Journal of Thermal Analysis and Calorimetry

, Volume 133, Issue 1, pp 337–354 | Cite as

CO2 adsorption on activated carbon prepared from mangosteen peel

Study by adsorption calorimetry
  • Liliana Giraldo
  • Juan Carlos Moreno-Piraján


In this work, four series of activated carbon (AC) were prepared from mangosteen peel by chemical activation. The effect of the ratio of the activating agent/char using H3PO4 and KOH and the effect of temperature were correlated with the CO2 adsorption capacity. The results show that KOH AC had a higher CO2 adsorption capacity of around 19.0 mmol g−1, while those activated with H3PO4 had CO2 adsorption capacity values around 13.5 mmol g−1. In this investigation, the adsorption of CO2 was monitored by adsorption microcalorimetry, and it was found that the samples with smaller micropores generated higher enthalpy values. The plots indicate homogeneity or heterogeneity in addition to the acidity or basicity of the surface.

Graphical abstract

CO2 adsorption isotherms at 25 °C up 50 bar for all samples prepared in this research. (a) MPP and Differential enthalpies of adsorption of CO2 on the samples prepared from MP: (b) MPK.


Adsorption calorimetry Mangosteen peel Activated carbons CO2 storage 


Thermodynamics is one of the most important branches in physicochemistry because it allows for establishing the main functions that characterize the fundamental states of matter. For example, characterizing the solid state or gas is usually very interesting, as are their respective interactions in the solid–gas interphase. This type of interaction is very common when it comes to establishing the characteristics of a porous solid, whatever it may be, for example, an MOF (metallic organic framework), mesoporous silica (i.e., SBA-15), activated charcoal cloth or pillared clays. When a gas is placed in contact with a solid, a thermal effect of a normally exothermic nature is generated. A thermodynamic study of this material in sufficient detail provides information on the textural and chemical characteristics of the solid. The final applications for the solid will depend on these characteristics. This type of interaction can be studied by the measurement of adsorption heats. Evaluating with sufficient precision, these thermal effects will provide important information about the energy involved in these interactions at the level of surface processes.

In the case of catalytic-type interactions, such as any chemical reaction, the thermal effects associated with these reactions (involving enthalpy changes) can be studied by means of a technique which, although not new, has been supported by the development of instruments and software that have modernized it to such an extent that nowadays it is one of the most important methods in thermodynamics: calorimetry.

Development in both electronics and materials has generated a great diversity of calorimeters, operating on different principles, that have been used effectively for this type of research, as shown by several publications [1, 2]. Of the various types of calorimeters, adsorption calorimeters [2, 3, 4, 5, 6] are particularly noteworthy because of their wide applicability, especially for studies in chemical catalysis and solid–gas interface studies (as in this research). They offer a number of advantages illustrated by selected examples.

The calorimetry technique used in conjunction with other techniques (adsorption isotherms of N2 at – 196 °C, X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy with energy-dispersive X-ray analysis (SEM-EDX) among others can be used to adequately characterize a porous solid such as an activated carbon. There are scientific reports that, for example, differential heats of ammonia adsorption can quantitatively describe the surface acidity in small-pore zeolites such as H-ZSF15 and thus complete the information provided by infrared spectroscopy [7]. Adsorption calorimetry can provide information on the energy of the bonds, deduced from the calorimetric results, to achieve a theoretical description of the adsorbate–adsorbent bond. The study of nickel–copper alloys by the adsorption of hydrogen shows that there is a correlation between adsorption heats and surface magnetic properties. The correlation indicates that the energy of the bond between hydrogen adsorbed and nickel atoms is regulated by the electron density of the states, near the Fermi level, on the metal surface [8].

Therefore, calorimetric investigations of the adsorptions of pure reagents on generally porous solids such as activated carbons can provide information about the surface itself and its characteristics [9, 10, 11, 12].

Thermodynamics of adsorption: fundamentals

Theoretical considerations [12, 13, 14, 15, 16]

It is to be expected that the thermodynamic relationships between the free energy, enthalpy, and entropy of adsorption should be derivable.

In particular, a very useful relationship is that between the temperature variation of the adsorbate pressure and the heat of adsorption, i.e., the appropriate version of the Clausius–Clapeyron equation.

It will be recalled that the Clausius–Clapeyron equation
$${\rm d}\frac{\ln P}{{{\text{d}}T}} = \frac{\Delta H}{{RT^{2} }}$$
is a specialization of the more general relationship in a one-component system.
$$\left( {\frac{\partial P}{\partial T}} \right)_{\text{V}} = \left( {\frac{\partial S}{\partial T}} \right)_{\text{P}}$$
Very commonly, investigators in the field of adsorption have used Eq. (1) and, by analogy with Eq. (2), have used the restriction that the number of adsorbed moles, n s, be held constant. Equation (1) then becomes
$$\left( {\frac{\partial \ln P}{\partial T}} \right)_{{{\text{n}}_{\rm s} }} = \frac{{q_{\text{st}} }}{{RT^{2} }}$$
here, q st is called isosteric heat of adsorption, and the relationship of this quantity to other thermodynamic quantities has been discussed in the literature [17, 18, 19, 20, 21] and will be considered only briefly here.

This definition of isosteric heat obtained by differentiating a series of adsorption isotherms at constant loading n s is still used today. Unfortunately, Eq. 1 applies only to a pure, perfect gas and its connection with the enthalpy of the adsorbed phase and its extension to the case of real gas mixtures have led to considerable confusion. The terminology “isosteric heat” intended to mean lines of constant loading is misleading, because the evolution of a heat of adsorption requires a change in loading. The existence of other heats of adsorption such as the “differential heat of adsorption” equal to q st y RT, the “equilibrium heat of adsorption” obtained by performing the differentiation in Eq. 1 at constant spreading pressure, and others [22] adds to the confusion. Agreement within the adsorption community on the thermodynamic definition and physical meaning of the energy of adsorption is long overdue [13, 14].

Differential enthalpy and isosteric heat [15, 16]

Confusion about the meaning of isosteric heat is widespread in the adsorption literature. For example, certain textbook gives the following relation between the differential entropy and the isosteric heat (q st).
$$\Delta \bar{s}_{\text{i}}^{\text{a}} = \frac{{q_{\text{st}} }}{T} = - RT\ln \frac{{f_{\text{i}} }}{{f_{\text{i}}^{\text{o}} }}$$
Equation 4 has a sign error: The first term should be (−q st/T). Errors like this one could be avoided by replacing the ill-defined isosteric heat with the differential enthalpy of adsorption. Differential enthalpy is not a heat of adsorption, the value of which would depend upon the path. Differential enthalpy is a state function that can be measured either by calorimetry (see below) or by differentiating a series of adsorption isotherms at constant loading
$$\Delta \bar{h}_{\text{i}}^{\text{a}} = - RT^{2} \left[ {\frac{{\partial { \ln }f_{\text{i}} }}{\partial T}} \right]_{{n_{\text{i}}^{\text{a}} ,n_{\text{j}}^{\text{a}} }}$$
For a pure, perfect gas
$$\Delta \bar{h}_{\text{i}}^{\text{a}} = - RT^{2} \left[ {\frac{{\partial { \ln }P}}{\partial T}} \right]_{{n^{\text{a}} }}$$

Except for the minus sign, this equation is identical to Eq. 3 [13, 14].

The isosteric heat is a positive quantity by definition, but the differential enthalpy of adsorption is negative exothermic.

