Separation of ethylbenzene and n-octane using deep eutectic solvents

1 Introduction

Aromatics are important chemicals in the petrochemical and chemical industries, the most important aromatics are benzene, toluene, ethyl benzene and xylenes (BTEX). Benzene is widely used for producing styrene, phenol and cyclohexane. Styrene is the raw material to make polymers and plastics, phenol is used for the production of resin and cyclohexane for the production of nylon.

Separation of aromatic from aliphatic hydrocarbons is a challenging process because of azeotropes occurrences and the close boiling points range. Liquid-liquid extraction (LLE) was found to be suitable for the range of 20–65 wt% aromatics, extractive distillation more applicable for the range of 65–90 wt% aromatics and azeotropic distillation for higher aromatic content (>90 wt%). However, until now, no separation processes is available for concentrations below 20 wt%.

Indeed, some petroleum processes, such as naphtha steam cracking, need the removal of low level aromatic hydrocarbons to achieve many benefits like products purity, energy and lower operation costs, etc. In this process, the cracker feed usually contains 10–25 wt% of aromatic components, but it may contain higher concentrations as well. During the cracking process, the aromatic compounds are not converted to olefins, so the aromatics occupy a large portion of the furnaces capacity and burden the separation section. If a large amount of aromatics can be separated before entering the furnace it will enhance the thermal efficiency and decrease the fouling and the energy demand and thus the effective cost of the process [1]. LLE is the most applied process for hydrocarbons mixtures with an aromatic content in the range of 20–65 wt%. It is energetically advantageous when compared with the extractive and azeotropic distillation techniques. Therefore, it would be easier to extend the range of aromatics content in which extraction might be profitably used.

The problem inherent to the techniques cited above is the use of a solvent, which is mixed with the feed and comes out of the extractor, or the column, that accompanies the raffinate and extract streams. Typical solvents in these processes are organic compounds such as sulfolane, ethylene glycols, dimethylsulfoxide, N-methylpyrrolidone, N-formylmorpholine [2]. The recovery of these solvents from the extract and raffinate streams must be carried out by distillation, thus resulting in an additional increase in the installation and operational costs of the processes [3].

Sulfolane is the most commonly used solvent in industry. The disadvantage of using this solvent is the need of an additional separation step to purify the raffinate, extract and solvent streams, which implied additional high investment and energy demand [4]. The regeneration of sulfolane also results in a high energy cost because of its high boiling point (278.3°C). It will be taken from the top column of the regenerator and returned to the bottom aromatic stripper as a vapor. The high temperature also gives another effect, that it will decompose and lead to corrosion in the paraffin and aromatic stripper if oxygen intrusion occurs [1].

Ionic liquids (ILs) have emerged as potential alternative exhibiting important properties such as low melting temperature and very low vapor pressures compared to industrial organic solvents. Their non-volatile nature makes possible the recovery of the solvent by using simple techniques such as flash distillation. This will reduce the investment in heat equipment to only 20% in comparison to the sulfolane process [1]. However, it is still a challenge for the large-scale applications of ionic liquids in industry, due to complicated synthetic processes and the expensive raw chemical materials. Hence, eutectic mixtures, so-called deep eutectic solvents (DESs), have been recognized as a low-cost alternative of ILs. The reason they have been termed DESs is because when the two components (a salt and a complexing agent (CA) usually a hydrogen bond donor) are mixed together in the correct ratio, a eutectic point can be achieved as the result of strong hydrogen bonding leading to a very large depression of the mixture freezing point [5].

DESs have properties comparable to ILs. They are known as tunable solvents because they can be customized to a particular type of chemistry [5]. Moreover, DESs are potential candidates to replace non-volatile ILs since they possess many benefits compared to traditional ILs: (1) simple to synthesize, the components of the eutectic can be easily mixed and converted to IL without further purification, (2) very cheap because of the low cost of raw materials and (3) better bio-compatibility, for example the quaternary ammonium salt such as choline chloride was used as an additive in chicken feed [6].

The experimental data on the extraction of aromatic hydrocarbon by DESs are scarce. Experimental data were reported using phosphonium-based DESs for toluene/heptanes [7], benzene/hexane [8] and tetrabutylammonium bromide based DES with BTEX aromatics [9]. In this work, we investigated the use of ammonium based salts (namely, benzyltrimetylammonium chloride and tetraethylammonium p-toluenesulfonate) combined with ethylene glycol and sulfolane as CA. Both ethylene glycol and sulfolane are commercially used as solvents to extract aromatics from naphta. The ethylene glycol showed good capability to extract aromatics at low temperatures with systems involving toluene; its selectivity, which reflects the ability for the extraction of aromatics from aromatics plus aliphatics mixture is high [10]. At the same time, the selectivity of sulfolane for separating benzene and hexane was reported between 7 and 20 with a very good distribution ratio [11].

2 Materials and methods

Ethyl benzene, n-octane, ethylene glycol, sulfolane, benzyltrimetylammonium chloride and tetraethyl-ammonium p-toluenesulfonate were obtained from Acros Organics (Belgium). All chemicals were of high purity (>99%) and used without any further purification. The DESs were prepared by mixing salts and either sulfolane or ethylene glycol as complexing agent. The ammonium salts:CA molar ratio for all the solvents were 1:4. Table 1 shows the DESs synthesized in this work and the abbreviations used in the following sections.

