Yonglin Li ,He’an Luo,2, *,Qiuhong Ai,2, *,Kuiyi You,2 ,Fei Zhao ,Wenlong Xiao

1 School of Chemical Engineering,Xiangtan University,Xiangtan 411105,China

2 National &Local United Engineering Research Center for Chemical Process Simulation and Intensification,Xiangtan University,Xiangtan 411105,China

Keywords:Amines Coal tar Liquid-liquid equilibrium Thermodynamic model

ABSTRACT In this paper,a method of extracting phenols from coal tar by amines aqueous solution was proposed.The effects of various amines on the extraction properties of phenols in coal tar were researched from the views of molecular structure.The parameters such as molar ratio,concentration,extraction time and temperature for the extraction of coal tar by the monoethanolamine and ethylenediamine aqueous solution were examined.The results show that the organic amine with more amino groups,hydroxyl structure and strong electronegativity exhibited better extraction performance.Under the optimal conditions,the extraction yields of phenols in coal tar by the monoethanolamine or ethylenediamine aqueous solution are above 80%,and the recovery yields of amines reach 99%.Furthermore,the probable geometries of complexes formed by the combination of phenols and organic amines were calculated by density function theory.In addition,several thermodynamic models were evaluated through comparing the relative deviation of simulation results by ASPEN PLUS to the experimental ones,which provide feasibility thermodynamic models for the simulation of extraction process.The present work affords a mild,efficient and green approach for the extraction of phenols from coal tar by an aqueous solution of amines in industry application.

1.Introduction

The low-temperature coal tar is a by-product in the process of coal pyrolysis under a lower temperature.It is an important organic chemical raw material containing thousands of compounds,in which phenols account for 20%–50% of its mass [1]and have a higher value for the production of synthetic resins,plant protection agents,preservatives,coatings,pesticides,antioxidants and other chemical products[2–4].Thus,direct extraction of phenols from the coal tar is a key step for their comprehensive utilization.On the other hand,it can reduce the hydrogen consumption during the hydrocracking to light oils of coal tar due to removing most of the phenolic compounds [5–7].

At present,the main method for extracting phenols from coal tar is still the traditional alkali method [8].After the coal tar contacts with an inorganic alkali solution,a phenolate solution can be formed from its phenols and alkali,and subsequently neutralized by a mineral acid to release the phenols.This method has to consume some inorganic acid and alkali as well as generate a plenty of wastewater containing salt and little phenols [9,10].Hence there have been an amount of investigation on improvement for the extraction processes since many years,such as using quaternary ammonium salts [11,12],urea [13] and the like to extract phenols by the principle of deep eutectic crystallization[14,15].However,owing to the high viscosity and high melting point of these materials,their industrial applications are limited[16,17].Someone also has proposed the use of ionic liquids [18–22] or adsorbent [23–25] to extract phenols from coal tar and its distillates,such as imidazole [26] and alcohol amine salts [27],but their hygroscopicity and economic problems hinder their industrial applications.

In order to take phenols out of the coal tar environmentalfriendly with a lower cost,the most important thing is to choose an extractant,which has not only a higher selectivity to pick up the phenols from the coal tar,but also a lower cost to regenerate the extractant and simultaneously gather the phenols with little environmental impact.The aqueous solution of some amines,such as ethanolamine,has shown an excellent selectivity to draw phenols from coal tar [28].This is probably a hydrogen bond formed between amidogen of the amines and the hydroxyl of the phenols,based on the Lewis acid-base theory and the Bronsted acid-proton theory [20].However,it is usually uneasy and needs a higher energy consumption to break the hydrogen bond for regeneration of the extractant [28].Hence,based on the characteristic of the amine aqueous solutions used for absorption and desorption of CO2in flue gases[29–32],we recently proposed an indirect method to restore the extractant with two step reactions[33]:(1)the aqueous solution of amines containing phenolic compounds is saturated with CO2through forming an amino-carbonate in aqueous phase,that makes phenols free from the amines to become an oil phase;(2) the aqueous phase containing amino-carbonate is heated to desorb CO2from the aqueous solution.According to the zwitterion reaction mechanism of CO2absorption by organic amine solution,it can be inferred that this method can release the phenols bound with organic amines and effectively lower the energy consumption for the regeneration of the extractant.In addition,the amine aqueous solution normally has a low viscosity,which is beneficial to the extraction processes.

