• No results found

Vapor–liquid behavior of phenol and cumene in ternary and quaternary mixtures

N/A
N/A
Protected

Academic year: 2021

Share "Vapor–liquid behavior of phenol and cumene in ternary and quaternary mixtures"

Copied!
10
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Received: 30 December 2020 Revised: 19 February 2021 Accepted article published: 19 February 2021 Published online in Wiley Online Library: (wileyonlinelibrary.com) DOI 10.1002/jctb.6704

Vapor

–liquid behavior of phenol and cumene

in ternary and quaternary mixtures

Lisette MJ Sprakel,

a

G Bargeman,

b,c

Lara G Sanchez

d

and B Schuur

a*

Abstract

BACKGROUND: Although phenol is a key intermediate in the plastics and polycarbonate industry, it is also a toxic component that requires removal from dilute aqueous streams, potentially by liquid–liquid extraction (LLX). For LLX, cumene is suggested as a solvent as it is already present in processes in the polycarbonate industry. For the recovery of cumene from phenol by dis-tillation, knowledge on vapor–liquid equilibrium (VLE) behavior is important, in combination with how this is affected by other components possibly present as an impurity or explicitly added as a solvent. This was investigated in this work.

RESULTS: The binary cumene–phenol system shows a tangent pinch in the binary VLE diagram. Addition of a range of impuri-ties and solvents showed that hydrogen bond accepting compounds strongly improve the relative volatility of the mixture, whereas dodecane, not capable of forming hydrogen bonds, has a negative effect on the relative volatility.

CONCLUSION: Addition of polar components with hydrogen bonding abilities, i.e. ketones or ethers, affected the relative volatility of cumene over phenol the most positively. Combining two types of components results in similar effects, and clear synergistic effects could not be shown based on current VLE measurements. Addition of an apolar component in combination with polar com-ponents with hydrogen abilities had only a minor effect on the relative volatility.

© 2021 The Authors. Journal of Chemical Technology and Biotechnology published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry (SCI).

Supporting information may be found in the online version of this article.

Keywords: renewable phenol; cumene; polycarbonates; vapor–liquid equilibrium; extractive distillation

INTRODUCTION

Phenol is a key intermediate in the production of polycarbonates and plastics in the chemical industry. It is produced by the cumene process: a process in which benzene and propylene are used as reac-tants to form cumene. The cumene is then oxidized to cumene hydro-peroxide, which is decomposed in the presence of an acid to form phenol and acetone. Phenol and acetone are further processed to polycarbonates and precursors for plastics.1,2Despite its industrial importance, phenol is a toxic chemical and treatment of phenol-containing wastewaters is of utmost importance. For removal of phe-nol from wastewater, several techniques have been applied, where the concentrations in the aqueous phase determine which technique is most applicable.3 For example, at low phenol concentrations,

biodegradation,3 various types of oxidation-based processes,3

solid-phase extraction,3liquid–liquid extraction (LLX),4,5ozonation3

and reverse osmosis/nanofiltration are applicable.3At higher

phe-nol concentrations LLX,3,6 membrane-based pervaporation and extraction,3,7adsorption and distillation are more applicable.3,8There are also various less commonly applied processes based on, for exam-ple, biological treatment and oxidation.3,9-11Next to the concentration in the aqueous stream, the maximum allowable concentration in the remaining water stream is a determining factor.3For surface water the maximum allowable concentration is less than 1 ppb of phenol.9Thus the vapor–liquid equilibrium (VLE) behavior of phenol and cumene is

industrially relevant in two cases: that is, either phenol can be present in relatively large concentrations in non-aqueous mixtures in indus-trial processes where phenol is formed from cumene, and where it is to be separated from cumene and other impurities; or phenol is pre-sent in highly diluted aqueous streams of several mass percent or less, with additional impurities present. The lower concentration aqueous solutions are typically observed in downstream separations at the industrial production from cumene.

Several extractants have been applied for liquid–liquid extraction of phenol from aqueous streams, e.g. sulfoxides,12cumene13,14and ethyl

* Correspondence to: B Schuur, Sustainable Process Technology Group, Faculty of Science and Technology, University of Twente, Drienerlolaan 5, Meander 221, 7522 NB Enschede, The Netherlands. E-mail: b.schuur@utwente.nl a Sustainable Process Technology Group, Faculty of Science and Technology,

University of Twente, Enschede, The Netherlands

b Expert Capability Group, Research Development & Innovation, Nouryon Che-micals B.V., Zutphenseweg 10, P.O. Box 10, 7400 AA Deventer, The Netherlands c Membrane Science and Technology Cluster, Faculty of Science and Technol-ogy, University of Twente, Drienerlolaan 5, 7522 NB Enschede, The Netherlands d SABIC Europe B.V., SABIC Technology Center Geleen, P.O. Box 319, 6160 AH

Geleen, The Netherlands

© 2021 The Authors. Journal of Chemical Technology and Biotechnology published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry (SCI).

