Accepted Manuscript

Coupled phase-reaction equilibrium for dihydromyrcene hydration system

Jingjing Chen, Jinliang Chen, Jun Liu, Suying Zhao, Huidong Zheng, Yao Gu

PII: S0378-3812(16)30554-4
DOI: 10.1016/j.fluid.2016.11.007
Reference: FLUID 11317

To appear in: Fluid Phase Equilibria

Received Date: 12 January 2016

Revised Date: 22 October 2016
Accepted Date: 3 November 2016

Please cite this article as: J. Chen, J. Chen, J. Liu, S. Zhao, H. Zheng, Y. Gu, Coupled phase-
reaction equilibrium for dihydromyrcene hydration system, Fluid Phase Equilibria (2016), doi: 10.1016/
j.fluid.2016.11.007.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Coupled phase-reaction equilibrium for

dihydromyrcene hydration system
Jingjing Chen, Jinliang Chen, Jun Liu, Suying Zhao, Huidong Zheng*, Yao Gu

PT
School of Chemical Engineering, Fuzhou University, Fuzhou, Fujian 350116, P. R. China
2 Xue Yuan Road, University Town, Fuzhou, Fujian 350116, P. R. CHINA

RI
Abstract
The solvent usually forms a large part of the heterogeneous hydration mixture in the

SC
biphasic hydration of liquid olefins, and solvent selection greatly affects the hydration
conversion. In this work, the hydration of dihydromyrcene (DHM) in two different solvents,
i.e., acetone and 1,4-dioxane, was investigated experimentally and mathematically. A

U
theoretical model coupling liquid–liquid phase equilibrium and reaction equilibrium is
proposed for evaluating the effects of the solvent on the heterogeneous hydration of DHM.
AN
Experiments were performed to obtain unreported parameters for the phase equilibrium and
reaction equilibrium and the conversion of DHM in biphasic hydration was then predicted by
the coupled model. The effects of the mass feed ratios on the total conversion X of DHM are
M

discussed and varying conversions were observed when different solvents were employed.
DHM was more soluble in the acetone system than in 1,4-dioxane, and the conversion of the
D

hydration reaction was higher. This predictive evaluation process based on the overall
equilibrium could provide a practical strategy for selecting the optimal solvent for the
TE

heterogeneous hydration of DHM and other liquid olefins.

Keyword:
Hydration; Liquid-liquid equilibrium; Reaction equilibrium; Solvent effect; Dihydromyrcene
EP

1. Introduction
The reversible hydration of liquid olefins with water is a typical liquid-liquid
C

heterogeneous reaction and yields many valuable alcohol products and intermediates, such as
AC

cyclohexanol and terpineol. However, the interphase mass transfer problems in this biphasic
system and the extremely slow reaction rates, both caused by poor miscibility between olefins
and water[1,2], seriously hinder industrial applications of these reactions.
The use of solvents can increase miscibility of the reactants and homogenize the reaction
mixture, thereby accelerate hydration and shorten the reaction time. However, great amounts

*
Corresponding author. E-mail: youngman@fzu.edu.cn.
 ACCEPTED MANUSCRIPT

of solvents are required, as well as large treatment capacity, to obtain a homogeneous

hydration mixture, which is commercially unfavorable[3-7]. Partly for this reason, many
modern-day industrial hydration processes are still performed in two liquid phases and suffer
from low reaction rates, and the solvents in these processes can still account for up to 60 % of
the reaction mixture[8-12].
However, biphasic reactions can also offer advantages. Similar to the olefins themselves,
their hydration products, i.e., the corresponding alcohols from liquid olefins, are usually

PT
immiscible with water[1,2,13-15]. During a biphasic reaction, these alcohols are synthesized in
the aqueous phase and extracted into the organic phase. The continuous removal of products
promotes forward hydration, which may increase the conversion by biphasic hydration.

RI
Evaluation of the advantages and disadvantages of monophasic and biphasic hydration can
help identify appropriate methods for increasing the hydration of liquid olefins.

SC
Most of studied on the hydrations of olefins were carried out in monophasic
systems[16-22], while biphasic reactions were seldom involved. It is worth investigating the
overall equilibrium of a heterogeneous hydration system in different solvents, which is a
combination of the reaction equilibrium and phase equilibrium, and the understanding of the

U
solvent effects on the overall equilibrium could provide a screening method for solvent
AN
optimization for a specific heterogeneous hydration.
This work focused on the direct hydration of dihydromyrcene (DHM) as a representative
biphasic hydration. In the reaction, one of the double bonds in DHM is hydrated and the
M

produced dihydromyrcenol (DHMOL) is one of the most popular perfumes worldwide.

