The Inuence of Microwave Irradiation on Physicochemical Characteristics of Nitrogen-Enriched Carbon and Its Performance in Biogas Desulfurization

In this work, palm shell activated carbon (PSAC) was functionalized with nitrogen groups via urea impregnation, followed by the synthesis of microwave (MW) and conventional heating (TH) at temperature of 950 °C, 500 ml/min of N 2 ow rate and 30 minutes of heating time. The effects of MW and TH heating on the modied PSAC adsorbent were analyzed and compared towards hydrogen sulde (H 2 S) removal. The eciency of H 2 S removal was calculated based on adsorption capacity of the adsorbent samples. Nitrogen-functionalized PSAC that was synthesized via MW heating (PSAC-MW) has shown better performance with adsorption capacity of 356.94 mg/g. The chemical and physical characteristics of all adsorbent samples were studied and analyzed by using CHNS elemental analysis, N 2 adsorption-desorption analysis, SEM, TGA and FTIR analysis. The adsorbent sample that was synthesized via MW heating showed signicant characteristics, such as high surface area with sponge-like structure, in which there are additional pores developed inside the existing pores. Instead of that, there was an observation on ‘hot spot’ appearance during the MW heating process, which is believed to be responsible for the development of physical and chemical characteristics of the PSAC-MW adsorbent. These characteristics indirectly contribute to the high removal of H 2 S.


Introduction
In recent years, Malaysia is listed as the second leading of palm oil producer in the world. With the increase in the worldwide demand, the production of palm oil products indirectly lead to the environmental pollution. In 2015, up to 75 million m 3 of palm oil mill e uent (POME) has been generated annually (Ahmed et al. 2015). To encounter this challenge, POME has been used in anaerobic digestion as a feedstock for biogas production (Ahmed et al. 2015;Foong et al. 2021). In general, POME biogas can be classi ed as the source of greenhouse gases (GHG), with a composition of 65% of methane (CH 4 ), 35% of carbon dioxide (CO 2 ) and traces of hydrogen sul de (1000 -3000 ppm of H 2 S) (Asis et al. 2016).
The GHG potential of CH 4 is incredibly high, which is predicted to a hundred years of global warming potential (Unece 2021). As a result, this concern urges the researchers to bene t the utilization of CH 4 as a renewable energy. Prior to recover high energy content in the biogas, the traces of H 2 S need to be eliminated.
H 2 S is a very toxic and harmful gas with unpleasant smell, that is not only dangerous to human health but also cause detrimental effect to the gas engine and metal parts, due to its corrosive behaviour (Ou et al. 2020). To accommodate this problem, numerous methods have been explored, such as chemical oxidation, biological treatment, electrochemical abatement, catalytic conversion, precipitation, chemical scrubbing, incineration, and adsorption (Wiheeb et al. 2013). Among these methods, adsorption onto lignocellulose adsorbent has been widely employed due to its effectiveness in removing gaseous pollutant (Supanchaiyamat et al. 2019; Georgiadis et al. 2020).
Signi cant interest on the use of lignocellulose derived materials in the adsorption application is due to it is cheaper and abundantly available. Activated carbon (AC) derived from lignocellulose biomasses have been utilized for various gas pollutant adsorption, such as H 2 S, CO 2 and SO 2 (Nor et al. 2013; Supanchaiyamat et al. 2019). AC is a promising adsorbent due to its high adsorption capacity in capturing target pollutants. This is contributed by its high surface area, well-developed pore structure, and availability of various functional groups. Palm shell biomass consists of physical and chemical characteristics that are suitable to be converted into AC. The selection of palm shell as a precursor in this research work due to it is relatively economic and abundantly available (Phooratsamee et al. 2014;Habeeb et al. 2017;Nor et al. 2018). Furthermore, the conversion of this biomass into palm shell activated carbon (PSAC) could contribute to zero carbon emission, as palm shell and biogas are byproducts that were produced from the palm oil mill.
To craft different types of functional groups onto AC's structures, chemical activation/modi cation need to be introduced. The modi cation is responsible for the oxidation of H 2 S to be converted into elemental sulfur during the adsorption process (Zhang et al. 2018

Adsorbent Pre-oxidation
The PSAC was ground and sieved to 1 -2 mm size. It was pre-oxidized by using 50 wt% of HNO 3 for 4 h at ambient temperature. Then it was washed out with distilled water to remove excess acid and watersoluble products. The purpose of pre-oxidization of the PSAC is to enhance surface oxidation state of the adsorbent (Adib et al. 2000). The pre-oxidized PSAC sample was then dried at 110°C for 24 h, to remove excess water.

