Combined Effect of Propyl Gallate and Tert-Butyl Hydroquinone on Biodiesel and Biolubricant Based on Waste Cooking Oil

Abstract

The current energy scenario requires the use of alternatives for petrol-based products. In this context, the role of biodiesel and biolubricants can be promising, offering multiple advantages but also some challenges like their low oxidation stability. The use of antioxidants can offset this disadvantage, improving the general performance of these products during storage or usage. The aim of this work was to assess the combined and separate use of two different antioxidants (propyl gallate, PG, and tert-butylhydroquinone, TBHQ) on biodiesel and biolubricant based on waste cooking oil (WCO), carrying out a thorough characterization of these products and quantification of antioxidants through voltammetry techniques. Thus, the innovation of this work consists in the use of combined antioxidants and its simultaneous quality assessment and quantification. As a result, the combined use of antioxidants did not offer a synergistic effect, and the use of PG at low concentrations (300 ppm) is recommended in the case of WCO biolubricant. Voltammetry was a suitable and fast method to quantify both PG and TBHQ.

1. Introduction

1.1. Current Environmental Issues

Currently, the role of the circular economy is more and more influential due to the problems related to the exploitation of oil products. Specifically, petroleum products have traditionally been a key factor in geopolitical tension, provoking the search for alternatives (by national and international agencies or institutions) like sustainable practices that allow less petroleum dependence [1]. Green chemistry is highly appreciated by consumers and society in general, as these processes usually decrease environmental and health impacts, reducing waste and largely avoiding pollution. Consequently, the role of biorefineries will be a key factor, as a wide range of different products (many of them with similar properties compared to petroleum-based products) with environmentally friendly properties such as high biodegradability and lower emissions can be produced [2,3]. Furthermore, these biorefineries present high atom economy values, implying a lower generation of waste and by-products with difficult environmental management. These biorefineries can be based on multiple feedstocks (including edible and non-edible biomass to produce different kinds of biofuels [4,5,6]). Specifically, in the case of oilseed crops, vegetable oil exploitation has been covered in the literature. Some wastes derived from vegetable oils, such as waste cooking oil (WCO), can be potential sources of interesting products such as biodiesel or biolubricants [7,8,9]. Thus, waste management is not only focused on environmental protection, but also on energy or material valorization, presenting interesting advantages such as a contribution to sustainable land usage.

1.2. Waste Cooking Oil

Waste cooking oil (WCO) is mainly produced when different kinds of vegetable oils, such as soybean, olive, or corn oil, are used for culinary purposes, especially frying (the selection of vegetable oils depends on factors like the kind of dish).

For instance, in the case of Spain, frying is the fifth most preferred method used to prepare dishes, with other cooking methods (like homemade dishes, microwave heating, or the direct consumption of fresh food) ranking above this option, and presenting a decreasing trend after the COVID-19 pandemic. In Spain, vegetable oil production decreased 14.8% in November 2023 (compared to the same month the year before), whereas this sector posted an increase in turnover of 1.9% due to a price increase. As a result, the consumption of the most valuable vegetable oil in the Mediterranean diet has decreased, depending on the kind of extraction process, from 22.6% for virgin olive oil to 52.5% for extra virgin olive oil [10]. Consequently, other vegetable oils, such as sunflower oil, have been more frequently used, even as frying oil, resulting in higher proportions of these products being used. WCO has considerable polluting potential, especially when accidentally spilled in the environment (specifically in aquatic ecosystems), which should be avoided through suitable management of this waste. In this sense, collection or clean points are currently being implemented for waste disposal from households or restaurants [9]. Alternatively, WCO can be used as the starting point for different industrial processes, such as soap, bioplastic, or biodiesel production. The latter, along with biolubricant production through double transesterification [7,8], could be an interesting proposal for the implementation of biorefineries, as explained in the following subsection.