(2)

ether.15In membrane-based extraction processes, combinations of methyl tert-butyl ether, cumene and hydrocarbon mixtures have been applied.7The process of extractive distillation with tricresyl phosphate was patented in the 1970s.8

Most of these solvents introduce new impurities into the process – for example, through byproduct formation as a result of the regeneration of solvent by stripping with a base.13-15Conversely, when cumene– an intermediate in the production process of phe-nol and polycarbonates– is applied as an extractant, the solvent is not an impurity in itself and cumene can be recycled in the produc-tion process.14The application of cumene in LLX of phenol resulted in high distribution ratios, although these are strongly affected by other polar impurities and contaminations.

For a viable process, successful regeneration of the cumene from the phenol–cumene mixture after extraction is important. Separation of phenol from the reactant cumene is also an important step in the production process of phenol, as the presence of cumene is inhibit-ing the oxidation of cumene to cumene hydroperoxide. Therefore, phenol is to be removed from the recycled product stream in the cumene process at high purity. This is generally performed by distil-lation, condensation and a wash with caustic soda to remove other impurities.6 Extractive distillation (ED) of the cumene and phenol mixture may be an advantageous alternative for this process.

The aim of this research is to obtain more insight into the VLE behavior of the cumene–phenol system, and to study how impurities affect their separation. Effects of the relevant Table 1. Chemicals used in this study

Chemical Compound name (abbreviation) Purity

Acetaldehyde 99%

Acetone Lichrosolv® ≥99.8%

Acetophenone (AcPhO) 99%

Alpha-methylstyrene (AMS) 98%

Cumene 98

Dimethyl benzyl alcohol (DMBA) 97

Cyclohexane 99.5

Diethylene glycol dibutyl ether (DEGDBE) ≥99%

Dodecane (DDec) ≥99%

1-Dodecanol (1 Dec) ≥98%

Glycerol 99%

Hexadecane (HexD) 99%

Hydroxyethyl ethylene diamine (Heediamine) 99%

Hydroxyacetone (HyAc) ≥95%

40-Hydroxyacetophenone (HyAcPhO) 99%

Mesityl oxide (MesOx) 98%

5-Methyl-2-hexanone (5-m-2-one) 98.0%

Methyl isobutyl ketone (MIBK) ≥99.7%

1 Octanol (1-Oct) ≥99%

2-Octanone (2-OctO) 98%

Polyethylene glycol (average molar weight 200 g mol−1) (PEG200) —

Phenol (PhOH) ≥99.5%

Note: All chemicals were provided by Merck KGaA/Sigma-Aldrich.

0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 . 0 0 0.2 0.4 0.6 0 8. 1 2 1.0 3 4 5 6 7 yCum ene (a) xCumene α (b) xCumene

Figure 1. Binary VLE data for the system phenol cumene at 300 (squares) and 1000 mbar (circles), showing (a) vapor and liquid composition and (b) relative volatility as a function of the liquid composition.

(3)

intermolecular interactions– for example, bonding mechanisms and interactions with dipole inducing components– were studied by experimental investigation of the VLE systems. With the increased fundamental understanding it is possible to design and develop separation processes for cumene–phenol separa-tions based on extractive distillation.

MATERIALS AND METHODS

Chemicals

The chemicals used in this study, the associated compound name (abbreviation) used in the rest of the article, purity and provider are listed in Table 1. All compounds were used as received. Vapor–liquid equilibria

