Prior to biodiesel preparation, the acid value of the raw oil was determined to establish whether transesterification or esterification would be used. To determine the acid value, 10 g of oil (croton/WCO) was dissolved in 50 ml of ethanol in a 150 ml beaker and heated to enhance the solubility. A few drops of phenolphthalein indicator were added, and the solution was titrated to the endpoint, which was characterized by the persistent pink colour of the mixture [24]. The low acid value of waste cooking oil (3.57 mg KOH/g) and Croton oil (4.48 mg KOH/g) signified that the transesterification process could be directly used for biodiesel production.

In this process, 200 ml (170.34 g) of refined WCO was poured into the reactor and heated to 60 °C. Then, 1% (w/w) of KOH (relative to the oil mass) was dissolved in 100 ml of methanol (50% v/v relative to the oil). The methanol-KOH mixture was added to the preheated WCO, and the reaction was maintained for 60 minutes under continuous stirring [17]. Croton biodiesel was produced by initially preparing a 1% w/v solution of KOH in methanol, which was then mixed with Croton oil.

The resulted mixture was heated for 60 minutes under continuous stirring at a constant temperature of 60 °C [17], [25]. Upon completion, all reaction products were transferred to a separation funnel and left overnight to allow the phase separation of biodiesel and glycerine. The biodiesel, containing impurities such as glycerine, methanol, and KOH catalyst, was then washed several times with warm distilled water at 50 °C to remove impurities. The remaining methanol and water in the biodiesel were removed by heating the product to 110 °C and holding it at this temperature until all water bubbles vanished. Eq. (1) was used to determine the biodiesel yield under varying transesterification conditions, while Figure 1 summarizes the process of biodiesel production [16], [17], [19]:

Figure 1.

Schematic diagram of the transesterification process for biodiesel production

The equation suggests that yield is highly sensitive to oil-methanol ratio, which influences transesterification kinetics, methanol-oil miscibility, and catalyst activity. Using this equation allows quantitative comparison of biodiesel output at different oil-methanol ratio, ensuring selection of the optimal operating condition.

To investigate blending effects, various fuel mixtures of petroleum diesel, WCO and Croton biodiesel, and nanoparticle blends outlined in Table 3 were studied. Different WCO and Croton biodiesel ratios were tested to find the optimal ratio for favourable properties, as shown in the supplementary information. To form WCB, an 80% WCO biodiesel and 20% Croton biodiesel blend was used. To produce B20, 20% WCB was mixed with 80% petroleum diesel and stirred at 1500 rpm for 20 minutes.

To study the effect of nanoparticles, 1 litter of B20 blend was mixed with different nanoparticle concentrations (50 ppm, 75 ppm, and 100 ppm) based on related previous studies [26]. The required nanoparticle (GNPs and MWCNTs) mass was accurately weighed using a high-precision ENTRIS224-1S analytical balance with an accuracy of 0.0001 g. To ensure uniform dispersion of nanoparticles in the fuel, the mixture was subjected to ultrasonication for 20 minutes at a frequency of 24 kHz using a Hielscher ultrasonicator (Model CL-188).

Gas Chromatography-Mass Spectrometry was performed with GCMS (Model-QP-2010 SE) to analyse the chemical composition of Croton oil and waste cooking oil (WCO). FTIR, was identified as functional groups of raw Croton oil, WCO, and biodiesel products, including 100% Croton (C100) and waste cooking biodiesel (W100), with/without nanoparticles and petroleum diesel [27].

The physicochemical properties of D100, B100, and B20, as well as B20 blends doped with graphene and multi-walled carbon nanotubes (MWCNTs) at concentrations of 50, 75, and 100 ppm, were evaluated. Density, viscosity using a Viscometer, Model AN-823 m, calorific value with a Bomb Calorimeter, Model C200/3/1, flash point using a Pensky Martens apparatus, Model K16270, water content, iodine number, and cold filter plugging point were also conducted. The measured properties were then compared with the standard specifications of EN 14214 and ASTM D6751 to determine the effects of nanoparticle additives on the fuel blends.