The conceptual design flowsheet is shown in Fig.1,where after the extraction agent(stream 2)extracts phenols in coal tar(stream 1)by the extraction tower,the extraction phase(stream 4)directly enters the absorption tower to absorb CO2(stream 17),and converts the organic amine in the extraction phase into amine salt to release phenols.Next,the crude phenols oil (stream 7) and organic amine carbonate solution (stream 8) are separated by phase separator,and then the organic amine carbonate solution(stream 8) is pumped into the desorption tower to desorb CO2and regenerate organic amine.Finally,the export streams of regeneration tower are recycled as extractant after mixing with fresh materials (stream 18).

In this paper,we will focus on the extractant selection,the extraction conditions optimization and the thermodynamic model selection to support the effective design or optimization for phenols extraction process by means of the aqueous solution of amines.Moreover,these works wills also contribute to increase the understanding of the development process and evaluate the potential economy of the development process.

2.Experimental

2.1.Reagents and analytical methods

Monoethanolamine,(MEA,AR,99%),diethanolamine,(DEA,AR,99%),triethanolamine (TEA,AR,98%),ethylenediamine (EDA,AR,99%),ethylamine (EA,68%–72% in H2O),diethylamine (DA,AR,99%),cyclohexylamine (CA,AR,98%),n-propylamine (NPA,AR,99%),isopropylamine (IPA,AR,99%),phenol (AR,98%),o-cresol(AR,98%)andm-cresol(GC,98%)were obtained from Macklin Biochemical Reagent corporation.Toluene (AR,99.5%) andn-hexane(AR,99.5%) were afforded by Kermel Chemical Reagent Corporation.The quantitative analysis of phenols such as phenol,o-cresol andm-cresol in the model oil and the raffinate was performed on Agilent gas chromatography equipped with HP-5 capillary Column(30 m×0.320 mm×0.25 μm)and hydrogen ion flame detector,while methanol and chlorobenzene were chosen as the solvent and the internal standard,respectively.The chromatographic conditions for the injector and detector were 563.15 K.The temperature programmed was as follows:initial oven temperature was 333.15 K and held for 1 min ramped at 5 K?min-1to 533.15 K.The amines in aqueous solutions,such as the extract phase,can be analyzed by acid-base titration.The sulfuric acid solutions(0.1 mol?L-1) and the methanol solution of 0.1 vol% bromocresol green were used as the standard solutions and indicator,respectively.

2.2.Method and procedure

The content of a coal tar is very complex,nevertheless,most of which are non-polar organic compounds except for phenols [34].For convenient illustration and analysis,this work used a model oil consisting of phenol,o-cresol,m-cresol and a solvent to represent for a coal tar in each equilibrium extraction.Considering some compounds in a coal tar are polar and they may behave a different way in the extraction process,this work prepared a model oil by using toluene with a weak polarity as the solvent,named as toluene,besides usingn-hexane with a weak non-polarity as the solvent as usual [24,26],named asn-hexane.The two model oils,n-hexane and toluene,all comprised 22.27% (mass) of phenolic compounds and 77.73% (mass) ofn-hexane or toluene,in which the contents of phenol,o-cresol andm-cresol were 11.14% (mass),5.56%(mass) and 5.57%(mass)respectively.In addition,the aqueous solutions with various amines and their concentrations were also prepared to be the extractants.

Fig.1.The conceptual design flowsheet of capturing phenols from coal tar by amines aqueous solution.Extraction tower(E-1),Absorption tower(E-2),Decanter(E-3),Pump(E-4),Desorption tower (E-5),Reboilet (E-6),Condenser (E-7) and Flash tank (E-8).

As an example of one equilibrium extraction,a MEA aqueous solution of 20% (mass) was used to extract phenols from model oil:40.0 g of the model oil and 27.0 g of MEA aqueous solution were added in a flask and stirred vigorously with a magnetic stirrer for 60 min at 293.15 K.After stopping the agitation,the mixture was poured into a separatory funnel where it was divided into the upper phase (oil or raffinate phase) and lower phase (aqueous or extract phase),and then they were taken for analysis,respectively.According to the analytical results,the extraction yield of phenols,Yp,and the recovery yield of amines,Ya,can be determined by the following definitions,respectively.

whereMpandRpare the mass quantities of phenols in the model oil and the raffinate phase;EaandRaare the mass quantities of amine in the extractant and raffinate phase,respectively.