The integral enthalpy can be calculated by differentiating the surface potential Φ(T, P, y i ). According to Eq. 7, or by
$$\Delta H^{\text{a}} = \sum n_{\text{i}}^{\text{a}} \left( {\bar{h}_{\text{i}}^{\text{g}} + h_{\text{i}}^{\text{o}} } \right) - T^{2} \frac{\partial }{\partial T}\left[ {\frac{\Phi }{T}} \right]_{{P,y_{\text{i}} }}$$
Consider the process for n s moles of species X
$$n_{\text{s}} X_{{({\text{gas,P,T}})}} = n_{s} X_{{({\text{adsorbed,}}\Gamma ,{\text{T}})}}$$
The integral heat of adsorption, Q i, is then given by
$$Q_{\text{i}} = E_{\text{g}} - E_{\text{s}}$$
or per mole
$$g_{\text{i}} = E_{\text{g}} - E_{\text{s}}$$
where the small capital denotes that the quantity is on a per mole basis, and thus \(E = E/n\); the quantity g i corresponds to a calorimetric heat measured in such a way that no PV work is done, as, for example if the adsorption is allowed to occur by opening a stopcock between the adsorbent and the gas phase, both vessels being immersed in the same calorimeter.
A second experimental quantity of importance is the differential heat of adsorption, q d, which given by
$$q_{\text{d}} = \left( {\frac{\partial Q}{{\partial n_{\text{d}} }}} \right)_{\text{A,T}}$$
where A now denotes the total surface area. Alternatively, qd may be written as
$$q_{\text{d}} = \left( {\frac{{\partial E_{\text{g}} }}{{\partial n_{\text{s}} }}} \right)_{\text{A,T}} - \left( {\frac{{\partial E_{\text{s}} }}{{\partial n_{\text{s}} }}} \right)_{\text{A,T}} = \overline{{E_{\text{g}} }} - \overline{{E_{\text{s}} }}$$
where the bar over a quantity indicates that it is a partial molar one.
One usually assumes ideal gas behavior for the adsorbate, so
$$q_{\text{d}} = E_{\text{g}} - \overline{{E_{\text{s}} }}$$
also, of course
$$q_{\text{d}} + RT = H_{\text{g}} - \overline{{E_{\text{s}} }}$$
It is now necessary to consider in more detail the nature of the q obtained when a Clausius–Clapeyron-type equation, such as Eq. (3), is used. The condition for equilibrium between X in the gas phase at (P, T) and on the adsorbent at (Γ, T) is
$$\mu_{\text{g}} (P,T) = \mu_{\text{s}} (\Gamma ,T)$$
and for any small variation in conditions
$$\left( {\frac{{\partial \mu_{\text{g}} }}{\partial P}} \right)_{\text{T}} {\text{d}}P + \left( {\frac{{\partial \mu_{g} }}{\partial T}} \right)_{\text{P}} {\text{d}}T = \left( {\frac{{\partial \mu_{s} }}{\partial \Gamma }} \right)_{\text{T}} {\text{d}}\Gamma + \left( {\frac{{\partial \mu_{s} }}{\partial T}} \right)_{\Gamma } {\text{d}}T$$
For the case of Γ held constant
$$\left( {\frac{\partial P}{\partial T}} \right)_{\Gamma } = \frac{{\left[ {\left( {\frac{{\partial \mu_{\text{s}} }}{\partial T}} \right)_{\Gamma } - \left( {\frac{{\partial \mu_{\text{g}} }}{\partial T}} \right)_{\text{P}} } \right]}}{{\left( {\frac{{\partial \mu_{\text{g}} }}{\partial P}} \right)_{\text{T}} }}$$
Alternatively, μ s may be regarded as a function of π an T rather than of Γ and T, and, correspondingly, one obtains
$$\left( {\frac{\partial P}{\partial T}} \right)_{\pi } = \frac{{\left( {\frac{{\partial \mu_{\text{s}} }}{\partial T}} \right)_{\pi } - \left( {\frac{{\partial \mu_{\text{g}} }}{\partial T}} \right)_{\text{P}} }}{{\left( {\frac{{\partial \mu_{\text{s}} }}{\partial P}} \right)_{\text{T}} }}$$
For the gas phase
$$T\left( {\frac{{\partial \mu_{\text{g}} }}{\partial T}} \right)_{\text{P}} = \mu_{\text{g}} - H_{\text{g}} = TS_{\text{g}}$$
$$\left( {\frac{{\partial \mu_{\text{g}} }}{\partial P}} \right)_{\text{T}} = V_{\text{g}}$$
For the surface phase from \(\left[ {{\text{d}}E^{\text{s}} = T{\text{d}}S^{\text{s}} + \sum\limits_{\text{i}} {\mu_{\text{i}} dm_{\text{i}}^{\text{s}} + \gamma A} } \right]\):
$${\text{d}}E_{\text{s}} = T{\text{d}}S_{\text{s}} - \pi {\text{d}}A + \mu_{s} {\text{d}}n_{\text{s}}$$
$$\overline{{E_{\text{s}} }} = TS_{\text{s}} + \mu_{\text{s}}$$
$$\left( {\frac{{\partial \mu_{\text{s}} }}{\partial T}} \right)_{\text{P}} = - \overline{{S_{\text{s}} }} = \left( {\frac{1}{T}} \right)\frac{{\left( {\mu_{\text{s}} - \overline{{E_{\text{s}} }} } \right)}}{{RT^{2} }}$$
On substituting Eqs. (19) and (22) into Eq. (17) and making the usual approximation that V g = RT/P, one obtains
$$\left( {\frac{\partial \ln P}{\partial T}} \right)_{\Gamma } = \frac{{\left( {H_{\text{g}} - \overline{{E_{\text{s}} }} } \right)}}{{RT^{2} }}$$
Since the left-hand side of Eq. (22) is by convention defined as giving the isosteric heat, q st, and divided by RT 2, it follows that
$$q_{\text{st}} = H_{\text{g}} - \overline{{E_{\text{s}} }} = q_{\text{d}} + RT$$

Notice that, strictly, it is Γ and not n s that is to be held constant; the difference involves the change in total surface area with temperature and is usually neglected. In actual calorimetric measurements, the heat of adsorption is believed to lie between q d and q st since there is usually some exchange of work between portions of the gas and not all of the gas lies within the calorimeter. The difference between these two quantities, RT, again frequently is neglected.

To evaluate Eqs. (16, 19) is integrated, giving
$$E_{\text{s}} = TS_{\text{s}} - \pi A + \mu_{\text{s}} + n_{\text{s}}$$
$$E_{\text{s}} = TS_{\text{s}} - \left( {\frac{\pi }{\Gamma }} \right) + \mu_{\text{s}}$$
On differentiating Eq. (27) and comparing it with Eq. (21), one obtains
$$0 = S_{\text{s}} \, {\text{d}}T - \left( {\frac{1}{\Gamma }} \right){\text{d}}\mu + {\text{d}}n_{\text{s}}$$
$$\left( {\frac{{\partial \mu_{\text{s}} }}{\partial T}} \right)_{\pi } = - S_{\text{s}} = \left( {\frac{1}{T}} \right){\text{d}}T\left( {\mu - E - \frac{\pi }{\Gamma }} \right)$$
on inserting Eqs. (29) and (19) into Eq. (18), one obtains
$$\left( {\frac{\partial \ln P}{\partial T}} \right)_{\pi } = \left( {\frac{{H_{\text{g}} - E_{\text{s}} - \frac{\pi }{\Gamma }}}{{RT^{2} }}} \right) = \frac{{q_{\pi } }}{{RT^{2} }}$$
$$q_{\pi } = q_{\text{i}} + RT - \frac{\pi }{\Gamma }$$
The above treatment shows that, depending upon whether dlnP/dT is evaluated at constant Γ or constant π, one obtains a measure of the differential or of the integral heat of adsorption. From these, one may obtain the corresponding entropies of adsorption, since
$$H_{\text{g}} - \overline{{E_{\text{s}} }} = T\left( {S_{\text{g}} - \overline{{S_{\text{s}} }} } \right)$$
$$H_{\text{g}} - E_{\text{s}} - \frac{\pi }{\Gamma } = T(S_{\text{g}} - S_{\text{s}} )$$

The quantities q st, \(\overline{{E_{\text{s}} }}\), and \(\overline{{S_{\text{s}} }}\) are easier to obtain since all that is needed is the equilibrium pressure for a given amount of adsorption at two temperatures. On the other hand, the quantities q st, E s, and especially Ss may be the more informative since these give the average or integral values. In the application of Eq. (30), however, the film pressure, π, must be obtained by integration according to the Gibbs equation, so that it is necessary to know the complete isotherm. These procedures have been discussed by Hill et al. [18, 19, 21].

In summary, from thermodynamics, it is possible to evaluate the immersion energy. Integral functions are needed for thermodynamic calculations and are useful for the characterization of adsorbents, as shown below. The integral functions for the adsorbed phase are defined relative to the perfect-gas reference state at the same temperature
$$\Delta G^{\text{a}} = G^{\text{a}} - \sum n_{\text{i}}^{\text{a}} \mu_{\text{i}}^{\text{o}} = \Delta H^{\text{o}} - T\Delta S^{\text{o}}$$
$$\Delta H^{\text{a}} = H^{\text{a}} - \sum n_{\text{i}}^{\text{a}} h_{\text{i}}^{\text{o}}$$
$$\Delta S^{\text{a}} = S^{\text{a}} - \sum n_{\text{i}}^{\text{a}} s_{\text{i}}^{\text{o}}$$
The quantities µ i , h i, and s i refer to the molar values in the perfect-gas reference state. The integral free energy G a and enthalpy H a are measured in joules per kilogram of adsorbent. Substituting for G a in Eq. 36 from the definitions of fundamental thermodynamics gives:
$$\Delta G^{\text{a}} = \sum n_{\text{i}}^{\text{a}} \left( {\mu_{\text{i}}^{\text{g}} + \mu_{\text{i}}^{\text{o}} } \right) + \Phi$$
$$\Delta H^{\text{a}} = \sum {n_{\text{i}}^{\text{a}} \left( {\bar{h}_{\text{i}}^{\text{g}} + h_{\text{i}}^{\text{o}} } \right)} - T^{2} \frac{\partial }{\partial T}\left[ {\frac{\Phi }{T}} \right]p,y_{\text{i}} .$$
$$\Delta S^{\text{a}} = \sum n_{\text{i}}^{\text{a}} \left( {\bar{s}_{\text{i}}^{\text{g}} + s_{\text{i}}^{\text{o}} } \right) - \left[ {\frac{\partial \Phi }{\partial T}} \right]_{{{\text{P}},{\text{y}}_{\text{i}} }}$$
The equations for ∆G a, ∆H a, and ∆S a contain two parts: (1) changes for isothermal compression of the gaseous adsorbates from their perfect-gas reference states to the equilibrium pressure; (2) changes for isothermal, isobaric adsorption, and its derivatives. Considering isothermal immersion of clean, evacuated adsorbent into the compressed gas; the free energy of immersion is:
$$\Delta G^{\text{imm}} = \left( {G^{\text{a}} - \sum n_{\text{i}}^{\text{a}} \mu_{\text{i}}^{\text{g}} } \right) + \left( {G^{\text{s}} - G^{{{\text{s}}*}} } \right)$$
G imm is the change in the free energy of the condensed phase caused by adsorption, measured relative to the compressed gas and evacuated adsorbent. The first term on the right-hand side of Eq. 40 is the free energy of adsorbed gas relative to bulk-compressed gas; the second term is the free energy of the adsorbent in its standard state (T, P) relative to the evaluated state (T, in acuo). Combining Eqs. 3740 and observing that \((G^{\text{s}} {-}G^{{{\text{s}}*}} ) = PV^{\text{s}}\) [22], we obtained as follows
$$\Delta G^{\text{imm}} = \Phi + PV^{\text{s}}$$
$$\Delta H^{\text{imm}} = - T^{2} \frac{\partial }{\partial T}\left[ {\frac{\Phi }{T}} \right]_{{{\text{P,y}}_{\text{i}} }} + PV^{\text{s}}$$
$$\Delta S^{\text{imm}} = - \left[ {\frac{\partial \Phi }{\partial T}} \right]_{{{\text{P,y}}_{\text{i}} }}$$