Urea Impregnation
Prior to the conventional thermal and MW heating, the pre-oxidized PSAC sample was impregnated with urea, to tailor nitrogen functional groups onto the adsorbent structures. The mixture containing preoxidized PSAC (30 g), urea (20 g), and ethanol (100 ml) was stirred (200 rpm) at room temperature for 5 h. Next, the mixture was boiled to evaporate the alcohol, and was dried at 110°C in a drying oven (Yang et al. 2019).

Conventional Thermal Heating
The conventional thermal heating of urea impregnated PSAC was synthesized by using a vertical-type furnace (Carbolite, CTF 1200). The heating parameters were set to temperature of 950°C with the heating rate of 15°C/min. The heating time was set to 30 min, under 500 ml/min of N 2 ow rate. The N 2 gas was supplied to provide inert condition for lignocellulose carbon throughout the heating process. To remove excess urea, the nal adsorbent sample was washed with boiling water and was dried in an oven at 110°C for 24 h. The sample was indicated as PSAC-TH.

Microwave Heating
MW heating was introduced to urea impregnated PSAC at temperature of 950°C with 500 ml/min of N 2 ow rate and 30 min of heating time. The sample was rst placed into a double-walled quartz annular column (inner diameter: 1.5 cm, outer diameter: 3.5 cm, height: 30 cm) prior to locate in a MW oven as illustrated in Fig. 1. The MW heating unit consists of a multimode domestic microwave oven (Cornell, CMO-EL17L) with a maximum power output of 1150 W and frequency of 2.45 GHz, a double-walled quartz annular reactor, mass ow meter, N 2 gas cylinder and temperature controlling unit. For MW's temperature controlling unit, a K-type thermocouple, 25 cm long with sheath diameter of 1.5 mm was placed inside the quartz annular column. The MW's heating temperature was maintained by modi cation of the power supply circuit that has been equipped with PID controller. The MW heating apparatus (magnetron) was modulated in an on-off cycle to maintain a constant MW's heating temperature. The nal product was washed and dried similar as PSAC-TH sample. The sample prepared by MW heating was indicated as PSAC-MW.
To study the effect of various MW heating temperature on the nitrogen-modi ed adsorbent, the MW temperature were varied from 500 to 900°C, with constant N 2 ow rate and heating time. The samples were indicated as MW500, MW600, MW700, MW800, and MW900.

H 2 S adsorption Analysis
A custom designed H 2 S desulfurization rig as shown in Fig. 2 was employed to study the performance of the modi ed adsorbents toward H 2 S removal. An amount of 0.75 g of the adsorbent sample was placed in the middle of a xed bed stainless-steel adsorption column (length: 450 mm, internal diameter: 12.5 mm) that was supported with approximately 0.5 g borosilicate glass wool. A simulated biogas that consists of 59.7 vol% of CH 4 , 40 vol% of CO 2 and 3000 ppm of H 2 S was passed through the column at constant ow rate and temperature of 200 ml/min and 30°C, respectively. The biogas mixture was simulated to imitate the actual composition of biogas collection pond in Felda Besout palm oil mill, Malaysia. The composition of the simulated biogas was manipulated by regulating the ow rate of individual gas using mass ow controllers (Aalborg, AFC26). To achieve 40% of relative humidity (RH) prior entering the column, the H 2 S gas was mixed with CH 4 and CO 2 gases that were passed through the humidi cation system at a water bath temperature of 40°C. The outlet concentration of H 2 S was analysed using an electrochemical H 2 S sensor (IMR500 gas analyser), that was calibrated for 0-5000 ppm range. To ensure the safety of workstation, a ventilation system equipped with exhaust fan was installed on top of the biogas desulfurization rig, to aerate the H 2 S gas to the fume hood scrubber.