1.3. Biodiesel and Biolubricant Production: Advantages and Challenges

As in the case of other vegetable oils, several research works have focused on biodiesel and biolubricant production from vegetable oils, presenting an interesting chemical route to produce different intermediate and final products, with great potential for energy purposes or other usages [11,12]. Normally, biodiesel production is carried out through transesterification of triglycerides with methanol (or ethanol), yielding fatty acid methyl esters (FAMEs, that is, biodiesel) and glycerol. FAMEs can be converted to biolubricants through a second transesterification with more complex alcohols such as neopentylglycol (NPG), trimethylolpropane (TMP) [13], or pentaerythritol (PE), normally under vacuum conditions [14]. In this process, methanol is evolved, which could be reused in the first transesterification. Consequently, this combined process shows different advantages: (1) A high atom economy is possible, as high yields are normally obtained in both transesterification processes, and the intermediate products are reused for further steps. Finally, some by-products like methanol can be reused in the first transesterification. Consequently, low amounts of pollutants can be evolved in the global process. (2) Many of the products obtained are biodegradable compared to petroleum-based products, reducing the environmental impact in hypothetical spills. (3) A wide range of feedstocks (vegetable oils, WCO, or animal fats) can be used, proving their versatility and adaptability to different contexts.

However, biodiesel and biolubricants obtained through this method have different challenges, such as their low oxidation stability and improvable cold properties [14,15]. The former is mainly due to the presence of double bonds in ester structures (derived from fatty acids), and using vegetable oil with low levels of poly-unsaturated compounds such as linolenic acid is recommended. Otherwise, auto-oxidation processes can be accelerated, implying quality loss in the final biodiesel or biolubricant, especially regarding increased viscosity and acidity due to polymerization and free fatty acid generation, respectively [14]. Nevertheless, there are alternatives such as antioxidant addition, explained in the following subsection.

1.4. Antioxidants

Antioxidants are chemical compounds that reduce or delay auto-oxidation processes, mainly through the interruption of free radical propagation. Specifically, antioxidants transfer electrons to free radicals generated in biodiesel and biolubricants, avoiding the typical chain reaction that provokes their degradation (mainly observed in increased viscosity due to polymerization and increased acidity due to free fatty acid generation) [16]. The antioxidant performance of these products depends on many factors such as the solubility in biodiesel or biolubricants, redox potential, chemical structure, or FAME profile. Both natural and synthetic antioxidants are mainly phenolic compounds. Regarding the former, many of them are included in natural extracts in fruits, vegetables, or plants [17]. Thus, a wide range of studies are focused on the use of multiple natural or renewable antioxidants (many of them including phenolic compounds) on the improvement of biodiesel performance (in terms of different aspects such as rheological behavior or performance and emissions) [18,19], including different samples such as soybean [20,21,22,23], WCO [24], daok [25], or sunflower biodiesel [26,27]. Even though there are fewer studies related to the addition of antioxidants to biolubricants, some interesting studies can be found in the literature. Indeed, some natural extracts (for instance, from rosemary) have been directly used in biolubricants, offering different results depending on the kind of biolubricant used (due to factors such as solubility) [28]. Another example is turmeric rhizome extract applied to biolubricants, which could act as an antioxidant and pour depressant [29]. Additionally, curcumin-extracted waste cooking oil could act as a sustainable option in biolubricant formulations, showing good tribological and rheological performance [30]. Equally, extracts from rice husks had a positive antioxidant effect on bio-oils, reflected in TGA analyses [31]. However, these compounds are currently an interesting research field as their application is still unknown in many cases. On the other hand, different synthetic antioxidants have been widely used (especially in the improvement of biodiesel properties [19,32]), with a wide range of compounds like those included in Figure 1.

As observed in this figure, these antioxidants share some points in common, as they all are phenolic compounds, which are effective electron donors. For this work, TBHQ and PG were selected for further studies, as they are efficient antioxidants, especially for the improvement in the oxidation stability of biodiesel, as observed in the literature [19,32]. In this sense, low amounts of these antioxidants (below 2000 ppm) are normally required to comply with the corresponding standard (for instance, UNE-EN 14214 [33]) or to considerably increase oxidation stability.

1.5. Voltammetry

Voltammetry is an electroanalytical technique that allows the quantification of synthetic antioxidants in samples such as biodiesel and biolubricant. Thus, there is a direct relationship between the magnitude of the electrical property and the concentration of analytes (in this case, antioxidants), allowing their quantification. Alternatively, other techniques can be used, such as chromatography, normally requiring longer analytical times and higher numbers of samples [34]. In detail, in this work, two different voltammetric techniques are used. The first of these is cyclic voltammetry (CV), where a linear potential scanning over time is carried out (from an initial to a final potential, and returning to the initial value). Intensity signals in the direct scanning (oxidation, for instance) are obtained, and another intensity signal for the indirect scanning (reduction) is obtained if the process is reversible. The sensitivity of this technique is relatively low (10⁻⁵ M), requiring, for some tests, the use of differential pulse voltammetry (with higher sensitivity). The most important parameters in CV are anodic and cathodic peak potential, along with the corresponding peak areas. Another technique is differential pulse voltammetry (DPV). In this case, the excitation signal consists of a series of pulses, where the potential gradually increases at low intervals (between 5 and 10 mV), whereas the amplitude is constant (between 10 and 150 mV). The signal is obtained as the difference in intensities obtained before and after the application of a certain pulse. In this case, sensitivity is higher (10⁻⁷ M), and the most important parameters are intensity and peak potential, in order to obtain peak height and area for analyte quantification.