At low pressure, the factor correction (PV s) for the enthalpy and free energy may be ignored. Thus, disregarding the (PV s) nuisance term, the surface potential is equal to the free energy of immersion. The free energy of immersion is negative because adsorption is spontaneous. The enthalpy of immersion (∆H imm) may be measured directly by calorimetry or indirectly by differentiating the surface potential according to Eq. 43. Since the heat of immersion is exothermic, ∆H imm is negative in sign. Since the adsorption process is associated with a loss of entropy, the entropy of immersion ∆S imm is also a negative quantity. The free energy and enthalpy of immersion in a pure liquid can be used to predict adsorption from liquid mixtures [15, 16].

In the scientific literature, there are articles presenting the preparation of porous solids with suitable storage characteristics for methane storage. Materials commonly used to prepare activated carbon are agricultural residues, i.e., coal, among others [23, 24, 25, 26, 27, 28, 29]. Recently, other materials such as MOFs, aerogels, xerogels, and others have been explored for methane storage [30, 31, 32, 33, 34, 35, 36]. Different orders of magnitude of CO2 adsorption capacity expressed in mmol g−1 have been reported ranging from 4 to 13 mmol g−1 using different materials, also reported at high pressures [37, 38, 39].

It is interesting to study the use of agricultural waste for the preparation of activated carbon and its application in different areas to reduce the environmental impact factor generated by these wastes. The mangosteen is a tropical fruit (Garcinia mangostana L.) known as the “queen of fruits” among other reasons for its different medical properties. India, Thailand, and Vietnam are the world’s largest producers. In Colombia (South America), this fruit has reached high levels of production, so its consumption generates a lot of waste from the husks, causing a problem in the community as they gradually ferment and release odors that affect health [25, 26, 27, 28].

It is for this reason that conducting research on the use of mangosteen peel (MP) to produce activated carbon (AC) and exploring its use in CO2 storage will give added value to these wastes and will reduce the cost of disposal [25, 26, 27, 28]. In this work, activated carbons were prepared by chemical activation from mangosteen peel (MP) by varying the chemical/char ratio (AA/char) and temperature to analyze its effect on CO2 storage capacity. In addition, the adsorption process was monitored by adsorption calorimetry to investigate the role of pore size and consistency with other properties established by conventional methods.


Raw materials

The starting material that was used in this research to prepare activated carbon was mangosteen Garcinia mangostana L. peel (MP). The material was collected at a market center in Valle del Cauca (Colombia). The MP was then subjected to thorough washing with distilled water in the laboratory in order to remove dirt and other impurities from the shells. In this step, pH control was performed in order to use it as a control parameter of washing, in such a way that the MP was washed until a constant pH was reached. The post-treatment of the MP has been widely reported in the scientific literature with a few variations [25, 26, 27, 28]. The samples were then dried at 105 °C for 48 h to remove the moisture content. The chemical reagents used in this research were analytical grade.

To the MP, a proximate and ultimate analysis according to the ASTM (ASTM D3172) standards was performed, while for the final analysis (to establish the content of C, H, N and O), an elemental analyzer (Carlo Erba model EA-1108) was used. The results (Table 1) show that MP has an adequate fixed carbon content and a low ash content, which makes it a suitable precursor for the preparation of activated carbons. The samples were crushed and ground to a size of 40 to 80 mesh and maintained in an oven at a temperature of 65 °C until use.
Table 1

Proximate and ultimate analysis of mangosteen peel (on a dry basis)

Proximate analysis


Ultimate analysis (dry basis)






Volatile matter








Fixed carbon




Using samples that had been cleaned and dried according to the procedure described in the previous step, the activated carbon samples were prepared by chemical activation. The detailed procedure for preparing activated carbons is described below. In brief, 50 g of MP was taken and placed in a quartz boat with a design such that all nitrogen was in contact with the entire mass of MP. The boat was placed in a quartz tube, and this assembly was placed in a Carbolite™ model CTF 12/TZF 12 oven. The furnace was then brought to 450 °C at a rate of 10 °C min−1, and this condition was maintained for 3 h (this variable was not modified throughout this research), and thus carbonization of the MP was obtained.

The carbonizates were treated with a 40% (mass/mass) solution of the respective activating chemicals: H3PO4 and KOH. To obtain each sample, 25 g of carbonized MP was placed in a beaker and impregnated at 75 °C for 6 h, making sure that the mass as a whole was wetted with the respective chemical reagent. After the time has elapsed, the solid was removed and activated.

The solids were activated for 180 min by varying the temperature (700, 800, or 900 °C) with a 10 °C min−1 heating ramp and a nitrogen flow of 150 mL min−1 (99.999% purity). The activated carbons were labeled as follows: MPP and MPK, corresponding to carbons activated with H3PO4 and KOH, respectively. Table 2 shows how each sample will be identified in this work depending on the preparation conditions. For illustration: sample MPP4 corresponds to an activated carbon impregnated with H3PO4 solution at an activating/carbonization ratio of 2.0 to 800 °C as the activation temperature for 180 min.
Table 2

Activation parameters and yield % and properties of the samples


Adsorbent characteristics

AA/char ratio/w/w




I2 (IN)/mg g−1









































































a The yield (%) of carbonization process was calculated according to the following formula; \({\text{Yield}}(\% ) = \left( {\frac{{W_{\text{after}} }}{{W_{\text{before}} }}} \right) \times 100\) where W before is mass of MP and W after is mass of the carbon product after carbonization

Once the activated carbons were obtained, they were washed according to the activating agent employed:
  1. a.

    Those activated with H3PO4 were washed with hot water several times until the filtrate had a constant pH.

  2. b.

    Samples activated with KOH were washed with deionized water several times and then treated with hydrochloric acid (100 mL HCl and 300 mL deionized water) to remove excess activating agent and material residues inorganic and then washed again with deionized water to a constant pH.


Characterization of activated carbons


The thermal stability of MP was evaluated by performing TGA–DTA between 30 and 900 °C with a heating ramp of 10 °C min−1 to analyze the behavior of the MP. A thermogravimetric analyzer (Hitachi TGA/SDTA Model 7200) was used. Thermal stability was measured under an N2 atmosphere with a flow rate of 100 cm3 min−1.

Textural and chemical characterization

The textural properties of the activated carbons prepared in this work were determined by recording the nitrogen adsorption–desorption isotherms at − 196 °C using an IQ2 (Quantachrome Instruments, Miami, Fl, USA) calculator and computer software to calculate the characteristics such as BET area and pore size distribution by functional density theory (both as NLDFT and QSDFT) [29, 40]. Before taking the adsorption isotherms of N2, activated carbon samples were degassed at 300 °C in an inert atmosphere for 12 h to remove moisture or adsorbed contaminants that may be pre-adsorbed on the activated carbons [41]. The nitrogen adsorption isotherm was evaluated obtained at a relative pressure P/P° for a pressure range of 10−6 < P/P° < 0.99 [42]. The BET area was estimated by the Brunauer–Emmett–Teller model [43] by the multipoint nitrogen adsorption method; as for microporous adsorbents, the linear range of the BET plot may be very difficult to locate. A useful procedure [9] allows one to overcome this difficulty and avoid any subjectivity in evaluating the BET monolayer capacity. In this research, the range 0.05 < P/P° < 0.15 was determined [41, 42, 43, 48]. The microporous surface (S mic) and the total volume (V tot) and volume of micropores (V mic) were evaluated by using the Dubinin/Astakhov (DA) equation [49, 50, 51].
$$\left( {\frac{W}{{W_{0} }}} \right) = \exp \left\{ { - \left( {\frac{R}{E}} \right)^{\text{n}} \left[ {T\ln \left( {\frac{P}{{P^{0} }}} \right)} \right]^{\text{n}} } \right\}$$

The mean pore diameter, D p, was calculated from the functional density theory (DFT) as applied to the nitrogen adsorption isotherm using the software supplied by Quantachrome Inc™. The pore size distribution was calculated by applying the Functional Theory of Solid Pore Density (DFT) [29, 42, 43, 44]. Also using the Dubinin/Astakhov model [49, 50, 51], the n parameter was evaluated in order to compare the homogeneity or heterogeneity of the micropores with the QSDFT predictions [40, 48].