Adsorbent Characterizations
The pH of the modi ed adsorbent was measured using a mixture of dried adsorbent (0.4 g) and water (20 mL) that was agitated for 24 h to reach equilibrium. The pH of the solution was measured by using desktop pH meter (Mettler Toledo). The elemental analysis was performed by using a Perkin Elmer 2400 CHNS Analyzer. N 2 adsorption-desorption test was done in an Autosorb-1-MP instrument (Micromeritics ASAP 2020) to determine the speci c surface area (S BET ), average pore diameter and pore volume of the adsorbent. The samples were heated at 120°C, followed by 5 h of degassing process. The proximate analysis (moisture, volatile organic matter, xed carbon, and ash) was carried out using thermogravimetric analyser (TA Instrument SDT Q600). Fourier Transform Infrared (FTIR) of nitrogenmodi ed samples were analysed using Thermo Scienti c (Nicolet iS10) equipment. The spectrum of the sample was displayed after 64 seconds with corrected background noise. The adsorbent surface morphologies were analysed by using of Scanning Electron Microscopy (SEM) (Quanta FEG 450) at an accelerating voltage of 3kV at magni cation of 5k.

H 2 S Adsorption Study
To understand the performance of nitrogen modi ed PSAC adsorbents towards H 2 S removal, the adsorbents that were synthesized via thermal heating (PSAC_TH) and microwave heating (PSAC_MW) were analysed by using H 2 S adsorption study. The physical and chemical characteristics of the modi ed adsorbents were investigated and will be discussed in detail in the following sections. Commercial raw palm shell activated carbon (PSAC) adsorbent was used as a baseline to evaluate the performance of the nitrogen modi ed adsorbents in this work. To be noted, the nitrogen modi ed adsorbents that were synthesized via conventional thermal and MW heating methods were performed under similar parameters.  Table 1.   The correlation between pore size distribution and their involvement onto the porosity of the PSAC-based adsorbents is shown in Fig. 6. The PSAC-MW has high pore volume for average diameter of < 20 Å of pore width, which exhibits well pore distributions. Based on the average pore diameter, the PSAC-MW pore width was narrower and had the highest pore volume compared to the PSAC and PSAC-TH. (Ren et al. 2020) suggested that carbon adsorbent that has small pore structure, preferably in micropore size are e cient for H 2 S adsorption to be happened. The micropore size of the PSAC-MW could con ne the H 2 S molecules within the pores whilst keeping the molecules from escaping from the carbon material. The H 2 S is a small molecule with a diameter of 3.6 Å (Shah et al. 2017), in which smaller pore structure is more preferrable for the physisorption and chemisorption to occur. This occurrence may possibly improve the ability of H 2 S adsorption, in which the reaction between the H 2 S and nitrogen functional groups could effortlessly happen via trapping the gas molecules into smaller and narrower pore structure. The shrinkage of pore width in the MW modi ed adsorbent provides better structures compared to the PSAC, where the H 2 S molecules could be trapped and adsorbed more. This theory can be applied for both PSAC-TH and PSAC-MW. The same trend was observed in the nitrogen modi ed PSAC-based adsorbents at different MW heating temperature (Fig. 7), where their micropore volumes observed were lesser than the PSAC. As increase in the MW heating temperature, the micropore volume of the adsorbents seem to be improved. Pore size distribution of the adsorbents at different MW heating temperatures indicates an increment in the micropore volume of pore widths < 40 Å.

