} \!\mathord{\left/ {\vphantom {{\left( {P_{i} - P_{f} } \right)V} {RTZ}}}\right.\kern-0pt} \!\lower0.7ex\hbox{${RTZ}$}}$$
(2)
Figure 2
figure 2

CO2 adsorption setup.

Full size image

where pi is initial pressure, pf is final pressure. w is the weight of the adsorbent. In addition, the adsorption capacity of adsorbent was measured by Eq. (3)22:

$$q_{e} = \left( {{\raise0.7ex\hbox{${(p_{i} - p_{{_{f} }} )VM_{{{\text{CO}}_{2} }} }$} \!\mathord{\left/ {\vphantom {{(p_{i} - p_{{_{f} }} )VM_{{{\text{CO}}_{2} }} } {RTZ}}}\right.\kern-0pt} \!\lower0.7ex\hbox{${RTZ}$}}} \right) \times 10^{3}$$
(3)

where V is the reactor volume, MCO2 is 44 g/mol, and R is the gas constant.

Characterization of adsorbent

The synthesis of Pg@GO was studied by various analysis methods to check structure, morphology, and composition. The structures and surface features of solid the adsorbent was examined using Scanning Electron Microscopy (SEM, Philips XL30 ESEM). X-ray power diffraction (XRD, Philips PW1730) was used to analyze the characteristics and recognition of a crystalline catalyst. The XRD result was recorded for a period of 2 h, covering a range of 5–70° with an increment of 0.05° and a duration of 1 s per step. The Brunauer–Emmett–Teller (BJH, BELSORP MINI II, and BEL) was utilized to identify the shapes and surface areas of non-porous adsorbents and a mesoporous solid with varying pore sizes. It means that the results of BET analysis assist in understanding adsorbent is macroporous (> 50 nm) or mesoporous (2–50 nm) or microporous (< 2 nm)23. The Fourier Transform Infrared spectroscopy (FTIR, Thermo, Avatar) provided information on the molecular vibrations of the catalysts within the 400–4000 cm-1 range. Data on the thermal stability of Pg@GO was obtained using Thermo gravimetric analysis (TGA, TA, Q600, USA) by measuring the change in weight of the adsorbent as it is heated at a constant rate.

Experimental design

To improve the process of CO2 adsorption by Pg@GO, the response surface methodology (RSM) based on the Box–Behnken design (BBD) was utilized. RSM-BBD as an experimental design statistical tool was applied to specific interactions between two independent parameters and optimized response of multiple parameter processes. Alhajabdalla et al.24 reported the main benefit of RSM-BBD. RSM-BBD ability is to analyze independent variables by performing few experimental runs than other RSM methods25. The study focused on three independent factors: pressure, time, and adsorbent weight. The dependent factor is CO2 adsorption capacity. Table 1 provides the design and range levels for each parameter. To determine the absolute errors, seventeen experiments with six duplicate runs were conducted at central points, including 16 factorial points, six central, and eight axial26. The results of these experiments are presented in Table 2.

$$q_{{{\text{CO}}2}} = \alpha_{0} + \times \sum\limits_{i = 1}^{k} {\alpha_{ii} X_{i} + \sum\limits_{i = 1}^{k} {\alpha_{ii} X_{i}^{2} + } \sum\limits_{i - 1}^{2} {\sum\limits_{j = 1 + 1}^{3} {\alpha_{ij} X_{i} X_{j} + \varepsilon } } }$$
(4)
Table 1 Experimental design levels of factors according to BBD.
Full size table
Table 2 Experimental results for CO2 adsorption capacity obtained from BBD.
Full size table

The value of qCO2, which represents the CO2 adsorbent capacity, is determined by the coefficients α0, αi, αii, and αij calculated through regression programming. These coefficients represent the constant, linear, quadratic, and interaction factors. Xi and Xj are independent variables, while ε represents error27,28.

