Influence of oil-to-glycerol molar ratioThe IV measurement helps determine the conversion of unsaturated fat into saturated fat. An increase in saturated fat indicates a more successful breakdown of C=C double bonds in palm olein during hydrogenation. An investigation on the glycerol ratio was conducted to check the decrease in IV which indicates the degree of hydrogenation. Investigations started with four variations of oil-to-glycerol molar ratio between 1:0.5, 1:1, 1:3, and 1:5. The total volume of the solution was fixed at 35 mL. The reaction was set with a gap size of 10 mm, 1 L min−1 He gas flow rate, 75 W of plasma discharge power, ambient temperature with an initial temperature of 27 °C, and stirring rate of 200 rpm for 2 h of reaction time. The influence of the oil-to-glycerol molar ratio on IV is shown in (Fig. 2). The feed referred to the refined palm olein oil fed into the reactor without glycerol mixed before plasma treatment. This was set as a control with the IV of 67.16 ± 0.70. The finding was that the IV of the molar ratio of 1:1 decreased the most from the initial feed value to 42.85 ± 4.84. For the ratios of 1:0.5, 1:3, and 1:5, the IV increased from the ratio of 1:1.Figure 2IV of oil in the investigation of oil-to-glycerol molar ratio (10 mm gap size, 1 L min-1 He gas flow rate, 75 W plasma discharge power, ambient temperature with initial temperature of 27 C, and 2 h of reaction time).A high glycerol content resulted in a high mixed viscosity. This was supported by the calculation proposed by Arrhenius’ and Graham’s model31,32, with the expression for the viscosity of a mixed solution highlighted in Sect. “Viscosity”. The calculated viscosity of each molar ratio is shown in (Table 2).
Table 2 Calculated viscosity of mixed solution for each molar ratio of oil to glycerol.When the ratio of 1:0.5 was applied, low viscosity allowed the mixed solution to move easily in the reaction chamber, enabling a thorough reaction with the plasma. However, due to the small amount of hydrogen source, IV only decreased by about 7 points from the feed value. When the ratio increased to 1:1, this condition seemed to match Kongprawes et al.’s collision hypothesis37, as proper glycerol content in a mixed liquid increased the probability of successful collisions between reactive species in plasma and glycerol molecules, resulting in hydrogen radicals (Ḣ) formation through dehydrogenation reaction. The radicals might then contribute to the C=C bonds in the oil molecules becoming single bonds.The large amount of glycerol from plasma dehydrogenation might produce Ḣ radicals very effectively. However, when higher molar ratios of 1:3 and 1:5 were applied, the higher viscosity may potentially disrupt the reaction by slowing down molecular movement and reducing the frequency of collisions between reactant molecules in the reaction chamber, leading to a decrease in reaction rate or even inhibit the interaction between the plasma-liquid surface/interface altogether. Then, the generated Ḣ radicals appear to have a challenging circumstance when interacting with the oil. In agreement with Al-Moameri et al.38, it was concluded that increased viscosity leads to a substantial decrease in diffusion rates and, subsequently, reaction rates, resulting in a slow reaction. Since the ratio 1:1 showed the optimal condition in this case, it was consequently fixed to investigate other parameters.Influence of gap sizeTo investigate the influence of the gap size, it was varied from 5 to 15 mm. The reaction took place at a 1:1 molar ratio of oil to glycerol, 1 L min-1 He gas flow rate, 75 W of plasma discharge power, ambient temperature with an initial temperature of 27 °C, and stirring rate of 200 rpm for 2 h of reaction time. The obtained results are demonstrated in (Fig. 3). The smallest gap (5 mm) offered the reduction of IV from the initial value of 67.16 ± 0.70 to 53.58 ± 3.48, and it further reduced to 42.85 ± 4.84 for the 10 mm gap. However, for the larger gap of 15 mm, the IV increased to 57.09 ± 2.66. It was found that a 10 mm gap offered the most optimal IV reduction.Figure 3IV of oil in the investigation of gap size (1:1 oil-to-glycerol molar ratio, 1 L min-1 He gas flow rate, 75 W plasma discharge power, ambient temperature with initial temperature of 27 C, and 2 h of reaction time).The atmospheric-pressure plasma discharge system was sensitive to the