The acid/base character of the surface groups of the carbon samples was determined by acid/base titration [45]. At least triplicate analyses were carried out for each carbon sample in order to obtain accurate data.

Raman spectra were collected in a Raman Jasco NRS-5100e Laser Raman Spectrometer, and SEM images were collected using a Hitachi S-3400 N scanning electron microscope following the procedure published by Fiuza et al. [45].

CO2 adsorption isotherm measurements at high pressure

Carbon dioxide adsorption experiments were carried out at 25 °C to 50 bar using high-pressure automatic adsorption equipment (HPVA II, Micromeritics). All samples were heated at 250 °C under vacuum for 6 h to remove adsorbed water and other contaminants before methane adsorption measurements. The temperature was adjusted to 20 °C for the cold volume measurement and then changed to 25 °C for hot volume measurement and for adsorption experiments. The pressure range for the adsorption was 0.15–50 bar. The pressure range for the desorption was 10−2 bar. The contribution of the empty cell was systematically measured and left to all data in order to improve accuracy. The amount of material within the sample cell was about 1000 g. The calibration of the empty space is realized with helium and the taking of each of the methane isotherms was performed according to procedures presented in the literature [46, 47, 48].

CO2 differential enthalpy measurement by adsorption calorimetry

A high-sensitivity Tian–Calvet heat flux microcalorimeter, suitable for the study of the gas–solid interactions, was used to perform the differential enthalpy measurements of CO2 adsorption on the activated carbon samples prepared in this work. This equipment has been described in the literature [13, 52, 53] and was designed and built in our laboratory. This instrument is not only less expensive than a commercial system, but is also simpler to use.

The microcalorimeter uses two calorimetric cells: one containing the sample under investigation, the other a usually empty cell, through which the gas or vapor in this case passes CO2 that acts as reference cell. These cells are designed and constructed of special stainless steel to withstand the working pressures and to avoid physical or chemical interactions between the adsorbate and the cell. Since it is a Tian–Calvet differential calorimeter, all non-measured processes are eliminated. The adsorption pressure in the measurements was monitored with two transducer type: a Baratron pressure transducers to record low pressure and Honeywell type Model HP to record high pressure. Prior to each experiment, the samples were degassed at 250 °C. Once 250 °C was reached, the sample was held at this temperature for 24 h. Subsequently, it was cooled to room temperature at the same rate. The experiments were carried out in an isothermal manner, feeding with increasing doses of CO2 on previously degassed samples (activated carbon from MP). The initial dose sent to the system corresponds to a pressure of 10 mbar. Since the initial dose is sufficiently small, the heat obtained can be considered as a differential adsorption heat. A new dose of CO2 (10 mbar) was then added, until equilibrium pressure was achieved. Subsequently, the CO2 dose was increased and the procedure was repeated until no change in pressure was observed, achieving this around 50 bar (5 MPa). The calorimeter allows for determining the heat generated by each dose added and the pressure drop in the volume of the cell which allows to obtain the amount adsorbed. For each dose, the thermal equilibrium was reached before the pressure Pi (initial pressure); the adsorbed amount δ nads and the generated integral heat were measured. The calorimetric adsorption experiments are terminated when at a relatively high pressure there is no significant heat generation, and the increase in adsorbed amount is insignificant [13, 52, 53]. The adsorption differential heat is obtained from the data obtained by microcalorimetry. The enthalpy is calculated by dividing the amount of heat by the moles adsorbed for each dose. The adsorption heats were determined by the integration of the calorimetric data and the amount of adsorbed molecules which are evaluated using each pressure value. The adsorbed amounts were expressed as mmol g−1 by gasified samples at 250 °C. The calorimetric data were reported here as differential enthalpy, ΔH diff = δQ int/δ nads. The results can also be reported as integral adsorption enthalpy, ΔH int [23, 24, 25, 26, 27]. The results can be presented as ΔH diff versus adsorbed amount of methane, ΔH diff versus pore size, or ΔH diff versus pressure.

Results and discussion

TGA analysis of mangosteen peel

Thermogravimetry is used to assess the thermal behavior of various materials, including biomass. Figure 1 shows the TGA–DTA curve of the MP. The mangosteen is classified as a lignocellulosic material composed mainly of lignin, cellulose, and hemicellulose that add up to 70%. The decomposition reactions for this type of material have been widely described in the literature [45, 54, 55, 56, 57]:
$${\text{Biomass}} \to {\text{solid}}\;{\text{residue}} + {\text{volatile}}$$
Fig. 1

Thermogravimetric differential curve of raw mangosteen peel

Our research group has already investigated these results (preparation paper). The decomposition of the MP occurred in four steps, as published in the literature [45, 54, 55]. There was an initial mass loss starting at about 85 °C and extending to a temperature of 200 °C. The change of slope is associated to a process of free water loss and to which is found unit to the matrix of the MP. The following change in the slope in the TGA is very strong and corresponds to a mass loss (which is the maximum for this material) and was observed between 200 and 335 °C; this corresponds to the rapid decomposition of cellulose and hemicellulose to volatile materials and tars [45, 56, 57]. Subsequently, mass loss was observed in the temperature range of 335–600 °C, and then a low rate of mass loss occurred at temperatures above 600 °C. The mass loss occurring above 350 °C was attributed mainly to the carbonization process of lignin and possibly to the cracking of CC bonds. The curve observation confirms the carbonization temperature chosen to obtain the carbonized MP [45, 54, 55, 56, 57]. Then some authors have suggested [45, 54] that lignin gradually decomposes to 740 °C, and at the same time the residual volatiles of the first stage are released, leaving carbon. Above 800 °C, the mass loss was small, indicating that the char structure of the material had formed at approximately this temperature [45, 57].

Effect of temperature and the ratio of activating agent/char (AA/char) of MP on textural and chemical properties

During the design and preparation of an activated carbon, it is necessary to determine among other variables the performance, BET surface, and pore volume, among other characteristics. This research investigated two widely known factors that influence the activation of this type of biomass [45, 57]: the ratio activating agent/char (AA/char) of MP, and the activation temperature, while the time was maintained at 3 h (180 min), since this constant value is in agreement with previously reported values [45, 52] for materials of the lignocellulosic type. In this way, it is possible to establish the influence of these two variables on the textural and chemical properties of the activated carbons obtained from MP, as well as their effect on the adsorption of CO2 and on the studies carried out by adsorption microcalorimetry. In the scientific literature, it is widely reported that there is a relationship between the number of iodines (IN) and the internal surface area in terms of m2 g−1 of activated carbon for various precursors [45]. The results presented in Table 2 correspond to the series of activated carbons prepared in this work activated with phosphoric acid and potassium hydroxide (MPP and KOH, respectively) which show that the IN increases as a function of the ratio AA/char; thus, MPP > MPK. The effect of the mass ratio AA/char on the textural properties of each AC increases with increasing the ratio to 1.5 and then a decrease occurs. This may be associated to the fact that, after a certain AA/char ratio, an excess of the activating agent can block part of the porous system and small traces of metals and/or salts produced during activation can block pores, generating a decrease in the properties or rupture of the porous structure.

This study was carried out by modifying the temperature (700 and 900 °C) at a fixed ratio AA/char (1.50) and at a fixed time (180 min). This result can be attributed to the fact that, at a higher temperature, the reaction rate between the char and the respective activating agent allows the liberation of more volatiles, generating more surface area and porosity.

It was observed that there was a decrease in the yield at the higher temperature, and this associated with an improvement in textural characteristics [45]. A higher temperature provides better conditions for producing carbon with higher adsorption properties. The results obtained in this work are in good agreement with studies carried out on similar materials [41, 45], including results previously obtained by our laboratory.

Activated carbon characterization

Figure 2 presents the adsorption–desorption isotherms of N2 and the pore distributions of the activated carbons prepared from MP under the experimental conditions presented in the experimental part. All isotherms were characterized by having strong adsorption at low relative pressures, forming a knee, which is characteristic of essentially microporous materials.
Fig. 2

Pore size distribution and nitrogen adsorption/desorption isotherms (inset). a Varying AA/char for MPP. b Varying AA/char for MPK (Preparation conditions: activation temperature 800 °C and activating time 180 min)

According to the latest IUPAC recommendation [48], the isotherms correspond to type I (b). These isotherms are typical of primarily microporous materials with pore size distributions in a relatively broad range of pores including wider micropores and sometimes narrow micropores (<~ 2.5 nm) [48].

As can be observed for all samples (see Fig. 2a, b) in the initial part, high nitrogen adsorption was observed, presenting concavity until reaching a limit value (a plateau) that is associated with the volume of the developed pores in each material, in particular the volume of micropores. None of the analyzed samples shows a hysteresis bubble, which shows that the prepared activated carbon samples only have micropores, which was intended in order to have samples that can present a good capacity of CO2 absorption at high pressure.

A detailed analysis of N2 isotherms at − 196 °C corresponding to the activated carbons obtained by varying the AA/char ratio from 0.5 to 2.0 (see Fig. 2a, b) (activated at 800 °C and 180 min) shows that an AA/char ratio of 1.5 in the activated carbon provided the best nitrogen adsorption capacity.

