Development of a low temperature adsorbent from karanja seed cake for CO₂ capture

Abstract

The applicability of karanja (Pongamia pinnata) seed cake as a low temperature CO₂ adsorbent was examined. The processing of the cake was done by extraction and hydrothermal treatments, followed by calcination. The physico-chemical characteristics of sorbent materials were determined by nitrogen adsorption, X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, scanning electron microscopy-electron diffraction X-ray analysis (SEM-EDXA) and C¹³ solid state nuclear magnetic resonance (C¹³-NMR) spectroscopy techniques. CO₂ uptake capacities of the sorbents were obtained by generating their breakthrough curves, in a fixed bed reactor connected to an on-line gas chromatograph. Hydrothermal treatment improved the surface area and afforded exposure of more sites for CO₂ adsorption. The hydrothermally treated cake showed higher adsorption capacity, easy thermal desorption and also good recyclability up to five consecutive cycles, as compared to its extracted analogue, proving it to be a promising adsorbent. The Yoon–Nelson rate constant values also supported the superior adsorption capacity of hydrothermally treated sample.

---

1. Introduction

Carbon dioxide is a major contributor to global warming. It is mostly released by fossil fuel combustion in thermal power plants,¹ by manufacturing units like steel, cement and other chemical plants and to some extent, by burning of wood and plant wastes for heating purposes in rural areas. Before the start of the industrial revolution, the CO₂ concentration in the atmosphere was about 280 parts per million by volume (ppmv). Currently, the CO₂ concentrations crossed 380 ppmv, increasing by about 1.5 ppmv per year. By the end of the 21st century, it is expected that the concentration would attain 500 ppmv levels, nearly double that of the pre-industrial value, if the emissions continue with the same rate. Thus, there is a need to contain the emissions of CO₂. Generally, there are three steps to limit carbon dioxide emissions: separation, transportation, and sequestration.² Carbon dioxide separation is necessary to carry out step 2 as well as step 3. In the total cost of carbon dioxide management the majority of expenditure goes to the separation process.³ For that reason, it is vital to identify an efficient and profitable method for the capture of CO₂. Cryogenic, absorption, adsorption, membrane separation and micro algal bio-fixation are important among the various separation technologies. Separation by gas adsorption using solid adsorbents is considered as an efficient method due to low capital cost, easiness to apply, better corrosion resistance and production of high purity CO₂ stream.^4–9 Therefore, more focus has been laid on the development of solid adsorbent materials that operate at low temperature and pressure with a greater number of consecutive capture cycles.

For CO₂ capture from flue gas streams both chemical and physical adsorbents are reported in the literature, among which metal–organic frameworks, zeolites, activated carbons, calcium oxides, hydrotalcites and supported amines make up the main classes of adsorbents. However, if the adsorbent is developed from plant waste material, it could also be used at the generation site in rural areas where biomass is burnt as fuel, eventually generating CO₂. The biogas plants could also adopt these materials for carbon dioxide capture and enhance the calorific value of the gas. With this objective, many wood based materials have been developed, drawing inspiration from the efficiency of active carbons which are proved to be excellent adsorbents for carbon capture due to their large surface area and porosity.

Pongamia pinnata, commonly called karanja, belongs to the leguminosae family, native in tropical and temperate Asia including parts of India, China, Japan, Malaysia and Australia. Karanja seed mainly contains 20–30% protein, 27–39% oil and a group of furano flavonoids that amount to 5–6% by weight of the oil.¹⁰ It is expected that in coming years an enormous quantity of expelled cake will be generated, as karanja oil is identified as the most suitable raw material for biodiesel production.^11,12 About two tons of expelled cake is generated for every ton of biodiesel produced. The expelled karanja cake contains 10–12% oil which is a good source of protein. The karanja cake is toxic, has bitter taste and pungent odour and is non-edible due to the presence of karanjin, pongamol, significant amino acids and glabrin. Some efforts have been made to remove the toxic components and make the cake safe. However, better methods that could add value to the cake are being explored. To the best of our knowledge, there have been no detailed reports on the modification of karanja seed cake suitable as a CO₂ adsorbent. The development of new carbonaceous adsorbents from abundantly available waste biomass, particularly the toxic seed cakes, is always advantageous.