VLE were measured using Fischer Labodest VLE602 ebulliometers, in which a magnetically stirred equilibrium cell is present that is connected to a Cottrell circulation pump. The cell is heated with an immersion heater. The pressure can be set between 100 mbar and either atmospheric pressure using an Edwards E2M1.5 or 4 bar using a Pfeiffer DUO 3 vacuum pump. The power supplied to the reboiler and the heating mantle temperature can be set. The equilibrium temperature of the mixture as well as the liquid temperature in the reboiler were measured using Pt-100 thermo-couples. For each sample composition the equilibrium cell was

filled with 80 mL of sample (including the solvent). Mixtures con-sisting of phenol, cumene and a single solvent were prepared on a specific weight-based solvent-to-feed (S/F) ratio, the exact ratio being described for each of the experiments when the results of that experiment are discussed in the Results and Discussion sec-tion. In the case of simultaneous application of multiple additives, including hydrogen bonding additives and dodecane, which does not act as hydrogen bond donor or acceptor, samples were pre-pared by applying a dodecane mass fraction in the total mix of 25 wt%. The other solvents were applied on a 1:1 mass basis. As a result of differences in molar mass and volume of the solvents, there are differences in the molar-based S/F ratios. Most of the applied hydrogen bonding additives have boiling points with at least a 20°C difference from the boiling points of cumene and phenol, which minimizes chances of azeotropic behavior, and downstream fractionation would be possible thermally. The impurities were selected so that their functional groups covered a large range. The applied impurity dodecane was chosen as model component with the largest anti-solving effect compared to other impurities expected. Specific compositions of mixtures, solvents and solvent mixtures in combination with the applied binary compositions are given in the following sections when the specific measurement data are presented. When the equilib-rium temperature was reached after 30–90 min, samples of both the liquid phase and the condensed vapor phase were taken.

0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 0.80 0.85 0.90 0.95 1.00 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 0.80 0.85 0.90 0.95 1.00 yCu m e ne xCumene (MIBK) (2-OctO) (AcPhO) (2-OctO) (5-m-2-one) (MesOx) (HyAcPhO) (MIBK) (5-m-2-one) (a) yCu m e ne xCumene (b) (DEGDBE) (PEG200) 0.65 0.60 0.70 0.75 0.80 0.85 0.90 0.95 1.00 0.80 0.85 0.90 0.95 1.00 yCumene xCumene (PEG200) (1-Oct) (HyAcPhO) (1-Dec) (DMBA) (c) (DDec) (AcPhO)

Figure 2. Binary (closed symbols) and (pseudo-)binary (open symbols) VLE data of cumene phenol at 300 mbar for (a) solvents with ketone functionality, (b) ethers and (c) solvents with other functionalities, all at a weight-based S/F ratio of 1.

(4)

Gas chromatography–mass spectrometry (GC–MS)

Liquid-phase and condensed vapor-phase samples from the ebul-liometer experiments were dissolved in acetone and analyzed with a 7890A Ms 5975C Agilent gas chromatograph–mass spec-trometer withflame ionization detector and an Agilent HP-5Ms, HP19191S-433 column. The carrier gas applied was helium. The oven temperature was increased from an initial temperature of 45°C up to 250 °C using a linear profile of 15 °C min−1.

RESULTS AND DISCUSSION

VLE data for binary mixtures of phenol–cumene

To validate the measurements, binary VLE data at 300 and 1000 mbar were obtained for the mixture of phenol (Tm= 41 °C,

Tb= 181.7 °C) and cumene (Tm= −96 °C, Tb= 152.4 °C), where

Tmis melting point and Tbis boiling point. The results are shown

in Fig. 1, where the VLE data are expressed in mole fraction of

0.0 0.2 0.4 0.6 0.8 1.0 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.6 0.8 1.0 (0.5 / 1.3) (0.5 / 1.2) (0.2 / 0.6) (1.1 / 2.6) (0.4 / 1.3) (0.4 / 1.3) (0.3 / 1) (0.8 / 2) (0.8 / 2) yCu me ne xCumene (0.2 / 0.6) yCu me ne xCumene (a) yCu me n e xCumene (b) yCu me n e (d) xCumene (c) yCum ene xCumene (-/0.2) (e) (-/0.2)

Figure 3. Binary (closed), pseudo-binary (open, with a solvent) and quaternary (grey, with a solvent and anti-solvent dodecane) VLE of the cumene– phenol system. Solvents applied are (a) PEG200, (b) DEGDBE, (c) MIBK and (d) 1-octanol, at 300 mbar at an S/F ratio of 1. Thefirst number in parentheses indicates the molar-based ratio of the tertiary component to the total number of moles of the cumene–phenol mixture. In (e), only dodecane was added for comparison. Dodecane was in all cases applied at a mass fraction of 25 wt% in the total mixture (at a molar based ratio of quaternary component dode-cane to the total of cumene–phenol mixture indicated by the second number in parentheses).