discharge gap distance. As the smallest and largest electrode gaps were constructed in this study, which were 5 and 15 mm, an insufficient or excessive distance separating the two barriers could potentially result in the instability of the gas layer and liquid surface located between them39, resulting in ineffective reduction of IV. A high discharge current, more streamers, as well as more active species, were formed because of the strengthened electric field of 1.01 kV/mm for the 5-mm case. In contrast, a small discharge current was obtained because of the weakened electric field (0.31 kV/mm), showing fewer active species formed for the 15 mm case (see Supplementary Data Table S1).The increased plasma current in the small gap (5 mm) can enhance the mutual interaction between the plasma and the facing liquid40. Nonetheless, impurities such as carbon residue from the burned oil vapor were formed and observed on the upper electrode when the gap size was too small, resulting in decreased plasma performance due to the impact of mutual interactions as illustrated in the previous research (Fig. 4). These impurities were present due to the limited space of active species formed in the plasma while the plasma intensity was increased by the smaller gap size19. Yoon et al.40 proposed the mutual interaction process between plasma and liquid in plasma-treated water (PTW). It was suggested that the conductivity of the PTW became higher after the concentration of the plasma column caused the discharge current to be higher. Then, PTW with higher conductivity increased the discharge current again as the feedback result. This process will cause an abnormal plasma discharge to transit from glow to arc. Similar phenomena of plasma interaction with liquid oil might be experienced with the same characteristics. The transition into the abnormal plasma discharge thus led to the non-uniform discharge, resulting in plasma active species and radicals only interacting with several spots of liquid oil. Due to higher plasma density of these spots, oil and carbon impurities may form on the upper electrode, and the impurities may reduce the efficiency of hydrogenation as was found in the IV reduction of (Fig. 3).Figure 4Mutual interaction between plasma and liquid (modified from Yoon et al.40).Conversely, the decreased plasma current and weakened electric field in the wide gap (15 mm) resulted in less effective ionization and excitation of plasma carrier gas41. It automatically reduced the production of reactive species, such as radicals, which were insufficient to interact with oil and glycerol. As reactive species have more space to diffuse, the longer the travel distances, the more chance of these species undergoing recombination before they can participate in the desired hydrogenation reactions. This results in a lower concentration of reactive species at the reaction site, reducing the overall reactivity and then diminishing the efficiency of hydrogenation. Therefore, a suitable electrode distance is critical. As shown in Fig. 3, the optimal gap size was fixed at 10 mm to investigate the next parameters.Influence of plasma discharge powerThe plasma discharge power was varied at 35, 50, and 75 W at the reaction condition of 1:1 molar ratio of oil to glycerol, gap size of 10 mm, ambient temperature with an initial temperature of 27 °C, 1 L min−1 He gas flow rate, and stirring rate of 200 rpm for 2 h. The results shown in Fig. 5 indicate that 75 W achieved the highest conversion rate of up to 36.2% for unsaturated fatty acids (67.16 ± 0.70 to 42.85 ± 4.84), while conversions at 35 and 50 W were comparatively lower at 15.7% (67.16 ± 0.70 to 56.61 ± 1.81) and 26.8% (67.16 ± 0.70 to 49.19 ± 1.85), respectively. The diminished yield observed at low power may be attributed to the relatively lower energy levels of excited helium, resulting in weaker chemical bond cleavage42. The limited formation of excited helium species at low power hindered the conversion of glycerol molecules into hydrogen atoms due to electron energy distribution43. A small fraction of the energetic electrons’ distribution might occur due to collisional quenching at atmospheric pressure plasma. Consequently, a significant number of charged particles and excited helium atoms can be lost due to this phenomenon44. Conversely, the broader spectrum of excited states at higher discharge power (75 W) provided