Result and discussion

Structure properties of Pg@GO

The crystal structure of raw Persian gum (Pg), GO and Pg@GO was studied by XRD analysis and the results are shown in Fig. 3. Persian gum has board diffraction peaks at 20–42° which shows semi-crystal micro-structure of Pg@GO29. Bashir and Haripriya reported that lacking of sharp peaks in the XRD pattern of Pg showed its amorphous character29. The diffraction peak at GO appeared 2θ of 10.01°, a crystal plane with a d-value of 8.75 A° correspond to the typical diffraction peaks of GO sheets30. The XRD pattern of Pg@GO has main peaks at 9.22°, 10.20°, and 20° related to the crystalline structure of GO and Pg. The accessibility of these peaks shows that the synthesis of Pg@GO was successful, however, the intensity of GO peak at 9.22° decreases. The thermal stability of Pg@GO was determined by TGA analysis and showed three stages of the weight loss (Fig. 4). The first step at < 100 °C related to the loss of adsorbent water and moisture. It was increasing temperature from 100 to 360 °C related to the thermal decomposition of oxygen-containing groups in GO and Persian gum to CO2, CO, and H2O31,32. The third stage occurs at 360 to 1000 °C because of oxidative pyrolysis of the carbon frame work of GO after the elimination of the oxygen-containing groups33.

Figure 3
figure 3

XRD patterns for raw GO, raw Persian gum, Pg@GO.

Full size image
Figure 4
figure 4

TGA analysis for Pg@GO as a solid adsorbent.

Full size image

The morphology structure of raw GO and Pg@GO was studied by FESEM analysis, and the result is shown in Fig. 5. The flat morphology of GO is shown in Fig. 5a, and it has oxygen functional groups that cause the surface of GO seem smooth and wrinkled34. GO’s thin layers structure displays that GO has a folded and rippled wavy shape. The oxidation process causes the edges of the exfoliated GO to crumple35. The unique structure of GO induced its porous location on the sheet or between sheets36. Based on Fig. 5b, after the immobilization of Persian gum on GO, the surface of GO creates a surface remarkably rougher37. Also, as Pg has a soft, and free pores structure38, Pg@GO has low amount of pores in its structure.

Figure 5
figure 5

FESEM images for (a) raw GO, (b) Pg@GO.

Full size image

The functional groups of raw GO, Persian gum, and Pg@GO as adsorbent were determined by FTIR analysis, and its results are shown in Fig. 6. The board peaks at 3200–3400 cm−1 which can be related to –OH stretching vibration. The characteristic peaks at 1035 cm−1 and 1167 cm−1 are related to either epoxy or alkoxy (C–O), and stretching vibration of C–O, respectively. The peak at 1415 cm−1 relates to the carboxy (C–O). The peak at 1720 is related to the carboxylic acid (C=O)39. The FTIR spectra of raw Persian gum have a board peaks at 3300–3400 cm−1 showed –OH stretching vibration. The main peaks at 2927, 2926, and 2924 cm−1 are related to the asymmetric –CH2– functional group. The peak at 2855 cm−1 corresponds to the symmetric stretching vibrations of the –CH2– functional group. The peaks at 1601 and 1602 cm−1 are assigned to the asymmetrical stretching of carboxylate groups or the intramolecularly bound water. The amid I in protein has peaks at 1600 and 1700 cm−1. The stretching vibration of alcoholic groups creates peaks at 1023 and 1024 cm−1. The carbohydrate fingerprint has peaks at 1500 and 500 cm−140,41. The FTIR spectra of Pg@GO has characteristic bands of Persian gum on the solid adsorbent which show the presence of Persian gum species without any ruin after loading on GO. However, the characteristic bands of Pg@GO are displacement and become visible very weaker comparing the FTIR spectra of the raw Persian gum.

Figure 6
figure 6

FTIR spectrum for raw GO, pure Persian gum, and GO@Pg.

Full size image

Adsorbent mechanism

The results showed that in low temperatures, the adsorption of CO2 on Pg@GO was physisorption. Both GO and Persian gum have hexagonal structures with hydroxyl, epoxyl, and carbonyl groups42,43. The interaction between CO2 molecules and the delocalized π-aromatic system of GO are seen because of the presence of oxygen groups in the interlayers. Based on the different reports, the high amount of oxygen groups on Pg@GO has a positive effect on CO2 adsorption. As both Persian gum and GO have oxygen functional groups on their chemical structures, CO2 adsorption capacity was performed with increasing pressures9,11,44. Figure 7 shows the suggested mechanism for CO2 adsorption via Pg@GO.