1.6. Scientific Interest

According to previous subsections, the application of antioxidants to biodiesel and biolubricants, as well as their quantification in these products, seems to be a promising research field, as observed in the literature included in

Table 1. Different works focused on the main research fields covered in this study.

Details	Ref.
TBHQ was quantified in biodiesel samples (WCO, cardoon and canola) by using cyclic voltammetry (CV).	[35]
TBHQ and Cu were quantified in biodiesel using squarewave voltammetry (SWV).	[36]
PG was quantified in cardoon biodiesel, improving its oxidation stability. PG concentration decreased during oxidation.	[37]
Different methods (including CV and DPV) were compared to HPLC, to simultaneously quantify TBHQ and BHA in biodiesel.	[38]
TBHQ and BHA were determined in biodiesel through a batch injection analysis and pulsed-amperometric detection.	[39]
Different antioxidants applied to biodiesel, including TBHQ and PG, were simultaneously determined by using DPV and artificial neural network.	[40]
TBHQ was quantified in a biolubricant from WCO, through CV and DPV methods.	[41]
TBHQ was added to a high-oleic safflower biolubricant produced through double transesterification with methanol and pentaerythritol, showing a high efficiency at low concentrations (500 ppm) and keeping the main properties of this biolubricant (mainly viscosity and acidity).	[42]
Different antioxidants, including PG, were applied to several biolubricants, proving the highest efficiency of propyl gallate.	[43]
Lignin-based additives were applied to castor oil in biolubricant formulations, increasing oxidation induction time.	[44]
CuO nanostructures were used as additives for biolubricants (Pongamia oil), with high antioxidant efficiency (above 70%) at low concentrations (50 µg·mL−1)	[45]

.

Thus, different studies have been carried out related to this research, with a wider variety of works focused on the addition of antioxidants to biodiesel samples. In this case, assessment of the effects of antioxidant addition (on the improvement in oxidation stability and combustion performance, among others), along with the detection of these antioxidants in biodiesel through different methods (like voltammetry), have been carried out, including simultaneous and fast quantification of different antioxidants. On the other hand, studies in this field about biolubricants are relatively scarce, and have not observed the simultaneous use or quantification of antioxidants to the best of our knowledge. In this sense, this research work tries to complement the research gap in both kinds of products (biodiesel and biolubricants) with the combined effect of two synthetic antioxidants and their effect and quantification.

1.7. Aim and Innovation of This Work

Consequently, the aim of this work was to assess the effect of combined and separate antioxidant addition (PG and TBHQ) on biodiesel and biolubricant production from WCO through double transesterification (an interesting process in a biorefinery context), especially concerning oxidative stability and biolubricant performance during extreme oxidation conditions. Additionally, a voltammetric method was used to determine PG and TBHQ concentrations in WCO biolubricant. This method is proposed as an alternative to traditional analytical tools such as HPLC, which is more expensive and time-consuming.

2. Materials and Methods

2.1. Raw Material

WCO was used as a feedstock for biodiesel and biolubricant production. The samples were collected from households and restaurants located in Badajoz (Spain), in November and December 2023. WCO was a heterogeneous mixture with 7–8% solid waste that could not be reused (mainly composed of residues from breaded foods, among others) and moisture. In this way, the sample was filtered to remove solid waste and heated at 110 °C for 20 min to remove moisture. Acidity was measured, being below 2%, and suitable for further treatments. As a result, the sample presented the following characteristics: density, 927.5 kg∙m⁻³; viscosity at 40 °C, 44.89 cSt; and acid number, 0.79 mg_KOH∙g⁻¹. Once purified, WCO was stored in 25 L opaque containers at room temperature.

2.2. Biodiesel and Biolubricant Production

First and second transesterification were carried out according to previous experiments, with modifications [35,46,47,48]. The main operating conditions are included in

Table 2. Operating conditions for WCO biodiesel production.

