Biodiesel production from marine macroalgae Ulva lactuca lipids using novel Cu-BTC@AC catalyst: Parametric analysis and optimization

Energy Conversion and Management: X 23 (2024) 100628 Available online 23 May 2024 2590-1745/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Biodiesel production from marine macroalgae Ulva lactuca lipids using novel Cu-BTC@AC catalyst: Parametric analysis and optimization Muhammad Zubair Yameen a , b , Dagmar Juchelkov´ a b , Salman Raza Naqvi a , c , * , Tayyaba Noor a , Arshid Mahmood Ali d , Khurram Shahzad e , Muhammad Imtiaz Rashid e , Aishah Binti Mahpudz f a Laboratory of Alternative Fuels and Sustainability, School of Chemical and Materials Engineering, National University of Sciences and Technology, Islamabad 44000, Pakistan b Department of Electronics, Faculty of Electrical Engineering and Computer Science, Vˇ SB – Technical University of Ostrava, 17. Listopadu 15/2172, Ostrava-Poruba 708 00, Czech Republic c Department of Engineering and Chemical Sciences, Karlstad University, Sweden d Department of Chemical & Materials Engineering, Faculty of Engineering, King Abdulaziz University, Jeddah, Kingdom of Saudi Arabia e Center of Excellence in Environmental Studies (CEES), King Abdulaziz University, Jeddah, Kingdom of Saudi Arabia f Department of Industrial Engineering, Jeddah International College, Jeddah, Saudi Arabia ARTICLE INFO Keywords: Biomass valorization Lipid-extracted algae Green catalyst Biofuel Macroalgal biorefinery Circular bioeconomy ABSTRACT The pursuit of renewable fuels for the transportation sector, particularly for combustion engines like diesel, is crucial in reducing greenhouse gas emissions. This study introduces an innovative strategy for biodiesel production utilizing marine macroalgae Ulva lactuca as the primary feedstock, emphasizing sustainability and resource efficiency. Lipids were extracted from the macroalgae via a Soxhlet process and characterized using GC–MS and FTIR to ascertain fatty acid composition and functional groups. The Cu–BTC@AC catalyst, synthesized from the lipid-extracted algae residue via pyrolysis and hydrothermal treatment, underwent characterization using SEM–EDS, XRD, and FTIR techniques. Subsequently, the Cu–BTC@AC catalyst was employed in the transesterification process to efficiently convert the extracted algal lipids into biodiesel, achieving a high yield of 92.56 % under RSM-optimized conditions: 65 ◦C temperature, 3.96 wt% catalyst amount, 15:1 methanol-to-lipid ratio, and 140 min reaction time. Kinetic and thermodynamic parameters for biodiesel production were calculated as follows: E a =33.20 kJ mol −1 , ΔH # =30.39 kJ mol −1 , ΔS # =–165.86 J mol −1 K −1 , and ΔG # =86.48 kJ mol −1 . GC–MS analysis identified a significant FAME content in the biodiesel, comprising 98.12 % of its composition. Notably, the Cu–BTC@AC catalyst exhibited excellent reusability, maintaining 80.21 % biodiesel yield after the third cycle. Moreover, physicochemical analysis of the biodiesel confirmed its compliance with ASTM D6751 specifications, underscoring its potential as a viable alternative fuel for the transportation sector. 1. Introduction The ever-increasing global population and rapid industrialization have led to a dramatic surge in demand for fossil fuels like coal, oil, and natural gas [1]. This escalating demand has not only depleted fossil fuel reserves but also raised significant environmental concerns [2]. Therefore, it is obligatory to establish clean and sustainable energy sources to effectively address the global energy crises and mitigate environmental issues [3]. Biofuels offer promising solutions to global challenges in the energy and environmental sectors [4]. Particularly, the utilization of marine macroalgae as a third-generation biomass feedstock has the potential to revolutionize the biofuel industry [5]. Marine macroalgae, which are multicellular plant-like species, can be cultivated in various aquatic environments, ranging from freshwater and seawater to even wastewater [6]. They play a pivotal role in decarbonization efforts owing to their carbon sequestration potential, as they can sequester approximately 1.83 t of CO 2 for every 1 t of biomass produced [7]. Additionally, these entities exhibit exceptional growth rates, leading to significantly higher biomass production compared to terrestrial plants [8]. Among these, Ulva lactuca, commonly known as sea lettuce, is abundant in coastal regions worldwide and represents a promising source for biodiesel production [9]. Ulva lactuca is rich in valuable biomolecules like lipids, proteins, carbohydrates, and minerals, all of which serve as precursors for various next-generation bioproducts, * Corresponding author at: Department of Engineering and Chemical Sciences, Karlstad University, Sweden. E-mail address: [email protected] (S.R. Naqvi). Contents lists available at ScienceDirect Energy Conversion and Management: X journal homepage: www.sciencedirect.com/journal/energy-conversion-and-management-x https://doi.org/10.1016/j.ecmx.2024.100628 Received 6 March 2024; Received in revised form 26 