D

+H+ +H2O
-H+
TE

OH
a a' b
EP

This hydration reaction has mostly been investigated in the monophase[23-25] and low-boiling
acetone has been suggested as a potential solvent as its use would help to reduce energy
consumption[26].
C

In this study, the overall hydration conversion of DHM in the equilibrated system, in
which phase equilibrium between the aqueous and organic phases and reaction equilibrium in
AC

the aqueous phase are both reached, was investigated. The coupled equilibrium was evaluated
using the NRTL model and the reaction equilibrium model. The predicted results in two
different solvents, i.e., acetone and the most commonly used solvent 1,4-dioxane, were
compared and the effects of the solvents on the solubility and activity coefficients were
discussed. Experiments on the liquid-liquid equilibrium (LLE) of the water + acetone + DHM
system and the reaction equilibrium of DHM hydration in acetone were performed to obtain
previously unreported parameters of the NRTL model and reaction equilibrium.
 ACCEPTED MANUSCRIPT

2. Phase-reaction equilibrium model

The industrial hydration of DHM is performed in a multiphase system that consists of
two liquid phases and a solid resin catalyst phase. Hydrophilic resin catalysts tend to swell in
water or other polar solvents[27] and become covered with a thin water layer during the
hydration. The reaction can then be considered to take place only in the aqueous phase,

PT
therefore only liquid phases in the reaction are counted in this work, i.e., “monophasic” or
“homogeneous” indicates one liquid phase, and “biphasic” or “heterogeneous” indicates two
liquid phases in this paper.

RI
During the reaction, the DHM in the organic phase transfers into the aqueous phase,
reacts with adsorbed water, and produces DHMOL; the DHMOL is continuously extracted by
the organic phase from water (Figure 1). The hydration reaction in water is therefore pushed

SC
forward until both phase equilibrium between the two liquid phases and reaction equilibrium
in the aqueous phase, which is the phase-reaction equilibrium, are achieved.

U
Organic Phase AN
solvent DHM water DHMOL
M

solvent DHM + water DHMOL

D

Aqueous Phase
TE

Figure 1. Mass transfer and reaction in heterogeneous mixture.

EP

When the phase-reaction equilibrium is reached, the following equations must be

satisfied:

Phase equilibrium xα ,iγ α ,i = xβ ,iγ β ,i , i = 1, 2,3, 4 1

C
AC

aβ ,4
Reaction equilibrium Ka = 2
aβ ,1 ⋅ aβ ,3
eq

where x is the mole fraction; γ is the activity coefficient; Ka is the reaction equilibrium
constant; subscript i (= 1, 2, 3, 4) represents the components water, solvent, DHM, and
DHMOL, respectively; and subscripts α and β represent the organic phase and aqueous phase,
respectively. The equilibrium concentrations of each component in the organic and aqueous
phases can be predicted by solving the two equations.
The LLE and reaction equilibrium data are needed to predict the phase-reaction
 ACCEPTED MANUSCRIPT

equilibrium of the DHM hydration system. The LLE data at 343.2-358.2 K for the water +
1,4-dioxane + DHM + DHMOL system have been reported[7,14,28], as have the data for the
reaction equilibrium of DHM hydration with 1,4-dioxane as the solvent[29]. The LLE data at
358-368 K for the water + acetone + DHMOL system have also been reported in our previous
work[13].
In this study, LLE experiments for water + acetone + DHM and the hydration of DHM
in acetone were performed to obtain the binary interaction parameters for the acetone + DHM

PT
system and the hydration equilibrium data for DHM hydration in acetone. These reported and
experimental data can be used to calculate the phase–reaction equilibria of the DHM
hydration in two different solvents, i.e., acetone and 1,4-dioxane.

RI
3. Experimental

SC
3.1. Materials

In this work, technical grade DHM (purity ≥ 89 wt%, CAS no. 2436-90-0, Huaian

U
Wanbang Aromatic Chemicals Industry Co., Ltd.) was used in experimental measurements so
AN
as to more closely reproduce an actual industrial reaction. The main impurities in the DHM
were isomers of DHM produced in the isomerization of α-pinane, the main method for
producing DHM. The other chemicals listed in Table 1 were used without further purification.
M

The hydration catalyst NKC-9 was purchased from the Catalyst Plant of Nankai University;
the particle size was 0.40-1.25 mm. Distilled water was used in all experiments.
D
TE

OH
EP

DHM Impurities in the DHM material DHMOL

C

Table 1. Chemicals and suppliers.

AC

Chemical CAS no. Purity (mass%) Supplier

Acetone 67-64-1 Analytical grade (>99.5) Sinopharm Chemical Reagent Co., Ltd
Cyclohexanone 108-94-1 Analytical grade (>99.5) Sinopharm Chemical Reagent Co., Ltd
DHMOL 18479-58-8 Technical grade (>99.5) Huaian Wanbang Aromatic Chemicals
Industry Co.,Ltd.
Ethanol 64-17-5 Analytical grade (>99.7) Sinopharm Chemical Reagent Co., Ltd
n-Propanol 71-23-8 Analytical grade (>99.8) Sinopharm Chemical Reagent Co., Ltd
 ACCEPTED MANUSCRIPT

3.2. LLE experiments

The apparatus for the LLE measurements, which was described in detail in our previous
work[13], consisted of a steel cell of capacity 50 mL equipped with a heating jacket and a
temperature control system. Prior to each experimental run, a mixture of water, acetone, and
DHM was added to the cell and the cell was sealed. The system was stirred for 8 h to allow
for sufficient contact between the two liquid phases and settled for 12 h to ensure complete

PT
phase separation. Duplicate samples for quantitative analysis were removed using syringes.
During the experiment, nitrogen was charged to keep the cell pressure at 0.8 MPa to avoid
boiling of the mixture; the system temperature was controlled by circulating water from a

RI
water bath and measured using a thermocouple with a precision of 0.1 K.