First Transesterification
Parameter	Conditions	Details
Alcohol used	Methanol	Pure, pharma grade, Panreac Applichem, (Castellar del Valles, Barcelona, Spain)
Reaction time, min	120	--
Reaction temperature, °C	65	Higher temperatures were not recommended to avoid methanol boiling
Oil/methanol ratio	1:6	Excess methanol ratio was used to ensure high conversion
Catalyst, %	MeONa, 0.5	30% in methanol, Merck (Darmstadt, Germany)
Purification	Separation funnel	Through decantation to remove glycerol and catalyst (by washing treatments)

and

Table 3. Operating conditions for WCO biolubricant production.

Second Transesterification
Parameter	Conditions	Details
Alcohol used	Pentaerythritol (PE)	Pure, Merck (Darmstadt, Germany)
Reaction time, min	120	--
Reaction temperature, °C	160	--
FAME/PE ratio	1:1/3	A slight excess in pentaerythritol was used to avoid problems during filtering
Catalyst, %	MeONa, 1%	30% in methanol, Merck (Darmstadt, Germany)
Pressure, mmHg	260	To promote methanol removal from the reaction medium
Purification	Filtration	A first gravity filtration was used, followed by different vacuum filtrations

:

Once purified, WCO biodiesel and biolubricant were stored in opaque containers at room temperature for further characterization. For further representations, these samples were labeled as “WCOBD” and “WCOBL”, respectively.

2.3. Characterization

Both WCO biodiesel and biolubricant were characterized, with the main analysis included in

Table 4. Main characterization tests for WCO biodiesel and biolubricant.

Property	Details	Ref.
Viscosity and Cold Filter Plugging Point (CFPP)	A Cannon-Fenske viscometer was used, controlling temperature at 40 °C. For CFPP, the corresponding standard was used.	[49,50]
Density	A densimeter was used for this determination.	[51]
FAME content	FAME content was analyzed by using a gas chromatograph (Varian 3900, Varian, Palo Alto, CA, USA) coupled to a flame ionization detector (FID). Main FAMEs, such as methyl oleate, linoleate, linolenate, palmitate, and stearate.	[52]
Acid value	According to UNE-EN 14104 standard.	[53]
Iodine value	According to UNE-EN 14111 standard.	[54]
Oxidation stability	Rancimat method was used, at 110 °C.	[55,56]
Flash and fire points	Cleveland open cup method was used.	[57]
Phytotoxicity test	Marvel of four seasons lettuce (Lactuca sativa) was selected for this test (lettuce is a recurring species in this kind of tests, as observed in references), adding 20 mL of water as control sample, 25% and 50% solution of WCO biolubricant. Afterwards, the lettuce samples were visually assessed, including chlorophyll content by using a chlorophyll meter (SPAD 502 Plus, Konica Minolta, Tokyo, Japan). Measurements were carried out in different leaves on different days for one week.	[58,59]

:

It should be noted that most of these tests were based on the UNE-EN-14214 standard. Additionally, viscosity and acid number evolution were carried out during extreme oxidation conditions, as explained in Section 2.6.

2.4. Antioxidant Addition

PG and TBHQ were added, combined or separated, to WCOBD and WCOBL. In this way, the corresponding amount of antioxidant was added to 3 g of sample, and the mixture underwent ultrasound treatment for 5 min. Once biodiesel or biolubricant were added with PG and/or TBHQ, the final sample was characterized through voltammetric analysis, or made ready for further treatments such as extreme oxidation conditions, explained in the following subsections.

2.5. Voltammetric Analysis

The analyses were carried out according to previous experiments [41]. The samples were prepared in 50 mL flasks, adding 8 mL of ethanol (96% v/v), the required volume of biolubricant, 4 mL of buffer (NaH₂PO₄/H₃PO₄ 0.5M, pH = 2.5), and 2 mL of tensioactive (0.8% cetylmethylammonium bromide, CTAB), to obtain the required PG and/or TBHQ concentrations. Afterwards, it was diluted to 50 mL with ultrapure water, treating the sample with ultrasound for 4 min for a suitable homogenization. Finally, the sample was placed in the electrochemical cell (with a previous deoxygenation for 3 min) and CV or DPV voltammograms were registered. The measure was carried out by using a voltammetric device (µAutolab, ECO Chemie, Utrecht, The Netherlands) coupled to a 663 VA-Stand Met Rohm unit (Herisau, Switzerland) with a three-electrode system: a working electrode (glassy carbon), a reference electrode (Ag/AgCl), and an auxiliary electrode (Pt) with a measuring cell. The glassy carbon was cleaned with a piece of cotton after each measurement, using N,N-dimethyl formamide (DMF) for 2 min and Milli-Q water for 30 s.