April 2024; Accepted 18 May 2024 Energy Conversion and Management: X 23 (2024) 100628 2 including biofuels and biocarbons [10,11]. Lipid extraction is a fundamental processing step in algal biodiesel production [12]. Several techniques can be employed for this purpose, including solvent extraction, enzyme-assisted extraction, expeller pressing, microwave extraction, and supercritical fluid (CO 2 ) extraction [13]. The solvent extraction method, using a Soxhlet apparatus, is widely recognized as a highly efficient process for extracting algal lipids [14]. The conversion of algal lipids into biodiesel via transesterification typically necessitates a suitable catalyst [15]. High levels of free fatty acids (FFA) in algal lipids pose a significant obstacle in biodiesel production [16]. When an alkali catalyst is used, FFA can react with it to form soap [17]. Indeed, the transesterification reaction using an alkaline catalyst requires a pretreatment step called esterification [18]. In this step, an acid, typically H 2 SO 4 , is utilized to react with the FFA, effectively removing them and preventing soap formation [19]. However, this additional esterification step does make the biodiesel production process more complex [20]. To overcome the challenges of saponification and simplify the biodiesel production process, the adoption of a highly active bifunctional catalyst is essential [21]. Biomass-derived activated biocarbons (AC) have proven exceptionally effective for biodiesel production, attributed to their high surface area, significant porosity, presence of heteroatoms, and abundant functional groups [22,23]. Similarly, copper-based metal–organic frameworks like Cu–BTC exhibit immense potential for catalyzing transesterification of oils with high FFA content due to their abundant active sites, tunable porosity, and tailored functionality [24]. The Cu–BTC@AC catalyst offers a synergistic combination of the advantages of both materials, potentially leading to efficient conversion of Ulva lactuca lipids into biodiesel. Significant gaps exist in prior research, particularly concerning the underexploited potential of residual biomass remaining after lipid extraction from macroalgae. These remnants hold promise as a valuable source material for synthesizing catalysts essential for converting algal lipids into biodiesel. Additionally, there is a lack of investigation into developing bifunctional catalysts capable of performing both esterification and transesterification reactions simultaneously, especially for macroalgal biodiesel production. Most importantly, the optimization of process parameters to maximize both lipid and biodiesel yields has not been adequately documented in existing literature. To fill the identified gaps, the present study introduces a novel approach for producing biodiesel from marine macroalgae Ulva lactuca lipids by employing a Cu–BTC@AC catalyst derived from the lipidextracted algae residue. The novelty of this research lies in its innovative approach to harnessing marine macroalgae as a dual source of lipids and catalyst for biodiesel generation, thereby creating an economically advantageous biodiesel technology. To this end, macroalgal lipids were extracted with the Soxhlet apparatus and characterized by GC–MS and FTIR techniques. The Cu–BTC@AC catalyst, synthesized from the lipidextracted algae residue, was characterized using SEM–EDS, XRD, and FTIR techniques. The Cu–BTC@AC catalyst was subsequently employed in the transesterification process to convert the extracted macroalgal lipids into biodiesel. The transesterification process was optimized by fine-tuning its parameters, including temperature (60–80 ◦C), catalyst amount (1–9 wt%), methanol-to-lipid (M/L) ratio (6:1–18:1), and time (60–180 min), employing response surface methodology (RSM), with the aim of achieving the highest possible biodiesel production yield. The kinetic and thermodynamic analyses were conducted to assess the energy efficiency of the transesterification process. The reusability performance of the Cu–BTC@AC was assessed through three consecutive cycles of biodiesel production. Analytical techniques, specifically GC–MS and FTIR, were employed to analyze the macroalgal biodiesel, aiming to identify the FAME compounds and the functional groups present. Moreover, the macroalgal biodiesel underwent physicochemical analysis, and the results were compared to the quality standards specified by ASTM D6751 for use in the transportation sector. 2. Materials and methods 2.1. Materials All the chemical substances consumed in the experiments, including methanol (CH 3 OH, 99.9 %), n-hexane (C 6 H 14 , 99.0 %), sulfuric acid (H 2 SO 4 , 98.0 %), ethanol (C 2 H 5 OH, 99.9 %), potassium hydroxide (KOH, 99.5 %), N,N-dimethylformamide (HCON(CH 3 ) 2 , 99.8 %), trimesic acid (C 6 H 3 (CO 2 H) 3 , 95.0 %), and copper (II) nitrate trihydrate (Cu (NO 3 ) 2 ⋅3H 2 O, 99.0 %), were obtained from Sigma-Aldrich. 