**1. Introduction**

Nowadays, so many countries have established a climate and energy policy framework, advancing in decarbonization to arrive towards a new friendly climate economy [1]. In this respect, an unprecedented effort is being made to implement energy alternatives (photovoltaic, wind, hydrogen, and nuclear energy) that allow the gradual replacement of natural gas, coal and fossil fuels in the field of electricity generation for reducing the high consumption of fossil fuels [2,3]. However, there is no

such equivalent in transport, since vehicles using fuel cells or electric motors cannot compete ye<sup>t</sup> with fossil fuel engines, especially in the field of heavy trucks [4], aviation [5], or the shipping sector. In this context, the incorporation of biofuels as fossil fuel substitutes is the strategy assumed to accomplish this necessary energetic progress since there is no need of modifying the compression-ignition (CI) diesel engines of the current car fleet [6,7]. Therefore, the use of biofuels leads to easy and gradual integration into the worldwide transportation logistics systems.

Among the existing biofuels, biodiesel has emerged as one of the best options. Biodiesel is obtained through homogeneous or heterogeneous alkaline transesterification of vegetable oil or animal fat with methanol, giving rise to a mixture of mono-alkyl esters of long-chain fatty acids (FAMEs) [8,9]. Over the past decade, biodiesel has been of interest by contributing to reduce energy dependence on fossil fuels and to minimize greenhouse-gas emissions from transportation [10].

Despite the fact the replacement of fossil diesel by biodiesel seems easy, this process is still considered economically unfeasible. The main reason is the high production cost associated, among other things, to the purification process needed because of the generation of glycerol as a by-product (10wt% of the total biodiesel produced). Considering the economic di fficulties in the production of biodiesel, the search for di fferent alternative biofuels is still mandatory.

In this regard, the transformation of vegetable oils into high-quality diesel fuels, avoiding the glycerol formation, has been deeply investigated [11]. These biodiesel-like biofuels, such as Gliperol [12], DMC-BioD [13,14] or Ecodiesel [15,16], avoid the generation of residues or by-products, integrating glycerol in the reaction products. Thus, these new biofuels can be obtained as soluble derivatives in transesterification processes, analogous to that for the production of FAME [17]. Besides, it has been reported as a high-quality diesel fuel known as "green diesel´´, which can be obtained by several treatments, such as cracking, pyrolysis, hydrodeoxygenation, and hydrotreating of vegetable oils [18].

Very recently, the use of straight vegetable oils (SVO) has become an interesting option for the replacement of fossil diesel. On one hand, vegetable oils can be obtained easily from agricultural or industrial sources, avoiding energetic costs associated with the transesterification required to obtain biodiesel. On the other hand, all the relevant physicochemical properties of vegetable oils are analogous to conventional diesel, except for the viscosity, which is much higher in oils. The high viscosities cause poor fuel atomization by premature injector contamination. To solve this issue without the need for carrying out the transesterification reaction, researchers have focused their attention on the reduction of high-viscosity oils by blending them with low viscous biofuels. Thus, it is possible to obtain double blends that comply with the requirements stipulated in the current diesel engines (EN 590 standard). Several blends have been reported employing plant-based sources such as vegetable oils and organic compounds [19,20]. These compounds exhibit a relatively short carbon chain, low viscosity values and also low cetane number since they have been identified as LVLC (Low Viscosity Low Cetane) fuels [21].

Employing this strategy, low viscous vegetable oils, such as pine oil or camphor oil, have also been studied in several blends with di fferent biodiesels [22,23], vegetable oils [19] or with fossil diesel [24–26], to improve performance in the CI engines. So far, the additives studied as LVLC are natural compounds (pine oil, eucalyptus oil, camphor oil or orange oil), obtained from crops, which may contribute to reducing the world's dependence on oil imports, providing benefits to local agricultural industries. Similarly, compounds obtained by chemical synthesis from renewable products, mainly alcohols (methanol, ethanol, and butanol), are also applied as LVLC [27–29]. This strategy has even been applied with a non-renewable compound such as gasoline, capable of reducing the viscosity of vegetable oils in double and triple blends with fossil diesel [30,31]. Following this methodology, high levels of fossil fuel substitution have been obtained in a technically and economically feasible way.