(5)

0.0 0.2 0.4 0.6 0.8 1.0 5 10 15 20 25 30 0 0. 0.2 0.4 0.6 0.8 5 1.0 10 15 20 25 0.0 0.2 0.4 0.6 0.8 1.0 5 10 15 20 25 0 0. 0.2 0.4 0.6 0.8 2 1.0 4 6 8 10 12 0.0 0.2 0.4 0.6 0.8 0 1.0 2 4 6 8 (0.2 / 0.6) (0.5 / 1.3) (0.5 / 1.2) (0.2 / 0.6) (1.1 / 2.6) (0.4 / 1.3) (0.4 / 1.3) (0.3 / 1) (0.8 / 2) (0.8 / 2) α (a) α xCumene (b) α xCumene (c) α xCumene (d) α xCumene (e) xCumene (-/0.2) (-/0.2)

Figure 4. Binary (closed), pseudo-binary (open, with a solvent) and quaternary (grey, with a solvent and anti-solvent dodecane) relative volatility⊍ of the cumene–phenol system. Solvents applied are (a) PEG200, (b) DEGDBE, (c) MIBK and (d) 1-octanol, at 300 mbar at an S/F ratio of 1. The first number in parentheses indicates the molar-based ratio of the tertiary component to the total number of moles of the cumene–phenol mixture. In (e), only dodecane was added for comparison. Dodecane was in all cases applied at a mass fraction of 25 wt% in the total mixture (at a molar-based ratio of quaternary com-ponent dodecane to the total of cumene–phenol mixture indicated by the second number in parentheses).

(6)

cumene. The results are similar to those published by Cepeda et al.16for the cumene–phenol system at atmospheric pressure. The equilibrium curve shows a tangent pinch (relative volatility ⊍=ycumene=x

cumene

yphenol=x phenol

close to 1) at low phenol fraction, where xiis liquid

phase mole fraction of i, and yiis vapor phase mole fraction of i.

Due to the tangent pinch, the use of distillation to obtain pure cumene is hindered and makes it an energy-intensive process. The influence of additives on the VLE behavior is of interest to evaluate both impact of impurities, and of potential entrainers in ED. The system is a case where ED is potentially interesting as a separation process.17 Since the tangent pinch is obtained at

low phenol fraction, the use of a solvent that increases the relative volatility of cumene over phenol is beneficial. Moreover, cumene is still (slightly) more volatile than phenol close to the tangent pinch point, and inversing the relative volatility is thus not needed.

Effects of ternary components on pseudo-binary VLE data mixtures of phenol–cumene

The effect of a selected range of ternary components with differ-ent functional groups on the VLE of the cumene–phenol mixture at 300 mbar is shown in Fig. 2. In this figure only the higher cumene mole fraction region of the VLE data is represented, since in this part of the diagram the effects on relative volatility are most important due to the presence of the pinch point. Results at a pressure of 1000 mbar show similar effects, and are shown in the Supporting Information, Fig. S1(a)–(c). For hydroxyacetone the color of the mixture changed significantly, indicating the occurrence of chemical reactions. Therefore, no results are pre-sented for this ternary component. In the vapor phase of the mix-ture with mesityl oxide no phenol was detected, as is also shown in Fig. 2. The presence of dodecane results in inversion of the rel-ative volatility, making phenol marginally more volatile than cumene, and a relative volatility close to 1 is obtained for the applied S/F= 1. All other ternary components increase the volatil-ity of cumene over phenol, as desired. The strongest effects are observed for the ketone group containing molecules mesityl oxide, methyl isobutyl ketone (MIBK), acetophenone and 5-methyl-2-hexanone (5-m-2-one), followed by 2-octanone (see Fig. 2(a)), which is due to their hydrogen bond accepting function-ality. Similar effects observed for MIBK, 5-m-2-one, acetophenone and 2-octanone, for example, are in line with their similar BF3

affinity (75.09 kJ mol−1 for MIBK, 74.84 kJ mol−1 for 5-m-2one, 74.52 kJ mol−1 for acetophenone and 74.47 kJ mol−1 for 2-octanone),18which is an indication of their hydrogen bond for-mation strength.19For hydroxyacetophenone the effect was con-siderably less strong, in line with its worse hydrogen bonding capability compared to acetophenone, as indicated by its lower pKBHX(0.56 for hydroxyacetophenone compared to 1.11 for