On the other hand, the adsorption graphs of N2 of the activated carbons prepared by varying the temperature (not shown here) at a fixed AA/char plot of 1.5 and an activation time of 180 min show that all activated samples treated at 900 °C adsorbed more nitrogen than those that activated at 700 °C.

Table 3 presents the results obtained for the textural and chemical characterization of the prepared samples. The S BET, S mic, and the relationship between these two variables, the pore volumes (total and micropores), the relationship between these volumes and the mean pore diameter of the prepared activated carbons, and the acidity and basicity of each of them are presented. From the results, it is clear that both H3PO4 and KOH were suitable activating agents to obtain activated carbons from MP with properties, from the point of view of the BET area, very similar to ACs obtained from other lignocellulosic residues, with MPP10 and MPK7 (with BET areas 1360 and 1270 m2 g−1, respectively) being the best.
Table 3

The effect of activation temperature and impregnation ratio on BET surface areas, pore volumes, and average pore sizes of carbons activated by H3PO4 and KOH


S BET a /m2 g−1

S mic b /m2 g−1

S mic/S BET/%

C c

V tot d /cm3 g−1

V mic e /cm3 g−1

V tot/V mic/%

N f

D p g /nm

%E h

Sites acid/mmol g−1

Sites basic/mmol g−1





























































































































































a S BET: BET specific surface area

b S mic : micropore specific surface area

c C (BET): the parameter C is exponentially related to the energy of monolayer adsorption

d V tot : total pore volume

e V mic: micropore volume

f n: exponent of the Dubinin–Astakhov equation

g D p: the mean pore (average pore diameter; average pore diameter is estimated using the DFT method)

h %E: error obtained in the fit of the proposed model (QSDFT cylinder slit)

The studies based on AA/char show that textural variables (S BET, S mic, S mic/S BET (%), V tot (cm3 g−1), V mic (cm3 g−1), and V mic/V tot increased with the AA/char ratio. For example, for samples MPP, a maximum value of surface area reached in MPP3 samples (S BET = 1340 m2 g−1) and then diminished in MPP4 sample (S BET = 1067 m2 g−1) where the ratio AA/char increases from 1.5 to 2.0. This same behavior was observed in the MPK samples. In the scientific literature, AA has been widely explained as a function of the carbonate ratio [43, 44, 45, 49] and in particular the effect of excess chemical agents. It has been suggested that the activating agent (AA) increases the rate of reaction between this and the starting material, accelerated by increasing porosity generation. When an excessive amount of an AA is added, the values of the textural properties decrease, due to the fact that there is an additional reaction between the AA and the carbon of the already formed microporous structure. That is, the addition of an excessive amount of AA can destroy the microporous structure and enlarge the volume of pores.

The pore size distribution (Fig. 2) shows that all AAs generated activated carbons from microporous MPs according to the IUPAC classification [48], i.e., around 2 nm (Table 3).

The pore diameter for each series increased as a function of the AA/char ratio for all activating agents. When performing the studies in function of the temperature (where the AA/char and time were kept constant), the values of the textural properties increased.

Table 3 also shows the values corresponding to the parameter C of the BET equation are related to the adsorbate–adsorbent interaction strength, and so to the heat of adsorption. These values of C (BET) have values above 100 (between 185.3 and 345.9), so the combination of linearity and C value suggests that the validity and reliability of the S BET data are acceptable in the range used for its evaluation. Activated carbons usually present heterogenous microporous structures [58, 59, 60]. The heterogeneity of the micropores may be assessed in principle using the exponent of the Dubinin/Astakhov equation (n), which indicates the degree of heterogeneity of a microporous system [50, 51, 58, 59, 60]. The heterogeneity in activated carbons is related to the carbonatic precursor and to the activation process, such that the narrower the micropore size range, the more homogenous the pore system. The value of n for carbonaceous adsorbents is usually found in the range of 1–4, namely, n > 2 for carbon molecular sieves or carbonaceous solids with homogenous micropores. In contrast, for heterogenous microporous carbon structures, n < 2 [45, 60, 61]. Thus, the results in Table 3 indicate that chemical activation using H3PO4 developed a homogenous micropore system, with n ≈ 3.5–4.5, while the activation with KOH led to a heterogenous micropore system, between n ≈ 1.4 and 1.9.

Porosity can be studied and analyzed by applying the thermodynamic principles on which the NLDFT and QSDFT models are based, which were used in this work to calculate the homogeneity and roughness of the materials using IQ2™ software (Quantachrome Inc. Boynton Beach, USA). The results show that the function that best describes the surface of the carbons is QSDFT (the results are shown in Table 3), which allowed us to conclude that the samples were rough and heterogeneous, in this case using a cylinder-slit pore model, which have low errors, whose ranges are between 0.02 and 0.27%. This aspect is interesting considering the results obtained using the Dubinin/Astakhov model.

CO2 sorption performance of MP

The adsorption isotherms of CO2 taken at 25 °C for the different AA/char ratios are shown in Fig. 3.
Fig. 3

CO2 adsorption isotherms at 25 °C up 50 bar for all samples prepared in this research. a MPP, b MPK

The results show that with increasing CO2 pressure, the adsorbed volume of adsorbed increased rapidly at low pressures. Subsequently, the increase was more softer at medium and high pressures (30–50 bar).

The activated carbon samples that had the highest adsorption capacity against CO2 were those activated with KOH, with the maximum value for the samples activated with H3PO4 reaching a maximum value at 13.5 mmol g−1, corresponding to sample MPK7 prepared at a constant temperature and constant time (800 °C and 180 min) and a AA/char ratio of 1.5. The isotherms in Fig. 3a present the following order with respect to adsorption capacity for CO2: MPP3 > MPP2 > MPP4 > MPP1. The adsorption isotherms corresponding to the MPK samples shown in Fig. 3b follow the same order. It is worth emphasizing that the values found in this research are within the same order of magnitude as others reported in the literature, and some activated carbons have even higher values [62, 63, 64, 65, 66]. A similar behavior was observed in the isotherms corresponding to samples MPP9, MPP10, MPK11, and MPK12 with an order of magnitude very similar to the adsorption capacity of CO2 (these isotherms are not shown here).

The results obtained in this research regarding the adsorption capacity of the CO2 on the ACs obtained from the samples prepared in the experiments performed here showed that they depend not only on the microporosity but also on the acidity and basicity of the sample. With respect to this latter aspect, as the CO2 molecule is slightly more acidic, it has more affinity toward the activated carbons with KOH as they had a superficial chemistry with a basic character, as shown by several publications [61, 62, 67].

By analyzing the results of the CO2 adsorption isotherms for the samples at a constant ratio (1.5) as well as time (180 min) by varying the temperature (not shown in this study), the result again is that the samples of activated carbons prepared with KOH had the highest CO2 adsorption values. A value of 19.0 mmol g−1 was reached, suggesting that MP debris is a suitable material to prepare activated carbon suitable for CO2 storage.

In order to evaluate the main variables that affect the adsorption capacity of CO2, some graphs are presented in Fig. 4. Figure 4a.1, a.2 shows that there was a linear correlation between the BET area for the carbons chemically activated with KOH, whereas those activated with H3PO4 did not present a linear correlation. This behavior was repeated for the correlations performed as a function of the total volume (Fig. 4b.1, b.2).
Fig. 4

Correlating the capacity of CO2 adsorption at 25 °C to the properties of the activated carbons derived from mangosteen peel for the MPK: a surface area BET, b total pore volume, c micropore surface area, d basic sites (mmol g−1), and e graphitization degree (I D/I G)

When studying the relationship between the volume of micropores and the adsorption capacity of CO2, there was relative linearity for both the MPK and MPP series (Fig. 4c.1, c.2). Consequently, the BET surface area and the total pore volume do not present an acceptable linear correlation with the adsorption capacity.

Two different behaviors occurred when analyzing the basicity parameter (Fig. 4d.1, d2). This result occurred in accordance with the acidic or basic functional groups generated on the surface of the carbon during the chemical activation process. Finally, when analyzing the relationship according to the degree of graphitization (See Fig. 4e) (I D/I G), determined by Raman spectroscopy, the results show that high CO2 adsorption capacities are present with a highly crystalline organization of the carbonaceous structure, both for acid- and base-activated samples. Here, it is necessary to remember that the ratio of the intensities of the D and G bands indicate the degree of graphitization of the material, so that the degree of graphitization is higher when I D/I G approaches zero [45].

Carbon dioxide differential enthalpy measurement by adsorption calorimetry: analysis

When making the respective injections of CO2 on each of the samples inside the microcalorimeter, two signals register based on how it is assembled: a signal corresponding to the thermometric signal that is identified by a strong peak due to the exothermicity of the process between the gas–solid interaction (red line), which subsequently returns to baseline (this signal is recorded by the thermopiles) and a series of “stairs” (green line) corresponding to the respective pressure changes that subsequently reach the state of equilibrium. These two parameters are taken as a function of time as shown in Fig. 5.
Fig. 5

Thermometric signal corresponding to the adsorption of CO2 on a sample of activated carbon (MPK7)

The adsorption capacities of the CO2 on the activated carbon samples were measured at 25 °C and pressures up to 20.0 bar, to be able to carry out comparative studies with the literature [68].