Therefore, if a renewable compound like diethyl ether (DEE) is employed, instead of a non-renewable one such as gasoline, we can make the process greener. In fact, DEE can be easily obtained from ethanol, which can be obtained from biomass. Despite the fact that DEE is known as a cold-start aid for engines, its potential as a transportation fuel in blends with vegetable oil and/or with fossil diesel has not been much investigated. This molecule has several favorable properties

for blending with diesel fuel, including very low kinematic viscosity, low autoignition temperature, high oxygen content, broad flammability limits, high miscibility with vegetable oils and diesel fossil, and very low values of cloud point (CP) and pour point (PP) that improves cold flow properties [29]. Therefore, it is expected that by blending the DEE with SVO, a notable improvement of some of the fuel properties will be obtained, such as a reduction of the CP and PP and so on. However, the calorific power of this compound is relatively low, so this fact could limit the percentage of substitution of fossil fuel by the DEE/oil blend to operate in today's internal combustion diesel engines, maintaining the appropriate parameters of EN 590 standard. In fact, we have recently reported that the low calorific power of ethanol and 2-propanol (27 and 33 MJ/kg, respectively) constitutes the greatest limitation for its use in double blends with oils [32]. This calorific power is very similar to that for DEE (34 MJ/kg) and, therefore, it is foreseeable it will show similar behavior.

Another aspect to take into consideration is the very low ignition quality of some diesel/DEE blends [27] that promote a high ignition delay, given the low heat of evaporation value that DEE exhibits [33]. This ignition delay can be overcome by the addition of vegetable oils to the blend. Therefore, a mutual benefit can be obtained by the use of DEE with vegetable oils in triple blends with diesel, i.e., DEE reduces the high viscosity of the oils, whereas the oils could compensate for the heat and evaporation of DEE. In fact, DEE has been reported as a low-emission renewable fuel and high-quality combustion improver in blends with diesel fossil [27,28], with biodiesel [27,29,34,35], with oils [27,29] or with diesel/oil [36–38]. Besides, better performance of compression-ignition engines operating with DEE/diesel/biodiesel triple blends has been achieved [39,40].

Waste cooking oil (sunflower oil) and castor oil have been selected as the vegetable oils for this work since they come from crops that are not destined to human or livestock food and they are easily available, not competing with oils for food uses. Sunflower oil has been selected as a standard reference to study the behavior of waste cooking oils because the use of any waste cooking oil from different sources, implies the difficulty of reproducing the results obtained. Castor oil is currently the only inedible vegetable oil available on an industrial scale.

The present study intends to advance the strategy of the substitution of fossil fuels by others of a renewable nature that can be used in current diesel engines in a viable way, not only from a technical point of view but also economically and even more importantly, applicable in the most immediate way. To do so, DEE has been employed as an oxygenated additive in blends with diesel and vegetable oil. In this respect, the optimum proportion of DEE/vegetable oil blend will be evaluated based on the most significant parameter, the kinematic viscosity, to meet with appropriate parameters of EN 590 standard that allow its use in internal combustion diesel engines. Additionally, flow cold properties (cloud and pour points) will be studied to ascertain the applicability of the fuel in cold climates. The diesel/DEE/oil triple blends obtained have been tested in a diesel engine. The most important parameters, such as fuel consumption and power generation have been studied to know the viability of the new fuel produced. The degree of pollution of all blends will be also evaluated from the generated smoke opacity values.

#### **2. Materials and Methods**

Some of the significant physicochemical characteristics of diesel, diethyl ether, and SVOs (sunflower oil and castor oil) have been collected in Table 1.


**Table 1.** Properties of diesel, sunflower oil, castor oil, and diethyl ether. All data collected in the Table were taken from the literature [36,41–43], except for the kinematic viscosity values which were experimentally measured in this work.

> 1Viscosity value errors were obtained from the average of 3 measurements.