2.6. Extreme Oxidation Conditions

Finally, WCOBL (both control samples and samples with TBHQ and/or PG optimum addition) underwent extreme oxidation conditions to assess its quality loss during oxidation or storage, according to previous studies where the same methodology was used in biodiesel [35]. In this sense, the experiment was similar to the Rancimat method, with slight modifications. Thus, air (500 mL∙min⁻¹) was bubbled into a certain amount of sample (9 g in this case) at 110 °C, with different parameters such as viscosity and acidity measured at different oxidation times (2, 4, 6, and 8 h). When the sample had antioxidants, the corresponding voltammetric analysis was carried out to quantify the quantity of antioxidants remaining in the sample. A comparison between control samples and those with antioxidant addition was carried out to evaluate the effectiveness of PG and TBHQ on WCOBL.

3. Results and Discussion

3.1. Biodiesel and Biolubricant Characterization

The FAME profile of WCOBD after the first transesterification of WCO is shown in Figure 2. As observed, methyl linolenate is the main FAME, accounting for 66.8%, followed by methyl oleate (22.3%), palmitate (3.9%), linolenate (3.0%), and stearate (2.9%). Other studies included similar FAME profiles, with the main composition differences related to the two main FAMEs (that is, methyl linoleate and oleate) [41,47]. This could be due to the fact that WCO composition depends on the culinary traditions of each area or circumstance; apparently, sunflower oil prevailed in this study (at the expense of olive oil) as the typical vegetable oil in this region.

A typical chromatogram was obtained in this case (as reported in previous works), with high peak resolution, which allowed the quantification of each kind of FAME [60]. As explained in the literature, the fatty acid profile of the vegetable oil (and the subsequent FAME profile for biodiesel and ester profile for the corresponding biolubricant) will play a central role in biodiesel and biolubricant characteristics [14,61].

Table 5. WCOBD and WCOBL characterization.

Property	WCOBD	WCOBL
Viscosity at 40 °C, cSt	4.50	62.89
Density, kg∙m−3	881	904
Yield, %	98.9	90.18
Acid value, mgKOH∙g−1	0.22	0.55
Iodine value, gI2∙100 g−1	98	Not determined
Oxidation stability, h	1.28	2.63
Flash and fire points, °C	173–179	242–249

shows the main characteristics, for this study, of biodiesel and biolubricant obtained from WCO.

Firstly, the suitability of most WCOBD properties should be noted, according to the standard (including CFPP in hot climates, which was 1 °C in this case) [33], except for oxidation stability, which is below the upper limit (8 h). This could be on account of the fact that the FAME profile had high quantities of methyl linoleate, which is a di-unsaturated component. Thus, double bonds are usually the starting point for auto-oxidation processes due to free-radical generation, and the presence of saturated or mono-unsaturated FAMEs (such as methyl palmitate and oleate, respectively) is desirable [15].

On the other hand, WCOBL presented a high viscosity at 40 °C, with viscosity at 100 °C of 11.24 cSt and a viscosity index of 173, which implies low changes in viscosity with temperature (a desirable property for industrial purposes). In this case, acidity increased compared to WCOBD possibly due to the additional heat treatment carried out for biolubricant production, with the possible degradation of the sample and the subsequent release of free fatty acids [62]. Additionally, the high flash and fire points found in this case should be noted, which ensure safety during storage and handling. Again, as in the case of WCOBD, WCOBL showed low oxidation stability (exceeding 2.5 h), due to the presence of linoleic acid in the original WCO, thereby transferring its instability to this final product.

In this sense, both biodiesel and biolubricant obtained from WCO required the addition of antioxidants to improve oxidation stability and the subsequent quality maintenance during storage or oxidation, as explained in further subsections.

Regarding WCOBL, its phytotoxicity applied to lettuce was assessed, including the main results in Figure 3 and Figure 4. Firstly, visual degradation of control samples was clear on Days 3 and 7 (due to the lack of watering since Day 0), with slight differences shown compared to samples watered with WCOBL (25% and 50%), which were more degraded (especially concerning turgor and yellowing of leaves).