Adsorbate–adsorbent and adsorbate–adsorbate interactions can be calculated with good precision by adsorption microcalorimetry, which can calculate the differential adsorption enthalpies that in some investigations are evaluated by the adsorption isotherms at different temperatures using the Van’t Hoff equation. From these curves (called curves or enthalpy graphs), it is possible to perform a study on the adsorbate filling mechanism and phase transitions, as well as any structural changes to the adsorbent.

Other information can be obtained by analyzing the form of the curve, associated with both textural and chemical characteristics. In this way, it is possible to relate the differential adsorption enthalpies of a molecule (in this case CO2) on the adsorbent with properties such as heterogeneity and energetic homogeneity of the surface and formation of a monolayer, as well as to determine the presence of Lewis acid sites. This is possible because when a solid–vapor system is analyzed by adsorption microcalorimetry, an increase in the amount of adsorbed gas on the sample leads to an increase in the interactions between the adsorbate molecules in the interior of the porous system. Also, with respect to adsorbate–adsorbent contributions, the interaction of an adsorbate molecule with an energetically homogeneous surface will give rise to a constant signal. Finally, if the adsorbent is energetically heterogeneous due to a pore size distribution and/or a variable surface chemistry (defects, cations, etc.), a relatively strong interaction between the adsorbent molecules and the surface would be expected. The strength of these interactions will then decrease as these specific sites are dealt with. Thus, for the energetically heterogeneous adsorbents, a gradual decrease of the calorimetric signal is observed. However, each curve of differential enthalpy varies from one sample to another, showing an enthalpic contribution according to the textural characteristics of each activated carbon.

Figure 6 shows the differential adsorption enthalpies of CO2 on each of the samples prepared from MP under the experimental conditions of this investigation.
Fig. 6

Differential enthalpies of adsorption of CO2 on the samples prepared from MP: a MPP, b MPK

Figure 6a shows the calorimetric isotherms corresponding to the differential enthalpy of CO2 adsorption on MPP1–MPP4, which show the same trend with three characteristic zones (Fig. 6a). Sample MPP3 has the highest differential enthalpy compared to the other samples prepared using H3PO4; however, Fig. 6b shows higher enthalpic values with the KOH-activated samples. A careful analysis of this figure reveals that, for the MPP samples, the highest enthalpy value was 50.0 kJ mol−1; MPP1 had the lowest value of 35.0 kJ mol−1. The enthalpies of MPK samples covered a range between 75.0 and 45.0 kJ mol−1, corresponding to MPK7 and MPK5, respectively, for an adsorbed amount of CO2 between 1.0 and 16.0 mmol g−1.

It is interesting to mention that the shape of the curves of Fig. 6 is associated with the homogeneity or energy heterogeneity of the surface, as well as the acidity and/or basicity. Additionally, the inspection of these allows us to infer if there are interactions of the catalytic type between the possible oxides and/or salts that remain after the respective activation, for example, phosphates and potassium oxides. Comparing the differential enthalpies generated during CO2 adsorption under the experimental conditions described in this study, they vary as follows: MPP3 > MPP4 > MPP2 > MPP1 for samples activated with H3PO4 and identical behavior for those activated with KOH. As mentioned above, the highest CO2 capacity was presented by the samples activated with KOH, which shows that the storage of this gas is not only a function of the textural properties (these samples presented high S BET areas and good development of V mic) and smaller diameter micropores (Fig. 2b), but also of the chemical character of the surface, which in the case of the MPK samples all turn out to be basic. For example, for sample MPK7, we can attribute the first zone (zone I) to the amount adsorbed on the basic centers generated during the chemical activation with KOH and the possible traces of metal oxides (K) and CO2.

Then a plateau for the MPK7 sample is presented which corresponds to filling the volume of micropores that were still empty. Zone II corresponds to the area of the surface which is energetically homogeneous once the heterogeneous sites of surface adsorption are occupied. In this sample, this plateau was observed at about 28.0 kJ mol−1, i.e., between 2.0 and 4.0 mmol g−1 sample.

Interesting behavior occurred in zone III (28.0 and 20.0 kJ mol−1 corresponding to 4.0–7.8 mmol g−1) where the differential enthalpies increased slightly. This behavior was associated with a possible interaction between molecules that are being adsorbed, in this case CO2.

In summary, the magnitude of the differential enthalpy values of adsorption is a function of the type of surface generated during the process of chemical activation because, as has been widely described in the literature, the use of these activating agents generates not only materials with good microporosity but usually leaves traces of species that interact with the adsorbent and generate additional thermal effects. In the case of H3PO4, for example, it can under certain conditions generate pyrophosphoric acid and polymetaphosphate, among others, which may eventually lead to the formation of C–O–P bonds, which in addition to the heterogeneity of the carbonaceous surface allows for additional interactions between these atoms and CO2 [69, 70, 71, 72, 73, 74, 75, 76, 77, 78]. For the case of KOH, the chemical activation process can also leave traces of metal, in this case K [79, 80, 81, 82, 83, 84, 85].

In summary, the adsorption of CO2 on the activated carbons prepared in this work showed a high differential enthalpy, associated with the development of high microporosity generated during the preparation of the carbons, surface heterogeneity, and traces left by the activating agents. Additionally, for the particular case of KOH ACs, these generate a basic surface chemistry that, when interacting with CO2 (a molecule with a slightly acidic character), generates higher enthalpy values than those obtained with H3PO4-activated samples.

These results show that adsorption calorimetry can be used to follow the interactions of CO2 adsorption on activated carbon at high relative pressures. Calorimetry can also establish the characteristics and the chemistry of the surface.


In this work, activated carbons were prepared from MP by chemical activation using H3PO4 and KOH solutions to study the adsorption capacity of CO2. Samples showed BET surface areas from 1035 to 1360 m2 g−1 (MPK5 and MPP10, respectively); the KOH-activated samples had similar textural parameters to the carbons prepared by chemical activation using H3PO4. However, the basic character of the KOH-activated samples meant that they presented a greater capacity for CO2 adsorption. Carbon dioxide storage was investigated by obtaining high-pressure adsorption isotherms and taking measurements of adsorption microcalorimetry, and the results were compared.

A systematic study between the AA/char ratio and temperature was carried out, finding that at an AA/char ratio of 1.5 and at a temperature of 800 °C, samples with good textural characteristics were obtained. The CO2 storage capacity was between 19.0 (MPK7) and 11.4 mmol g−1 (MPP3). This research shows that activated carbon obtained from MP (a waste generated in large quantity) presents good properties for CO2 storage under the experimental conditions of our investigation.

The results obtained by adsorption microcalorimetry are in good agreement with previous studies. The differential adsorption enthalpies were determined for each of the samples by the application of CO2 pulses, generating well-defined curves and allowing for the calculation of the different areas under the curve of each peak to determine the adsorption of CO2 on the activated carbon. Finally, the results obtained by adsorption microcalorimetry were used to associate the differential heats with the characteristics of the activated carbons prepared in this research from mangosteen peel.



The authors wish to thank the framework agreement between the National University of Colombia and the Andes (Bogotá, Colombia). The authors also wish to thank the multinational project EraNet-LAC ELAC2014/BEE-0367, BioFESS (Universitat Hohenheim, Germany, project leader) and Colciencias Contract No. 217-2016 (Colombia) for funding to carry out this investigation.