3.Results and Discussion

3.1.The effect of molecule structure on extraction performance

To investigate the effect of different functional groups of amines on the extraction of phenols from the model oils,this work chose the aqueous solutions of MEA,DEA,TEA,EDA,EA,DA,NPA,IPA and CA as the extractants respectively.The extraction yields of phenols for an equilibrium extraction under different conditions,Yp,had been measured as listed in Fig.2.It can be seen from the Fig.2 of graphs (a) and (b),when the mole ratio of the amine to the phenols,ξ=0.1,Ypof phenol,o-cresol andm-cresol fromnhexane can reach 80% except for CA.Moreover,due to the effect of phenols liquid–liquid equilibrium,Ypby an amine solution of 20% (mass),orC=20% (mass),is higher than that by one of 40%(mass) at the same ξ.The comparison between graphs (b) and(c) of Fig.2 shows us thatYpby using the aqueous solutions of MEA,DEA,TEA,EDA and EA as the extractants fromn-hexane,may increase with increasing ξ from 0.1 to 1,but obviously decrease by using other alkyl-amines at the same way.This may be caused by that DA,NPA,IPA and CA usually have a higher solubility in an organic solvent,liken-hexane,and then may have a higher partition coefficient of oil phase to water phase than the others do.Besides,comparing the Graphs (c) with (d) of Fig.2,it can be found thatYpfor toluene are much smaller than those forn-hexane.It indicates,despite the total content of polar organic compounds is minor in coal tar,that they may result in an unnegligible and negative effect on extracting phenols by an aqueous solution of amines from coal tar,especially by that of DA and CA.Fig.2 also clearly shows us an order of extraction selectivity to phenols for the amines,that is,Yphenol>Ym-cresol>Yo-cresol.The reason may be the methyl group in the phenols not only makes its hydroxyl group difficult to impact on amidogen in the amines,but also may influence electronic distribution in the molecules of phenols that can reduce the acidity of phenols.

The recovery yields of amines,Ya,had been measured as shown in Fig.3.It can be seen that the recovery yields of alcohol amines are generally higher than that of alkyl amines in the extraction process.With the increase of molar ratio and concentration,the recovery yields of alcohol amines basically unchanged,while the recovery yields of alkyl amines are manifestly become smaller except for EA and EDA.It may be that the amino and hydroxyl groups in organic amines are hydrophilic groups,which can improve the solubility of organic amines in water.In addition,theYafor toluene are much smaller than those forn-hexane,the reason may be that the solubility of organic amine in polar solvent higher than non-polar solvent.Comparing Fig.3 with Fig.2,it can be found thatYawas positively correlated withYp,which indicates that the partition coefficient of amine has a significant effect on the extraction yield of phenolic compounds.

In general,with the exception of EDA and EA,the alkyl-amines usually exhibit an inferior extraction performance to the alcoholamines.This phenomenon can be well explained from the perspective of the molecular structure.The amines’electron distribution inFig.4 shows that the electronegativity of N in alcohol-amines follows an order of MEA >DEA >TEA which results in a same order for their extraction yields of phenols[35].The number of amidogen in an amine can also enhance its extraction yields of phenols and correspondingly EDA with a double–NH2gives EDA the highest extraction yields of phenols.The electronegativity of N atom in EA,NPA and IPA are significantly stronger than that in the remaining alkyl-amines,it improved the ability of amine combine with phenols.However,the solubility of EA inn-hexane and toluene is lower than that of NPA and IPA,resulting in a higherYpby EA in graphs (c) and (d) of Fig.2.Although the electronegativity of N atoms in DEA and TEA is obviously weaker than that in EA,their extraction performance are better than EA as shown in graphs(b),(c) and (d) of Fig.2.This may be the hydroxyl group in alcohol-amines plays a surface activation role in an extraction process,resulting in a significant decrease of interfacial tension.Besides,the electropositivity of H atom of the phenolic hydroxyl group for the phenols follows an order of phenol >m-cresol >ocresol,that is the same as their selectivity extracted by amine aqueous solution.

Fig.2.The extraction yields of phenols for various amines solutions from the model oil.Graphs (a)-(c):extracting from n-hexane,(d):extracting from toluene;Graph (a):C=20% (mass),(b)-(d): C=40% (mass);Graphs (a) and (b):ξ=0.1,(c) and (d):ξ=1; T=293.15 K; t=60 minutes with stirring.