The morphology of carbon-based nanoparticles (CBNPs) was examined using scanning electron microscope, and transmission electron microscope, SEM (JEOL, JSM-6010 LV) and TEM (JEOL JEM-2100 F), while their structural analysis was performed using X-ray diffraction, XRD (Shimadzu XRD-6100). The average particle sizes of the graphene nanoparticles and MWCNTs were evaluated using the Scherrer equation as illustrated in eq. (2) [28], [29]:

It indicates the effective surface area of nanoparticles, which dictates their thermal conductivity and catalytic enhancement of combustion, depends on crystallite size. Since biodiesel combustion and viscosity are temperature-dependent, this equation indirectly links nanoparticle structural stability to their performance in biodiesel blends under varying thermal conditions.

The phase composition and crystal structure of the nanoparticle are examined using X-ray diffraction as described in Figure 3. Figure 3A and Figure 3B depict the XRD pattern of the graphene nanoparticles and MWCNTs, respectively.

Figure 3.

The XRD pattern of the graphene nanoparticles (A), and MWCNTs (B)

The formation of -OH and -COOH groups is responsible for the intense peak (2 theta) at around 25.7° ‒ 26.3° indexed to (002) hkl values in both XRD spectra, confirming the highly graphitic structure of the Graphene and MWCNT particles [12]. Graphene sheets are stacked in a concentric cylindrical shape, and the nanotubes are multi-walled. The peak at 44.3° ‒ 46.3° in both spectra, which corresponds to (100) hkl values, is associated with JCPDS fl.no: 41 ‒ 1487, which validates the preservation of graphite structure following the reduction procedure process [27]. The average particle sizes of the Graphene nanoparticles and MWCNTs determined by the Scherrer equation were 24.2 nm and 1.5 nm, respectively.

Figure 4 shows the TEM images of graphene nanoparticles and MWCNTs. The GNPs are spherical and sheet-like (Figure 4A) while MWCNTs are tubular in shape (Figure 4B).

Figure 4.

TEM images of Graphene nanoparticles (A), and MWCNTs (B)

The special qualities and benefits: of graphene nanoparticles include exceptional mechanical and electrical characteristics, exceptional charge carrier mobility, high thermal conductivity, high surface area, and exceptional mechanical strength [28]. The properties of high thermal conductivity and high surface area have been reported to enhance the fuel quality and improve engine combustion characteristics [29]. Using high-magnification TEM, the outer diameters of MWCNTs were measured using the point-to-point inbuilt measuring tool and found to be in the range of 12mm to 35 nm.

Figure 5 shows the surficial characteristics of the nanoparticles. Figure 5A shows the SEM images of graphene nanoparticles, while Figure 5B shows the SEM images of the MWCNTs.

Figure 5.

SEM images of G (A), and MWCNTs (B)

The numerous lamellar layer structures of Graphene nanoparticles are visible, and the SEM images clearly show the margins of separate sheets as shown in Figure 5A. The asymmetric images show confined spaces stacked on top of one another. Additionally, it is observed that the GNP sheets have thicker edges, which could be attributed to the oxygen-containing functional groups that are joined at the margins of graphene nanoparticles [30]. The MWCNTs are tubular in shape and are arranged in bundles with smooth surfaces (Figure 5B) [31].

The chemical compositions of fatty acid methyl esters (FAMEs) derived from WCO and Croton oil are presented in Table 5. The results reveal clear differences in their saturation levels, which strongly influence fuel properties and stability. WCO FAME is dominated by monounsaturated fatty acids, with oleic acid methyl ester (C18:1) as the major component (66.8%), followed by palmitic acid methyl ester (C16:0, 22%) and stearic acid methyl ester (C18:0, 9.1%). These saturated and monounsaturated fractions account for more than 97% of the composition. The relatively high proportion of oleic acid provides WCO FAME with enhanced oxidative stability and resistance to rancidification, while the presence of saturated C16:0 and C18:0 esters contribute to a higher cetane number and improved combustion stability, albeit at the cost of poorer cold flow behaviour due to crystallization at low temperatures. Croton oil FAME exhibits a markedly different profile, being highly unsaturated, with linoleic acid methyl ester (C18:2) as the predominant component (70%), along with notable fractions of oleic acid (C18:1, 9.2%) and stearic acid (C18:0, 10%). The high degree of polyunsaturate enhances cold flow properties but makes Croton biodiesel more prone to oxidative degradation, reducing storage stability and increasing susceptibility to rancidification. Interestingly, Croton oil also contains small amounts of non-standard fatty acids such as C19:0 (2.80%) and long-chain unsaturated C20:1 (1.9%), which may influence fuel lubricity and thermal behaviour.