3.3. DHM hydration experiments

SC
The hydration experiments were performed in a 200 mL stainless-steel batch reactor
(Figure 2). The reactor was equipped with a mechanical agitator and a temperature control

U
system. A heating jacket was provided to keep the reaction temperature constant. The reactor
was charged with nitrogen to keep the pressure at 0.8 MPa. Known amounts of reactants
AN
(DHM and water), solvent (acetone), and ion–exchange resin catalyst (NKC-9) were added to
the reactor. Agitation was started when both the pressure and temperature reached the set
points. Samples were taken at regular intervals for analysis.
M

5
D

1 6
4
TE

2 7
8
9
3
C EP
AC

Figure 2. Reaction apparatus. 1 - nitrogen inlet, 2 - pressure gauge, 3 - heating jacket, 4 -

nitrogen outlet, 5 - stirrer, 6 - thermocouple, 7 - sample valve, 8 - reactor, and 9 - temperature
controller.

3.4. Sample analysis

The organic components in the LLE samples and hydration samples were analyzed using
a GC system (Varian 3900) equipped with a flame ionization detector and a capillary column
(AT OV-101). The column temperature program was as follows: 333 K hold 1 min; heat at 10
 ACCEPTED MANUSCRIPT

K/min for 10 min; and 433 K hold 2 min. Nitrogen was used as the carrier gas. The internal
standards were n-propanol for acetone and cyclohexanone for DHM and DHMOL.
The water contents of the LLE samples were determined using a GC system (GC1690,
Kexiao Chemical Equipment Co., Ltd.) equipped with a thermal conductivity detector and
two packed columns (Porapak Q). The column temperature was kept constant at 413 K and
the current was 150 mA. Hydrogen was used as the carrier gas. Ethanol was used as the
internal standard for the quantitative analysis of water.

PT
Each LLE sample was analyzed three times to ensure that the error of the measured
mass fraction was less than 3% for DHM in the aqueous phase and 0.5% in all other cases.
All measured mass fractions were converted into mole fractions for the model correlation.

RI
The uncertainty for DHM in the aqueous phase, as a mole fraction, was 0.0004 (absolute
value) when the mole fraction was less than 0.005, and 5% when the mole fraction was

SC
greater than 0.005; the uncertainty for acetone in each phase was 3% when the mole fraction
was less than 0.05 and 1% when the mole fraction was greater than 0.05; the uncertainties for
the mole fractions of water in each phase, and DHM in the organic phase, were 1%.

4. Results and discussion

U
AN
4.1. LLE of water + acetone + DHM
M

LLE experiments on the ternary system water (1) + acetone (2) + DHM (3) were
performed at 358, 363, and 368 K at 0.8 MPa. No significant amounts of DHMOL were
detected in the LLE samples, indicating that DHM hydration barely occurred in the system
D

without catalysts under the experimental conditions.

The experimental data are presented in Table 2. The tie lines and calculated binodal
TE

curves of the system are shown in Figure 3. All compositions are expressed as mole fractions.
The experimental data show that the ternary system behaves as a type-I system, which
has only one immiscible pair (water + DHM) and large biphasic area are found in the phase
EP

diagrams. The addition of acetone clearly increases the miscibility of the olefin and water.
The system temperature, however, has little effect on the miscibility of the ternary system
within the experimental temperature range.
C

The reliability of the measured data was ascertained using the Othmer-Tobias
correlation[30] (equation 3) and the Hand correlation[31] (equation 4) at all three temperatures
AC

as follows:
1 − xα’ ,3 1 − x’β ,1
log = A1 ⋅ log + B1 3
xα’ ,3 x’β ,1

xα’ ,2 xβ’ ,2
log = A 2 ⋅ log + B2 4
xα’ ,3 xβ’ ,1
 ACCEPTED MANUSCRIPT

0.00 0.00
1.00 1.00

0.25 0.25
0.75 0.75
e

e
en

en
yrc

yrc
Ac

Ac
rom

rom
eto

eto
0.50 0.50
0.50 0.50
yd

yd
n

ne
e
D ih

D ih

PT
0.75 0.75
0.25 0.25

1.00 1.00
0.00 0.00

RI
0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00
Water Water

(a) (b)