2.2. Collection and preparation of marine macroalgae The marine macroalgae Ulva lactuca was harvested from Rawal Lake, Islamabad, Pakistan (33◦41 ′ 59.99 ″ N 73◦06 ′ 60.00 ″ E) in March 2022. The physicochemical conditions of seawater at the collection site, including temperature, pH, turbidity, and salinity, were recorded as 26 ◦C, 7.8, 5.5 NTU, and 30.9 PSU, respectively. The pretreatment procedures for the Ulva lactuca involved washing with tap water, followed by a drying process that included sun-drying for 3 d and further drying in a laboratory oven at 105 ◦C for 10 h. The dried macroalgae underwent grinding using an Electronica 400-A grinder and then were subjected to a sieving process to achieve a uniform particle size. Finally, the macroalgae were exposed to ultrasonication at 24 kHz for 7 min, maintaining a constant temperature of 50 ±1 ◦C, aimed at disrupting rigid cell walls and enhancing accessibility to intracellular lipids. 2.3. Lipid extraction from marine macroalgae The Soxhlet apparatus (FAT-4, BEGER) was utilized to extract macroalgal lipids using n-hexane as the solvent. The Soxhlet process was conducted within a temperature range of 60 to 80 ◦C, with varying solvent-to-solid ratios from 3:1 to 7:1, particle sizes ranging from 0.05 to 0.25 mm, and extraction times spanning 60 to 180 min. The solvent was reclaimed by a rotavap under the conditions of 54 ◦C temperature, 60 kPa pressure, and 250 rpm rotation. The lipid extraction yield was calculated using Eq. (1). Lipidextractionyield(%) = Weightoflipids(g) Weightofmacroalgae(g)×100 (1) 2.4. Characterization of macroalgal lipids The fatty acid (FA) profile of the extracted macroalgal lipids was evaluated using GC–MS (QP-2020-NX, Shimadzu). The functional group analysis of the macroalgal lipids was conducted through an FTIR spectrometer (IRTracer-100, Shimadzu). The titration method was employed to calculate the FFA value of macroalgal lipids, as per Eq. (2). FFA(%) = VKOH ×NKOH ×28.2 WLipids (2) 2.5. Catalyst synthesis from lipid-extracted algae residue 2.5.1. Synthesis of AC The biochar (BC) was produced from lipid-extracted algae residue using a pyrolysis process. The residual algae were subjected to pyrolysis in a tube furnace under an N 2 atmosphere at 550 ◦C, employing a heating rate of 5 ◦C/min and a residence time of 60 min. After the pyrolysis process, the BC was cleansed with ethanol to omit any contaminants and dried in an air oven at 90 ◦C overnight. To synthesize activated biocarbon (AC), 1 g of BC was introduced into 10 mL of sulfuric acid (H 2 SO 4 , 98 % purity). The resulting mixture was then stirred continuously at 180 ◦C for 240 min. Once the temperature had normalized, 100 mL of DI water was included in the mixture and stirred at 120 rpm for 30 min. The mixture was subsequently filtered and rinsed M.Z. Yameen et al. Energy Conversion and Management: X 23 (2024) 100628 3 with hot DI water until the pH reached a neutral level. Ultimately, the material was subjected to drying in an oven at a temperature of 110 ◦C overnight to obtain the porous AC. 2.5.2. Synthesis of Cu–BTC The Cu–BTC was fabricated using a hydrothermal synthesis technique. To initiate the synthesis, 3.1 g of Cu (II) nitrate salt was added to 22.5 mL of DI water while maintaining continuous stirring. Another solution was made by dissolving 1.5 g of benzene-1,3,5-tricarboxylate (BTC) in 45 mL of equal parts of DMF and ethanol, also with continuous stirring. The prepared solutions were delicately introduced into a Teflon-lined autoclave and set inside an air oven at 100 ◦C overnight. Subsequently, the resulting mixture was filtered using vacuum filtration, cleansed with ethanol and DI water, and dried under vacuum at 60 ◦C for 240 min. Finally, blue crystals of Cu–BTC were obtained. 2.5.3. Synthesis of Cu–BTC@AC For the synthesis of Cu–BTC@AC, 0.1 g of Cu–BTC and 0.9 g of AC were dispersed in 50 mL of equal parts of DMF and ethanol using sonication for 30 min. After sonication, the blend was continuously stirred for 60 min at ambient temperature. Following that, it was filtered under a vacuum, cleansed with ethanol and DI water, and later dried under vacuum at a temperature of 60 ◦C overnight. Finally, the Cu–BTC@AC composite was successfully obtained and utilized in the transesterification process to efficiently convert the extracted macroalgal lipids into biodiesel. 2.6. Catalyst characterization The surface morphology and elemental profile of the samples, namely AC, Cu–BTC, and Cu–BTC@AC, were explored by SEM–EDS using a JEOL JSM-6490A microscope with an EDAX METEK Z2-i7 detector. Crystalline structures were identified by collecting XRD patterns of these catalytic materials using a Bruker D8 Advance instrument. Additionally, FTIR spectra of the synthesized samples were captured within the range of 4000 to 400 cm −1 using a Shimadzu IRTracer-100 instrument. 