These compositional differences indicate that WCO FAME, with its higher oleic acid content, offers better oxidative stability and combustion efficiency, while Croton oil FAME, due to its high linoleic acid proportion, contributes superior cold flow performance but at the expense of storage stability. Thus, blending the two feedstocks balances these complementary properties, improving overall biodiesel quality by combining the oxidative stability of WCO with the cold flow benefits of Croton oil.

FTIR analysis was conducted on Croton oil, WCO, C100, W100, B20, and B20 blends with varying concentrations of nanoparticles (50 ppm, 75 ppm, and 100 ppm) in the 400 – 4000 cm⁻¹ range as shown in Figure 6. FTIR analysis, as shown in Figure 6A, presents the spectra of diesel, WCO, C100, W100, and B20 to identify functional groups of the fuels. A distinct peak at ~3741 cm⁻¹ was observed in diesel, C100, WCO, W100, and B20, which corresponds to O–H stretching vibrations (COOH groups), while Croton oil lacked this peak, showing no evidence of hydrogen bond formation [32]. All fuels displayed strong peaks at 2929 and 2858 cm⁻¹, indicating the presence of C–H asymmetric stretching vibrations of CH₂ and CH₃ groups, characteristic of the alkane family [19]. Additional peaks were observed at 1745 cm⁻¹ (C=O in esters [33]), between 1451 – 1460 cm⁻¹ (CH₂ bending [34]), and at 1165 –1175 cm⁻¹ (C–O vibrations [35]). WCO and C100 also showed a prominent peak at 720 cm⁻¹, related to C–H out-of-plane deformation and C-S stretch [36].

Figure 6.

FTIR spectra of diesel, Croton oil, WCO, C100, W100, B20 (A), B20 with GNPs (B), and B20 with MWCNTs (C)

Figure 6B and Figure 6C represent the FTIR spectra of biodiesel blends with varying concentrations of nanoparticles. The addition of graphene and MWCNTs at different levels (50 ppm, 75 ppm, and 100 ppm) influenced the composition of the B20 blend, as evidenced by variations in peak intensities. Notably, peaks appearing between 3731 cm⁻¹ and 3740 cm⁻¹ for B20, graphene, and MWCNT-doped biodiesel blends are more appropriately assigned to O–H stretching vibrations of COOH groups, rather than to C–H asymmetrical stretching. Significant peaks were also observed between 2856 cm⁻¹ and 2936 cm⁻¹, corresponding to the C–H stretching vibrations of methylene (CH₂) and methyl (CH₃) groups, which are typical of alkanes [16]. This observation aligns with previous results reported by Gawonou et al. [19] and Nespeca et al. [36].

Additionally, a strong peak was recorded at 1745 cm⁻¹, characteristic of C=O stretching in ester groups [37]. Nandiyanto et al. [38] reported that peaks in the 1800 – 1700 cm⁻¹ range correspond to ester carbonyl vibrations, consistent with our findings and those of Wambui et al. [21]. Slightly pronounced peaks were also observed between 1374 cm⁻¹ and 1460 cm⁻¹, attributable to C–H bending vibrations of CH₂ in alkanes [39]. Weak peaks around 1165 cm⁻¹ represent C–O stretching in esters, which El-seesy et al. [40] indicated occur between 1125 –1195 cm⁻¹. Furthermore, C–H out-of-plane bending vibrations were observed in all fuels between 720 – 734 cm⁻¹ [21].