a wider range of energy options, facilitating a more diverse set of reactions.Figure 5IV of oil in the investigation of plasma discharge power (1:1 oil-to-glycerol molar ratio, 10 mm gap size, 1 L min-1 He gas flow rate, ambient temperature with initial temperature of 27 C, and 2 h of reaction time).Helium exceeds the energy of 24.6 eV with its highest-energy excited state23. The higher plasma power offered the greater dissociation of He gas, leading to a higher concentration of reactive species in the system to collide with the glycerol molecules to generate hydrogen radicals. To extract hydrogen, the estimation of C-H bonds dissociation energy (BDE) within the hydroxyl groups in glycerol compounds ranges from approximately 350 to 450 kJ/mol which equals 2.19 to 2.81 eV45. The surface discharge on the liquid surface typically exhibits an electron density of approximately on the order of 1016 cm-3 and electron energy > 10 eV23. These energies are more than sufficient to break down this bond in the plasma-liquid interaction. From the mechanisms of the hydrogenation activity, it was reported that the energy of the C=C double bond was ∼6.36 eV23. Afterward, the sum of metastable helium and electron energy levels (> 24.6 eV) can drive Ḣ radical with the C=C bond to become a stable single bond37.Influence of reaction temperatureThe effect of initial reaction temperature was investigated at 20 (low-temperature case), 27 (ambient temperature case), and 50 °C (high-temperature case). The plasma power was tuned at 75 W at the reaction condition of 1:1 molar ratio of oil to glycerol, 10 mm gap size, 1 L min−1 He gas flow rate, and stirring rate of 200 rpm for 2 h. For the cooling system, two peristaltic pumps were installed to deliver chilled water circulation surrounding the plasma chamber. For the heating system, external heating from the hot plate stirrer was utilized. Figure 6 presents the temperature history of the mixture during hydrogenation. For the low-temperature case, the steady-state temperature of about 50 °C was reached. For the ambient temperature case, the steady-state temperature of about 80 °C was obtained. For the high-temperature case, the steady-state condition of about 120 °C was reached.Figure 6Temperature history from external cooling, ambient, and external heating conditions.Figure 7 presents the IV of the oil-glycerol mixture. It was found that the low-temperature condition reduced the IV from 67.16 ± 0.70 to 46.44 ± 6.42. For the ambient temperature investigation, the largest IV reduction to 42.85 ± 4.84 was achieved. The cooler environment makes the oil more viscous (27.06 cST at 50 °C and 11.61 cST at 80 °C)19,20, making physical interchange between the oil underneath and at the surface layer more difficult. Plasma hydrogenation only occurred at the oil surface; hence, mass transfer limited the reaction. For the high-temperature case, the IV was reduced to only 55.93 ± 2.79. This aligns with the basic principle that an exothermic reaction should be conducted at a reduced temperature to promote the continuation of the reaction19. In conclusion, performing the experiment at ambient temperature (initially) without any external heating or cooling yielded the most favorable result, offering the simplest experimental configuration.Figure 7IV of oil in investigation of reaction temperature (1:1 oil-to-glycerol molar ratio, 10 mm gap size, 1 L min-1 He gas flow rate, 75 W plasma discharge power, and 2 h of reaction time).Influence of He gas flow rateTo examine the influence of carrier gas flow rate on the reduction of IV, variations of the He flow rate were conducted at 0.8, 1, and 1.5 L min–1. The lowest flow rate was 0.8 L min−1 because it generated a stable plasma. The reaction took place at a 1:1 molar ratio of oil to glycerol, fixed gap size of 10 mm, 75 W of plasma discharge power, ambient temperature with an initial temperature of 27 °C, and stirring rate of 200 rpm for 2 h of reaction time. The obtained results are presented in (Fig. 8). For the cases with 0.8, 1, and 1.5 L min−1, the IV reduction was from 67.16 ± 0.70 to 54.30 ± 0.99, 42.85 ± 4.84, and 57.24 ± 0.28, respectively. The IV reduction of the 1 L min−1 scenario was optimal, with the hydrogenation rate calculated to be 24.31 IV units/120 min or 12.15 IV units/h. Using 0.8 and 1.5 L min−1, the hydrogenation rates were 6.41 and 4.96 IV units/h, respectively.Figure 