2.7. Biodiesel production through transesterification process The transesterification reaction of macroalgal lipids was performed in a 3-neck flask furnished with a heating source, magnetic stirrer, condenser, and thermometer. For each experimental run, 10 g of macroalgal lipids were added to the flask, which was subsequently heated to 60 ◦C. Following that, the mixture of Cu–BTC@AC catalyst and methanol was introduced into the flask through one of the necks, and the reaction was conducted for the desired duration. After the accomplishment of the transesterification process, the resulting products were subjected to separation using a laboratory centrifuge (Z-366-K, HERMLE) operating at 5500 rpm. The centrifugation led to the creation of three distinct layers. The top layer comprised biodiesel, the central layer contained glycerol, and the undermost layer was composed of Cu–BTC@AC catalyst. The biodiesel was isolated and purified through successive washings with hot DI water. Subsequently, it was heated to 105 ◦C to remove any surplus water and residual alcohol from the biodiesel. Finally, purified biodiesel was obtained, and the yield was determined by employing Eq. (3). Biodieselyield(%) = Weightofbiodiesel(g) Weightoflipids(g)×100 (3) 2.8. Optimization of transesterification parameters using RSM–CCD The reaction parameters, including temperature (60–80 ◦C), catalyst amount (1–9 wt%), M/L molar ratio (6:1–18:1), and reaction time (60–180 min) were optimized by applying RSM–CCD to accomplish the highest possible biodiesel production yield. Design-Expert® V13 software was utilized for the RSM–CCD investigation. The symbols and levels of the reaction variables are displayed in Table 1. The design of experiments (DOE) was produced using RSM–CCD, which included 30 runs of experiments: 8 axial, 6 central, and 16 factorial experimental runs. The obtained biodiesel production yields were modeled using a quadratic polynomial equation, represented as Eq. (4). YBD =β0+∑ k i=1 βiXi+∑ k i=1∑ k j=i+1 βijXiXj+∑ k i=1 βiiX2 i(4) The interactions between reaction parameters were analyzed using 3D surface graphs. After that, the reaction parameters were optimized through a numerical approach to achieve the highest possible biodiesel yield. 2.9. Biodiesel characterization The FAME compounds of the macroalgal biodiesel were identified by analyzing it using a GC–MS (QP-2020-NX, Shimadzu). The GC settings included a 150 ◦C injection temperature, a 10:1 split ratio, a 10 ◦C/min ramping rate, a 280 ◦C final temperature, and a 15 min final hold time. The MS conditions comprised a 210 ◦C ion source temperature, a 260 ◦C interface temperature, a 5 min solvent cut time, a 70 eV EI energy, and a 40–1000 m/z scan range. The functional groups within the macroalgal biodiesel were analyzed by FTIR (IRTracer-100, Shimadzu). 3. Results and discussion 3.1. Lipid extraction and analysis The Soxhlet process yielded a maximum lipid content of 10.17 % from the marine macroalgae Ulva lactuca. The optimal extraction conditions were identified as 70 ◦C temperature, 5:1 (v/w) n-hexane to macroalgae ratio, 0.15 mm macroalgae particle size, and 120 min extraction time. Fig. 1 illustrates how each factor influenced the lipid extraction yield. Notably, temperature emerged as the most influential factor affecting lipid yield, followed by extraction time. The lipid yield obtained in this study (10.17 %) closely aligns with the findings of Suganya and Renganathan [25], Kalavathy and Baskar [9], Gurusamy et al. [26], and Binhweel et al. [10], who reported 10.88 %, 10.54 %, 8.50 %, and 7.60 %, respectively. The fatty acid (FA) composition of the extracted lipids was determined by GC–MS, and it showed a high proportion of mono-unsaturated FAs (78.85 %), followed by saturated FAs (17.12 %) and poly-unsaturated FAs (3.02 %). The prominent FAs found in macroalgal lipids were oleic acid (C18:1) with a composition of 77.11 %, palmitic acid (C16:0) with a composition of 14.12 %, and stearic acid (C18:0) with a composition of 2.06 %. Indeed, macroalgal lipids consist of rich unsaturated FAs with a total % of ~82 %. The functional group analysis of macroalgal lipids was conducted using FTIR (depicted in Fig. 7a). The prominent peaks around 2926 cm −1 are indicative of the C–H stretching vibrations of aliphatic hydrocarbons. This strong peak is a characteristic feature of all macroalgal lipids due to the abundant methylene (–CH 2 ) and methyl (–CH 3 ) groups in their fatty acid chains. The peak at 1713 cm −1 signifies the presence of carbonyl groups (C =O), which serve as the main linkage between fatty Table 1 Transesterification parameters and their corresponding symbols and levels. Parameters Symbols Levels – α –1 0 +1 + α Temperature (◦C) A 60 65 70 75 80 Catalyst amount (wt.%) B 1 3 5 7 9 M/L molar ratio C 6 9 12 15 18 Time (min) D 60 90 120 150 180 M.Z. Yameen et al. Energy Conversion and Management: X 23 (2024) 100628 4 acids and glycerol in triglycerides. Additionally, a distinct peak at 1461 cm −1 can be attributed to the C–H bending vibrations of methylene (–CH 2 ) groups, further supporting the presence of aliphatic hydrocarbons within macroalgal lipids. The observations regarding the fatty acid composition and chemical functional groups of macroalgal lipids are in full agreement with the findings reported by Nair et al. [27] and Pravin et al. [28]. Notably, the detection of carboxylic acid species by FTIR was in good agreement with the titration approach, which revealed a noticeable FFA content (7.58 %) in macroalgal lipids. The findings from the physicochemical analysis of macroalgal lipids are presented in Table 2. 