#### *2.1. Diethyl Ether*/*Vegetable Oil Double Blends, and Diesel*/ *Diethyl Ether* /*Vegetable Oil Triple Blends*

Sunflower oil was bought from a local market and castor oil (Panreac, Castellar Del Valles, Spain) was purchased from a local commercial representative. DEE (≥99.5% purity) was procured from Sigma-Aldrich Chemical Company. First of all, sunflower oil (food grade) and castor oil were mixed with DEE in di fferent concentrations to find out the optimum DEE/SVO double blends. The best double blends, which comply with the established requirements by European petrodiesel standard EN 590 for being employed as biofuels, were selected to be mixed with conventional diesel fuel (Repsol service station) in di fferent proportions, from 20% to 100% by volume, denoted as B20, B40, B60, B80, and B100. The percentage of biofuel (DEE/SVO blend) added to fossil diesel is expressed as B, where B0 corresponding to 100% of fossil diesel and B100 means 100% of renewable DEE/oil biofuel. The components of all blends were manually mixed at room temperature. Additionally, all components were completely miscible with petroleum diesel, allowing the blending of these in any proportion. The obtained diesel/DEE/SVO triple blends were investigated as biofuels in this work.

#### *2.2. Characterization of the Biofuel Blends*

Fuel reformulation can a ffect the physicochemical and safety properties of the fuel. Hence, knowledge of these properties is especially important. In this work, some of the most crucial properties to evaluate the suitability of biofuels have been determined either experimentally or using specific equations to predict them.

As aforementioned, kinematic viscosity and cold flow properties are the rheological properties more influenced by blends of vegetable oils with fossil diesel and other additives. In fact, these properties play a crucial role in the correct performance of conventional diesel engines [32]. Kinematic viscosity significantly a ffects the quality of fuel atomization and the combustion process. Low viscosity can cause leakage in the fuel system while high viscosity can lead to incomplete combustion because of premature injector contamination [44].

Cold flow properties, such as cloud point and pour point, are responsible for solidification of fuel, causing operability problems as solidified material clogs fuel lines and filters. The temperature at which the crystals become visible (diameter ≥ 0.5 mm) is defined as the cloud point (CP), whereas the pour point (PP) is defined as the temperature at which the liquid ceases to flow. The CP usually occurs at higher temperatures than the PP [45]. Crystallization of fuel takes place when fuel molecules condensate forming a gel at low temperatures. Crystallization occurs in two general steps: The first step is the nucleation and crystal growth and the second step is the organization of the molecules creating a stable nucleus in a crystalline network. The continuous growth of the crystalline network generates the interruption of fuel flow, causing fuel starvation and incomplete combustion, which leads to starting problems in the vehicle in cold weather [46].

The flash point (FP) is a crucial property for production, handling, transportation and storage of fuels. This parameter provides an indication of the fire hazard of fuel under ambient conditions. The flash point is defined as the lowest temperature at which a liquid produces enough vapors to ignite in the presence of a flame or spark. Generally, a lower flash point is related to higher vapor pressure, so this property provides information on both flammability and volatility. The values of the flash point can be predicted from Kay's mixing rule:

$$\text{T}\_{\text{FP}} = \sum\_{i} \text{y}\_{i} \text{Ti} \tag{1}$$

where T is the temperature corresponding to the flash point of the blend (◦C), yi is the volume fraction of each component in the blend and Ti is the flash point of each component [47].

Another important property to define the e fficiency of fuels is the calorific value (CV), also called the heat of combustion or calorific power, which is the quantity of heating energy released during complete combustion of a unit mass of the fuel, usually expressed in kilojoules per kilogram. The calorific value increases with increasing chain length and decreases with increasing unsaturation, and it is important for estimating the fuel consumption, the greater the calorific value the lower the fuel consumption. Calorific value is usually determined experimentally by a bomb calorimeter, but a theoretical value can be calculated, according to the volumetric concentration of each component in the blend, from the following equation:

$$\text{CV} = \sum\_{i} \text{CV}i\text{Xi} \tag{2}$$

where CVi is the calorific value of each component and *Xi* is the percentage of each component in the blend [37].