acet-ophenone).18Consequently, in ketones of similar size, the pres-ence of other functional groups, such as a double bond, an aromatic group or an alcohol group attached to a linear carbon chain, only seems to have a (small) secondary effect on the VLE of phenol and cumene. Only when the alcohol group is situated on the aromatic ring structure (for hydroxyacetophenone) does this seem to lead to a weaker interaction with phenol. Dimethyl benzyl alcohol, containing an alcohol functionality in the branched organic part of the molecule attached to the aromatic ring, also shows a strong interaction with phenol (see Fig. 2), despite its relatively low expected pKBHX, which is not expected

to exceed 1 as it is expected in the range between 0.86 for benzyl

alcohol and 0.97 for phenethyl alcohol.18 The relatively strong effect of dimethyl benzyl alcohol may be due to the relatively high molar-based S/F ratio for this component (due to its relatively low molecular weight). The effect for 1-octanol seems to be smaller than for benzyl alcohol and seems to be marginally lower than the effect of 2-octanone (see Fig. 2(a) and (c)), indicating a similar but slightly worse effect of an alcohol group compared to a ketone group in a similar linear organic molecule. This effect is fur-ther supported by the relatively small effect of 1-dodecanol on phenol–cumene VLE. Both alcohols have a lower pKBHXvalue than

the linear ketone (1.04 for 1-octanol compared to 1.18–1.21 for 2-octanone, assuming similar pKBHXfor similar ketones).18 The

ethers polyethylene glycol (PEG) and diethylene glycol dibutyl ether (DEGDBE) have a positive effect on the relative volatility as well. This effect is comparable to the molecules containing ketone and alcohol functional groups. For DEGDBE, with a BF3affinity of

78.57 kJ mol−1(based on similar molecules)19slightly higher than that for the ketones, this is somewhat surprising, but the effect may be caused by the relatively high molecular weight of this molecule, resulting in a relatively low molar-based S/F ratio. Another explanation may be that the three ether groups can form a crown structure, resulting in a chelating effect, reducing the hydrogen bonding susceptibility. The difference in the relative volatility effect between DEGDBE and PEG200 cannot be explained by a difference in molar weight. At xCumene= 0.83 for

PEG200 and xCumene = 0.87 for DEGDBE (see Fig. 2), the

mole-based solvent-to-phenol ratios are approximately 3.1 and 4.3 for PEG200 and DEGDBE, respectively, and PEG200 has on average 4.4 ether groups and 2 OH groups per mole, whereas DEGDBE only has 3 ether O atoms and no OH groups. Despite the lower amount of effectively available hydrogen bond accepting groups for the ether compared to the glycol ether, and the lower molar-based solvent-to-phenol ratio, the relative volatility increase caused by the presence of the glycol ether is higher than for the ether. Therefore, it is suggested that the increased effect of PEG200 compared to DEGDBE could be caused by differences in

0.0 0.2 0.4 0.6 0.8 0.6 1.0 0.8 1.0

y

Cu men e

x

Cumene

Figure 5. Effect of combined solvents on binary (closed) and pseudo-binary (open) VLE of cumene–phenol system at 300 mbar for mixtures of solvents at a molar-based S/F ratio between 0.2 and 0.4 (exact number in parentheses): ◁ MIBK + DEGDBE (0.3); 5 MIBK + PEG200 (0.3); ▷ PEG200 + 1-octanol (0.2); ? MIBK + 1-octanol (0.3);  1-octanol + DEGDBE (0.3); 4 PEG200 + DEGDBE (0.2); ♢ glycerol + DEGDBE (0.3); X glycerol + PEG200 (0.4);☐ glycerol + MIBK (0.4).

(7)

intermolecular interactions as a result of the larger polarity of the glycol ether compared to the ether.

Quaternary mixture effects for compounds with opposite effects on binary VLE behavior of phenol–cumene

In industrial processes, several impurities and contaminations can be present in the process streams. Next to impurities that may be expected in traditional phenol processes, renewable phenol is expected to contain a series of additional impurities that may complicate the process. These impurities may also be present in the phenol–cumene separation step, affecting the relative volatil-ity. Dodecane was the single component that reduced and even (marginally) reversed the relative volatility in the phenol–cumene system. To investigate the effect of the presence of such a compo-nent among compocompo-nents that have an opposite (positive) effect on the relative volatility, quaternary mixtures including dodecane were studied. The results for the combined presence are shown as pseudo-binary VLE data in Fig. 3(a)–(d) and expressed as the rela-tive volatility in Fig. 4(a)–(d), together with the results of the effect of dodecane as single impurity in Figs 3(e) and 4(e). For these experiments an S/F ratio of 1 was applied. This implies a molar-based S/F ratio of 0.47–0.60 for PEG200, 0.43–0.55 for DEGDBE, 0.94–1.2 for MIBK and 0.72–0.92 for 1-octanol. The results in Figs 3 and 4 show that the negative effect of the quaternary compound