The aim of this work is to develop the karanja cake into an efficient adsorbent for CO₂ capture. A detailed study was carried out to make it suitable for CO₂ adsorption. The activation of the expelled cake was affected in two ways. It was subjected to (i) Soxhlet extraction followed by calcination and (ii) hydrothermal treatment followed by calcination. The carbonaceous materials thus obtained were thoroughly characterised by different techniques. The CO₂ adsorption capacities were determined by generating break-through curves to ascertain their suitability as adsorbents. The reason for the difference in adsorptive capacity was explained in terms of their physico-chemical properties.

2. Experimental

2.1 Materials

Locally available karanja seed cake was used as the starting material for the synthesis of the adsorbents. Doubly distilled water was used in the preparation.

2.2 Preparation of adsorbent materials

Hexane was used as the solvent for Soxhlet extraction.¹³ In this method 30 g of the karanja seed cake was treated in 300 ml of hexane in Soxhlet apparatus for about 6 h. The residual cake was then dried followed by calcination at 650 °C for 5 h under inert gas flow. The sample is designated as EKC. In the case of hydrothermal treatment, 20 g of the cake was added to 50 ml of de-ionized water and stirred for about 30 min at room temperature. The solid was separated, another 50 ml of water was added to it and the mixture was transferred into a Teflon lined sealed autoclave and kept at 140 °C for 48 h. The cake was then washed with distilled water, kept in an air oven at 100 °C for about 12 h and then subjected to calcination at 650 °C under inert atmosphere for 5 h. The sample is designated as HKC. For the sake of comparison, another sample was prepared by directly calcining the karanja cake under nitrogen flow at 650 °C. It is denoted as KC.

2.3 Characterization

The data on surface area, pore volume and average pore diameter were obtained by nitrogen adsorption at liquid N₂ temperature (−196 °C) using Autosorb-1 (Quanta chrome, USA) instrument using a standard procedure. The powders were first out-gassed at 200 °C to ensure a clean surface prior to construction of the adsorption isotherm. X-Ray powder diffraction patterns of the samples were recorded on a Rigaku Miniflex diffractometer using Cu Kα radiation. The measurements were obtained in steps of 0.045° with a scanning time of 0.5 s and in the 2θ range of 10–80°. The FT-IR spectra were recorded on a Biorad (DIGILAB Excalibur series) spectrometer using KBr disc method, with a resolution of 1 cm⁻¹. Scanning electron microscopic pictures of the catalysts were obtained on a Hitachi S-520 electron microscope running at an accelerated voltage of 10 kV. The samples were mounted on aluminum stubs using double-adhesive tape and gold coated in a Hitachi HUS-5GB vacuum evaporator. C¹³ NMR spectra were recorded on a Gemini Varian 200 MHz, Bruker AV 300 MHz and unity 400 MHz spectrometer by using TMS as an internal standard.

2.4 CO₂ adsorption measurements

CO₂ adsorption studies were conducted in a dynamic adsorption flow system (Fig. 1). The adsorption was conceded in a fixed bed reactor (SS, 410 mm length: 9 mm id). 1 g of adsorbent mixed with 1 g of glass beads was suspended in the middle of the reactor between two quartz plugs and activated under helium atmosphere at 200 °C for 1 h. After cooling the reactor to the required temperature, a gas mixture of 10% CO₂ balanced by He was passed through the bed at the required flow rate with the help of mass flow controllers. The effluent was analyzed on-line using an Agilent Technologies 7820A gas chromatograph equipped with a thermal conductivity detector and a Porapak Q column. The adsorption data were collected at 50–70 °C. Then, the sample was flushed with pure helium gas for 30 min, and desorption proceeded at 140 °C.

Fig. 1 Schematic of the experimental setup.

3. Results and discussion

3.1 Structural and textural properties

3.1.1 Surface area and pore volume. The BET surface area values of the adsorbents are reported in Table 1. HKC has a higher surface area, 5 times more than that of EKC.

Table 1 Textural characteristics of the adsorbent materials^a

	SBET (m^2 g^−1)	VT (cm^3 g^−1)	DA (nm)
a SBET – surface area; VT – total pore volume; DA – average pore diameter.
HKC	10.2	0.350	13.6
EKC	2.1	0.003	3.3
KC	1.1	0.001	3.1

---

This observation is in agreement with the results of Saksule et al.,¹⁴ who have prepared a carbon adsorbent from karanja seed cake by various chemical and physical activation processes and used it for the adsorption of dyes and for waste water treatment. The pore diameter of the adsorbents, as determined by BJH method, varied in the range of 2–14 nm.

