Towards Biowastes Valorization. Peanut Shell as Resource for Quality Chemicals and Activated Biochar Production


 The purpose of this work was to valorize a regional biowaste through a modest thermo-catalytic pyrolysis process. ZSM-11 zeolites modified by Ni incorporation (1–8 wt%) where synthesized and characterized by means of X-Ray Diffraction, Inductively Coupled Plasma Atomic Emission Spectroscopy, Infrared Fourier Transform Spectroscopy, UV–vis Diffuse Reflectance Spectra and Temperature Programmed Reduction. Results demonstrated that Ni was mainly incorporated as oxide. These porous materials were evaluated as heterogeneous catalysts to improve biooil composition. In this sense, higher hydrocarbon yields and quality chemicals were obtained and oxygenates were diminished. The deactivation of the most active material was studied over six cycles of reaction. In order to achieve the circular bioeconomy postulates, the obtained biochar (usually considered a residue) was further valorized through a physicochemical activation. The obtained activated biochars were extensively characterized.


Introduction 25
In its 17 Sustainable Development Goals (SDGs), the United Nations promotes the 26 principle of optimum and responsible usage of resources for purposes leading to a convincing 27 transition to a circular economy. The main idea of the so-called circular economy consists in 28 advanced redesigning and technological breakthroughs to minimize waste [1]. From these 29 ideas, the concept of "circular bioeconomy" is proposed to be a more efficient biobased 30 renewable resource management. This term integrates the circular economy principles with 31 the bioeconomy theory [2]. 32 Biomass is a promising alternative to fossil resources since its carbon richness. But 33 considering its importance for food production, a competitive situation should be avoided. In 34 this sense, agricultural wastes are ideal candidates for fuels and chemicals production. Peanut 35 is an important grain legume cultivated worldwide with a growing trend of production of 36 about 48 million tonnes [3]. Argentina is the 8 th world peanut producer, with around 1,3 37 million tonnes produced in 2019/2020 campaign. In view of these high quantities, it is 38 important to consider the wastes that peanut processing generates. A 25 wt% of the 39 production corresponds to the shell, a by-product not properly used. Keeping in mind the 40 concept of circular economy, peanut shell (PS) could be consider as the feed of an integral 41 process. 42 PS contains around 50% cellulose, 20% hemicellulose and 30% lignin [4] that make it 43 difficult to digest and unable as feedstock for animals. Furthermore, biodegradation is 44 extremely low, resulting in the need of severe treatments. Hence, fast pyrolysis seems to be 45 a good alternative in order to treat and give value to this biowaste. In a previous paper [5] we 46 have shown the importance of this thermal process. 47 As proposed by Dai et al. [6] the use of zeolites as catalysts in pyrolysis reactions could 48 improve hydrocarbons and aromatics yields. Particularly, they have found that Ni modified 49 zeolites promoted pyrolysis vapor deoxygenation and selectivity to aromatic compounds. 50 The use of Ni promoted catalysts in biomass pyrolysis has been studied by many researchers. 51 Most studies focused on ZSM-5, MCM or different oxides [7], but the use of Ni modified 52 ZSM-11 zeolite has hardly been investigated. 53 Infrared Fourier Transform Spectroscopy (FTIR) studies were done to determine 95 acidity of catalysts, employing a Thermo Nicolet iS10. Pyridine was first adsorbed to the 96 materials at room temperature under vacuum conditions and was further desorbed at 250 °C 97 and 10 -4 Torr. Acids sites quantification was done from the bands of 1545 cm -1 (Brønsted) 98 and 1450 cm -1 (Lewis) using the literature data of the integrated molar extinction coefficients 99 [10]. 100 The amount of coke deposited on the spent catalysts was measured by 101 thermogravimetric analysis (TGA) using a Mettler Toledo thermobalance 102 (TGA/SDTA851e/SF/1100•C). The sample was heated from 25 to 900 °C at a heating rate 103 of 10 °C min -1 under 75 mL min -1 of air flow. Eq. (1) was used to calculate the relative 104 amount of coke [11]. 105 Eq. (1) 106 Where 100° and 900° correspond to the catalyst mass at 100 °C and 900 °C, 107 respectively. 108 Functional structure of spent catalysts was characterized by XRD and FTIR using the 109 equipment depicted above. In the case of FTIR, the KBr technique was employed. 110