Fig.3.The recovery yields of amines for various amines solutions from the model oil.(a)-(c):extracting from n-hexane,(d):extracting from toluene;(a): C=20%(mass),(b)-(d):C=40%(mass);(a)and(b):ξ=0.1,(c)and(d):ξ=1;T=293.15 K;t=60 minutes with stirring.

3.2.The effect of various factors on extraction performance

To further study the extraction performance and mechanism of amines,MEA and EDA are selected to examine the effects of their extraction time,t,concentration,C,mole ratio,ξ,and extraction temperature,T,on the extraction yield of phenols and the recovery yield of amines,YpandYa,since they may have the best selectivity to extract phenols from coal tar among the amines mentioned above.

Considering that the dynamic behavior of extraction and especially its equilibrium state for phenols by aqueous solutions of amines are necessary for an industrial practice,the effect of extraction time onYpandYawas examined in this work firstly.Fig.5 shows the variation ofYpandYafor the extraction of phenols fromnhexane by MEA or EDA aqueous solution of 20% (mass) with the time under the conditions of ξ=0.1 andT=293.15 K,where the extraction timet=0,1,2,5,10 and 20 minutes were selected respectively.The extraction witht=0 was in fact that an extractant and a model oil were added into a separatory funnel to stand for 5 minutes without stirring.The extraction with any other time was under sufficiently stirring during the extraction time.As shown in Fig.5,the extraction rates of phenols by the MEA or EDA aqueous solution fromn-hexane are very fast and seem to be able to reach an equilibrium within 1–2 minutes,but the latter rate is much quicker than the former one.This indicates it is feasible to use the method of liquid-liquid equilibrium to their industrial application.However,to ensure the experimental examinations for the other effect factors had reached their equilibrium states,this work kept a fixed time of 60 minutes for all of them.

Additionally,it can found,from Fig.5,thatYpfor the phenols by EDA is distinctly higher than those by MEA and the differences ofYpamong phenol,o-cresol andm-cresol are quite small by EDA,not like MEA.This should also be caused by its double amidogen,as mentioned above.

Fig.4.The electrostatic potential map of amines and phenols were calculated by Gaussian-03 package,version 3.6.(Method:Job type:opt+freq;Ground state:DFT;Default Spin:B3LYP;Basic Set:6-311G++(d,p),isoval=0.0004,scale spans from -6×10-2(red) to 6×10-2(blue)).

Fig.5.Variation of Yp and Ya for MEA (left) or EDA (right) in n-hexane with t at C=20% (mass),ξ=1 and T=293.15 K with stirring.

The variation ofYpandYawith the concentrations of MEA and EDA in their aqueous solutions is shown in Fig.6.It can be found thatYpof phenols andYaof amine generally decrease with increasing the concentrations in range of 20% (mass) to 60% (mass) at ξ=0.1 for both MEA and EDA.This is because that a higher concentration of an amine means a smaller amount of aqueous solution to accommodate the phenols at a given value of ξ and so may turn downYpof the phenols.In addition,a higher concentration can afford a higher mass transfer driving force for the amine from the water-phase to oil-phase and makesYaof the amine lower slightly.It also can be seen from Fig.6 that the decrease extent ofYpof the phenols for EDA is smaller than MEA,since EDA has a double -NH2that means it has a higher capability to combine with more phenols at the same value of ξ.

On the other hand,when keeping a fixed concentration of MEA or EDA such as 20% (mass) and raising ξ,as shown in Fig.7,Ypof the phenols for both MEA and EDA will increase as ξ <1 and reaches a constant of 90%and more as ξ>1,whileYaof the amines changes little.This is an understood case.In fact,at a given concentration of amines,raising ξ also means increasing the ratio of the aqueous solution to the model oil,which is increasing the content of the extractant for phenols.In other words,no matter which concentration when ξ is changed,the amount of aqueous solution,the phase ratio of extractant to model oil (W/O),will vary with its change.Consequently,the concentration of amines,the mole ratio of amine to phenols and the W/O are correlative in reality and so their optimization for an extraction process is complicated.However,as a result discussed above,ξ=0.5–1.0 and theC=20%(mass)–40% (mass) should be desirable,that can give a suitable mass ratio of W/O in the practical application.

Fig.6.Variation of Yp and Ya for MEA (left) or EDA (right) in n-hexane with C at ξ=0.1, T=293.15 K and t=60 minutes with stirring.