  1. 1.
    Sedov IA, Muhametzyanov TA, Solomonov BN. A procedure for calibration of differential scanning calorimeters. Thermochim Acta. 2016;639:10–3.CrossRefGoogle Scholar
  2. 2.
    Zimmermann W, Keller JU. A new calorimeter for simultaneous measurement of isotherms and heats of adsorption. Thermochim Acta. 2003;405:31–41.CrossRefGoogle Scholar
  3. 3.
    Wadsö I. Neither calorimeters nor calorimetrists are what they used to be. Thermochim Acta. 1997;300:1–5.CrossRefGoogle Scholar
  4. 4.
    Rouquerol J, Zielenkiewicz W. Suggested practice for classification of calorimeters. Thermochim Acta. 1986;109:121–37.CrossRefGoogle Scholar
  5. 5.
    Kusano K. Micro conduction calorimeters to measure enthalpies of vaporization. Thermochim Acta. 1985;88:109–20.CrossRefGoogle Scholar
  6. 6.
    Elsayed AM, Askalany AA, Shea AD, Dakkama HJ, Mahmoud S, Al-Dadah R, Kaialy W. A state of the art of required techniques for employing activated carbon in renewable energy powered adsorption applications. Renew Sust Energy Rev. 2017;79:503–19.CrossRefGoogle Scholar
  7. 7.
    El-Hendawy AA, Samra SE, Girgis BS. Adsorption characteristics of activated carbons obtained from corncobs. Colloid Surf A. 2001;180:209–21.CrossRefGoogle Scholar
  8. 8.
    Dai XD, Liu XM, Zhao G, Qian L, Qiao K, Yan ZF. Treatment of activated carbon for methane storage. Asia Pac J Chem Eng. 2008;3:292–7.CrossRefGoogle Scholar
  9. 9.
    Dai XD, Liu XM, Qian L, Yan ZF, Zhang J. A novel method to synthesize super-activated carbon for natural gas adsorptive storage. J Porous Mater. 2006;13:399–405.CrossRefGoogle Scholar
  10. 10.
    Burchell T, Rogers M. Low pressure of natural gas for vehicular applications. SAE. 2000;. doi: 10.4271/2000-01-2205.CrossRefGoogle Scholar
  11. 11.
    Lozano-Castello D, Alcaniz-Monge J, de la Casa-Lillo MA, Cazorla-Amoros D, Linares-Solano A. Advances in the study of methane storage in porous carbonaceous materials. Fuel. 2002;81:1777–803.CrossRefGoogle Scholar
  12. 12.
    Adamson AW. Physical chemistry of surface. London: Interscience Publishers Inc.; 1960.Google Scholar
  13. 13.
    Garcia-Cuello V, Moreno-Piraján JC, Giraldo-Gutiérrez L, Sapag K, Zgrablich G. A new microcalorimeter of adsorption for the determination of differential enthalpies. Microporous Mesoporous Mater. 2009;120:239–45.CrossRefGoogle Scholar
  14. 14.
    Ruthven DM. Principles of adsorption and adsorption processes. New York: Wiley; 1984.Google Scholar
  15. 15.
    Myers AL, Monson PA. Physical adsorption of gases: the case for absolute adsorption as the basis for thermodynamic analysis. Adsorption. 2014;20:591–622.CrossRefGoogle Scholar
  16. 16.
    Myers AL. Thermodynamics of adsorption in porous materials. AIChE J. 2012;48:145–64.CrossRefGoogle Scholar
  17. 17.
    Brunauer S. The adsorption of gases and vapors, vol. I. Berlin: Springer; 1945.Google Scholar
  18. 18.
    Hill TL. Statistical mechanics of adsorption. IX. Adsorption thermodynamics and solution thermodynamics. J Chem Phys. 1949;17:520–34.CrossRefGoogle Scholar
  19. 19.
    Hill TL. Statistical mechanics of adsorption. IX. Adsorption thermodynamics and solution thermodynamics. J Chem Phys. 1950;18:246–54.CrossRefGoogle Scholar
  20. 20.
    Everett DH. The thermodynamics of adsorption. Part III. Analysis and discussion of experimental data. Trans Faraday Soc. 1950;46:957–73.CrossRefGoogle Scholar
  21. 21.
    Hill TL, Emmett PH, Joyner LG. Calculation of thermodynamic functions of adsorbed molecules from adsorption isotherm measurements: nitrogen on graphon. J Am Chem Soc. 1951;73:5102–10.CrossRefGoogle Scholar
  22. 22.
    Valenzuela DP, Myers AL. Adsorption equilibrium data handbook. Englewood Cliffs: Prentice-Hall; 1989.Google Scholar
  23. 23.
    Biloe S, Goetz V, Mauran S. Characterization of adsorbant composite block for methane storage. Carbon. 2001;39:1653–62.CrossRefGoogle Scholar
  24. 24.
    Lozano-Castello D, Lillo-Rodenas MA, Cazorla-Amoros D, Linares-Solano A. Preparation of activated carbons from Spanish anthracite. I. Activation by KOH. Carbon. 2001;39:741–9.CrossRefGoogle Scholar
  25. 25.
    Ahmad MA, Alrozi R. Optimization of preparation conditions for mangosteen peel-based activated carbons for the removal of Remazol Brilliant Blue R using response surface methodology. Chem Eng J. 2010;165:883–90.CrossRefGoogle Scholar
  26. 26.
    Okonogi S, Duangrat C, Anuchpreeda S, Tachakittirungrod S, Chowwanapoonpohn S. Comparison of antioxidant capacities and cytotoxicities of certain fruit peels. Food Chem. 2007;103:839–46.CrossRefGoogle Scholar
  27. 27.
    Zien R, Suhaili R, Earnestly F, Indrawati Munaf E. Removal of Pb(II), Cd(II) and Co(II) from aqueous solution using Garcinia mangostana L. fruit shell. J Hazard Mater. 2010;162:43–50.Google Scholar
  28. 28.
    Chen Y, Huang B, Huang M, Cai B. On the preparation and characterization of activated carbon from mangosteen Shell. J Taiwan Inst Chem Eng. 2011;42:837–42.CrossRefGoogle Scholar
  29. 29.
    Cychosz KA, Guo X, Fan W, Cimino R, Gor GY, Tsapatsis M, Neimark AV, Thommes M. Characterization of the pore structure of three-dimensionally ordered mesoporous carbons using high resolution gas sorption. Langmuir. 2012;28(34):12647–54.CrossRefPubMedGoogle Scholar
  30. 30.
    Gedrich K, Senkovska I, Klein N, Stoeck U, Henschel A, Lohe MR, Baburin IA, Mueller U, Kaskel S. A highly porous metal-organic framework with open nickel sites. Angew Chem Int Ed. 2010;49:8489–92.CrossRefGoogle Scholar
  31. 31.
    Seki K. Design of an adsorbent with an ideal pore structure for methane adsorption using metal complexes. Chem Commun. 2001;16:1496–7.CrossRefGoogle Scholar
  32. 32.
    Ma SQ, Sun DF, Simmons JM, Collier CD, Yuan DQ, Zhou HC. Metal-Organic framework from an anthracene derivative containing nanoscopic cages exhibiting high methane uptake. J Am Chem Soc. 2008;130:1012–6.CrossRefPubMedGoogle Scholar
  33. 33.
    Ma SQ, Zhou HC. Gas storage in porous metal-organic frameworks for clean energy applications. Chem Commun. 2010;46:44–53.CrossRefGoogle Scholar
  34. 34.
    Li B, Wen HM, Zhou W, Xu JQ, Chen B. Porous metal-organic frameworks: promising materials for methane storage. Chem. 2016;1:557–80.CrossRefGoogle Scholar
  35. 35.
    Beckner M, Dailly A. A pilot study of activated carbon and metal–organic frameworks for methane storage. Appl Energy. 2016;162:506–14.CrossRefGoogle Scholar
  36. 36.
    Wang H, Zhu QL, Zou R, Xu Q. Metal-organic frameworks for energy applications. Chem. 2017;2:52–80.CrossRefGoogle Scholar
  37. 37.
    Himeno S, Komatsu T, Fujita S. High-pressure adsorption equilibria of methane and carbon dioxide on several activated carbons. J Chem Eng Data. 2005;50:369–76.CrossRefGoogle Scholar
  38. 38.
    Casas N, Schell J, Pini R, Mazzotti M. Fixed bed adsorption of CO2/H2 mixtures on activated carbon: experiments and modeling. Adsorption. 2012;18:143–61.CrossRefGoogle Scholar
  39. 39.
    Bae Y-S, Mulfort KL, Frost H, Ryan P, Punnathanam S, Broadbelt LJ, Hupp JT, Snurr RQ. Separation of CO2 from CH4 using mixed-ligand metal–organic frameworks. Langmuir. 2008;24:8592–8.CrossRefPubMedGoogle Scholar
  40. 40.
    Neimark AV, Lin Y, Ravikovitch PI, Thommes M. Quenched solid density functional theory and pore size analysis of micro-mesoporous carbons. Carbon. 2009;47:1617–28.CrossRefGoogle Scholar
  41. 41.
    Rodríguez-Reinoso F, Garrido J, Martín-Martínez JM, Molina-Sabio M, Torregrosa R. The combined use of different approaches in the characterization of microporous carbons. Carbon. 1989;27:23–32.CrossRefGoogle Scholar
  42. 42.
    Rouquerol J, Llewellyn P, Rouquerol F. Is BET equation applicable to microporous adsorbents? Stud Surf Sci Catal. 2007;160:49–56.CrossRefGoogle Scholar
  43. 43.
    Brunauer S, Emmett PH, Teller E. Adsorption of gases in multimolecular layers. J Am Chem Soc. 1936;407:309–13.Google Scholar
  44. 44.
    Lippens BC, de Boer JH. Studies on pore systems in catalysts: V. The t-method. J Catal. 1965;4:319–23.CrossRefGoogle Scholar
  45. 45.
    Liu Q, Guan JS, Li J, Li C. SO2 removal from flue gas by activated semi-cokes 2. Effects of physical structures and chemical properties on SO2 removal activity. Carbon. 2003;41:2225–30.CrossRefGoogle Scholar
  46. 46.
    Fiuza RA, de Jesus Neto RM, Correia LB, Andrade HMC. Preparation of granular activated carbons from yellow mombin fruit stones for CO2 adsorption. J Environ Manage. 2015;161:198–205.CrossRefPubMedGoogle Scholar
  47. 47.