Figure 7
figure 7

Proposed mechanism for adsorption of CO2 molecules using Pg@GO.

Full size image

Adsorption modeling

Isotherm modeling

Adsorption isotherm is used for understating how CO2 molecules as adsorbed interact with the active sites of adsorbent in a variation of gas pressures at constant temperature45,46. This interaction usually is determined by adsorption isotherm models including Langmuir, Freundlich, Dubinin–Radushkevich (D–R), Sips, and Temkin. Equation (5) shows Langmuir isotherm model. Based on Langmuir model as a simple and empirical model, the rate of adsorption and desorption of CO2 molecules on solid surface is equal. The adsorption process happens in limited specific sites and adsorb only one molecule47.

$$q_{e} = \frac{{q_{m} K_{L} P_{{CO_{2} }} }}{{1 + K_{L} P_{{CO_{2} }} }}$$
(5)

where qe is the maximum adsorption capacity (mmol g−1), qm is the amount of CO2 absorbed at equilibrium (mg g−1), and KL is constant Langmuir (bar−1). PCO2 is the equilibrium pressure of the gas. The Freundlich model is calculated based on Eq. (6). This model is used for inhomogeneous and homogenous surfaces. It can use for chemical and physical adsorption48.

$$q_{e} = K_{F} P^{{{\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 n}}\right.\kern-0pt} \!\lower0.7ex\hbox{$n$}}}}$$
(6)

Kf and is the Freundlich constant (g bar mmol−1).

Dubinin–Radushkevich model (D–R) is applied for the recognition of either chemical or physical sorption48. D–R model is calculated by Eq. (7).

$$\ln q_{e} = \ln q_{m} - \beta .\varepsilon^{2}$$
(7)

where qm shows a single-layer adsorption capacity, β and ε are the constant related to the adsorption energy (mol2 KJ−2). The sign of ε or the Polanyi potential (KJ2 mol−2) is calculated by Eq. (8):

$$\varepsilon = RT\ln \left[ {1 + \frac{1}{{P_{{CO_{2} }} }}} \right]$$
(8)

where R is gas constant (J mol−1 K−1), and T is the absolute temperature (k). Temkin isotherm model is determined by Eq. (9). According to the Temkin isotherm assumption, the adsorption heat related to the adsorption process correlated to all adsorbent molecules and reduced linearly rather than logarithmic with coverage49.

$$q_{e} = B\ln (K_{T} p)$$
(9)

In this equation, B, and KT (atm−1) are Temkin constant. By mixing Langmuir and Freundlich isotherm models, the Sips model is obtained, while the main difference between Sips and Langmuir models is ns. The ns is the heterogeneity parameter. Usually, it is lower than 1 which shows more heterogeneity of the surface of adsorbents. If ns is equal to 1, the adsorbent surface is homogenous, and the Sips model reduces to Langmuir model50. Sip isotherm model is determined by Eq. (10).

$$\begin{gathered} b_{s} = b_{0} .\exp \left( {\frac{Q}{{R.T_{0} }}} \right).\left( {\frac{{T_{0} }}{T} - 1} \right);n_{s} = n_{0} + \alpha \left( {1 - \frac{{T_{0} }}{T}} \right); \hfill \\ q_{ms} = q_{m0} .\exp (\chi (1 - \frac{T}{{T_{0} }})) \hfill \\ \end{gathered}$$
(10)

where b0 is adsorption affinity, n0 is heterogeneity coefficient, qm0 describes maximum adsorption capacity, R and T0 are gas constant and reference temperature which assumed 298 K, respectively. Two signs including χ and α are Sips constants.

Figure 8, and Table 3 display the CO2 adsorption isotherm models. The isotherm experiments were performed at constant temperature of 298 K and at pressure in range of 1–9 bar. For each isotherm model, the high value of R2 proves which isotherm model is well fitted to experimental data. According to Table 3, the Sips model has a high value of R2 and has the highest accuracy. This means that CO2 adsorption process was carried out heterogeneous and multi-layered on Pg@GO.