Regarding chlorophyll levels, Figure 4 shows the main results, where there were no considerable differences between control samples (keeping the highest chlorophyll index throughout this experiment) and the addition of 25% and 50% WCOBL. In any case, there was a continuous decrease in this parameter, possibly due to the gradual degradation of lettuce on account of the lack of watering. However, the relatively worse results obtained for WCOBL could be due to the generation of a waterproofing layer. Nevertheless, the higher biodegradability of vegetable oil-based products could explain the low phytotoxicity observed in this case.

3.2. Effect of Antioxidants on Oxidative Stability of Biodiesel and Biolubricant

As previously stated, WCOBD and WCOBL required antioxidant addition. Thus, the effect of the separated addition of PG and TBHQ in WCOBD is included in Figure 5. In this sense, both antioxidants presented a similar efficiency, with similar oxidation stability increases at the same antioxidant concentrations. Consequently, both antioxidants allowed compliance with the standard at around 600 ppm. In the literature, there are different results concerning the efficiency of different antioxidants, with better results for PG in some cases [63], whereas TBHQ had a better antioxidant performance in other studies [64].

However, this effect was different when PG and TBHQ were added to WCOBL. As observed in Figure 6, PG presented a higher efficiency compared to TBHQ, considerably increasing oxidation stability of WCOBL at lower concentrations. Oxidation stability values of 8 h were achieved at around 200 ppm of PG, whereas the same result was obtained with around 1800 ppm of TBHQ. This specific behavior in biolubricants was observed in previous studies for similar products. Regarding the different efficiency depending on the kind of sample (biodiesel or biolubricant), this could be due to the different nature of these products, where different viscosity, polarity, and acidity values could alter the suitable dissolution and distribution of antioxidants. Indeed, previous studies have covered the possibility of TBHQ chemical modification to improve solubility in biodiesel due to its polarity [65], and this fact could be decisive at this point.

Equally, the mixture of antioxidants showed interesting results, as observed in Figure 7 for the combined effect of PG and TBHQ on oxidation stability of WCOBD. The data included in this experiment were fit to a third-degree polynomial (R² = 0.98857), according to ten samples with variable PG and TBHQ concentrations (at high, medium, and low concentrations, within 0–1000 ppm for each case). Possibly due to the similar efficiency of both antioxidants (as previously explained), a neutral or additive effect was observed (in general) when PG and TBHQ were combined at different concentrations.

In other words, strong synergistic or antagonistic effects were not observed in this case. Thus, the separate addition of antioxidants according to other criteria such as economic factors is recommended. Other studies including the combination of TBHQ and BHA in biodiesel from soybean and pork fat showed a favorable synergistic effect, with a 75–25% ratio recommended in this case [66].

However, the combination of antioxidants applied to WCOBL offered different results, as observed in Figure 8. The data included in this experiment were fit to a third-degree polynomial (R² = 0.99629), covering samples at different TBHQ and PG concentrations (both separate and combined) within the maximum and minimum concentrations covered in this work (from 0 to 800 ppm for PG and from 0 to 1000 ppm for TBHQ). In this way, an antagonistic effect was observed, obtaining lower or shorter oxidation stability values than expected according to the separate effect of PG and TBHQ.

As previously explained, this effect could be due to the lower antioxidant performance of TBHQ (due to the different acidity and viscosity of the medium) or the possible combination of both antioxidants during oxidation, thereby generating compounds with a lower antioxidant effect.

As a result, separate addition of PG (at around 300 ppm) was recommended in this case, as its efficiency was considerable compared to TBHQ addition or the mixture of both antioxidants.

3.3. Voltammetric Quantification of PG and TBHQ

This kind of quantification presents a challenge in biolubricants, as this technology has been widely studied in biodiesel samples (where the particularities of each sample could also affect the operating conditions of voltammetry). However, it needs adaptation for the different physico-chemical conditions found in biolubricants, especially concerning the variable viscosity values found for each specific product. That is the reason why an initial study should be accomplished to optimize voltametric response.

Firstly, preliminary studies were carried out to register different samples, including buffer solution, and biolubricant without antioxidant addition and with antioxidant addition (in this case PG), in order to assess the voltammetric (CV) response. Figure 9 shows the main results.