3.2. Characterization of catalyst The surface morphology of the catalytic materials (AC, Cu–BTC, and Cu–BTC@AC) was observed through SEM analysis, and the findings are presented in Fig. 2. The AC exhibited an irregular porous structure, which can be ascribed to the liberation of volatile components during the thermochemical processing of residual algae. The high porosity of the AC can significantly enhance catalytic performance by facilitating efficient mass transfer between lipid molecules and the catalyst [29]. Cu–BTC possessed a bipyramidal crystal structure. Nevertheless, those structure changes occurred upon the incorporation of AC, suggesting that the Cu atoms are embedded onto the surface of AC. EDS analysis was conducted to ascertain the elemental compositions of these samples (Fig. 2). The AC sample demonstrated a significant C content of 58.3 wt % and O content of 40.4 wt%. This finding strongly affirms that oxygencontaining groups are the major functional groups on the AC surface. Interestingly, a trace amount of Cu (1.3 wt%) was detected, which was attributed to its origin from the macroalgae feedstock. EDS spectra also showed a peak corresponding to S, indicating the presence of –SO 3 H groups incorporated into AC through the activation process. The composition of Cu–BTC was found to be 37.9 wt% C, 23.0 wt% O, and 39.0 wt% Cu. Cu–BTC@AC was primarily composed of 50.6 wt% C, 44.5 wt% O, and 4.9 wt% Cu. This composition confirms the successful fabrication of the Cu–BTC@AC catalyst. The XRD patterns of the catalytic materials are illustrated in Fig. 3 Fig. 1. Influence of Soxhlet extraction parameters on lipid yield. Table 2 Characteristics of Ulva lactuca lipids. Parameter Value Acid value (mg KOH/g lipids) 15.16 ±0.82 FFA (%) 7.58 ±0.41 Saponification value (mg KOH/g lipids) 193.78 ±0.63 Kinematic viscosity (mm 2 /s) 11.85 ±0.14 Density (kg/m 3 ) 884 ±2.0 Iodine value (g I 2 /100 g lipids) 87.20 ±0.35 Average MW (g/mol) 863 ±1.10 Fatty acid composition (%) Myristic acid, C14:0 0.15 ±0.04 Palmitic acid, C16:0 14.12 ±0.08 Palmitoleic acid, C16:1 0.38 ±0.05 Stearic acid, C18:0 2.06 ±0.01 Oleic acid, C18:1 77.11 ±0.09 Linoleic acid, C18:2 1.21 ±0.01 Linolenic acid, C18:3 1.81 ±0.03 Arachidic acid, C20:0 0.79 ±0.01 Eicosenoic acid, C20:1 0.42 ±0.02 Erucic acid, C22:1 0.94 ±0.06 M.Z. Yameen et al. Energy Conversion and Management: X 23 (2024) 100628 5 (a). The characteristic peaks of AC were recorded at 2θ =25.5◦and 40.8◦, with corresponding miller indices (hkl) of (002) and (100), respectively, evidencing the existence of graphitic carbon in the AC sample. The characteristic peaks of Cu–BTC were observed at 2θ =6.7◦, 9.5◦, 11.6◦, 13.4◦, and 19.1◦, with their corresponding miller indices (hkl) of (200), (220), (222), (400), and (440), respectively, indicating the exceptional crystallinity present in the Cu–BTC. The Cu–BTC@AC sample exhibited a combination of characteristic peaks from both the Cu-BTC and AC components with varying intensities. Specifically, Cu–BTC peaks were observed at 6.6◦(200), 9.4◦(220), 12.1◦(222), and 18.9◦(440), while the AC peaks were detected at 25.4◦(002) and 41.2◦(100), confirming the coexistence of both components within the composite. All the observations regarding the XRD patterns are in full agreement with the findings reported by Zhang et al. [30]. FTIR analysis of functional groups in catalytic materials revealed distinct peaks indicative of specific functionalities (Fig. 3b). O–H groups were identified by peaks at 3416 cm −1 , while C–H stretching and bending modes of aromatic structures were observed at 2925 cm −1 and 1371 cm −1 , respectively. The presence of carbonyl groups (C =O) on aromatic rings was indicated by peaks around 1695 cm −1 . Notably, the formation of Cu–O bonds, essential for understanding the material’s catalytic activity, was confirmed by peaks at around 483 cm −1 . These observations align well with the findings reported by Rizvi et al. [31], validating the functional group composition of the prepared samples. 3.3. Optimization of biodiesel production using RSM–CCD 3.3.1. Development and analysis of regression model The design of experiments (DOE) for optimizing the biodiesel production process was conducted, and the results are included in Table 3. RSM recommended the quadratic model as the most suitable fit for statistical analysis based on the following key factors: sequential pvalue =<0.0001, LOF p-value =0.2596, adjusted R 2 =0.9782, and predicted R 2 =0.9454. RSM developed a coded equation to predict biodiesel yield utilizing a quadratic model, as shown in Eq. (5). YBD =81.52+2.75A+1.75B+3.62C+6.61D−0.22AB−1.65AC −2.26AD −0.28BC −0.77BD −1.12CD −2.57A2−2.31B2 +2.28C2−3.06D2 (5) Table 4 presents the ANOVA investigation results of the regression model. The quadratic model was considered statistically significant, Fig. 2. SEM–EDS analysis of catalytic materials. Fig. 3. (a) XRD and (b) FTIR spectra of catalytic materials. M.Z. Yameen et al. Energy Conversion and