The incorporation of graphene and MWCNTs caused subtle shifts in peak positions. For B20 blends with graphene, signal intensities increased with higher nanoparticle concentrations (50–100 ppm). Similar results were observed in MWCNT-doped fuels, where the peak wavenumbers were slightly higher than in graphene-doped blends. Additionally, MWCNT-doped fuels showed progressively increased peak intensity with higher doping levels. These shifts and changes in peak characteristics highlight interactions between nanoparticles and biodiesel, suggesting possible catalytic or surface-mediated mechanisms. This supports the role of carbon-based nanoparticles in improving the composition and functional group characteristics of biodiesel – diesel blends.

The cold filter plugging point (CFPP) of diesel, biodiesel, and their blends is presented in Table 6 as a critical indicator of cold flow properties. Pure diesel exhibited the most favourable CFPP of –15 °C, highlighting its excellent operability under low-temperature conditions. In contrast, Croton biodiesel (C100) and WCO biodiesel (W100) showed poorer cold flow performance, with CFPP values of –3 °C and +2 °C, respectively. The higher CFPP of W100 is attributed to its larger proportion of saturated fatty acids, which tend to crystallize at low temperatures, thereby increasing the likelihood of fuel filter blockage. These results confirm that biodiesel generally suffers from inferior cold flow behaviour compared to diesel, posing potential challenges for reliable operation in cold climates.

The blended biodiesel (B20) significantly improved cold flow characteristics, with a CFPP of –10 °C, due to dilution with diesel. Incorporation of graphene and MWCNT nanoparticles further enhanced these properties. B20 doped with GNPs at 75 – 100 ppm reduced CFPP to –11 °C, while MWCNT-doped blends achieved even better cold flow, reaching – 12 °C at 75 and 100 ppm. This trend suggests that nanoparticles may interfere with crystal formation by promoting uniform dispersion and altering intermolecular packing within the fuel matrix. Improved cold flow properties are practically important, as they reduce the risk of fuel line blockages and filter plugging in cold conditions, ensuring reliable ignition and stable combustion during winter operation.

Water content values across all biodiesels and blends were below the EN 14214 limit of 0.05%, indicating effective purification and minimizing risks of microbial growth, phase separation, and corrosion in storage tanks and fuel lines. Similarly, iodine numbers reflected the degree of unsaturation: Croton biodiesel (92 g I₂/100 g) was highly unsaturated and more prone to oxidation, while WCO biodiesel (78 g I₂/100 g) exhibited comparatively better oxidative stability. Blends (B20 and nanoparticle-doped variants) showed much lower iodine numbers (35 – 38 g I₂/100 g), indicating reduced susceptibility to oxidative degradation and extended storage stability. The results indicate that nanoparticle doping not only enhances energy content but also improves low-temperature operability, oxidation resistance, and safety margins. This makes Croton – WCO biodiesel blends viable alternatives that comply with major biodiesel standards while minimizing operational risks in diesel engines, improving efficiency, and reducing pollutant formation.

Figure 7 presents the evaluated density and viscosity of the different fuel samples to evaluate the effect of blending biodiesel with nanoparticles. Figure 7A presents a pure diesel (D100) had the lowest density, while pure biodiesel samples of Croton biodiesel (C100) and waste cooking biodiesel (W100), and the biodiesel blend (B20) showed higher densities. This rise could be attributed to the higher molecular weight and more complex chemical makeup of biodiesel, as noted by Keera et al. [41]. Adding graphene nanoplatelets (GNPs) to B20 blends increased the density, in the range of 0.85413 g/cm³ to 0.85731 g/cm³ at concentrations of 50 ppm, 75 ppm, and 100 ppm. These densities exceeded those of both B20 and diesel fuels, showing a clear trend where higher GNP concentrations result in slightly higher fuel density [42]. On the other hand, the B20 blends containing multi-walled carbon nanotubes (MWCNTs) exhibited slightly higher and more consistent density values, ranging from 0.85717 g/cm³ to 0.85979 g/cm³ at concentrations of 50 ppm, 75 ppm, and 100 ppm. These densities were marginally higher than those of the graphene-enhanced blends.