Fig.7.Variation of Yp and Ya for MEA (left) or EDA (right) in n-hexane with ξ at C=20% (mass), T=293.15 K and t=60 minutes with stirring.

The extraction temperature should have a significant effect on an extraction process,since changing the extraction temperature can affect not only its thermodynamic equilibrium but also the interphase and intraphase mass transfer in it.This work measured the variation ofYpandYafor MEA or EDA inn-hexane with the temperature from 273.15 K to 313.15 K and the results are summarized in Fig.8,which shows that as the temperature increased,Ypat equilibrium state moves in the negative direction,especially for EDA,while the recovery yield of amines,Ya,basically unchanged with the elevated temperature.The variation ofYpwith temperature indicates that the combination between the amines and the phenols is exothermic.Nevertheless,it is not the lower the temperature,the better for the extraction yields in practice,because the lower the temperature,the slower the rates of interphase and intraphase mass transfer,and then the longer the time to reach extraction equilibrium.In addition,the electric energy consumption will becomes serious if the coolant used to keep a lower extraction temperature has to take a very low temperature.Hence there is an optimization temperature for an industrially process of extracting phenols from a coal tar by an aqueous solution of amines.

3.3.Possible geometry conjecture

Although the extraction mechanism for the organic amine and phenolsviahydrogen bond has been reported in many literatures[28],the possible geometries between organic amine and phenols had not been studied in detail.There are two different geometries between amines and phenols by the form of hydrogen bonds,as shown in Fig.9.To further demonstrated the stable geometries between phenols and amines,all possible geometrical configurations formed by hydrogen bonds were calculated by DFT method in Gaussian 09.The energies of formed complexes with different type of hydrogen bonds were shown in Table 1.The result shows that the energy of the complex formed by hydrogen bond between N atom in amine and H in phenolic hydroxyl is lower than that of between O atom in phenolic hydroxyl and H in amine,indicating the geometrical configuration of type (Ι) is more stable than that of type (Π) in Fig.9 [35,36].Therefore,the binding way between phenols and amines can be expressed as type (Ι) in Fig.9.

Table 1The energies of the formed complexes between amines and phenols by hydrogen bonds①

3.4.Selection for the thermodynamic models

As well known,it is very important to select a suitable thermodynamic model with better predictability for simulating such an extraction process of phenols from coal tar by amines aqueous solution,in which system the models based on the liquid activity coefficient methods,particularly NRTL,UNIQUAC,UNIFAC and the UNIFAC with modification,are commonly recommended.However,it is truly a very complicated work how to use a well predictable thermodynamic model and model parameters for this extraction system.Hence,the above recommended models are simply evaluated through comparing the simulated results by ASPEN PLUS to the measured ones by this work for the singlestage extraction equilibrium of phenols from a model oil and a coal tar respectively,and all of binary interaction parameters are estimated by the group contribution method.

Fig.8.Variation of Yp and Ya for MEA (left) or EDA (right) in n-hexane with T at C=20% (mass),ξ=1 and t=60 minutes with stirring.

Fig.9.The possible formed geometrical configurations between amines and phenols.

The simulation of a single-stage extraction equilibrium process by ASPEN PLUS is shown as Fig.10,where the extractant and the model oil or the coal tar are mixed,and then directly separated into the water and oil phases through a decanter module.

Since a real coal tar is a mixture of aromatic and non-aromatic components and highly non-ideal,the model oil,n-hexane,may be more representative and was chosen to participate in the comparison.The averaged relative deviations of the simulatedYpto the measuredYpfor extracting phenol,o-cresol andm-cresol fromnhexane by MEA and EDA are shown in Tables S2 and S3 of SI.The averaged relative deviation is defined as

Fig.10.The single-stage extraction process simulated by Aspen Plus.

wherenis the run number of the extraction.

As found in the Tables 2 and 3,the built-in property methods of UNIQUAC and UNIFAC in ASPEN PLUS appear a quite good applicability for the extraction systems fromn-hexane by MEA or EDA.Although the averaged relative deviations by UNIF-LBY and UNIFDMD for the systems are tolerable in engineering,they are in fact the system deviations on the low side.However,NRTL seems unsuitable for the systems by MEA,but behaves very well for those by EDA.This may be caused by the characteristics of MEA molecules with both -NH2and -OH.