Management: X 23 (2024) 100628 6 supported by its substantial F-value (93.85) and p-value (<0.0001). The term “D,” representing reaction time, exhibited the most significant impact on biodiesel production yield, as indicated by its highest SOS value (1048.35) and F-value (606.27). The individual parameter contributions in the transesterification reaction were found to be 64.80 % for time (D), 19.40 % for M/L molar ratio (C), 11.24 % for temperature (A), and 4.56 % for catalyst amount (B). Among the interaction terms, the term “AD,” representing the combination of temperature (A) and reaction time (D), emerged as the most influential factor impacting biodiesel production yield. This is evident from its significant F-value (47.42) and p-value (<0.0001). The LOF was not found to be significant, as evidenced by its F-value =1.84 and p-value =0.2596. The statistical analysis, shown in Table 5, revealed a standard Table 3 Design of experiments (DOE) for biodiesel production. Run Factors Response Residual temperature (◦C) Catalyst amount (wt.%) M/L molar ratio Time (min) Actual biodiesel yield (wt.%) Predicted biodiesel yield (wt.%) 1 65 7 9 150 82.86 82.84 0.02 2 65 7 9 90 63.45 64.38 –0.93 3 70 5 12 120 81.30 81.52 –0.22 4 70 5 12 120 82.97 81.52 1.45 5 75 7 15 90 70.06 71.87 –1.81 6 75 3 15 150 75.39 75.82 –0.43 7 75 7 9 150 75.19 75.65 –0.46 8 70 9 12 120 77.80 75.81 1.99 9 70 5 6 120 74.45 75.42 –0.97 10 70 5 12 120 80.76 81.52 –0.76 11 60 5 12 120 77.52 76.74 0.78 12 70 5 12 180 82.50 82.49 0.01 13 65 7 15 150 89.09 90.56 –1.47 14 65 3 9 90 58.98 58.34 0.64 15 70 5 12 60 56.85 56.05 0.80 16 70 5 12 120 81.43 81.52 –0.09 17 65 3 9 150 80.31 79.86 0.45 18 65 7 15 90 76.19 76.59 –0.40 19 70 1 12 120 67.61 68.79 –1.18 20 75 7 9 90 66.60 66.25 0.35 21 80 5 12 120 65.75 65.72 0.03 22 75 3 9 90 61.21 61.10 0.11 23 70 5 18 120 91.67 89.89 1.78 24 70 5 12 120 80.17 81.52 –1.35 25 75 7 15 150 76.69 76.79 –0.10 26 75 3 9 150 74.51 73.56 0.95 27 70 5 12 120 82.51 81.52 0.99 28 65 3 15 150 88.90 88.70 0.20 29 75 3 15 90 68.36 67.84 0.52 30 65 3 15 90 70.77 71.67 –0.90 Table 4 ANOVA investigation of the regression model. Source SOS DF MS F-value p-value Remarks Model 2272.04 14 162.29 93.85 <0.0001 significant A-Temperature 181.94 1 181.94 105.22 <0.0001 B-Catalyst amount 73.78 1 73.78 42.67 <0.0001 C-M/L molar ratio 313.78 1 313.78 181.46 <0.0001 D-Time 1048.35 1 1048.35 606.27 <0.0001 AB 0.7921 1 0.7921 0.4581 0.5088 AC 43.43 1 43.43 25.11 0.0002 AD 81.99 1 81.99 47.42 <0.0001 BC 1.25 1 1.25 0.7254 0.4078 BD 9.39 1 9.39 5.43 0.0341 CD 20.12 1 20.12 11.63 0.0039 A 2 181.66 1 181.66 105.06 <0.0001 B 2 145.86 1 145.86 84.35 <0.0001 C 2 2.19 1 2.19 1.27 0.2779 D 2 257.43 1 257.43 148.87 <0.0001 Residual 25.94 15 1.73 Lack of fit 20.40 10 2.04 1.84 0.2596 not significant Pure error 5.54 5 1.11 Cor. total 2297.98 29 Table 5 Fit statistics of the regression model. Statistic Value Standard deviation 1.31 Mean 75.39 CV (%) 1.74 R 2 0.9887 Adjusted R 2 0.9782 Predicted R 2 0.9454 Signal-to-noise ratio 37.1142 M.Z. Yameen et al. Energy Conversion and Management: X 23 (2024) 100628 7 deviation of 1.31 and a mean biodiesel yield of 75.39 %. The regression model demonstrated a CV value of 1.74 %. It achieved a high R 2 value of 0.9887, along with an adjusted R 2 of 0.9782 and a predicted R 2 of 0.9454. With a signal-to-noise ratio of 37.1142, the model is capable of effectively navigating the design space. Fig. 4(a) shows good agreement between predicted and experimental biodiesel yields, with data points following the line of equality (x =y) [32]. Fig. 4(b) further confirms the accuracy and reliability of the quadratic model, with residuals randomly distributed around zero within ±3, indicating a strong positive correlation. 3.3.2. Impact of transesterification parameters on biodiesel production yield To assess the influence of transesterification parameters on biodiesel production yield, 3D surface plots were used for graphical analysis. Fig. 5(a) shows how temperature (A) and catalyst amount (B) influence the biodiesel production yield. The results indicate that increasing the temperature (A) from 60 ◦C to 68 ◦C, while keeping the catalyst amount (B) at 5 wt%, improved the biodiesel production yield from 76.73 % to 82.24 %. The other variables were fixed at their central values (12:1 M/L molar ratio and 120 min time). Fig. 5(b) illustrates the relationship between temperature (A) and M/L molar ratio (C) on biodiesel production yield. The findings indicate that elevating both the temperature (A) from 60 ◦C to 68 ◦C and the M/L ratio (C) from 6:1 to 18:1 resulted in a notable increase in biodiesel production yield. Previous studies have reported that higher reaction temperatures are likely to enhance collisions between reactant molecules by increasing their energy state [33]. Additionally, elevated temperatures can lead to an improved molecular diffusion rate due to the higher solubility and reduced viscosity of the reactants. As a result, these factors contribute to increased FAME conversion in the reaction [34]. At a temperature of 68 ◦C, biodiesel production yield