Figure 7.

Effect of nanoparticles on fuel: (A) Density, and (B) Viscosity

A similar pattern of marginal increase in fuel density was reported by Wambui et al. [21], who studied Croton-oleander biodiesel blended with diesel (OCB20) enhanced with GNPs with the same concentration, recording densities of 0.850 g/cm³, 0.852 g/cm³, 0.854 g/cm³, and 0.856 g/cm³, for OCB20, OCB20+50G, OCB20+75G, and OCB20+100G, respectively. Similarly, Debbarma et al. [26] observed that a diesel and palm biodiesel blend (B30) doped with GNPs exhibited densities of 0.848 g/cm³, 0.849 g/cm³, and 0.850 g/cm³ at the same concentrations. A comparable trend was also reported by Sadhik and Anand [43], who investigated carbon nanotube-enhanced Jatropha Methyl Ester blends at concentrations of 25 ppm, 50 ppm, and 100 ppm, reporting densities of 0.8972 g/cm³, 0.8978 g/cm³, and 0.8994 g/cm³, respectively. A consistent rise in fuel blend density with increasing nanoparticle concentration has been observed for graphene and multi-walled carbon nanoparticles. The GNPs increase density due to their high specific surface area, mass, and planar structure, which promotes uniform dispersion and molecular packing [44]. The MWCNTs exhibit a slightly greater and more stable increase due to their tubular shape and high aspect ratio, which allow for stronger molecular interactions, the formation of an entangled network within the fuel matrix, and reduced intermolecular gaps [45]. In practice, higher densities can influence injection timing and spray penetration in diesel engines, potentially altering combustion phasing and pollutant formation.

Viscosity analysis as illustrated in the Figure 7B revealed a consistent trend where pure diesel exhibited the lowest viscosity at 2.99 mm²/s, followed by B20 at 3.77 mm²/s. However, C100 and W100 showed higher viscosities of 5.96 mm²/s and 5.42 mm²/s, respectively, reflecting the typical higher molecular weight and polarity of biodiesel components than petroleum diesel. The introduction of nanoparticles into B20 further influenced the viscosity of the blends. Specifically, B20 blended with graphene nanoplatelets recorded viscosities of 3.79 mm²/s (G50), 3.76 mm²/s (G75), and 3.85 mm²/s (G100), while B20 blended with multi-walled carbon nanotubes exhibited values of 3.92 mm²/s (MWC50), 3.94 mm²/s (MWC75), and 3.62 mm²/s (MWC100). El-Seesy [46], conducted a comprehensive study that investigated carbon nanotube-enhanced Jatropha Methyl Ester blends at concentrations of 10 mg/l, 20 mg/l, 30 mg/l, 40 mg/l, and 50 mg/l of MWCNTs, reporting viscosities of 4.1 mm²/s, 4.19 mm²/s, 4.25 mm²/s, 4.31 mm²/s, and 4.35 mm²/s, respectively. The rise in viscosity with increasing nanoparticle concentration results from agglomeration and microstructure formation, all of which restrict fluid flow. MWCNTs, due to their tubular and entangled structure, slightly increase viscosity more than graphene nanoplatelets, which have a flat, layered form that allows better dispersion and lower resistance [47].

Viscosity is one of the most critical parameters for fuel performance, as higher viscosity can delay injection, reduce atomization quality, and increase the risk of incomplete combustion, leading to carbon deposits, injector coking, and higher emissions of particulates and CO. Despite the marginal increases observed, all viscosities in this study remained within ASTM D6751 and EN 14214 standards, confirming the suitability of the blends for engine application while minimizing risks to efficiency and emission compliance.