First, each mixture was heated up to 100°C and mixed until the formation of a clear liquid [6]. Mixtures of ethyl benzene and n-octane were prepared for ten concentrations (10, 15, 20, 30, 40, 50, 60, 70, 80 and 90 wt%) by mixing weighed amounts of pure ethyl benzene and pure n-octane. The feed sample was then mixed with the DES with mass ratio 1:1. Each set of experiment was conducted at 25°C. The vials were placed in an incubator shaker. The shaking time was 6 h and followed by settling time of about 12 h to guarantee that the equilibrium state was completely attained. Samples were taken from the top and bottom layers and then analyzed using HPLC.

The samples were withdrawn using syringe and diluted using 2-propanol. These samples were analyzed using HPLC Agilent 1100 series with zorbax eclipse xdb-c8 column. The temperature of the column oven was set at 30°C. The mobile phase was acetonitrile: distilled water with volume ratio 3:1. The flow rate in the column was 1.4 ml min^-1 with a pressure of 120 bars.

3 Results and discussion

The feasibility for the potential application of these solvents to perform the extraction of aromatic hydrocarbons from mixture with n-octane was evaluated using the parameter distribution ratio (D) and the selectivity (S), calculated from experimental data. These parameters are defined by the following expressions:

Where w is the mass fraction; superscripts I and II refer to hydrocarbon rich phase and DES rich phase, respectively, while subscripts 1 and 2 refer to ethylbenzene and n-octane, respectively. The measured composition at equilibrium in both liquid phases for each DES are reported in Tables 2 and 3, respectively, and also plotted in the ternary diagrams given by Figures 1 and 2.

Figure 1:

Experimental and calculated (dashed) tie lines for the ternary mixture ethyl benzene+n-octane+DES1.

Figure 2:

Experimental and calculated (dashed) tie lines for the ternary mixture ethyl benzene+n-octane+DES2.

The analysis results of the top layer revealed the non-existence of both ethylene glycol and sulfolane. This is a remarkable finding that confirms the potential advantage of applying these DESs for extracting aromatics. Especially, when we know that ethylene glycol and sulfolane, currently used as commercial solvent in this process, appears in the top layer as reported by Mohsen-Nia et al. [12] for ethylene glycol and by Santiago and Aznar [4] and Awwad et al. [13] for sulfolane. The absence of these components will reduce the process steps and cost. As a consequence, the top layer consisted exclusively of ethyl benzene and n-octane. In the bottom layer, the quantity of DES was calculated by applying the mass balance based on the quantity of ethyl benzene and n-octane determined by HPLC.

3.2 Distribution coefficient and selectivity

Generally, the distribution ratio increases with aromatic content, the same trend is observed for sulfolane and ethylene glycol as solvent for aromatics separation [4, 12, 13]. The comparison between ethylene glycol and sulfolane as CA showed that the sulfolane is more capable to absorb the aromatic, as can be seen in Figure 3. This could be attributed to the higher aromatic character of sulfolane.

Figure 3:

The distribution ratio at 25°C as a function of the mass fraction of the aromatic in the raffinate for DES1 (diamond symbols) and DES2 (square symbols).

The selectivity decreases with increasing aromatic content in the feed as shown in Figure 4. The same trend was also reported in references [4, 8, 10, 16, 17]. However, again, sulfolane as CA showed better result compared to ethylene glycol. The higher selectivity at lower concentrations of aromatics in the feed is an important factor for the solvent selection.

Figure 4:

The selectivity at 25°C, as a function of the mass fraction of the aromatic in the raffinate for DES1 (diamond symbols) and DES2 (square symbols).

A comparison between the values of the distribution ratio and selectivity for DES studies in this work and those reported in the literature for propylene carbonate organic solvent [18] and different types of ILs ([Emim][Tf2N] [19], [MBPy][BF4] [20] and [Omim][SCN] [21]) is shown in Figures 5 and 6. The values considered in these figures are those obtained at an aromatic concentration of 20 wt% which do not correspond to the highest selectivity and distribution ratio but provide a good approximation of the average values for the separation purpose of this work. All these data are given at 25°C except for [MBPy][BF4] IL which selectivity and distribution ratio are given at 40°C. The comparison shows that the selectivity of our best solvent (DES2) is higher than that of propylene carbonate and [Omim][SCN] IL, but less than that for other ILs. On the other hand, the distribution ratio for this DES is lower than that for all other solvents. Furthermore, the distribution ratio values are less than unity for both DESs.

Figure 5:

The selectivity obtained in this work compared with the selectivity reported for other solvents for the same system.

Figure 6:

Comparison between the distribution ratio obtained in this work and that reported for other solvents for the same system.

This means that a large amount of the DES should be used as a solvent for the separation. However, as the DES can be recovered and reused, this problem should not be considered as a serious disadvantage. In spite of this, DESs are advantageous because they can be easily prepared in high purity and at much lower cost than ILs.

4 Conclusions

The separation of aromatic from aliphatic hydrocarbons is a challenging process. In this work, DESts synthesized from ammonium salt with ethylene glycol and sulfolane as CAs were used for the LLE of ethylbenzene from n-octane.

The ability of these solvents to selectively extract aromatics from a mixture of aromatics and aliphatics was proven. The solvent formed from tetraethylammonium p-toluenesulfonate:sulfolane showed the best performance, i.e., the distribution ratio and selectivity are comparable with the results obtained by commercial solvents used in industry. The CAs, i.e., ethylene glycol and sulfolane, were not found in the raffinate phase. This remarkable fact will facilitate the separation process protocol and hence reduce the cost of the process. Finally, the experimental data was satisfactorily correlated using the NRTL model, showing that this classical model can easily be used with systems including DESs.