Figure 8
figure 8

Isotherm modelling of CO2 adsorption on Pg@GO at 298 K.

Full size image
Table 3 Parameters of isotherm models for CO2 adsorption process at 25 °C.
Full size table

Kinetic modeling

Both the adsorption rate and mechanism are calculated by adsorption kinetic51,52. In this study, four nonlinear kinetic models are used and they are listed in Table 4. The adsorption rate based on adsorption capacity is known as a first-order equation. This phenomenon is seen when the adsorption process happens in different layer by diffusion53.

$$q_{t} = q_{e} (1 - \exp ( - k_{1} .t))$$
(11)

where qt, qe, and K1 are adsorption capacity at time (mg g−1), equilibrium condition (mg g−1), and first order rate constant (min−1), respectively. Another kinetic model is a second-order kinetic model that is based on solid-phase adsorption. It shows that the rate of chemical adsorption is slow and it can control adsorption process51,54. Equation (12) is used for determining second-order kinetic model.

$$q_{t} = K_{2} \cdot q_{e}^{2} \frac{t}{{(1 + K_{2} \cdot q_{e} \cdot t)}}$$
(12)

where K2 is second order rate constant (min−1), and t is adsorption time (min). The theory of Ritchie’s second-order equation (Eq. 13) is that each adsorption process is performed on two surface sites47.

$$q_{t} = q_{e} - \frac{{q_{e} }}{{(1 + K_{2} t)}}$$
(13)
Table 4 Kinetic parameters of CO2 adsorption on Pg@GO at different temperature.
Full size table

The Elovich model is suggested for chemical adsorption process55.

$$q_{t} = \beta \cdot \log (\alpha \cdot \beta ) + \beta \cdot \log (t)$$
(14)

α and β are the initial adsorption rate (mg g−1 min−1), and the value dependent on the activation energy and amount of surface coverage (g mg−1), respectively. Parameters related to the different kinetic models that are used in this study for Pg@GO are listed in Table 4. Figure 9 shows CO2 adsorption capacity (mg/g) against time (sec). Based on Fig. 9, the highest value of R2 is related to the Elovich model which shows this model is best fitted to the empirical data, and then the second-order model has a high value of R2. Elovich's model proves that the CO2 capture process is not constant during the adsorption process and Pg@GO has a uniform surface because surface coverage increased hence adsorption rate decreased, and the uniform surface area of the adsorbent causes the active sites for CO2 adsorption are not constant.

Figure 9
figure 9

Kinetic modelling of CO2 adsorption on Pg@GO at different pressures.

Full size image

Thermodynamic modeling

The properties of the CO2 adsorption mechanism were investigated by thermodynamic factors of enthalpy change (ΔH°), change in entropy (ΔS°), and Gibbs free energy in change (ΔG°) in physical and chemical adsorption. The parameters and magnitude are very important. In other words, the negative value of ΔH° shows an exothermic reaction, in contrast, the positive value ΔH° illustrates an endothermic reaction. In addition, if ΔH° is lower than 20 kJ/mol, absolute physisorption will happen in the process, while if ΔH° is more than 40 kJ/mol chemical adsorption will accrue4,56. The value of randomness of adsorption capture in the organization at the interface of gas/solid is determined by the positive and the negative signs of ΔS°. On one hand, when ΔS° is more than zero (ΔS° > 0), the process is more random. On the other hand, if ΔS° is lower than zero (ΔS° < 0), the process is less random. Moreover, the spontaneity of the process determines with signs of the Gibbs free energy change57. If the ΔG° is more than zero (ΔG° > 0), the process is not happen and is nonspontaneous, conversely, the negative value of ΔG° (ΔG° < 0), the process is possible and spontaneous, according to Fig. 10a, CO2 adsorption process via Pg@GO was possible and spontaneity58,59. The following equations assist thermodynamic factors to be calculated. The thermal adsorption enthalpy is calculated by Van’t Hoff equation (Eq. 17), which was obtained by Eqs. (15) and (16). Kd as the distribution coefficient calculated by Eq. (18).