The effect of nanoparticles on the calorific value and flash point is illustrated in Figure 8. The calorific value, which indicates the energy content of the fuel, shows significant variation among the tested samples as illustrated in Figure 8A. Diesel fuel (D100) has the highest calorific value at 42.401 MJ/kg, making it the most energy-dense fuel. On the other hand, the B20 blend (39.403 MJ/kg) has a higher calorific value than pure biodiesels (W100 and C100), which recorded lower values of 36.052 MJ/kg and 35.318 MJ/kg, respectively, due to the presence of the energy-rich diesel fraction. This decrease is mainly caused by the higher oxygen content in biodiesel, which increases the energy used in vaporizing and breaking down oxygenated compounds during combustion, resulting in lower overall heat release [48]. From an engineering perspective, fuels with lower calorific values can lead to higher brake-specific fuel consumption and reduced thermal efficiency, while higher calorific values are advantageous for maintaining engine power output and reducing fuel use.

Figure 8.

Effect of nanoparticles on fuel Calorific value (A), and Flash point (B)

Incorporating nanoparticles into B20 further increased the calorific value, with B20+GNPs ranging from 39.417 MJ/kg to 40.104 MJ/kg and B20+MWCNTs from 39.571 MJ/kg to 40.517 MJ/kg. This trend can be linked to improved combustion efficiency due to the catalytic activity of nanoparticles, which enhance atomization, oxygen availability, and heat transfer. MWCNTs surpass GNPs in calorific value possibly due to their higher carbon content, better combustion efficiency, and lower oxygen content, all of which help produce more heat in biodiesel blends. The highest calorific value was observed in B20+MWCNT100 ppm (40.517 MJ/kg), indicating that MWCNTs are more effective than graphene in increasing fuel energy density at similar concentrations. A higher calorific value not only improves energy efficiency but also reduces incomplete combustion, thereby lowering emissions of CO and unburned hydrocarbons. A similar pattern was reported by Gad et al. [47], who investigated biodiesel produced from waste cooking oil (WCO), with an energy content of 39.4 MJ/kg, and its B20 blend (41.50 MJ/kg) doped with carbon nanotubes (CNTs) and graphene nanosheets (CNSs) at 25 ppm, 50 ppm, and 100 ppm of concentrations. The highest energy values were observed for B20CNT100 and B20CNS100, which recorded 41.480 MJ/kg and 41.497 MJ/kg, respectively.

Figure 8B shows that the flash points of pure biodiesels, C100 and W100, are significantly higher, at 80 °C and 70 °C, respectively, compared to diesel, which has a flash point of 45 °C. This indicates that biodiesels are less volatile, making them safer to store and handle due to their lower ignition risk at reduced temperatures. The B20 blend has a lower flash point of 40 °C, indicating increased volatility compared to pure diesel (45 °C), likely due to the presence of residual methanol in the biodiesel component. Adding graphene nanoparticles results in a non-linear trend: a decrease to 35 °C at 50 ppm, followed by increases to 45 °C and 65 °C at 75 and 100 ppm, suggesting improved thermal stability at higher concentrations. In contrast, MWCNTs provide a more consistent improvement, with flash points rising to 75 °C (50 ppm) and peaking at 90 °C (75 ppm), then slightly dropping to 65 °C (100 ppm). This shows that MWCNTs are more effective at stabilizing the blend, likely due to their dispersion stability and stronger interaction with fuel molecules. From a practical viewpoint, higher flash points enhance operational safety by lowering fire hazards during storage and refueling, while also reflecting reduced volatility that contributes to steadier combustion. Conversely, low flash points may raise risks of premature ignition or vapor formation, which can destabilize combustion and increase pollutant formation.

A similar trend was reported by Gawonou et al. [19], who investigated diesel – biodiesel blends with a flash point of 40 °C alongside blends doped with graphene nanoparticles (GNPs) and graphene oxide (GO) at concentrations of 25 ppm, 50 ppm, and 100 ppm. The flash points for the GNP-enhanced blends were observed as 64 °C, 62 °C, and 51 °C, while those for the GO-enhanced blends were 36 °C, 39 °C, and 59 °C, respectively. These findings highlight that nanoparticle additives can strongly influence volatility and safety characteristics, and when carefully optimized, can support both safe handling and efficient, cleaner engine operation.