SC
0.00
1.00

0.25
0.75

U
e
en
yrc

Ac
rom

0.50 eto
AN
0.50
yd

en
D ih

0.75
0.25
M

1.00
0.00
0.00 0.25 0.50 0.75 1.00
Water
D

(c)
exp. cal.
TE

Figure 3. Experimental tie lines and calculated binodal curves for ternary system water +
acetone + DHM at 358 K (a), 363 K (b), and 368 K (c) at 0.8 MPa.
EP

where x' is the mass fraction; A1 and B1 are the Othmer-Tobias correlation constants; and A2
and B2 are the Hand correlation constants. Both correlations give good linearities (Figure 4),
C

and high correlation coefficients (R2 ≥ 0.99) are obtained (Table 3), which suggests that the
measured data in this work are reliable.
AC

The tie lines were correlated using the NRTL model[32] and the values of the non-
randomness parameter αij were set at 0.3 for the miscible pairs (water + acetone and acetone
+ DHM) and 0.2 for the immiscible pair (water + DHM).
Details of the method used for estimating the NRTL binary interaction parameters were
given in our previous work[13]. The parameters for the NRTL model were determined using
the simplex optimization method. The rule of thermodynamic equilibrium, which is the
isoactivity criterion, was used to deal with multi-roots problem in correlating process[33] and
the objective function for the correlation is defined as
 ACCEPTED MANUSCRIPT

Table 2. Experimental data for ternary system water (1) + acetone (2) + DHM (3) at 358,
363, and 368 K at 0.8 MPa a,b.

Organic phase Aqueous phase

T/K
xα,1 xα,2 xα,3 xβ,1 xβ,2 xβ,3
358 0.0235 0.0627 0.9139 0.9832 0.0159 0.0009
0.0515 0.2082 0.7403 0.9486 0.0504 0.0011

PT
0.0767 0.3046 0.6187 0.9191 0.0785 0.0025
0.0971 0.3690 0.5339 0.8906 0.1048 0.0046
0.1344 0.4492 0.4164 0.8473 0.1472 0.0055

RI
0.2069 0.5027 0.2905 0.7910 0.2029 0.0061
0.2438 0.5222 0.2340 0.7608 0.2317 0.0076

SC
0.2815 0.5452 0.1733 0.7086 0.2811 0.0103

363 0.0406 0.1409 0.8185 0.9662 0.0319 0.0018

0.0487 0.2288 0.7225 0.9440 0.0536 0.0024

U
0.0750 0.2733 0.6517 0.9141 0.0832 0.0027
0.0769 0.3523 0.5709 0.8787 0.1176 0.0037
AN
0.1137 0.4324 0.4539 0.8179 0.1775 0.0046
0.1723 0.4785 0.3492 0.7477 0.2467 0.0056
0.2149 0.5196 0.2655 0.6939 0.2989 0.0072
M

0.2532 0.5476 0.1992 0.6114 0.3676 0.0210

368 0.0482 0.0650 0.8868 0.9818 0.0165 0.0017

D

0.0592 0.1390 0.8018 0.9631 0.0348 0.0021

0.0664 0.2012 0.7324 0.9494 0.0482 0.0024
TE

0.0723 0.2839 0.6438 0.9128 0.0839 0.0033

0.0898 0.3572 0.5530 0.8673 0.1285 0.0042
0.1271 0.4282 0.4448 0.8252 0.1680 0.0069
EP

0.1833 0.4723 0.3445 0.7481 0.2433 0.0086

0.2950 0.5326 0.1724 0.6079 0.3671 0.0250
a
x: mole fraction; α, β: organic phase and aqueous phase, respectively.
C

b
The uncertainty is 0.1 K for the temperature, 0.01 MPa for the pressure. The
uncertainty for x is 5% for the DHM in the aqueous phase and 1 % for the others.
AC

2
N 3 xα ,ik γ α ,ik 
OF1 = ∑∑  − 1 / 3 N 5

j =1 i =1  xβ , ik γ β , ik



With the determined model parameters, predictions for the LLE were performed using
the procedure proposed by Zeng[34] and the objective function is defined as
 ACCEPTED MANUSCRIPT

0.4
1.0
0.2
0.5
0.0
0.0
-0.2
-0.5
log((1-x'α,3)/x'α,3)

-0.4

log(x'α,2/x'α,3)
-0.6 -1.0

-0.8 -1.5
-1.0 -2.0
-1.2

PT
-2.5
358 K 358 K
-1.4 363 K
363 K -3.0
-1.6 368 K 368 K
-3.5
-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5
log(x'β,2/x'β,1)

RI
log((1-x'β,1)/x'β,1)

(a) (b)

SC
Figure 4. Othmer-Tobias correlation (a) and Hand correlation (b) for measured data.

Table 3. Othmer-Tobias and Hand correlations parameters.