Pyrolysis reactions 111
PS pyrolysis and biooil upgrade were simultaneously carried out in a fixed bed glass 112 reactor (23 mm I.D., 290 mm length) under N2 flow (60 mL min -1 ). For a typical experiment, 113 PS (1 g) was loaded in a glass sample carrier over a catalytic bed. The bed consisted of 114 catalyst (1 g) and milled quartz (7 g). The reactor was introduced in an electric furnace once the pyrolysis temperature (500 °C) was reached. Condensable vapors were collected at the 116 reactor output in a condenser (<-10 °C). The reactions lasted 10 min. Three repeats of every 117 experimental run were done. Average values are reported. 118 Eqs.
(3) 121 Where, is the initial mass of the biomass sample (g), ℎ is the mass of the 123 solid product (g) after the reaction and is the mass of the liquid product (g). 124 Catalyst stability was assessed by using one catalyst sample for six consecutive 125 reaction cycles. Every new reaction was carried out employing the catalyst from the 126 preceding cycle (i.e., the partially deactivated catalyst) maintaining the 1:1 biomass to 127 catalyst mass ratio. Biooil was collected and analyzed after each reaction cycle. The heating rate was 20 °C min -1 until 750 °C, maintaining that temperature for 3 h. The as-145 prepared materials, named AB 1:1 and AB 3:1, were then neutralized and washed with 146 distilled water until pH=7 and dried in an oven at 120 °C until constant weight. 147 AB yields were calculated employing Eq. (6), where ℎ corresponds to the mass 148 of the biochar (g) after the pyrolysis reaction, is the mass of the as-prepared activated 149 biochar (g) after the synthesis procedure. 150 Eq. (6) 151 The crystalline structure, FTIR spectra and SBET of AB were determined in the 152 equipment depicted above. For FTIR analysis, the samples were prepared by blending a few 153 milligrams of the AB sample with KBr. 154 Morphological analyses were done by Scanning Electron Microscopy (SEM) 155 employing a microscope FE-SEM Σigma. It was operated at an acceleration voltage of 5 kV. 156 Prior to analysis, the samples were coated with gold. Proximate analysis of AB was also 157 performed using the thermobalance depicted above and following Saldarriaga et al., [13] 158 protocol. 159 A LabRam spectrometer (Horiba-Jobin-Yvon) coupled to an Olympus confocal 160 microscope was used to obtain the Raman spectra of AB samples. The spectrometer was 161 equipped with a CCD detector at ~ 200 K, the excitation wavelength was 532 nm and the 162 laser power was set at 30 mW. 163

Catalysts characterization 165
The physicochemical properties of the catalysts are shown in Table 1. It can be seen 166 that surface area of fresh materials varied as a function of Ni content. In general, when metal 167 is loaded on porous supports, the surface area decreases due to the pore blockage and metal 168 sintering during the calcination step [14]. SBET for spent catalysts showed a 23-35% decrease, 169 compared to the fresh samples. It was observed that the reduction in SBET was bigger when 170 greater the initial SBET was. 171 The ICP analysis of the materials confirmed the theoretical Ni quantities, which were 172 in agreement with the experimental results. 173 The FTIR spectra of the Ni-zeolites samples desorbed at 250 °C are shown in Figure  175  spectra increased with Ni contents. The peak centered at lower temperatures corresponds to 201 the highly dispersed Ni species, which have weak interactions with the support [20]. This 202 peak can be attributed to the reduction of NiO to metallic nickel. It is noticeable that at higher 203 metal content the signal became wider as consequence of diffusional limitations for bigger 204 oxide particles. The second peak, centered at higher temperatures, can be assigned to Ni 205 species with stronger interaction with the zeolite surface [18]. It is possible to infer that these 206 species are also present in Ni(5)Z, but they are more prominent at higher metal content. 207 Al, while the shoulder at 250 nm has been assigned to highly ordered structures with 211 octahedral symmetry [21]. All Ni modified samples present the characteristic band of NiO 212 around 300 nm. In the case of Ni(8)Z there is a shift in this band to higher wavelength that 213 may correspond to higher oxide particles [22], in accordance with XRD and TPR results. 214 This observation confirmed the previous analysis made for the acid sites of the materials. 215 Ni (8)