8IV of oil in the investigation of He gas flow rate (1:1 oil-to-glycerol molar ratio, 10 mm gap size, 75 W plasma discharge power, ambient temperature with initial temperature of 27 C, and 2 h of reaction time).Owing to this system being in semi-batch operation, the He gas was continuously fed into and withdrawn from the system while the liquid mixture of glycerol and oil was installed in the chamber. The helium flow rate might represent the number of the reactive species generated in the plasma. The gas was associated with helium atoms and ions near the substrate surface during plasma treatment. The extraction of hydrogen atoms from glycerol molecules was initiated by the highly energetic electrons and He* for further adsorption on the material being hydrogenated37.Therefore, the hydrogenation rate occurring less at a lower flow rate of 0.8 L min−1 might be attributed to a smaller number of reactive species. A higher flow rate of 1.5 L min−1 induced high intensity of reactive species to give rise to the recombination of reactive species to lower the chance of hydrogen atom extraction from glycerol while minimizing the extent of conversion to the desired hydrogenation. It was also suggested that a higher flow rate than 1.5 L min-1 was not preferable due to the increasing cost of reactions/process. The reactant temperature history for each flow rate was illustrated in (Fig. 9) to support this hypothesis.Figure 9The mixture temperature history of oil-glycerol mixture under different He flow rates.For the optimal flow rate of 1 L min−1, the system reached its steady-state temperature of 120 °C within the first 15 min. This rapid stabilization resulted from a rising slope of 8 °C/min to the steady-state value. In contrast, the gas flow rates of 0.8 and 1.5 L min−1 exhibited lower slopes of 3.12 and 2.16 °C /min, respectively. The faster temperature ramp rate indicated a higher rate of exothermic hydrogenation as derived from heat released from the system. This assisted in verifying that 1 L min−1 was the most suitable condition. These findings strongly suggested a successful conversion of unsaturated fatty acids into more saturated forms due to the favourable reaction rate.Influence of reaction timeThis study investigated the effect of reaction time, up to 12 h, on the conversion of unsaturated fatty acids into saturated forms, shedding light on the process kinetics. The other process parameters were a 1:1 molar ratio of oil to glycerol, 10 mm gap size, 75 W plasma discharge power, ambient temperature with an initial temperature of 27 °C, 1 L min−1 He gas flow rate, and stirring rate of 200 rpm. The result of IV over the reaction time is presented in (Fig. 10a). The IV decreased from 67.16 ± 0.70 to 42.85 ± 4.84 with a 2 h reaction time, indicating that the fatty acid chains became more saturated. At 4 and 6 h of hydrogenation with IV of 38.65 ± 0.95 and 34.23 ± 1.31, the color of the hydrogenated oil-glycerol mixture with semisolid emulsion particles became paler. At 12 h of reaction time with the IV of 34.05 ± 1.54, the mixture became very viscous as presented in (Fig. 10b). The samples in the Eppendorf tubes were placed upside-down to indicate that they had become semi-solidified/solidified characterized by certain/non displacement of the bubbles at the ends of the tubes.Figure 10(a) IV of oil in the investigation of reaction time, and (b) pictures of samples (1:1 oil-to-glycerol molar ratio, 10 mm gap size, 1 L min-1 He gas flow rate, 75 W plasma discharge power, and ambient temperature with initial temperature of 27 °C).Fatty acids compositionTable 3 outlines the fatty acids composition of the palm olein feed and hydrogenated oil-glycerol mixture (1:1 glycerol-to-oil molar ratio, 10 mm gap, 1 L min−1 He flow rate, ambient temperature initially, 75 W plasma discharge power, 12 h reaction time). The feed contained 12 fatty acid types, comprising 40.56% saturated fatty acids with palmitic acid (C16:0) being predominant. The unsaturated fatty acids comprised 57.83%, with oleic acid (C18:1 n-9 cis) being the most prevalent at 42.98%, followed by linoleic acid (C18:2 n-6 cis) at 14.85%. The hydrogenated product showed 9 fatty acid types, with palmitic acid (C16:0) being the highest at 27.95%, followed by oleic acid (C18:1 n-9 cis) at 32.02%, linoleic acid (C18:2 n-6 cis) at 11.26%, and trans-fat detection of C18:1 n-9 trans at 10.01%.