Table 4 shows the compared results between the simulated and the measured for the single-stage extraction by the aqueous solutions of three amines from a distilled coal tar that is the cut fraction from 393.15 K to 613.15 K of a Yulin rude coal tar with a total 45.5%(mass)of phenols.The main components and their content for the distilled coal tar,used for the single stage simulation by Aspen Plus,are listed in Table S1 of Supplementary Material,which is mainly determined by the actual measurement results of some components and the analysis result by a GCMS-QP2010Plus in Table S2 of SI.

As seen in Table 4,the predictability of UNIF-DMD and UNIFLBY seems still acceptable for the extraction by MEA or EDA in despite of there being some slight over-valuation,while UNIQUAC and UNIFAC is not suitable for the systems although they behave quite satisfactory for then-hexane systems.The main reason for the difference is that the distilled coal tar has a much more complicated composition in both the number and classes of components thann-hexane,hence too many thermodynamic parameters may be needed to estimate.Accordingly,for the application of thermodynamic methods in the simulation to an extraction system of phenols from a coal tar by amines,it should be preferable to choose UNIF-LBY or UNIF-DMD or the improved ones on the basis of them unless there have been the thermodynamic data related.

Table 2The averaged relative deviations of the simulated Yp to the measured Yp by MEA

Table 3The averaged relative deviations of the simulated Yp to the measured Yp by EDA

Table 4The relative deviations of Yp simulated to that measured from the distilled coal tar

3.5.Cycle experiment

In order to further verify the feasibility of the design flowsheet,a cycle experiment with 20% (mass) EDA aqueous solution as extractant was investigated in detail.The experimental results were as shown in the Fig.11,with the number of cycles increases,theYPin extraction process decreased firstly until reach stable,while that in the absorption process increased until reach stable.When the circulation experiment reached equilibrium,theYpwas higher than 95% in the extraction process and reached 100%in the absorption process.TheYEDAis close to 100% in the extraction process and 89% in the absorption process.The regeneration yield of EDA was 54.6%.Although the regeneration yield of EDA is low,the cyclic extractant containing amino carbonate can stably and efficiently extract phenols fromn-hexane.

Fig.11.Variation of Yp and Ya for 20% (mass) EDA aqueous solution in n-hexane with the number of cycles at ξ=1 with stirring.(a):extraction process,(b):absorption process).(The reaction temperature of desorption process was 403.15 K,and that of the other processes was 303.15 K.)

4.Conclusions

As discussed above,it is a mild and highly efficient method to extract phenolic compounds from coal tar by the aqueous solution of amines.The alkanol amines have a better extraction performance for phenols in model oil than alkyl amines except for EDA and EA.The experiment results indicate that the molecular structure of organic amines,such as the types and the quantity of functional groups,has a significant effect on the extraction performance.This effect may be manifested by the electrostatic potential and the O/W partition coefficient of amines.Therefore,it is possible to improve the extraction performance of organic amines for phenols in coal tar by adding additives,optimizing extractant and so on.The operating parameters such asT,Cand ξ also have an important influence on the systems.In views of the industrial application,the MEA and EDA aqueous solutions withC=20%(mass)–40%(mass)can be selected to extract phenolic compounds fromn-hexane under ξ=0.5–1.0 at atmospheric temperature.In addition,the relative deviations between the simulation result and the experiment result were used to evaluate the reliability of several recommend thermodynamic models.The results show that the UNIF-LBY and UNIF-DMD thermodynamic models are not as good as the UNIQUAC and UNIFAC thermodynamic models for the extraction systems by MEA or EDA fromnhexane.However,they have better applicability for the extraction system of real coal tar.The reason maybe that the UNIFAC modified models are more suitable for complex system considering the temperature and molecular sizes [37,38].Finally,a cycle experiment with 20% (mass) EDA aqueous solution as extractant was carried out,which verified that the design process of extraction phenols by organic amine aqueous solution is feasible and has a good potential of industrial application.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We are grateful for the financial support by the National Natural Science Foundation of China(21676226,21606186),the Natural Science Foundation for Distinguished Young Scholars in Hunan Province (2018JJ1023),Key Research and Development Program in Hunan Province (2019GK2041),and Collaborative Innovation Center of New Chemical Technologies for Environmental Benignity and Efficient Resource Utilization.

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2021.01.008.

Nomenclature

CConcentration,% (mass)

DThe average relative deviation

TTemperature,K

tTime,min

YYields,%

ξ Mole ratio of amine to the phenols

β Mass ratio of extractant to oil

Subscript

a Amines

p Phenols