As can be inferred from this figure, the signal corresponding to the buffer solution and biolubricant without antioxidant addition is similar, with no response being observed. Nevertheless, the biolubricant with 300 ppm PG shows an anodic response that is increased with PG addition, with a signal appearing at the same potential. Thus, PG in doped biolubricants could be quantified.

Afterwards, different studies for WCOBL with PG and TBHQ addition at different concentrations were carried out to assess the effect on antioxidant signals in voltammetry. Consequently, calibration lines for each antioxidant could be obtained. For this purpose, different standards were prepared, as explained in Section 2, at different PG or TBHQ concentrations; this test was conducted in duplicate for each case (see Figure 10 and Figure 11 for PG and TBHQ addition, respectively).

As expected, the signal corresponding to PG (an anodic peak at 0.5 V) increased with its concentration in WCOBL (both peak area and height), with no signal shown in the inverse scanning.

Concerning TBHQ, anodic and cathodic peaks (at 0.5 and −0.02 V, respectively) were observed, with a linear trend when the peak and height signal were represented. In this case, the anodic peak is more intense than the cathodic one, with PG and TBHQ having very similar anodic signals that are very close to each other. As a consequence, mixtures of both antioxidants in biolubricant could be determined, where two peaks (one anodic due to the mixture and the other cathodic due to TBHQ) were obtained. In both cases (PG and TBHQ), the quality parameters for the calibration curves are included in

Table 6. Curve calibration for PG and TBHQ in WCOBL.

Parameter	Units	PG	TBHQ (Anodic Peak)	TBHQ (Cathodic Peak)
Number of standards	--	7	7	7
Slope	nA·L·mg−1	74.1421	104.2100	52.5280
Standard deviation (slope)	nA·L·mg−1	3.7465	5.5110	3.0450
Intercept	nA	23.6430	28.2500	31.6660
Standard deviation (intercept)	nA	35.3190	61.3200	33.8810
R2	--	0.9727	0.9675	0.9612
Linearity	%	94.2030	94.1820	94.9700
Analytical sensitivity	γ−1	1.0790	0.5485	0.7900
Detection limit (Long–Winefordner)	ppm	1.94	1.22	1.42
Detection limit (Clayton)	ppm	3.15	1.83	2.31

.

The high quality of the parameters included in this table should be noted (especially with high linearity, obtained at different PG and TBHQ concentrations and conducting the experiments in triplicate), similar to previous studies where TBHQ was quantified in a biolubricant based on WCO, obtained through double transesterification with methanol and pentaerythritol [41].

Figure 12 includes the main signals obtained when PG and TBHQ were added to WCOBL. Thus, obtaining the anodic peak of PG + TBHQ, it is possible to correlate the peak area of this figure to the TBHQ signal (Figure 11) to obtain the signal corresponding to PG according to Equations (1) and (2):

$$\left|{\displaystyle \frac{I_{pA, TBHQ}}{I_{pC,TBHQ}}}\right|=2.14 \pm 0.16$$

(1)

$$\left|{\displaystyle \frac{A_{pA, TBHQ}}{A_{pC,TBHQ}}}\right|=2.98\pm 0.32$$

(2)

where I_p is the peak intensity, A_p is the peak area, and subscripts A and C correspond to anodic and cathodic, respectively.

As a result, the anodic peak for TBHQ has an intensity 2.14 times bigger (and a peak area 2.98 times bigger) than the cathodic one. In this sense, once the peak areas and heights (or intensities) are obtained, the anodic peak of PG can be calculated, proving that this method could be suitable for the simultaneous quantification of both antioxidants.

3.4. Effect of Extreme Oxidation Conditions on Biodiesel and Biolubricant Properties

In order to assess the separate effect of PG (the most effective antioxidant for WCOBL) along with the combined effect of PG and TBHQ, different measurements were carried out during WCOBL extreme oxidation. Thus, the selected samples, according to the previous results, were the following:

• WCOBL (control). This was WCOBL without antioxidant addition.
• WCOBL (PG + TBHQ). In this case, 300 ppm PG + 1000 ppm TBHQ were added to the sample. This was one of the points where the antagonistic effect (as previously mentioned) was evident, with lower oxidation stability than expected shown according to the additive effect of both antioxidants.
• WCOBL (PG). An addition of 300 ppm PG was carried out.