dodecane on the relative volatility is easily compensated by the presence of a small amount of ternary compound having a posi-tive effect on the relaposi-tive volatility. DEGDBE and PEG200, in partic-ular, have a strong improving effect on the relative volatility, and as a result only a minor effect of the presence of dodecane is observed. It can be concluded that this is a result of the amount of hydrogen bonding groups per molecule relative to the amount of phenol present, as DEGDBE has three ether oxygen atoms, and already at a molar ratio of 0.5 shows a strongly improved relative volatility, even at dodecane-ratio of 1.2 (both ratios relative to phenol + cumene). Similarly, PEG200 has approximately four ether oxygen atoms and one hydroxyl group, and also shows strong improvement of the relative volatility at a molar ratio of 0.3 for the ternary component while at the same time dodecane was present at a ratio of 1 (both ratios relative to phenol + cumene). The differences between DEGDBE and PEG200 in inter-molecular interactions and availability for hydrogen bond forma-tion as a result of their structure and polarity are expected to be enlarged in the presence of dodecane. This most likely explains that DEGDBE shows less effect of the addition of dodecane than PEG200 does.

For application in a process the relatively small effect of dode-cane that could function as an anti-solvent, compared to the larger effect of the hydrogen bonding solvent, shows the 1.0 0.6 0.8 0.4 0.2 0.0 0.6 0.8 1.0 1.0 0.8 0.6 0.4 0.2 0.0 0.6 0.8 1.0 1.0 0.8 0.6 0.4 0.0 0.2 0.6 0.8 1.0 yCum ene (a) xCumene yCum ene (b) xCumene yCume ne xCumene (c)

Figure 6. Effect of S/F ratio (S/F ratio= ☐ 1, ♢ 0.5, 4 0.2 or0.1) on pseudo-binary VLE of cumene–phenol at 300 mbar for (a) MIBK, (b) DEGDBE and (c) PEG200.

(8)

potential of these solvents for extractive distillation processes for separation of cumene and phenol. These effects are particularly strong for hydrogen bond accepting solvents.

Combined effect of multiple added components with similar effects on the pseudo-binary VLE of phenol and cumene

The effect of adding combinations of components with similar effects was studied to investigate possible synergistic effects, i.e. combining two components to reach an effect that is stronger than the sum of the separate effects. The possibility of synergistic effects may be expected as a result of combining functional groups, increasing polarity. Figure 5 shows pseudo-binary VLE data for the combined effect of multiple added components. It is clear that most binary combinations of MIBK, PEG200, 1-octanol and glycerol have a similar positive effect on the relative volatility (see those points in Fig. 6 that have an S/F ratio of 0.2). Only mixtures containing DEGDBE, MIBK + DEGDBE and 1-octanol + DEGDBE, and possibly also glycerol + DEGDBE (although at higher cumene fraction), show a significantly lower effect. The

mixture of MIBK + DEGDBE is the only mixture in which there are no hydroxyl groups present, resulting in the lowest solvent effect, followed by 1-octanol + DEGDBE, which is the mixture with hydroxyl groups present, but at the lowest concentration. On top of that, 1-octanol is also the least polar solvent of the selection. Although competition with the interaction with phenol could be expected, it appears that the presence of hydroxyl groups is not directly negative for the solvent effectiveness. The increased polarity of the mixture by the addition of hydroxyl groups may have a sufficiently improving effect on the solvent effectiveness to increase the relative volatility of cumene over phenol. Next to that, the effect of chain length and functional group density, and therewith London dispersion forces, for example, could play a role. However, although in all cases an improving effect of the addition of solvent is clear, an undoubtedly clear difference between the combined mixtures of solvent and the application of the pure solvents could not be shown in these results and sig-nificant synergistic effects could not be determined. This would require more data in several regions of the composition (pseudo-)binary mixture. 0.0 0.2 0.4 0.6 0.8 5 1.0 10 15 20 25 0.0 0.2 0.4 0.6 0.8 1 1.0 2 3 4 5 6 7 α xCu (a) α (b) xCu 0.0 0.2 0.4 0.6 0.8 5 1.0 10 15 20 25 30 35 α (c) xCu

Figure 7. Effect of S/F ratio (S/F ratio= ☐ 1, ♢ 0.5, 4 0.2 or0.1) on pseudo-binary relative volatility of cumene–phenol at 300 mbar for (a) MIBK, (b) DEGDBE and (c) PEG200.