Catalytic activity 223
In order to determine Ni loading effect in ZSM-11 catalytic activity for biomass 224 pyrolysis, a series of experiments were performed.

Biooil upgrade 244
The effects of the Ni modified zeolites on the product distribution of PS pyrolysis is 245 presented in hydrocarbons during the catalytic improvement of the pyrolysis vapors [28]. 267 It should be noted that the hydrocarbon fraction was mainly composed by aromatics 268 (Figure 3b). It is well known that aromatization reactions require synergies of Brønsted and 269 Lewis acid sites. Although BAS are active sites for aromatization reactions, dehydrogenation 270 or hydrogen atom transfer reaction occurs on the LAS [29]. As presented in Table 1 Ni (5) is evident that 5-HMF was promoted by all catalysts since a complete absence of this 293 compound was observed in the thermal reaction. Approximately 50% of furans corresponded 294 to furfural and 5-HMF. Similarly to hydrocarbons, Ni(5)Z was the most active catalyst in 295 favoring the formation of 5-HMF. In the chemical pathway for cellulose transformation to 5-296 HMF, Brønsted acidity is believed to promote the depolymerization of oligosaccharides to 297 monomeric anhydro-sugars [32]. The isomerization to fructose seems to occur in the presence 298 of LAS and the dehydration is thought to be promoted by BAS [33]. 299

Catalyst deactivation 300
Coke nature of spent catalysts was studied by FTIR spectroscopy (Figure 5a). An 301 intense band at 3434 cm -1 could be observed in all samples. That band is attributed to bridging 302 hydroxyl groups, while the smaller band at 3234 cm -1 is attributed to H-bonding between 303 acidic hydroxyl groups and adsorbed molecules. The bands between 2800-3000 cm -1 are 304 attributed to the CH stretching modes (symmetric and asymmetric) of CH3 groups. The bands 305 from 1300-1700 cm -1 can be assigned to the CH bending of paraffinic groups and the CC 306 stretching modes of unsaturated groups. The bands around 1450 -1700 cm -1 are mainly 307 attributed to aromatics structure vibration [34]. Figure 5b shows FTIR spectroscopy of fresh 308 HZ where a total absence of carbon signals was observed. 309 Catalyst stability was evaluated with six consecutive reaction cycles. Since products 310 yields (biooil, biochar and gases) did not suffer significant variations, six main compounds, 311 including furans and aromatics, were chosen to test the zeolite deactivation. The most active 312 catalyst -Ni(5)Z-was selected and its re-uses are presented in Figure 6. Results showed 313 partially deactivation just in the second cycle. This behavior can be explained by coke 314 deposition. From Table 1, it could be observed how SBET decreased in all spent catalysts. In 315 the case of Ni(5)Z, an area loss of about 100 m 2 g -1 was obtained after the first use. 316 From Figure 6 it is possible to confirm an increment on furfural selectivity as reuses 317 cycles advanced. However, the other compounds analyzed were reduced (Selectivities < 1 318 wt%). This behavior could be explained by the surface area reduction previously commented. 319 When zeolites pores were physically blocked by coke deposition, pyrolysis products could 320 not get in contact with the catalyst active sites. Thus, reaction products resulted to be quiet 321 similar to non-catalytic reactions. Besides, Renzini et al. [19] found that coke deposition over 322