enriched from 75.13 % to 91.68 % with an increase in M/L ratio from 6:1 to 18:1, while maintaining other variables at their central values (catalyst amount =5 wt% and time =120 min). Indeed, a larger amount of methanol provides greater diffusion and improved miscibility between the reactant phases [35]. Fig. 5(c) demonstrates the dynamic impact of transesterification temperature (A) and transesterification time (D) on biodiesel production. It was noted that when the time (D) was prolonged from 60 min to 148 min while keeping other variables at their central values, the yield at a temperature (A) of 68 ◦C upgraded from 55.11 % to 87.32 %. Longer reaction time improved diffusion between phases, promoting the diffusion of lipids to the active sites and accelerating the reaction rate, leading to a higher biodiesel yield. Fig. 5 (d) displays the combined impact of catalyst amount (B) and M/L ratio (C) on biodiesel production yield. It was noted that at a catalyst amount (B) of 5.5 wt%, the yield upgraded from 75.59 % to 89.98 % as the M/L ratio (C) enhanced from 6:1 to 18:1 while maintaining other variables at their central points. Fig. 5(e) displays the impact of catalyst amount (B) and time (D), and it was noted that with a catalyst amount (B) of 5.5 wt %, the biodiesel production yield improved from 56.81 % to 85.39 % as the time (D) extended from 60 min to 150 min. As expected, larger quantities of catalyst enhanced reaction rates by providing more catalytic sites in transesterification reaction [36]. Fig. 5(f) displays the synergistic impact of M/L ratio (C) and time (D), and it revealed that as the time progressed from 60 min to 148 min, while maintaining other variables at 70 ◦C temperature and 5 wt% catalyst amount, the biodiesel yield at an M/L ratio of 18:1 improved from 68.21 % to 90.97 %. 3.3.3. Optimization of transesterification parameters and model validation Transesterification parameters, which include temperature (A), catalyst amount (B), M/L molar ratio (C), and time (D), were optimized utilizing a numerical optimization tool built in RSM. The tool indicated that the optimum parameters to achieve a biodiesel yield of 90.69 % are as follows: 65 ◦C temperature, 3.96 wt% catalyst amount, 15:1 M/L molar ratio, and 140 min time. To validate the suggested parameters, experimental tests were carried out by performing transesterification reactions under the optimized conditions. Based on the experimental findings, it was remarked that the average biodiesel yield obtained under these optimized conditions was 92.56 %. The experimental data demonstrated a small error of 1.87 % when compared to the predicted yield, thereby confirming the efficacy of the optimized parameters in achieving a high biodiesel yield. Noteworthy to mention, the proven optimal temperature of 65 ◦C is also advantageous from an energy point of view. 3.4. Kinetic and thermodynamic investigations In the thermo-kinetic analysis of biodiesel production, the assumption was made that the reaction rate was predominantly governed by the concentration of triglycerides, considering the reversibility of the reaction [37]. A surplus quantity of methanol was utilized to promote the conversion of FAME while mitigating the potential impact of methanol on the overall process [38]. Therefore, the transesterification reaction was intended to adhere to the pseudo-first-order kinetic model, which can be represented by Eq. (6) [33]. Fig. 4. (a) Plot comparing predicted biodiesel yields against experimental yields, and (b) error values corresponding to the run order. M.Z. Yameen et al. Energy Conversion and Management: X 23 (2024) 100628 8 −d[TG] dt =k[TG](6) Fig. 6(a) illustrates the conversion of biodiesel at distinct temperatures across varying reaction times. Eq. (7) was utilized to construct a plot correlating the variables t and –ln(1 – X FAME ) at three distinct temperatures. From the plot depicted in Fig. 6(b), the rate constant (k) values were determined to be 0.2587 min −1 , 0.3068 min −1 , and 0.3669 min −1 at reaction temperatures of 333.15 K, 338.15 K, and 343.15 K, respectively. Furthermore, the obtained R 2 values of 0.9883, 0.9997, and 0.9803 at reaction temperatures of 333.15 K, 338.15 K, and 343.15 K, respectively, demonstrated the excellent accuracy of the model [39]. −ln(1−XFAME) = kt (7) The Arrhenius graph was constructed using Eq. (8) by plotting 1/T against ln k. From the plot depicted in Fig. 6(c), the E a value was noted to be 33.20 kJ mol −1 from the slope. Furthermore, the obtained value of E a was found to be very low compared to the findings by Huang et al. [40] and Mawlid et al. [41], indicating the energy efficiency of this process. ln(k) = − Ea RT +ln(Ar)(8) The Eyring-Polanyi plot was constructed employing Eq. (9) with 1/T plotted against ln k/T. From the plot depicted in Fig. 6(d), the ΔH # value of the transesterification reaction was attained as 30.39 kJ mol −1 from the slope, indicating the endothermic nature of the reaction [42]. The ΔS # value was determined to be –165.86 J mol −1 K −1 from the intercept of the plot, confirming the ordered mechanism of