Table 3 Summary of fatty acid compositions.When comparing the compositions, it is evident that palmitic acid (C16:0), oleic acid (C18:1 n-9 cis), linoleic acid (C18:2 n-6 cis), and α-linolenic acid (C18:3 n-3) decreased, while lauric acid (C12:0), myristic acid (C14:0), palmitoleic acid (C16:1), stearic acid (C18:0), and C18:1 n-9 trans increased. This shift indicates a reduction in mono-, di-, and tri-unsaturated fatty acids alongside an increase in saturated chains, being typical outcomes of hydrogenation affecting unsaturated carbon atoms. Remarkably, C12:0 and C14:0 also increased. Unlike C16:0 which substantially decreased, it is possibly due to the energetic energy of plasma radicals and reactive species provided by plasma that could break longer fatty acid chains into shorter ones. These findings align with the previous investigation by Yepez and Keener25.FTIR analysisThe functional groups of the oil-glycerol mixture before and after hydrogenation are presented in (Fig. 11a). The feed exhibited the characteristic peak at 3325 cm−1 indicating O–H stretching of glycerol46, and the two peaks at around 2924 and 2853 cm−1 indicating C–H stretching originating from palm olein which is oleic and palmitic acid46, and C-H stretching at 2878 cm−1 originating from glycerol47. The peak at 1746 cm−1 indicated C=O stretching46.Figure 11FTIR spectra of feed palm oil-glycerol and hydrogenated palm oil-glycerol. (a) Full spectra, (b) Zoom-in peak at 2924–2853 cm−1, (c) Zoom-in peak at 1032–723 cm−1.The hydrogenated oil-glycerol mixture after the reaction indicated a distinctly different peak at 3265 cm−1 representing O–H stretching. The O–H group became much less intense because the H atoms were extracted becoming H radicals and cooperating with the C=C bonds of the oil. The decreasing peaks (Fig. 11b) at around 2878–2853 cm−1 (C–H stretch from oleic and palmitic acids) could indicate a reduction of oleic acid and palmitic acid48. The oleic acid appeared to be converted into saturated fatty acids by hydrogenation, while the palmitic acid might transform into shorter hydrocarbon chains by plasma scissoring. The decreased peak at 1746 cm−1 (C=O group) could indicate changes in the FFA content from triglyceride fats48, while the –HC=CH (trans-) group49 in the hydrogenated oil initially appearing at 968 cm−1 was still slightly observed as detected 10% trans-fat in GC analysis (Table 3).Notably, the peak at 1465 cm−1 associated with C–H bending emerged50, while the presence of the peak at 1163 cm−1 indicated the stretching of the C–O ester group37. The formation of this peak possibly stemmed from the disintegration of the hydrogen atom from C–OH in glycerol. Consequently, the intensity of the O–H peak at 921 cm−1 and the C–O stretching peak at 1032 cm−1 (Fig. 11c) appeared diminished37.Optimal plasma hydrogenation conditions to prevent trans-fat formationPrevious investigations revealed that one of the factors determining the predicted formation of trans-fat in conventional hydrogenation is the deodorization process51. The process has a wider range of times and temperatures to remove pesticides and light polychromatic hydrocarbons from vegetable oil. In plasma hydrogenation, the temperature factor is also the cause of the trans-fat formation as shown in the FTIR determination at a wavelength of 968 cm−149. This temperature factor is very dependent on the applied plasma discharge power, as it was revealed in the present study that the difference in plasma discharge power resulted in a difference in the reaction temperature. The initial temperature for