Othmer-Tobias correlation
U Hand correlation
AN
T/K 2
A1 B1 R A2 B2 R2
358 1.2098 -0.0145 0.9984 0.8449 0.0514 0.9994
363 0.9612 -0.2203 0.9970 1.0701 0.5737 0.9957
M

368 1.0120 -0.2082 0.9956 0.9935 0.4783 0.9949

D

2 3
OF2 = ∑∑ ( xijexp − xijcal ) 2 6
TE

j =1 i =1

The root-mean-square deviation (RMSD) was calculated as follows to show the coincidence
EP

between the measured data and the predicted values.

2
 xijk
exp
− xijk
cal

C

N 2 3
RMSD = ∑ ∑∑   / 6 N 7
 x exp
k =1 j =1 i =1  
AC

ijk

The obtained binary interaction parameters for the NRTL model are listed in Table 4.
and the calculated binodal curves are shown in Figure 3. The calculated results show that the
NRTL model can precisely predict the data for the liquid-liquid phase equilibrium of the
water + acetone + DHM system.
 ACCEPTED MANUSCRIPT

Table 4. NRTL binary interaction parameters for water (1) + acetone (2) + DHM (3), and
RMSD values.

i-j ∆gij / J·mol-1 ∆gji / J·mol-1 RMSD

1-2 6679.16 259.39 0.0175
1-3 16368.68 5766.43
2-3 4229.40 -156.89

PT
4.2. DHM hydration reaction equilibrium

RI
DHM hydration was performed homogeneously to obtain the reaction equilibrium data.

SC
Experiments were performed at four different temperatures, ranging from 358 to 373 K. The
obtained equilibrium concentrations of DHM and DHMOL, and the DHM conversions, at
each temperature are shown in Figure 5. As the temperature rises, the concentration of DHM
increases, whereas that of DHMOL decreases; this suggests that hydration of DHM is

U
exothermic and reversible. The amount of water in the reaction mixture was much greater
AN
than that of DHM in the experiments and the water concentration was almost constant (x1 =
0.653) during the reaction.
M

5.5 0.55

5.0
D

0.50
3

TE
x×10

4.5
0.45
X

4.0
0.40
EP

3.5 DHM
DHMOL 0.35
X
C

3.0
355 360 365 370 375
AC

T/K

Figure 5. Equilibrium concentrations of DHM and DHMOL, and DHM conversion (X) at
different temperatures (agitation speed: 600 rpm, catalyst loading: 15 wt%,
water:acetone:DHM mass ratios: 9:15:1, pressure: 0.8 MPa).
 ACCEPTED MANUSCRIPT

The reaction equilibrium constant Ka can be calculated from the activity of each
component involved in the reaction as

a4
Ka = = K x ⋅ Kγ 8
a1 ⋅ a3 eq

x4 γ4

PT
where K x = and Kγ = . The reaction constants Kx at each temperature,
x1 ⋅ x3 eq
γ1 ⋅ γ 3 eq

based on the mole fraction, were calculated from the experimental data. The activity

RI
coefficients were calculated using the obtained NRTL model parameters (Table 4.). The
calculated constants for each temperature are listed in Table 5.

SC
Table 5. Reaction equilibrium constants for hydration.

U
T/K Kx Kγ Ka
358 1.262 0.223 0.282
AN
363 1.139 0.233 0.265
368 1.051 0.242 0.255
373 0.923 0.252 0.233
M

As the reaction temperature rises from 358 to 373 K, the concentration constant Kx
decreases, whereas the activity coefficient constant Kγ increases slowly. The equilibrium
D

constant Ka decreases slightly at higher temperatures. This suggests that DHM hydration is
exothermic and a lower temperature favors the production of DHMOL.
TE

-1.20
EP

-1.25

-1.30
C ln Ka

-1.35
AC

-1.40

-1.45

-1.50
0.00265 0.00270 0.00275 0.00280
1/(T/K)

Figure 6. Reaction equilibrium constant Ka as function of temperature.

 ACCEPTED MANUSCRIPT

According to the van’t Hoff equation (equation ), there is a linear relationship between
the logarithm of Ka and the reciprocal of the temperature (T), as shown in Figure 6.

∆rGΘ ∆ H Θ ∆r S Θ
ln K a = − =− r + 9
RT RT R

The linear function correlates to

PT
1642.5
ln K a = − 5.848 10
T

RI
The calculated enthalpy ∆rHΘ and entropy ∆rSΘ of the reaction are -13.66 kJ·mol-1 and 48.63
J·mol-1·K-1, respectively.