Regarding antioxidant content in WCOBL during extreme oxidation conditions, there was a continuous decrease in its content with time, as observed in Figure 13 for PG concentration evolution with oxidation time. In this case, concentration decreased from 300 ppm to 30–40 ppm at 4 h, slightly decreasing from then on. At 8 h of study, there was 20 ppm of PG in WCOBL, proving the suitability of the added quantity, which was enough for this test. PG was gradually degraded due to extreme oxidation in favor of WCOBL, whose oxidation was clearly decelerated as explained in further experiments.

This decrease was also observed in different antioxidant samples (PG and TBHQ) during oxidation of biodiesel [37] and biolubricants [41], highlighting the effectiveness of these products in avoiding the oxidation of the original samples.

Concerning the effect of viscosity during extreme oxidation conditions, Figure 14 shows the evolution of this parameter in different samples. Clearly, control samples presented an increase in viscosity at 4 h of extreme oxidation (from 71 to 370 cSt), with a moderate increase in the case of the sample with combined antioxidant addition (where viscosity increased from 63 to 65 cSt). Nevertheless, separate PG addition seemed to be the most effective option, as viscosity was maintained during the whole experiment, even at 8 h of extreme oxidation. An increase in viscosity is due to oxidation processes, where polymerization of biolubricant could have taken place, with the subsequent increase in viscosity and the corresponding quality loss due to the impossibility of selecting this product for specific industrial uses. In this sense, the use of antioxidants had a positive effect, and avoided the considerable increase observed for control samples. Specifically, the combined effect of antioxidants (PG + TBHQ) showed a slight increase in viscosity at the end of extreme oxidation (especially at 8 h), possibly due to the antagonistic effect of their combination (which implied lower oxidation stabilities in biolubricant than expected). That is the reason why, as cited in previous subsections, PG at 300 ppm is recommended, whose effectiveness was clear in this case, where a viscosity increase was almost negligible at 8 h. The same trend was observed for biodiesel and biolubricant control samples, where polymerization caused this undesirable effect [41,67], which was avoided by the addition of synthetic antioxidants.

These differences between control, PG + TBHQ, and PG samples were clear when the effect of extreme oxidation on acidity was assessed (see Figure 15). In this way, for control samples, the acid number considerably increased from 4 to 8 h, due to the generation of free fatty acids during the oxidation processes in biolubricant. In this case, the PG and TBHQ mixture showed an intermediate increase in this parameter, with PG addition showing the lowest increase in acidity. Again, the antagonistic effect of the mixture provoked the lower efficiency of antioxidants, with worse results obtained compared to the separate addition of PG. Equally, previous studies showed a considerable increase in acidity when oxidation took place in biodiesel and biolubricants based on vegetable oils, with the effective performance of synthetic antioxidants [37,41,67].

As a result, PG addition (300 ppm) was recommended to retain the main properties of WCOBL, at the expense of the PG and TBHQ combination, whose positive effect was reduced and, therefore, its combined use is not recommended for analytical and economic reasons.

4. Conclusions

In this work, double transesterification with methanol and pentaerythritol was studied as a possible management approach for waste cooking oil (WCO) to produce biodiesel and biolubricant.

Specifically, WCO biolubricant was obtained with a high yield, and had a low phytotoxicity and properties similar to a SAE 10W30 lubricant for conventional engines. However, oxidation stability (1.28 h and 2.63 h for biodiesel and biolubricant) was relatively low, requiring the use of antioxidants (like PG or TBHQ, covered in this work) to retain their main properties (viscosity and acidity) during storage or oxidation processes.

PG and TBHQ addition was effective at increasing the oxidation stability of biodiesel and biolubricant based on WCO. However, the combined effect of PG and TBHQ was less efficient. Thus, in the case of WCO biodiesel, the optimum antioxidant concentration was around 650 ppm of TBHQ, whereas in the case of WCO biolubricant, the optimum addition was around 300 ppm of PG. The different behaviors observed in these cases can be due to the different acidity and viscosity values of the product.

Consequently, regarding WCO biolubricant, individual PG addition contributed to retaining viscosity and acid number during extreme oxidation conditions, with the mixture of PG and TBHQ showing an antagonistic effect in this context.

The proposed voltammetric method can be used to analyze, simultaneously, both PG and TBHQ. Concerning the addition of PG to biolubricant, this method was quick, effective, simple, and low-cost compared to other analytical methods such as HPLC.

For further studies, the performance of WCO biodiesel and biolubricant in engines or industrial facilities, respectively, could be an interesting point to expand the results found in this work.