(9)

Quinary systems were also studied, in which a binary solvent and an anti-solvent were combined with the cumene–phenol mixture. The study of these systems showed similar results on the strong improving effects of solvents consisting of multiple components, e.g. ketones and alcohols, and the minor effect of the addition of the anti-solvent dodecane. Since the results were similar, they were not further interpreted, and are shown in the Supporting Information, Fig. S2.

Whether synergistic effects occur when binary solvents are applied to extractive distillation processes could not be deter-mined; however, related to the application in a process this means that a mixture of solvents is also not underperforming compared to a single solvent. Thus, depending on specific impurities and tar-get solutes present in an industrial process, mixtures of solvent could be used to tune the overall effect of the solvent with respect to specific conditions. Next to the studied contaminants, in indus-trial processes water contamination could also be expected; as its boiling point is well below that of cumene and phenol, only minor effects on the VLE behavior are expected.

Effect of S/F ratio on relative volatility of the cumene– phenol system

For application in a process, the effect of S/F ratio on the relative volatility⊍ is of importance, and thus the effect of S/F ratio was studied for different solvents in combination with the cumene– phenol mixture, as displayed in Fig. 6 as liquid and vapor phase compositions and in Fig. 7 as relative volatility as a function of the cumene fraction in the liquid phase. Although the actually applied S/F ratio has an effect on the pseudo-binary VLE for the three well-performing solvents MIBK, DEGDBE and PEG200 (Figs 6 and 7), they already perform very well at low S/F.

Based on these results solvents such as ketones and alcohols or mixtures of those are promising for application in an ED process, especially when they show an effect at low S/F ratio and effects of other components that might function as an ‘anti-solvent’ (e.g. dodecane) are less significant. However, for application in such an ED process other factors should also be taken into account, such as boiling point, toxicity and availability. Thus, for separation of a mixture of phenol and cumene, components such as MIBK cannot be applied based on its low boiling point, which does not fulfil the criterion for ED that the added solvent boiling temperature needs to be 40–50 °C higher than the high boiling component from the mixture to be separated (phenol in this case).20 Suggested solvents are, for example, DEGDBE and PEG200.

CONCLUSIONS

For recovery of phenol from cumene by distillation, knowledge on VLE behavior is important; therefore, this was studied, including how impurities affect the binary vapor–liquid equilibrium behav-ior of cumene-phenol mixtures. It was found that the addition of polar components with hydrogen bonding abilities, i.e. ketones or ethers, affected the relative volatility of cumene over phenol most positively. Combining two types of components results in similar effects, and clear synergistic effects could not be shown based on current VLE measurements. Addition of an apolar com-ponent in combination with polar comcom-ponents with hydrogen abilities had only a minor effect on the relative volatility, even up to a weight percentage of 25 wt%. The larger effect of the hydrogen bonding solvents, especially for the hydrogen bond

accepting types, compared to other components that could potentially act as anti-solvent, shows the potential of these sol-vents for extractive distillation processes for separation of cumene and phenol. For application in an extractive distillation process for separation of cumene and phenol, application of ketones and alcohols is promising, especially since there was already a significant effect at low S/F ratio, which allows for oper-ation at low S/F ratio.

ACKNOWLEDGEMENTS

This work is a project of the Institute for Sustainable Process Tech-nology (ISPT; TEEI314006/BL-20-07), co-funded by Topsector Energy by the Dutch Ministry of Economic Affairs and Climate Policy.

SUPPORTING INFORMATION

Supporting information may be found in the online version of this article.

REFERENCES

1 Fiege H, Voges HW, Hamamoto T, Umemura S, Iwata T, Miki H et al., Phenol derivatives, in Ullmann's Encyclopedia of Industrial Chemistry. Wiley, Hoboken, NJ (2000).

2 Coker A, Phenol, acetone, cumene, in Nexant's Chemsystems Perp Pro-gram. Nexant, White Plains, NY, pp. 1–6 (2012).

3 Mohammadi S, Kargari A, Sanaeepur H, Abbassian K, Najafi A and Mofarrah E, Phenol removal from industrial wastewaters: a short review. Desalin Water Treat 53:2215–2234 (2015).

4 Burghoff B, Goetheer ELV and de Haan AB, Solvent impregnated resins for the removal of low concentration phenol from water. React Funct Polym 68:1314–1324 (2008).