It may be understood that during the activation process, the spaces between the crystallites are cleared, or loosely bound carbonaceous material is formed and channels are created through the graphitic regions. This has resulted in the creation of a porous structure with a larger internal surface area in the material. Ganvir et al.¹⁵ also have prepared activated carbon from karanja seed cake by sulphuric acid chemical activation. The effects of the preparation conditions on the yield of activated carbon are reported.¹⁵

3.1.2 XRD results. Fig. 2 shows XRD patterns of HKC and EKC adsorbents. The large and broad peak around 2θ ∼ 20.1° represents the amorphous nature of the carbon formed. However, peaks visible at 2θ between 20 and 30° indicate the existence of aromatic layers in the sorbent matrix.^16,17

Fig. 2 XRD patterns of (a) HKC (b) EKC.

3.1.3 C¹³-NMR analysis. The ¹³C NMR spectra of karanja cake adsorbent materials are presented in Fig. 3. The spectra can be divided into a few common chemical shift regions: alkyl C (0–45 ppm), O-alkyl C (45–110 ppm), aromatic C (110–160 ppm) and carboxyl C (160–212 ppm).¹⁸ The peaks between 0–90 ppm indicate that the karanja cake consists of aliphatic carbons. The main peak cantered at ∼30 ppm further suggests the existence of an aliphatic straight chain polymethylene structure, more so in the case of HKC. In addition, aliphatic carbons bound to oxygen by different bonds (as present in aliphatic alcohols and ethers) could be present, as seen by some strong resonance signals appearing in the region of 50 and 90 ppm.

Fig. 3 ¹³C spectra of (a) HKC (b) EKC.

The presence of aromatic carbons can be noticed from the peaks appearing between 90–165 ppm.¹⁹ The peak at 174 ppm corresponds to the carbonyl carbon. The HKC spectrum more clearly displays the aliphatic and aromatic nature of the carbon as compared to that of EKC.

3.1.4 FTIR analysis. The FTIR spectra of carbons generally show the presence of bands in the regions 3000–2700, 1750–1630, and 1600–1450 cm⁻¹, which are assigned to aliphatic C–H stretching modes, C=O (carboxylic and lactones) and C=C bonds, respectively.²⁰ However, the spectra of the cake samples (Fig. 4) show some interesting features. The peaks at 2926 and 2854 cm⁻¹ represent asymmetric and symmetric stretching of the methyl groups. The main peak centered at ∼30 ppm in the NMR spectrum further supports this observation. Thus, the presence of aliphatic carbon chains is indicated by the spectra. The 1625 cm⁻¹ band corresponds to the aromatic ring mode,²¹ as also reflected in their XRD patterns. In addition, the IR spectra also show a peak at 1534 indicating the N–H bending vibration.²² The area of this peak in HKC is bigger than that of the EKC sample, giving credence to the existence of more N–H bonds in the case of the former.

Fig. 4 FT-IR spectra of (a) HKC (b) EKC.

In the case of HKC, a separate peak can be seen at 1642 cm⁻¹ corresponding to the stretching mode of the C=N bond.²³ Thus, the FTIR data reiterate the presence of aliphatic and aromatic carbons in the samples, as observed from NMR studies. In addition, the existence of N-containing species is also indicated.

3.1.5 SEM analysis. Scanning electron microscopy is a useful tool for the study of carbons morphology. From the SEM images shown in Fig. 5 it can be observed that the hydrothermal treatment has led to the formation of a more fine structure with crevices, whereas a flake-like structure is seen in the extracted material. This change in morphology might have helped in increasing the surface area of the solid after hydrothermal treatment.

Fig. 5 SEM images of (a) HKC (b) EKC.