all plasma power varieties was at room temperature, and the final temperature after 2 h rose to 40, 60, and 80 °C at 35, 50, and 75 W, respectively, of plasma discharge power (see Supplementary Data Fig. S1). A study by Puprasit et.al.20 observed that a temperature rise was attributed to the exothermic nature of the hydrogenation reaction, which supports the present investigation. Although prolonging the plasma exposure time to 12 h reduced the IV substantially and produced a desirable solid fat (Sect. “Influence of reaction time”), the GC-MS analysis revealed a remarkably high trans-fat composition (Table 3). The saturated fatty acids including C12:0, C14:0, and C18:0 increased, while the unsaturated fatty acids including C18:1 n-9 cis, C18:2 n-6 cis, and C18:3 n-3 decreased. Notably, there was trans fatty acid formation of 10.01%. To reduce the amount of trans-fat, which was also the primary goal of the present investigation, another experiment was performed with a low plasma discharge power of 35 W, resulting in the mixed liquid temperature rise of only 40 °C. The other process parameters were a 1:1 molar ratio of glycerol to oil, 10 mm of gap size, 1 L min−1 of He gas flow rate, and ambient temperature initially. The compared results are shown in (Fig. 12a).Figure 12(a) Comparative IV result after plasma discharge power adjustment from 75 to 35 W, and (b) hydrogenation rate analysis from the exponential model.The finding was that the IV reduction at 2, 4, 6, and 12 h at 35 W was 56.61 ± 3.23, 50.61 ± 0.54, 44.61 ± 2.76, and 31.61 ± 1.11, respectively. This finding is consistent with those reported by Liu and Lu52, which observed similar trends in the temperature effect to the trans-fat production using p-toluenesulfinic acid catalyst. It was explained that reaction temperature of 100 °C and above have been significantly result in 79.6% trans-fat production. It was stated that cis–trans isomerization caused by free radical intermediates from the catalyst has a definite temperature. This consistency across different studies reinforces the understanding that high temperature catalyzes the formation of trans fats, highlighting the need for controlled hydrogenation conditions to minimize trans-fat production. In terms of the plasma process, it offered the released heat at the electrodes. To reduce the trans-fatty acid and accelerate the reaction rate, several factors need to be considered, including the discharge power to provide the lowest operational temperature, the stirring rate to allow the most suitable mass transfer, the modified DBD power supply to decrease self-heating.Figure 12b illustrates that the reduction in IV occurred more rapidly at 75 W compared to 35 W. Specifically, by following the exponential fitting model fitted by OriginLab software, the rate of IV reduction for each case was proposed. According to the model, the hydrogenation rate can be calculated from the first derivative of the exponential plot. The rate of IV reduction along the time via hydrogenation rate for the 75 W case was −16.508e−0.49×, while the rate for the 35 W case was −4.67e−0.09x. The R2-value obtained from both cases was 0.99, indicating that both cases have a 0.99 variation in the dependent variable (IV) that is statistically predicted by the exponential model as similar to the rate of reactant concentration disappearing along the reaction time (−dCA/dt). Although it was found that the 35 W condition exhibited a slower IV reduction, the 12 h reaction time still gave the same IV result with zero quantities of trans fat as shown in (Table 4).