SC
4.3. Coupled phase-reaction equilibrium

In a phase-reaction equilibrated hydration mixture with an initial overall feed amount of

U
100 g [water (1) + solvent (2) + DHM (3); F0 is the corresponding initial molar feed amount
AN
and zi0 is the initial composition], equations 1 and 2 should be satisfied, and the conversion

X of DHM in the phase-reaction equilibrated mixture is defined as

M

Fz3
X = 1− 11
F 0 z30
D

where F is the equilibrated total amount and zi is the composition of the equilibrated mixture.
TE

The amount of each composition can be calculated based on mass conservation:

F = F 0 − F 0 z30 ⋅ X 12
EP

Fzi = F 0 zi0 +ν ⋅ F 0 z30 ⋅ X , i = 1, 2,3, 4 13

C

where ν is the stoichiometric number (= -1 for water and DHM, 1 for DHMOL, and 0 for
solvent). In a biphasic system, equation 13 is equivalent to
AC

Fα xα ,i + Fβ xβ ,i = F 0 zi0 +ν ⋅ F 0 z30 ⋅ X , i = 1, 2,3, 4 14

where Fα and Fβ are the amounts of organic and aqueous phases, respectively.
The coupled equilibria and conversions X for different mass feed ratios were calculated
in acetone and in 1,4-dioxane, using the reported and experimental data. A simple method
was used to determine X for DHM in the biphasic phase-reaction equilibrated mixture. For a
given initial value of X, the equilibrated total amount F and composition zi of the equilibrated
mixture can be calculated by equations 12 and 13. The equilibrium compositions xα,i and xβ,i,
 ACCEPTED MANUSCRIPT

and the amounts of the organic and aqueous phases Fα and Fβ, were calculated using the
procedure proposed by Zeng[34]. The reaction equilibrium in the aqueous phase (equation 2)
was calculated as a convergence criterion.
Qa is defined as the quotient of the activity of the product (DHMOL) to that of the
reactants (water and DHM) in a reaction mixture with any component ratio:

a4
Qa = = Qx ⋅ Qγ

PT
15
a1 ⋅ a3

γ4

RI
x4
where Qx = and Qγ = . Obviously, when the reaction mixture is equilibrated,
x1 ⋅ x3 γ1 ⋅ γ 3

SC
the Qa in the aqueous phase should be equal to Ka:

Qa ,eq = K a 16

U
The reaction will move forward when Qa < Ka, whereas the reverse reaction takes place when
AN
Qa > Ka. The Qa of the aqueous phase was therefore calculated and the initial value of X was
changed based on a comparison of Qa and Ka.
For monophasic mixtures, the initial feed concentration is in the monophasic area of the
M

LLE diagram for the water-solvent-DHM system and Qa was calculated based on zi. For some
initial ratios, Fα or Fβ became zero during the cyclic iteration of increasing X values, which
means that the biphasic mixture became homogenized with increasing production of
D

DHMOL. Under these conditions, Qa was calculated based on zi as well.

All the predictions were at 358 K, at which LLE data and reaction equilibrium constants
TE

are available for both solvents. The effects of the mass feed ratios of water, solvent, and
DHM on X were evaluated.
EP

4.3.1. Effect of solvent/reactant ratio

The effect of the solvent/reactant mass ratio Rs [= solvent/(water + DHM)] on the

conversion of DHM was estimated for a DHM/water mass ratio Rr = 0.5 (Figure 7); different
C

changes in X were observed in acetone (S1) and 1,4-dioxane (S2). Three different types of
AC

phase behavior were predicted for the reaction system: (1) monophasic throughout (T1), (2)
biphasic during feeding and homogeneous during reaction (T2), and biphasic throughout (T3).
For each point in T3, a fictitious conversion in a pseudo-homogeneous system (T4), which is
a hypothetical homogeneous hydration system with the same temperature and feed ratio, was
calculated to show the effect of heterogeneity.
Figure 7(a) shows that for the acetone system (squares), the conversion X in the biphasic
reaction system increases with increasing solvent concentration. The reaction system
becomes homogeneous when Rs ≥ 1.33. In this case, X in the monophasic reaction system
reaches an almost constant value. The first few biphasic points are higher than the
 ACCEPTED MANUSCRIPT

pseudo-homogeneous values (blue solid line), indicating that the heterogeneous reaction has a
positive effect in these feed ratios, whereas the rest are somewhat lower. At a low solvent
feed, the biphasic reaction is thermodynamically favored because of product extraction in the
organic phase; at a higher solvent feed, the increase in the solubility of DHMOL in the
aqueous phase weakens the extraction and inhibits the total conversion in water. As a result, X
becomes lower than in the pseudo-homogeneous system.
0.00

PT
1.00 S1T1
S1T3

0.25
0.75

L
MO

RI
Ac
DH

eto
0.50

M/
0.50

n
e
DH
0.6
S1T1 S2T1
S1T3 S2T2 0.75

SC
0.25
0.5 S1T4 S2T3
S2T4
1.00
0.4 0.00
0.00 0.25 0.50 0.75 1.00
Water

U
0.3
X

(b)a
AN
0.2
0.00
1.00 S2T1
S2T2
0.1 S2T3
0.25
0.75
L
M
MO

1,4
0.0
DH

-D
0 1 2 3 4 5 6

iox
M/

0.50
0.50

ane
DH

Rs

(a)
D

0.75
0.25
TE

1.00
0.00
0.00 0.25 0.50 0.75 1.00
Water

(c)a
EP

a
: Binodal curves: black solid line for water + solvent + DHM; black dot line for water + solvent + DHMOL.