$$\Delta G^{0} = \Delta H^{0} - T\Delta S$$
(15)
$$\Delta G^{0} = - RT\ln K_{d}$$
(16)
$$\ln K_{d} = \frac{{\Delta S^{0} }}{R} - \frac{{\Delta H^{0} }}{RT}$$
(17)
$$K_{d} = (P_{i} - P_{e} ) \times (V/W)$$
(18)

where T and R are absolute temperature and the value of gas constant (8.314 J mol−1 K−1), respectively. When ln Kd against 1/T was plotted the values of ΔH° (slop) and ΔS° (intercept) were calculated. However, the value of ΔG° is measured by Eq. (12). Kd is the distribution coefficient value calculated by Eq. (15). Pi (bar) is the initial pressure, Pe (bar) is the reduced pressure, V (cm3), and W (g) are, the volume, and weight of the adsorbent, respectively.

Figure 10
figure 10

CO2 captures (a) Van’t Hoff plot. (b) In different temperatures.

Full size image

Based on Fig. 10b, and Table 5, temperature has a significant effect on CO2 adsorption. When temperature increases, CO2 adsorption decreases. Rashidi et al.60 reported that since the type of bonds between CO2 molecules and solid adsorbents were van der Waals as weak bonds, increasing temperature cause these bonds destroyed. As solid adsorbent faced High CO2 molecules penetration, large surface adsorption energy, and instability of adsorption CO2 molecules to the solid adsorbent surface, the desorption process by Le Chatelier’s principle is an endothermic type and suitable.

Table 5 Thermodynamic variables for CO2 adsorption on Pg@GO adsorbent.
Full size table

RSM statistical analysis

RSM-BBD was used to optimize the CO2 adsorption process on Pg@GO. Based on the BBD, the experimental design was used to extend an RSM by the Quadratic model. Three independent factors such as Adsorbent weight (A), Pressure (B), and Temperature (C), was used in the RSM model while the adsorbent capacity (Y) was used as response. The quadratic equation on the design model was shown in Eq. (19). On the below equation negative coefficient will decline the CO2 adsorption capacity, while the positive equation will increase the CO2 adsorption capacity.

$$q_{{{\text{CO}}_{2} }} = 3.04 + 0.63A + 0.78B - 0.69C + 0.30AC - 0.15BC + 0.22A^{2} + 0.047B^{2}$$
(19)

To specify the coefficients of the quadratic of the empirical results, Analysis Of Variance (ANOVA) was applied. The coefficient of determination (R2) was used to calculate the accuracy of the suggested model, and then F-test was applied to check statistical significance61,62. According to Table 6, the quadratic model has an F-value equal to 1404.32 low error probability value ((Prob > F) < 0.0001) displayed that the mathematical models can be statistically illustrating the achieved experimental results. The F-value is obtained by dividing two mean squares, which allows for the assessment of the ratio between explained variance and unexplained variance. The calculation of the p-value is dependent on the sampling distribution of the test statistic assuming the null hypothesis, the data collected from the sample, and the specific type of test conducted (lower-tailed, upper-tailed, or two-sided)63.

Table 6 ANOVA for Response Surface Quadratic model.
Full size table

Since the R2 value was 0.9994, therefore, experimental results have good agreement with the model. The significance of each factor was introduced by p-value and F-test. The smaller p-value and larger F-test show the significant effects of independent factors64. It was clear that the order of priority among those factors on the CO2 adsorption capacity is the quadratic terms of adsorbent weight (F = 3221.16), pressure (F = 4914.41), and temperature (3819.57). The interaction effect between factors, however, has less impact on the adsorption process.

The effect of independent factors

Figure 11 shows theinteraction between independent factors. Figure 11a shows the interaction between adsorbent weights (A) and pressure (b), while temperature (C) is considered constant. According to the results, by increasing the adsorbent weight to 0.4 g and pressure to 7.88 bar, the CO2 adsorption capacity was raised to 4.80 mmol/g. That is because the interaction between active sites of solid adsorbent and CO2 molecules increased significantly. Therefore, the yield of CO2 adsorption capacity improved65. The growing active sites decrease adequate space for CO2 capture, therefore, both agglomeration and closure of pores are observed66. According to the mentioned reasons 0.4 g adsorbent weight was chosen.