Table 4 Summary of fatty acid compositions.Table 4 presents the fatty acid compositions of palm olein feed, 75-W hydrogenated oil-glycerol, and 35-W hydrogenated oil-glycerol maintaining other fixed parameters of 1:1 molar ratio of glycerol to oil, 10 mm gap size, 1 L min−1 He flow rate, and ambient temperature initially. The 35-W case contained 8 fatty acid types, comprising 58.61% saturated fatty acids, predominantly palmitic acid (C16:0) at 44.51%. The unsaturated fatty acids comprised 39%, with oleic acid (C18:1 n-9 cis) being the most prevalent at 35.24%, followed by linoleic acid (C18:2 n-6 cis) at 3.86%. The 35-W case exhibited a polyunsaturated fatty acid content reduction of 10.99% and a total monounsaturated fatty acid content decrease of 7.74%. This suggests that the hydrogenation of mono- and polyunsaturated fatty acids involves the cleavage of the double bond by the provision of a hydrogen atom from glycerol. An absence of elaidic acid (C18:1 n-9 trans) was observed (0%). In comparison with margarine production conducted by Puprasit et al.19 with a small amount of trans-fat present (1.44%), the present work can be a novel reference and competitive method as it does not involve the use of hydrogen gas.Table 4 indicates a small difference in the total fatty acid composition of the two cases. The total fatty acid percentage of the 75-W case was 95.71% while it was 97.71% for the 35-W case. The mass of the feed oil-glycerol before and after the reaction was also investigated and is presented in (Table 5). For the 75-W case, the liquid mixture lost around 0.78 g, whereas the 35-W case lost around 0.31 g after plasma hydrogenation. For the change in fatty acids composition, C16:0 was notably increased, while there was a small number of C16:1 and without unsaturated sixteen carbon chain presence in feed. This is because the bond-cracking process of oil molecules could occur in the plasma process53,54. The C16:0 was generated by the bond breaking of C18:0 at the ethyl group (C2H5–nCiH2i+1), which is the weakest bond in alkene (bold letter stands for the cracking group)55. The process of becoming C16:0 released the small products, including ethyl and methyl radicals, in the gas phase and moving out of the reaction chamber with the plasma gas. This is related to the mass loss that took place in this study. Higher mass loss was observed when using a higher power of 75 W. The higher applied plasma power offered stronger plasma filaments with higher energy of generated reactive species, resulting in a higher chance of bond breaking. Also, the direct contact of the strong filaments with the mixed solution might cause some palm oil to mix in the glycerol layer. Therefore, the amount of oil changed after the separation process.
Table 5 Summary of mass balance.The thermal image processing of the DBD plasma and the oil-glycerol mixture is presented in (Fig. 13a–d) It was found that the highest temperature in one spot of the plasma streamer was about 225° C (Fig. 13d). This observation suggests that localized regions within the plasma streamer could cause the mass change in palm oil. With heat, oil degradation may also produce gel/solid residues56. These residues can accumulate on the equipment surfaces and lead to an imbalance between the mass of the exiting feed material and that of the material entering the reaction chamber57. Another contributing factor for mass loss could be degradation or chemical modification during sample preparation or analysis58, leading to the formation of compounds not detected by the GC-MS as fatty acids or that cannot be quantified accurately.Figure 13Observed plasma hotspot during experiment via thermal image processing: (a) RGB visible-light image of feed oil-glycerol, (b) Jet-thermal colormaps, (c) Plasma-thermal colormaps, and (d) Gray-thermal colormaps with temperature.Formation of plasma reactive species during reactionThe reactive species generated during the reaction were detected using optical emission spectroscopy (OES) in the range of 200–1100 nm, as illustrated in (Fig. 14). When the plasma was on, the energetic electrons forming in the plasma could collide with gas molecules to create more He species, as there were eight detected peaks corresponding to He plasma at 356, 492, 501, 587, 668, 706, 778, and 846 nm59. Hydrogen species with the highest intensity in the range of Balmer and Paschen regions, named H(A, B, C, D, and E) were also observed25. This implies that the plasma successfully attacked the bonded hydrogen atoms in the molecules of glycerol becoming free hydrogen species in the system.Figure 14Optical emission spectrum of He plasma during hydrogenation.Physical characteristics of optimal hydrogenated oil-glycerol mixture: AV, SMP, and textureThe AV, SMP, spreadability, and hardness values of the synthesized product based on the 1:1 molar ratio oil to glycerol, 1 L min−1 H2 flow rate, 35 W plasma discharge power, 12 h reaction time, and ambient initial reaction temperature are shown in (Table 6).