Figure 7. Predicted total conversions at different solvent/reactant ratios (Rs). S1: acetone; S2:
C

1,4-dioxane; T1: monophasic throughout; T2: biphasic during feeding and homogeneous
during reaction; T3: biphasic throughout; and T4: pseudo-homogeneous.
AC

For 1,4-dioxane [triangles in Figure 7(a)], the biphasic ratio range is wider (Rs ≤ 4) than
that for the acetone system because of the poorer solubility. The total conversion X in the
biphasic system increases steadily, whereas that in the monophase decreases at higher solvent
concentrations (Rs > 4). The pseudo-homogeneous values (blue dotted line) are much higher
than those for the biphasic system, except the first three points. This indicates that the
biphasic reaction in 1,4-dioxane is inhibited by the heterogeneity of the system at most times,
which may result from the higher solubility of DHMOL than that in the acetone.
Some T2 points appear in the 1,4-dioxane system that are not found in the acetone
 ACCEPTED MANUSCRIPT

system. The initial points for T2 fall in the wide gap between the ternary binodal curves of
the water-solvent-DHM and water-solvent-DHMOL systems [Figure 7(c)] and are close to
the former. During the reaction, the olefin in the system is gradually replaced by the much
more water-soluble alcohol and the mixture becomes homogeneous when a certain
conversion is achieved. In the acetone system, the gap between the two curves is narrow,
which means that the solubilities of DHM and DHMOL are similar, therefore type T2 is not
observed in this case.

PT
Similar conversion variations have been reported in the biphasic hydration of DHM in
isopropanol as the solvent[12]: the equilibrium conversion increases from 29 % to 35 % with
increasing isopropanol/reactant ratio Rs from 0.33 to 1 for Rr = 2; from 7 % to 38 % with

RI
increasing Rs from 0 to 2 for Rr = 1. The DHM conversion achieved with isopropanol is
between those achieved with acetone and 1,4-dioxane.

SC
The ratio of solvent to reactants mainly affects the biphasic hydration system on the
concentrations and, as a consequence, the activity coefficients of water, DHM, and DHMOL.
Further discussion can be made by looking over the variations of compositions and activity
coefficients.

U
The addition of extra solvent into a phase-reaction equilibrated system, which is
AN
equivalent to the increase of the initial solvent/reactant ratio Rs, will result in new phase
equilibrium between the two liquid phases and disruption of the original reaction equilibrium
in the aqueous phase. The Qa (= Ka) in the original aqueous phase changes and becomes Qa'
M

(≠ Ka) as the new phase equilibrium is established. The effects of solvent concentration on the
conversion X can be evaluated based on the change in Qa of the system.
The addition of extra solvent to the equilibrated hydration system leads to an increase in
D

the solubility of DHM in water and a larger increase with the more water-miscible DHMOL,
and the concentration of water decreases as a result of dilution, resulting in an increase in Qx
TE

(Figure 8). The activity coefficients of these components, on the other hand, show contrary
variations (Figure 9): the activity coefficient of water decreases slightly with increasing Rs,
whereas those of DHM and DHMOL decline significantly. The activity coefficient of
EP

DHMOL drops faster than that of DHM, resulting in a decrease in Qγ.

6 1.5
C

3 25
5

20
4
AC

1.0
Qa×10
3

Qa×10

2
Qx×10
Qx×10

3 15

10
3

2
5

0.5
1
1 5

0 0 0.0 0
0.0 0.5 1.0 1.5 0 1 2 3 4 5
Rs Rs
(a) (b)

Figure 8. Qx and Qa in aqueous phase with acetone (a) and 1,4-dioxane (b) as solvent.
 ACCEPTED MANUSCRIPT

10 10
γ1 γ1
3 3
10 γ3 10 γ3
γ4 γ4
2 2
10 Qγ 10 Qγ
1

Qγ
1 1

Qγ
γι

10

γi
1
10
0
0 10
10 0.1

PT
-1
10
-1
10 0.1
0.0 0.5 1.0 1.5 0 1 2 3 4 5
Rs Rs

RI
(a) (b)

SC
Figure 9. Activity coefficients of water (1), DHM (3), and DHMOL (4), and Qγ in aqueous
phase with acetone (a) and 1,4-dioxane (b) as solvent.

U
The competition between the increasing Qx and decreasing Qγ leads to an overall decline
in Qa (Figure 8), which becomes smaller than Ka. As a consequence, the reaction is driven
AN
forward and a higher conversion is expected.