3.2 CO₂ adsorption properties

The adsorption uptake of carbon dioxide can be measured by static as well as dynamic methods. The breakthrough methodology is applicable for evaluation in fixed bed adsorbers, eventually for the design of pressure swing adsorption (PSA) systems. During the adsorption process in fixed beds the top layer is first exposed to the adsorbate. From that point of time, the concentration of the adsorbate in the outlet is zero. The situation continues until all layers are exposed to the adsorbate. The breakthrough capacity (BTC) is estimated by the time elapsed when the concentration in the outlet just becomes measurable. The outlet concentration slowly increases and equals that of the initial concentration of the adsorbate when the bed is fully saturated. This capacity is called saturation capacity. The efficiency of the adsorbate is judged by both capacities. However, the PSA systems depend more on the breakthrough capacity. The break-through curves generated on the adsorbents are shown in Fig. 6A. For comparison, the adsorption capacity of KC is also determined. The CO₂ adsorption capacity (shown in Table 2) of KC, EKC and KC is obtained as 2.52, 1.92 and 1.78 mmol g⁻¹, respectively. HKC exhibits the highest capacity and the EKC shows the lowest value indicating that the hydrothermal method is advantageous. The difference in adsorption behaviour can be explained as follows.

Fig. 6 (A) Break-through curves of different adsorbents (B) CO₂ adsorption capacity of adsorbents for 5 consecutive cycles at 70 °C and 1 atm pressure conditions.

Table 2 Yoon–Nelson kinetic parameters for CO₂ adsorption on modified karanja cake^a,c

Material	q0^b (mmol g^−1)	kYN (min^−1)	τ (min) (EXPER)	τ (min) (EMD)	R^2
a Adsorption conditions; temperature 70 °C and pressure1 bar.b q0 – breakthrough adsorption capacity.c EXPER = experimental; EMD = empirical.
HKC	2.52	0.362	32.0	32.07	0.98
EKC	1.92	0.272	25.5	25.51	0.99
KC	1.78	0.259	25.0	24.93	0.99

---

The N-containing groups in the samples might have imparted a basic nature to the carbons and form the sites for adsorption of CO₂, possibly by weak interaction. In the case of HKC the oil present in the cake during hydrothermal treatment undergoes hydrolysis forming free fatty acids. These acids, being lower boilers than the original oil, upon calcination helped create channels in the structure while emitted as vapours.

It can be understood that this way of increasing surface area exposes a greater number of basic sites in the solids. This observation is substantiated by the results obtained from SEM-EDXA analysis data (Table 3). The N content in HKC can be seen more than that of EKC and KC. In the case of EKC as the oil is mostly removed by extraction there is less scope for increase in the surface area and in turn an increase in the exposed basic groups. Without extraction, when heated to a high temperature, the high boiling oil present in the cake might have undergone degradation. The viscous liquid phase subsequently transforms into char and blocks the channels leading to decreased exposure of the adsorption sites, in the case of KC.

Table 3 SEM-EDXA elemental percentages of adsorbents

Element (atom.%)	HKC	EKC	KC
C	38.53	17.71	32.76
N	34.19	20.89	28.20
O	26.94	61.01	38.54

---

3.3 Multicycle adsorption experiments

Since the adsorption capacity of KC is anyway inferior to that of HKC, further work is devoted to estimating the reproducibility of the breakthrough capacities of HKC and EKC only. Fig. 6B depicts the results obtained in these studies. The breakthrough capacity of HKC starts at 2.5 mmol g⁻¹ and is stabilized at 2.1 mmol g⁻¹. When there is a 15% drop in adsorption capacity after five cycles in the case of HKC the drop is quite considerable, with 65% loss, in the case of EKC. The reason for non-reproducibility of the adsorption uptake of EKC needs to be established.

3.4 Yoon–Nelson model for CO₂ adsorption

The linear Yoon–Nelson equation is frequently used for the study of kinetics of adsorption in packed columns.²⁴ It is represented by the following equation.

In the above equation k_YN stands for the Yoon–Nelson rate constant in min⁻¹; τ is the time required for 50% adsorbate breakthrough time (min) and t is the sampling time (min). The parameter k_YN is determined from the plot of ln(C/(C₀ − C)) versus t. C₀ is the initial concentration of CO₂ and C is the concentration at any time during the evaluation. The values of k_YN and τ for the three adsorbents are reported in Table 2. A higher value of the Yoon–Nelson rate constant (k_YN) is obtained in the case of HKC adsorbent than that of EKC and KC due to the greater number of adsorption sites available for HKC than that of others.²⁵ These results are comparable with those reported in the literature.²⁶

4. Conclusions

The hydrothermal method appears to be better for the preparation of CO₂ adsorbents from karanja cake. A higher surface area can be achieved for the adsorbent prepared by the hydrothermal method as compared to the sample prepared after extraction of the cake. Hydrolysis of the oil generates fatty acids and helps obtain a more efficient adsorbent. The basic sites are better exposed and utilized by this material. The adsorbent prepared by hydrothermal treatment shows reproducibility in its adsorption capacity.