Table 6 Characterization of synthesized product from optimal process parameters (1:1 molar ratio oil to glycerol, 1 L min - 1 H2 flow rate, 35 W plasma discharge power, 12 h reaction time, and ambient initial reaction temperature).The AV serves as a fundamental indicator of product quality, reflecting the acid content in fats or oils because of hydrolytic deterioration, commonly known as rancidity, of the triacylglycerol. The results indicate a slight increase in AV from 0.47 to 0.51, an 8.51% gain. It was within the acceptable range, which should not exceed 0.6 mg KOH/g fats60. This increase can be attributed to the presence of hydrogen radicals from glycerol, which possess sufficient energy to catalyze the hydrolysis of triglycerides into free fatty acids and glycerol, particularly during the hydrogenation process. Consequently, this hydrolysis leads to an elevation in the concentration of free fatty acids, thus contributing to the observed increase in AV.The SMPs of the feed and product were recorded as 32 and 35.7 °C, respectively. A longer hydrogenation duration resulted in a higher SMP as the oil became more hardened due to a greater content of saturated fatty acids, agreeing with the decreasing IV21. The product exhibited a stable texture at room temperature. This stability is useful for prolonging the storage shelf-life. Supporting the agreement from Rajah et al.21, the investigation showed the range of melting point of hydrogenated palm oil of 32–37 °C. For the SMPs of margarine, the reported SMP for bakery & pastry products was 31.2–34.9 °C in Turkey61, and 32–41.31 °C in Iran62. Therefore, the produced product exhibiting the SMP of 35.7 °C was within the usable range for bakery purposes.Spreadability is primarily recognized as the primary textural characteristics of margarine, which significantly impact consumer acceptance and render it acceptable for various applications. The spreadability of margarine can be predicted by the quantification of hardness value. It is quantified by the maximum force attained upon initial compression. The margarine with enhanced hardness exhibited less spreadability63. The spreadability and hardness values were recorded as 0.027 ± 0.001 Nm and 8.71 ± 0.56 N, respectively.Comparison margarine production over plasma treatmentIn this section, comparative studies on fatty acids conversion by plasma hydrogenation methods is described. The summarized results based on the basic product properties and several parameters are shown in (Table 7). There are scarce sources of scientific papers available that converting fatty acids by cold plasma hydrogenation methods, despite of the use of glycerol includes in research studies. Therefore, the comparison between the plasma hydrogenation with glycerol and the conventional plasma hydrogenation (with H2 gas) methods was only done based on four parameters: product properties, energy consumption, time consumption, and cost production.
Table 7 Recent investigations of compared method for margarine production.Based on Wongjaikham et al.’s microwave plasma approach21, the energy consumption was found higher as compared to this study with 4.23% trans-fat product, while Puprasit et al.19,64 proposed 1.44% trans-fat product with a slightly competitive energy consumption of 0.437 kWh. Although Yepez & Keener25 found a successful zero trans-fat product with the pioneering research on this process, the energy consumption was also higher compared to this study. The summarized results showed that this frontier novel method for fatty acids conversion has significant product properties that have not been previously discovered. However, additional research is required to explore the pathway mechanism and optimization parameters to enhance energy efficiency while maintaining cost-effective production.
Enhanced cold plasma hydrogenation with glycerol as hydrogen source for production of trans-fat-free margarine