    Luo J, Liu YF, Jiang CF, Chu W, Jie W, Xie HP. Experimental and modeling study of methane adsorption on activated carbon derived from anthracite. J Chem Eng Data. 2011;56:4919–26.CrossRefGoogle Scholar
  48. 48.
    Luo J, Liu Y, Sun W, Jiang C, Xie H, Chu W. Influence of structural parameters on methane adsorption over activated carbon: evaluation by using D-A model. Fuel. 2014;123:241–7.CrossRefGoogle Scholar
  49. 49.
    Thommes M, Kaneko K, Neimark AV, Olivier JP, Rodriguez-Reinoso F, Rouquerol J, Sing KSW. Physisorption of gases, with special reference to the evaluation of surface area and pore size distribution (IUPAC Technical Report). Pure Appl Chem. 2015;. doi: 10.1515/pac-2014-1117.CrossRefGoogle Scholar
  50. 50.
    Do DD. Pure component adsorption in microporous solids, chapter 4. In: World Scientific Publishing Co, editor. Adsorption analysis: equilibria and kinetics. London: Imperial College Press; 1998.Google Scholar
  51. 51.
    Dubinin MM, Astakhov VA. Description of adsorption equilibria of vapors on zeolites over wide ranges of temperature and pressure. Adv Chem Ser. 1970;102:69–85.CrossRefGoogle Scholar
  52. 52.
    Dubinin MM, Astakhov VA. Development of the concepts of volume filling of micropores in the adsorption of gases and vapors by microporous adsorbents 1. Carbon adsorbents. Izv Akad Nauk SSSR Ser Khim. 1971;1:5–11.Google Scholar
  53. 53.
    Moreno-Piraján JC, García-Cuello VS, Giraldo L. Characterization of mordenite-supported Pd, Pt, and Ir determined by CO adsorption microcalorimetry and the dehydrogenation reaction of C3 alkanes. Top Catal. 2011;54:146–52.CrossRefGoogle Scholar
  54. 54.
    Moreno-Piraján JC, Bastidas-Barranco MJ, Giraldo L. Preparation of activated carbons for storage of methane and its study by adsorption calorimetry. J Therm Anal Calorim. 2017;. doi: 10.1007/s10973-017-6132-8.CrossRefGoogle Scholar
  55. 55.
    Correia LB, Fiuza RA, Jr. RC, Andrade HMC. CO2 capture on activated carbons derived from mango fruit (Mangifera indica L.) seed shells. J Therm Anal Calorim. 2017;. doi: 10.1007/s10973-017-6542-7.CrossRefGoogle Scholar
  56. 56.
    Bazan A, Nowicki P, Póilrolniczak P, Pietrzak R. Thermal analysis of activated carbon obtained from residue after supercritical extraction of hops. J Therm Anal Calorim. 2016;125:1199–204.CrossRefGoogle Scholar
  57. 57.
    Nowicki P. Effect of heat treatment on the physicochemical properties of nitrogen-enriched activated carbons. J Therm Anal Calorim. 2016;125:1017–24.CrossRefGoogle Scholar
  58. 58.
    Rodríguez-Estupiñán P, Giraldo L, Moreno-Piraján JC, A microcalorimetric study of methane adsorption on activated carbons obtained from mangosteen peel at different conditions. J Therm Anal Calorim. 2017 (Aceppted for publication).Google Scholar
  59. 59.
    Bansal RC, Goyal M. Activated Carbon Adsorption. New York: CRC Press; 2005.CrossRefGoogle Scholar
  60. 60.
    Mattson JS, Mark HB. Activated carbon: surface chemistry and adsorption from solution. New York: Marcel Dekker; 1971.Google Scholar
  61. 61.
    Hu HU, Ruckenstein E. Application of Dubinin–Astakhov equation to CO2 adsorption on single-walled carbon nanotubes. Chem Phys Lett. 2006;425:306–10.CrossRefGoogle Scholar
  62. 62.
    Aaron D, Tsouris C. Separation of CO2 from flue gas: a review. Sep Sci Technol. 2005;48:321–48.CrossRefGoogle Scholar
  63. 63.
    Figueroa JD, Fout T, Plasynski S, McIlvried H, Srivastava RD. Advances in CO2 capture technology—the US Department of Energy’s carbon sequestration program. Int J Greenh Gas Control. 2008;2:9–20.CrossRefGoogle Scholar
  64. 64.
    Sevilla M, Fuertes AB. Sustainable porous carbons with a superior performance for CO2 capture. Energy Environ Sci. 2011;4:1765–71.CrossRefGoogle Scholar
  65. 65.
    Hu X, Radosz M, Cychosz KA, Thommes M. CO2-filling capacity and selectivity of carbon nanopores: synthesis, texture, and pore-size distribution from quenched-solid density functional theory (QSDFT). Environ Sci Technol. 2011;45:7068–74.CrossRefPubMedGoogle Scholar
  66. 66.
    Schell J, Casas N, Pini R, Mazzotti M. Pure and binary adsorption of CO2, H2, and N2 on activated carbon. Adsorption. 2012;18:49–65.CrossRefGoogle Scholar
  67. 67.
    Belin T, Epron F. Characterization methods of carbon nanotubes: a review. Mater Sci Eng B Solid. 2005;119:105–18.CrossRefGoogle Scholar
  68. 68.
    Labus K, Gryglewicz S, Machnikowski J. Granular KOH-activated carbons from coal-based cokes and their CO2 adsorption capacity. Fuel. 2014;118:9–15.CrossRefGoogle Scholar
  69. 69.
    Giraldo L, Bastidas-Barranco M, Moreno-Piraján JC. Adsorption calorimetry: energetic characterization of the surface of mesoporous silicas and their adsorption capacity of non-linear chain alcohols. Colloid Surf A. 2016;496:100–13.CrossRefGoogle Scholar
  70. 70.
    Li W, Peng J, Zhang L, Xia H, Li N, Yang K, Zhu X. Effects of carbonization temperatures on characteristics of porosity in coconut shell chars and activated carbons derived from carbonized coconut shell chars. J Ind Crops Prod. 2008;28:73–8.CrossRefGoogle Scholar
  71. 71.
    Devi AS, Latif PA, Tham YJ, Taufiq-Ya YH. Physical characterization of activated carbon derived from mangosteen peel. Asian J Chem. 2012;24:579–83.Google Scholar
  72. 72.
    Kalderis D, Bethanis S, Paraskeva P, Diamadopoulos E. Production of activated carbon from bagasse and rice husk by a single-stage chemical activation method at low retention times. Bioresour Technol. 2008;99:6809–16.CrossRefPubMedGoogle Scholar
  73. 73.
    Williams PT, Besler S. The pyrolysis of rice husk in a thermogravimetric analyzer and static batch reactor. Fuel. 1993;72:151–9.CrossRefGoogle Scholar
  74. 74.
    Ahmadpour A, Do DD. The preparation of activated carbon from macadamia nutshell by chemical activation. Carbon. 1997;35:1723–32.CrossRefGoogle Scholar
  75. 75.
    Kalderis D, Bethanis S, Paraskeva P, Diamadopoulos E. Adsorption of polluting substances on activated carbons prepared from rice husk and sugarcane bagasse. Chem Eng J. 2008;144(1):42–50.CrossRefGoogle Scholar
  76. 76.
    Girgis BS, Ishak MF. Activated carbon from cotton stalks by impregnation with phosphoric acid. Mater Lett. 1999;39:107–14.CrossRefGoogle Scholar
  77. 77.
    Guo J, Lua AC. Textural and chemical properties of adsorbent prepared from palm shell by phosphoric acid activation. Mater Chem Phys. 2003;80:114–9.CrossRefGoogle Scholar
  78. 78.
    Girgis BS, Yunis SS, Soliman AM. Characteristics of activated carbon from peanut hulls in relation to conditions of preparation. Mater Lett. 2002;57:164–72.CrossRefGoogle Scholar
  79. 79.
    Adinata D, Daud WMAW, Aroua MK. Preparation and characterization of activated carbon from palm shell by chemical activation with K2CO3. Bioresour Technol. 2007;98:145–9.CrossRefPubMedGoogle Scholar
  80. 80.
    Yu-Peng G, Shao-Fang Y, Jing-Zhe Z, Zi-Chen W, Mu-Yu Z. Preparation of activated carbon with high specific surface area from rice husks. Chem J Chin U. 2000;21:335–8.Google Scholar
  81. 81.
    Cao Q, Xie KC, Lv Y-K, Bao WR. Process effects on activated carbon with large specific surface area from corn cob. Bioresour Technol. 2006;97:110–5.CrossRefPubMedGoogle Scholar
  82. 82.
    Kan Y, Yue Q, Li D, Wu Y, Gao B. Preparation and characterization of activated carbons from waste tea by H3PO4 activation in different atmospheres for oxytetracycline removal. J Taiwan Inst Chem Eng. 2017;71:494–500.CrossRefGoogle Scholar
  83. 83.
    Suárez-García F, Martínez-Alonso A, Tascón JMD. Activated carbon fibers from Nomex by chemical activation with phosphoric acid. Carbon. 2004;42:1419–26.CrossRefGoogle Scholar
  84. 84.
    Liu H, Wang X, Zhai G, Zhang J, Zhang C, Bao N, Cheng C. Preparation of activated carbon from lotus stalks with the mixture of phosphoric acid and pentaerythritol impregnation and its application for Ni(II) sorption. Chem Eng J. 2012;209:155–62.CrossRefGoogle Scholar
  85. 85.
    Benaddi H, Legras D, Rouzaud JN, Beguin F. Influence of the atmosphere in the chemical activation of wood by phosphoric acid. Carbon. 1998;36:306–9.CrossRefGoogle Scholar

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2017

Authors and Affiliations

  1. 1.Departamento de Química, Facultad de CienciasUniversidad Nacional de ColombiaBogotáColombia
  2. 2.Departamento de Química, Grupo de Investigación en Sólidos Porosos y Calorimetría, Facultad de CienciasUniversidad de los AndesBogotáColombia

Personalised recommendations