Acknowledgement

The authors gratefully acknowledge financial support from the Department of Science and Technology (DST), India, and thank Dr. B. Jagadeesh for the solid state NMR data.

References

1. C. Clastoskie, Science, 2010, 330, 595–596.
2. V. Zelenak, D. Halamova, L. Gaberova, E. Bloch and P. Llewellyn, Microporous Mesoporous Mater., 2008, 116, 358.
3. X. Xu, C. S. Song, J. M. Andresen, B. G. Miller and A. W. Scaroni, Microporous Mesoporous Mater., 2003, 62, 29.
4. A. R. Millward and O. M. Yaghi, J. Am. Chem. Soc., 2005, 127, 17998.
5. D. Dasgupta, K. Mondal and T. Wiltowski, Int. J. Hydrogen Energy, 2008, 33, 303.
6. F. Y. Chang, K. J. Chao, H. H. Cheng and C. S. Tan, Sep. Purif. Technol., 2009, 70, 87.
7. M. Ryden and A. Lyngfelt, Int. J. Hydrogen Energy, 2006, 31, 1271.
8. L. Y. Meng and S. J. Park, J. Colloid Interface Sci., 2010, 352, 498.
9. B. J. Kim, K. S. Cho and S. J. Park, J. Colloid Interface Sci., 2010, 342, 575.
10. N. V. Bringi, Non-traditional oilseeds and oils in India, Oxford and IBH Publishing Co. Pvt. Ltd, New Delhi, 1987, pp. 143–166.
11. B. K. De and D. K. Bhattacharyya, Fett/Lipid, 1999, 101, 404.
12. L. C. Meher, S. S. Vidya, S. N. Dharmagadda and S. N. Naik, Bioresour. Technol., 2006, 97, 1392.
13. S. N. Bobade and V. B. Khyade, Res. J. Chem. Sci., 2012, 2, 16.
14. A. S. Saksule and P. A. Kude, Int. J. Eng. Appl. Sci., 2012, 2(3), 13.
15. V. N. Ganvir, M. L. Meshram and A. P. Dwivedi, Preparation of adsorbent from karanja oil seed cake and its characterization, IJCA Proceedings on International Conference on Emerging Frontiers in Technology for Rural Area (EFITRA-2012), by IJCA Journal EFITRA – Number 3 Year of Publication, 2012.
16. H. Takagi, K. Maruyama, N. Yoshizawa, Y. Yamada and Y. Sato, Fuel, 2004, 83, 2427.
17. J. A. Thote, K. S. Iyer, R. Chatti, N. K. Labhsetwar, R. B. Biniwale and S. S. Rayalu, Carbon, 2010, 48, 396.
18. N. J. Mathers, Z. H. Xu, T. J. Blumfield, S. J. Berners-Price and P. G. Saffigna, Forest Ecol. Manage., 2003, 175, 467.
19. M. Kanmi, J. K. Gordon, S. M. Althaus and M. Pruski, Solid State Nucl. Magn. Reson., 2012, 47–48, 19.
20. Z. Zhang, M. Xu, H. Wang and L. Zhong, Chem. Eng. J., 2010, 160, 571.
21. Q. Li, J. Yang, D. Feng, Z. Wu, Q. Wu, S. S. Park, C. S. Ha and D. Zhao, Nano Res., 2010, 3(9), 632.
22. M. Vinoba, M. Bhagiyalakshmi, S. K. Jeong, Y. I. I. Yoon and S. C. Nam, Colloids Surf., B, 2012, 90, 91.
23. H. S. Yee, V. Aswin and J. Martin, Z. Anorg. Allg. Chem., 2012, 638, 1804.
24. Y. H. Yoon and J. H. Nelson, AIHAJ, 1984, 45, 509.
25. J. T. Nwabanne and P. K. Igbokwe, Int. J. Appl. Sci. Eng. Tech., 2012, 2, 5.
26. K. Upendar, A. S. H. Kumar, N. Lingaiah, K. S. Rama Rao and P. S. Sai Prasad, Int. J. Greenhouse Gas Control, 2012, 10, 19.

---