the transesterification reaction [43]. ln(k T)= − ΔH# RT +ΔS# R+ln(kB h)(9) Using Gibbs-Duhem Eq. (10), the ΔG # values at reaction temperatures of 333.15 K, 338.15 K, and 343.15 K were attained as 85.65 kJ mol −1 , 86.48 kJ mol −1 , and 87.31 kJ mol −1 , respectively, indicating that the reaction is nonspontaneous at a transition state and necessitates external actions, including heating and stirring, to form the FAME product [44]. ΔG#=ΔH#−TΔS#(10) 3.5. Biodiesel characterization and fuel properties The quantification of FAME compounds in macroalgal biodiesel was conducted through GC–MS. The predominant FAME constituents identified in the macroalgal biodiesel were methyl palmitate (C16:0) with a composition of 14.22 % and methyl oleate (C18:1) with a composition of 76.98 %. Table 6 presents the chemical composition of the macroalgal biodiesel, revealing that it predominantly consists of FAME compounds, which make up 98.12 % of the total composition. Among the identified FAME compounds, the macroalgal biodiesel was found to contain 79.17 % mono-unsaturated FAME, 15.82 % saturated FAME, and 3.13 % polyunsaturated FAME. The FTIR spectrum of macroalgal biodiesel, depicted in Fig. 7(a), exhibited a slight peak at 3460 cm −1 , representing O–H groups in residual alcohols from the transesterification process. Prominent peaks around 2928 cm −1 specify the C–H stretching vibration of aliphatic hydrocarbon chains present in the biodiesel. The presence of a strong peak at around 1742 cm −1 indicates the presence of ester group (C =O), confirming the successful conversion of triglycerides in macroalgal lipids to FAMEs in the biodiesel. The peak around 1459 cm −1 corresponds to the bending vibration of C–H bonds, indicating the presence of methyl groups (–CH 3 ) typically found in FAMEs. Additionally, the characteristic peak at 1168 cm −1 signifies the presence of ester linkages (C–O) within the fatty acid chains of the biodiesel. The observations regarding the FAME composition and functional groups of macroalgal Fig. 5. 3D surface plots illustrating the interactions between (a) temperature and catalyst amount, (b) temperature and M/L molar ratio, (c) temperature and time, (d) catalyst amount and M/L molar ratio, (e) catalyst amount and time, and (f) M/L molar ratio and time, in relation to biodiesel production yield. M.Z. Yameen et al. Energy Conversion and Management: X 23 (2024) 100628 9 biodiesel are consistent with previous investigations by Nair et al. [27] and Ravichandran et al. [45]. The HHV of the macroalgal biodiesel was measured as 42.75 MJ/kg using a bomb calorimeter (Parr, 6200). The observed flash point of the macroalgal biodiesel was recorded as 158 ◦C, while the fire point was recorded as 172 ◦C, utilizing a closed cup flash point tester (Seta, 13661–4). The abundance of unsaturated C18:1 acid chains in macroalgal biodiesel is responsible for its elevated flash point. It is worth noting that the flash point of macroalgal biodiesel meets the specifications outlined in ASTM D6751, and in most cases, it exceeds that of conventional diesel. This makes macroalgal biodiesel a safer fuel option compared to petroleum diesel. The cloud point of the macroalgal biodiesel was observed to be 2.5 ◦C, while the pour point was determined as –1.9 ◦C, utilizing a cloud and pour point apparatus (Koehler, K46100). The winterization properties of macroalgal biodiesel are higher than those of petroleum diesel, primarily due to the existence of oxygenated species in macroalgal biodiesel. Nevertheless, the low-temperature properties of macroalgal biodiesel still meet the acceptable ASTM D6751 specifications. All the observations related to the quality analysis of macroalgal biodiesel are displayed in Table 7. The fact that the synthesized macroalgal biodiesel meets the requirements of international biodiesel standards, its potential for commercialization appears to be very promising. 3.6. Catalyst reusability Cu–BTC@AC was isolated from the product mixture using a laboratory centrifuge operating at 5500 rpm for 30 min. The recovered catalyst underwent an extensive washing process using DI water and ethanol to eliminate any organic content and other impurities. After the washing Fig. 6. Thermo-kinetic analysis of FAME conversion over time at distinct temperatures: (a) X FAME versus time plot, (b) –ln(1 – X FAME ) versus time plot, (c) Arrhenius plot for determining E a , and (d) Eyring-Polanyi plot for determining ΔH # , ΔS # , and ΔG # . Table 6 FAME composition of macroalgal biodiesel. Sr. no. Carbon chain length FAME component Relative % 1 C14:0 Methyl tetradecanoate 0.10 ±0.02 2 C16:0 Hexadecanoic acid, methyl ester 14.22 ± 0.02 3 C16:1 9-Hexadecenoic acid, methyl ester, (Z)- 0.66 ±0.05 4 C18:0 Methyl stearate 1.34 ±0.00 5 C18:1 9-Octadecenoic acid (Z)-, methyl ester 76.98 ± 0.01 6 C18:2 9,12-Octadecadienoic acid (Z,Z)-, methyl ester 1.16 ±0.03 7 C18:3 Methyl linolenate 1.97 ±0.08 8 C20:0 Eicosanoic acid, methyl ester 0.16 ±0.00 9 C20:1 cis-11-Eicosenoic acid, methyl ester 0.67 ±0.02 10 C22:1 13-Docosenoic acid, methyl ester, (Z)- 0.86 ±0.01 M.Z. Yameen et al.