5 Burghoff B, Zondervan E and de Haan AB, Phenol extraction with Cya-nex 923: kinetics of the solvent impregnated resin application. React Funct Polym 69:264–271 (2009).

6 Bokhove J, Schuur B and de Haan AB, Equilibrium study on the reactive liquid–liquid extraction of 4-cyanopyridine with 4-nonylphenol. Chem Eng Sci 82:215–222 (2012).

7 Kujawski W, Warszawski A, Ratajczak W, Porębski T, Capała W and Ostrowska I, Removal of phenol from wastewater by different sepa-ration techniques. Desalination 163:287–296 (2004).

8 Murtha TP, Extractive distillation separation of cumene and phenol. US Patent, US4166772A (1979).

9 Villegas LGC, Mashhadi N, Chen M, Mukherjee D, Taylor KE and Biswas N, A short review of techniques for phenol removal from wastewater. Curr Pollut Rep 2:157–167 (2016).

10 Hussain A, Dubey SK and Kumar V, Kinetic study for aerobic treatment of phenolic wastewater. Water Resour Ind 11:81–90 (2015). 11 Bokhove J, Schuur B and de Haan AB, Solvent design for trace removal

of pyridines from aqueous streams using solvent impregnated resins. Sep Purif Technol 98:410–418 (2012).

12 Wang S, Shi D, Yang R, Xu Y, Guo H and Yang X, Solvent extraction of phenol from aqueous solution with benzyl 2-ethylhexyl sulfoxide as a novel extractant. Can J Chem Eng 93:1787–1792 (2015). 13 Liu J, Xie J, Ren Z and Zhang W, Solvent extraction of phenol with

cumene from wastewater. Desalin Water Treat 51:3826–3831 (2013). 14 Saien J and Razi Asrami M, Liquid–liquid equilibrium of the ternary sys-tem of water + phenol + (propan-2-yl) benzene at several sys- tempera-tures. J Chem Eng Data 62:3663–3670 (2017).

15 Shah Z, Renato CV, Marco AC and Rosangela DS, Separation of phenol from bio-oil produced from pyrolysis of agricultural wastes. Mod Chem Appl 5:1000199/1000191–1000199/1000198 (2017). 16 Cepeda E, Gonzalez C and Resa JM, Isobaric vapor–liquid equilibrium

for the cumene–phenol system. J Chem Eng Data 34:270–273 (1989). 17 Blahušiak M, Kiss AA, Babic K, Kersten SRA, Bargeman G and Schuur B, Insights into the selection and design offluid separation processes.

(10)

18 Laurence C and Gal JF, Lewis basicity and affinity measurement: defini-tions and context, in Lewis Basicity and Affinity Scales. Wiley, Hobo-ken, NJ (2009).

19 Jongmans MTG, Londoño A, Mamilla SB, Pragt HJ, Aaldering KTJ, Bargeman G et al., Extractant screening for the separation of

dichloroacetic acid from monochloroacetic acid by extractive distil-lation. Sep Purif Technol 98:206–215 (2012).

20 Lei Z, Chen B and Ding Z, Extractive distillation, in Special Distillation Processes, ed. by Lei Z, Chen B and Ding Z. Elsevier Science, Amsterdam, pp. 59–144 (2005).

Referenties

GERELATEERDE DOCUMENTEN

In de projectieve meetkunde gaat het alleen om incidentierelaties, zoals het op één rechte lijn liggen van drie of meer punten, en het door één punt gaan van drie of meer

Ook in Dilsen vinden we dergelijke voorbeelden (fig. Uit dit kort overzicht mogen we besluiten dat het hier besproken ensemble typologisch het best kan omschreven worden als

To explore cellular changes related to brain maturation in children we describe the longitu- dinal trajectories of absolute metabolite levels in gray matter, white matter and BG

termined as a function of several parameters (viz. angular velocity of disc electrode, potential scan rate, concentration of substrate, composition of supporting

The presence of significant positive relationships between genetic and geographic distance in species where no such relationship was evident for mtDNA data was particularly clear

Om tot een meer flexibele conceptualisering van het topsysteem te komen (betreffende waterafvoer en -inlaat, ontwatering, verdamping en grondwateraanvullling,

For a granule impacted on a liquid film on the wall, the increase of liquid viscosity decreases the restitution coefficient and thickness of liquid layer at

30 dependent variable intention to enroll and the mediator variable attitude as well as the extent of knowledge about Brexit, favorite country, visited the UK, and studied in the