Effect of Surface Chemistry
The results of proximate and elemental analysis of the PSAC-based adsorbents are presented in Table 3, whereas the elemental distribution of the adsorbents is illustrated in Fig. 8. (1000.62 m 2 /g) is another factor that allows the accessibility of nitrogen functional groups to be integrated into carbon's crystalline edges. This nding strengthens the theory that the use of MW heating possibly will enhance the AC characteristics that consists of high content of heteroatom (in this case nitrogen functional groups). The signi cant amount of nitrogen functional groups in the adsorbents helps in the adsorption of H 2 S. Consequently, adsorbent that has high nitrogen content is relatively promotes to high H 2 S adsorption capacity, as shown by the PSAC-MW (Table 1).  Fig. 8, a substantial amount of nitrogen content and an increase in carbon content were identi ed in all nitrogen-modi ed adsorbents, which consistent with nitrogen to carbon (N/C) ratio. The increase in carbon content in the nitrogen-modi ed adsorbent is relative to the raise of MW temperature. This happened due to the high degree of carbonization during the pyrolysis process of urea impregnated PSAC that experiences both thermal and MW heating, where the adsorbent encounters twofold of high temperature, where the rst one is during the production of AC (before urea impregnation). Nevertheless, as for hydrogen content, the amounts are signi cantly reduced as the MW temperature raised.
To investigate the in uence of urea impregnation and MW heating method on the surface chemistry of the PSAC-MW, FTIR analysis has been performed as illustrated in Fig. 9. All adsorbents show signi cant spectrums between bands 600 to 900 cm −1 , which indicates various positions of hydrogen in aromatic rings, mostly out of plane deformation of C-H group that was located at the edge of the aromatic group (Simons 1978

Effect of Thermal Analysis
Thermal stability analysis of the PSAC-based adsorbents is shown in Fig. 10 and Fig. 11. The DTG (differential thermal gravimetry) curve can be classi ed into three different ranges, which are from 80 -120°C, 150 -450°C, and 500 -950°C. The rst range of the DTG curve denotes the weight loss of moisture in the material, where it typically contributed to substantial percentage of weight loss (Bazan et al. 2016). The weight loss of absorbed water for the PSAC-TH is the highest (4.8 %/°C) followed by the PSAC (4 %/°C), and the PSAC-MW (3 %/°C). This implied that the PSAC-based adsorbents contain hydrophilic characteristic that tend to absorb water. Hence, the PSAC-based adsorbent is recommended to be stored in a desiccator that contains silica gel, to prevent the humidity in the air affecting or deteriorating the adsorbents. The second range of the DTG curve represents the decomposition of volatile organic matters that are easily volatile, presence on the surface of activated carbon such as carboxylic groups (Bazan et al. 2016). Fig. 10 displays that both the PSAC-TH and PSAC-MW experience higher weight loss of volatile organic matters compared to the PSAC. This is due to both the PSAC-TH and PSAC-MW were impregnated with urea, which had high intensity of carboxylic group (refer to Fig. 9). The nitrogen modi ed adsorbents were exposed at high temperature and experienced signi cant weight loss of 5 -6 %/°C. This happened because of rapid decomposition of urea-derived species and other organic components in the range of 500 -700°C as indicated in the DTG curves. These nitrogen functional groups were identi ed as amide, imide, nitrile groups and pyridine nitrogen. This explains the theory studied by (Seredych and Bandosz 2008), where the nitrogen-modi ed adsorbent that was heated at high temperature is very stable. As increases in the MW heating temperature from 500 -900°C for, the weight loss in the nitrogen modi ed adsorbents shows rapid decomposition of urea-derived species (nitrogen functional groups) and other organic components in the range of 500 -700˚C in the DTG curve. The DTG curve for the PSAC adsorbents at lower MW temperatures (MW500, MW600, MW700) shows signi cant weight losses compared to the MW900.