Figure 11
figure 11

The interaction between (a) adsorbent weight and pressure factors, (b) adsorbent weight and temperature factors, and (c) pressure and temperature factors.

Full size image

Figure 11b illustrates the interaction between adsorbent weight (A) and temperature (C). It shows that, by increasing temperature and adsorbent weight to 27 C and 0.4 g, CO2 adsorption capacity improved to 4.80 mmol/g. This phenomenon happened since CO2 adsorption is exothermic in nature, by increasing temperature the interaction between active sites and gas molecules augments, because of this molecular interaction grows and accessible and effective active sites on the surface of adsorption decline67. The highest adsorption capacity observed at lowest temperature59. In this study optimum temperature selected was 27 °C.

Figure 11c displayed the interaction between temperature (C) and pressure (B). According to the experimental results, a rising pressure factor from 1 to 7.88 bar causes CO2 capacity to improves to 4.80 mmol/g. Khajeh and Ghaemi68,69 reported that molecular movement and rate of reaction increase in high pressure. They reported that the adsorbents with different pore sizes show various behavior. Consequently, the optimum pressure was chosen at 7.88 bar.

Model variation and condition optimization

The RSM-BBD as a numerical method was applied for optimization of operating conditions in the selected range factors by considering the standard error28. The precision between the purposed solutions and actual results was determined by the optimal solutions. The optimum conditions was obtained at adsorbent weight of 0.4 g, pressure of 7.88 bar, and temperature of 27 C. Under the optimum conditions, the highest CO2 capture was 4.80 mmol/g.

Adsorbent regeneration

As an economical view, the regeneration of solid adsorbents is very important. The regeneration process is performed by three methods such as changing both of them resulting in a hybrid regeneration (VTSA/PTSA), changing the temperature (temperature swing; TSA), and changing the pressure (vacuum/pressure swing; VSA/PSA)70,71. In this study, the TSA method was used for regeneration of Pg@GO. At the end of each cycle, Pg@GO as a solid adsorbent was separated and regenerated in the oven at 40 °C for 10 h. As can be seen in Fig. 12, the adsorption capacity declined from 100 to 90% after 8 cycles. Because Pg@GO is economical and high-value adsorbent, it can be used in industrial gas adsorption applications.

Figure 12
figure 12

Recycling performance of Pg@GO for CO2 capture.

Full size image

Comparison of the absorbent results with other absorbents

Table 7 compares the adsorption capacity of Pg@GO used in this research with other research that used GO as an adsorbent for CO2 capture. According to the results, it is clear that Pg@GO can adsorb CO2 at 7.88 bar and 300 C with the maximum capacity of 4.80 mmol/g. As Pg@GO has many active sites, it can adsorb CO2 molecules in the presence of low amount of solid adsorbent (0.4 g). At the end of the CO2 adsorption process, Pg@GO recycled eight times without decreasing the adsorption capacity.

Table 7 Comparison of operation condition and CO2 adsorption capacity.
Full size table

Conclusion

Carbon dioxide is one of the greenhouse gases that is produced via burning fossil fuels, and it has a negative impact on the environment. Pg@GO was synthesized as a solid adsorbent for CO2 capture. The RSM-BBD method was used to optimize operation conditions. The maximum CO2 adsorption capacity was 4.80 mmol/g at 0.40 g adsorbent weight, 300 K, and 7.88 bar. The Freundlich isotherm model has a good agreement with experimental data. Hence, the adsorption process is heterogeneous. According to kinetic model results, the Elovich model was able to describe CO2 adsorption data because of the highest R2 value and showed the interaction between CO2 molecules and the adsorbent’s surface is chemisorption. The negative value of ΔG° in the thermodynamic study proved that the process was exothermic and spontaneous. The regeneration of Pg@GO was tested in optimum conditions. The solid adsorbent was able to be reused eight times without a significant loss on the CO2 adsorption capacity.