ZMS-11 zeolites caused a decrease of both BAS and LAS. 323
However, it is noteworthy that upon deactivation Ni(5)Z could be easily regenerated 324 by calcination. After that, catalytic activity was completely recovered. Figure S1 presents 325 XRD pattern of the regenerated catalyst, confirming ZSM-11 structure and crystallinity. 326 Coke deposition resulted in a temporarily catalyst poisoning and acid sites blockage that 327 could be thermally solved. 328 Proximate analysis revealed that fixed carbon in the activated biochars was higher than 334 in the starting biomass, which resulted from an effective carbonization process [36]. As the 335 temperature increases, it is expected that the carbon content increases and that of the volatiles 336 decreases, since devolatilization processes predominantly occur [37]. From the analysis of 337 activated materials, it could be observed that the fixed carbon content decreased as the surface 338 area increased, thus registering AB 3:1 sample the lowest fixed carbon value. Moisture 339 content in this sample was almost 4% higher, suggesting that it is slightly more hygroscopic 340 than AB 1:1. 341  (Figure 7). All specimens presented a broad band located at 2θ = 20-30° 344 which is characteristic of amorphous materials, suggesting the existence of amorphous 345 carbon caused by incomplete carbonization [38]. The peak at 2θ =23.5° corresponds to the 346 (002) graphite planes. From Figure 7a it can be observed that this broad peak is stronger in 347 biochar and weaker in the AB samples, proving an increase in the degree of graphitization. 348 the one at 1750 cm -1 resulted from the presence of -COOH groups. While the band at 2800-361 2900 cm -1 was responsible for the stretching of aliphatic CH, the broad band at 3400 cm -1 362 was assigned to the vibration of -OH, corresponding to adsorbed water molecules [35,40]. 363 SEM micrographs of the as-prepared materials are presented in Figure 9. AB 1:1 364 sample presented a rib-like structure (Fig. 9a) on which the presence of numerous dispersed 365 pores could be observed. As the KOH content increased, the precursor (biochar) continued 366 decomposing, and the formation of many new macropores could be observed in AB 3:1 367 sample (Fig. 9b). Thus indicating that the ribs observed in AB 1:1 were transformed into 368 channels in AB 3:1. Consequently, the opening of the macropores possibly contributed to the 369 formation of new micro and mesopores on the internal surfaces. 370

Conclusions 371
With aims for a circular bioeconomy, the valorization and recycling of PS in two 372 subsequent product systems was studied. The first system consisted of catalytic fast pyrolysis 373 of the residual biomass where a group of Ni-ZSM-11 matrices was tested. When varying Ni 374 content, a surface area reduction was obtained, but acid sites were improved. From the 375 evaluated materials, Ni(5)Z showed the best results in terms of hydrocarbons and platform 376 molecules selectivities. This catalyst was further measured in terms of stability over several 377 reaction cycles. The temporary poisoning was easily solved by calcination, after which the 378 material recovered its pristine crystallinity and catalytic behavior. Thus, it is possible to 379 affirm that nickel ZSM-11 zeolites with a 5 wt% of loading are ideal catalysts for the 380 pyrolytic conversion of peanut shell to interesting platform molecules. 381 The second system consisted of synthetizing AB employing the residual biochar from 382 the previous pyrolysis as the precursor. The materials were produced by a simple thermo-383 chemical procedure, employing KOH as activation agent. Surface area (SBET) could be 384 significantly increased upon activation, from 215 m 2 /g in biochar to 1645 m 2 /g when 385 KOH:biochar ratio was 3:1. Considering the type and quantity of surface functional groups 386 found on these materials, they could be used for a variety of sorption processes. 387 These results proved to be an example of an efficient biobased renewable resource 388 management. 389        Selectivity towards furans, furfural and 5-HMF in biooils from PS pyrolysis catalyzed by Ni-zeolites.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Fermanellietal.SupportingInformation.docx