Effect of Surface Morphology
The surface morphology of the nitrogen-modi ed PSAC-based adsorbents is shown in Fig. 12. All adsorbents had been subjected to the SEM analysis at a magni cation of 5kX. Apparently, the nitrogen modi ed PSAC-based adsorbents showed different pattern compared to the PSAC. The surface morphology of the PSAC-TH in Fig. 12 (b) displays the pore structure had been clogged by excess urea/impurities and the size of the pores seems smaller compared to the PSAC structure in Figure 12 (a). As for the PSAC-MW in Figure 12 (c), the pore structure was well-developed and fewer pores had been clogged with urea/impurities as in the PSAC-TH. Furthermore, the carbon structure in the PSAC-MW established the sponge-like structure, where there is another pore were being constructed inside the existing pores and the pore size observed was smaller compared to the PSAC-TH and PSAC. As discussed earlier, narrow pore size could assist in H 2 S adsorption activity. Thus, the surface morphology of the PSAC-MW justi ed the relationship between physical surface characteristics data in Table 2 and its surface morphology, where it contributes to high H 2 S adsorption capacity. As reported by (Halasz et al. 2010), the use of MW heating method on AC could signi cantly change the amorphous structure of AC, where the cavities were deepened, and the inner structure of the adsorbent was remarkable. The development of surface area and narrow pore structures in the PSAC-MW were clearly noticed and this con rmed the hypothesis, where additional pores have been developed inside the pores (as observed from the surface morphology). These new added pores were assumed to have additional pore created inside the structure (sponge-like structure) as the mechanism of MW heating itself is irradiated from the inner to the outer surface of the material. Having evaluated the thermal heating method on the PSAC-TH, it appears to have a slight identical effect as the PSAC-MW on the morphology of the AC.
The surface morphology of the PSAC adsorbents that have been modi ed with urea under different MW heating temperatures is shown in Fig. 13. For the nitrogen modi ed PSAC adsorbents that undergo different MW heating temperatures, as increases in temperature, the fouling of urea on the adsorbent surface seems to be better, as well as the development of the pore structure. This caused in the increases of H 2 S adsorption capacity as increases in the MW heating temperature. Fig. 14

Conventional Thermal Heating versus MW Heating Method
Based on the above discussions, MW heating method shows promising performance on the H 2 S adsorption capacity compared to conventional thermal heating method. The great performance of the MW heating method can be related to its heating mechanism, where the theory was deduced by (Garcia Reimbert et al. 1996). Referring to the theory, a carbon material that was exposed to induced currents (origin from electromagnetic eld) would be heated due to electrical resistances within the material.
These electromagnetic and thermal properties of the material were non-linear temperature dependent. Thus, if the rate of material's absorption of MW non-linearly increased with heating temperature, a nonuniform heating would be created within the material, and indirectly formed regions of very high temperature known as hot spot.
The formation of the hot spot shown in Fig. 5 throughout the MW heating can be de ned as sparks or electric arcs, where it could be regarded as plasma at the microscopic level. According to (Rodriguez-Fernandez et al. 2020), these plasmas could be considered as microplasmas and they were observed as the very tiny sparks that could last for a fraction of a second. Activated carbon is a good MW absorber and it could enhance the hot spot formation. Since the temperatures of these hot spots were considerably higher, this occurrence increased the kinetic energy of π-electrons (make the electrons jump out of the substance) and ionized the surrounding atmosphere of the material (Menéndez et al. 2010). During the MW heating process, the hot spot was believed to be responsible for the development of physical and chemical characteristics of PSAC-MW. Referring to all the reasons, the MW heating method was chosen over the conventional thermal heating method.

Conclusions
The impregnation of urea onto the PSAC adsorbent shows remarkable in uences on the H 2 S adsorption, where the nitrogen functional groups were tailored. The use of the MW heating method is recommended, where it provide a great assist in the H 2 S adsorption activity compared to the performance of PSAC-TH that undergo conventional thermal heating method. Based on the H 2 S breakthrough capacity, the performance of the PSAC-MW is 100% better than the PSAC and PSAC-TH adsorbents. The development of physical and chemical characteristics of PSAC-MW adsorbent is primarily important and give a remarkable in uence in the H 2 S adsorption performance. From the ndings, the nitrogen modi ed PSAC via MW heating method could be a benchmark to further enhance the H 2 S adsorption activity.

Declarations
Ethical approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
acquisition, supervision: Abdul Rahman Mohamed. All authors have read and agreed to the published version of the manuscript.