4.3.2. Effect of reactant ratio

M

The effect of the reactant mass ratio Rr (= DHM/water) was evaluated with the
solvent/reactant mass ratio Rs = 1. The trends in the changes in X in the two solvents with
D

increasing amount of water in the feed (decrease in Rr) differed [Figure 10(a)].
Because of the good solubility in acetone (squares), the system is homogeneous at most
TE

calculated ratios (Rr ≥ 3.33 or ≤ 0.13). The conversion X increases with increasing amount of
water in the feed to a maximum of 0.457 at Rr = 2.86, before slowly decreasing. Most of the
conversions in the biphasic system are higher than those in the monophasic system, but a
EP

little lower than the calculated pseudo-homogeneous values (blue solid line).
In 1,4-dioxane (triangles), increasing the reactant ratio Rr gives biphasic ratio ranges
from 0.017 to 12.5, much larger than those for acetone, and the conversions at the same ratios
C

in 1,4-dioxane are lower. The conversions in the biphasic region are far lower than the
calculated pseudo-homogeneous values (dashed line) and there is no obvious change in X
AC

with increasing Rr from 0.2 to 12.5. This can be explained from the positions of the initial
points on the phase diagram [Figure 10(c)]. The variations in the initial points in 1,4-dioxane
(Rc = 1) are almost parallel with the tie lines. This means that the initial compositions of the
aqueous and organic phases at each initial point are similar; variations in Rr therefore have
little effect on the biphasic equilibrium of the starting mixture. As proof, the data for a higher
solvent feed (Rs = 2) were calculated (diamonds and blue dotted line) and a greater and faster
increase in X was observed as Rr decreases. These calculated results show that the total mass
ratio of the two phases may have little effect on X when the initial compositions in each phase
are constant.
 ACCEPTED MANUSCRIPT

0.00
1.00
S1T1
S1T2
S1T3
0.25
0.75

L
MO

Ac
DH

eto
0.50

M/
0.50

n
e
DH
0.5
0.75
0.25

PT
0.4

1.00
0.00
0.3 0.00 0.25 0.50 0.75 1.00
Water
X

RI
0.2 (b)a
0.00
1.00 S2T1-1
S1T1 S2T1-1 S2T1-2 S2T2-1
0.1

SC
S1T2 S2T2-1 S2T2-2 S2T3-1
S1T3 S2T3-1 S2T3-2 0.25 S2T1-2
0.75 S2T2-2
S1T4 S2T4-1 S2T4-2

L
S2T3-2

MO
0.0

1,4
DH

-D
0.01 0.1 1 10 100

io
0.50

M/

xan
Rr 0.50

DH
U

e
(a) 0.75
0.25
AN
1.00
0.00
0.00 0.25 0.50 0.75 1.00
Water

(c)a,b
M

a
: Binodal curves: black solid line for water + solvent + DHM; black dot line for water + solvent + DHMOL.
b
: Tie lines: black dash line.
D

Figure 10. Predicted total conversions for different reactant mass ratios (Rr). S1: acetone; S2:
TE

1,4-dioxane; T1: monophasic throughout; T2: biphasic during feeding and homogeneous
during reaction; T3: biphasic throughout; T4: pseudo-homogeneous. For system S2, 1 and 2
indicate Rs = 1 and Rs = 2, respectively.
EP

Several T2 points are observed for both tested solvents in the gaps between the ternary
binodal curves for DHM and DHMOL [Figure 10 (b) and (c)]; this confirms that
C

homogenization of the biphasic system is caused by the difference between the solubilities of
DHM and DHMOL rather than the properties of the solvents.
AC

Similar to Section 4.3.1, the variations of concentrations and activity coefficients caused
by proportionally addition of extra water-solvent mixture into a phase-reaction equilibrated
system, which is equivalent to the decrease of the initial DHM/water ratio Rr, to evaluate the
effect of Rr on the conversion.
As reactant ratio Rr increases, the solubilities of DHM and DHMOL in the aqueous
phase changes, as do the activity coefficients. As the differences in solvent properties, the
variations in Qx and Qγ in acetone or 1,4-dioxane are in opposite directions, and different
variation trends for Qa are also observed (Figure 11). In acetone, Qa decreases at higher
DHM/water ratio, which results in increased conversion. In contrast, the Qa increases with
 ACCEPTED MANUSCRIPT

increasing Rr in the 1,4-dioxane system, leading to a decline in conversion. This explains the
different variations in the conversions in the two solvents.

3.9 0.88
1.05 0.67
3.8 0.86

2
2

(Qγ or Qa)×10
0.66
(Qγ or Qa)×10

3.7 1.00 0.84

PT
3.6 0.65

Qx
Qx
0.95 0.56
3.5 Qγ Qγ 0.64
Qx 0.90 0.55 Qx

RI
3.4
Qa Q 0.63
a
3.3 0.85 0.54
0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0

SC
Rr Rr
(a) (b)

Figure 11. Effect of reactant ratio on Qx and Qγ in aqueous phase with acetone (a) and
1,4-dioxane (b) as solvent.
U
AN
5. Conclusions
M

In this work, the hydrations of DHM in acetone and 1,4-dioxane, respectively, were
predicted by coupling the NRTL phase equilibrium model and the reaction equilibrium model.
The LLE data for the ternary system water + acetone + DHM at 358-368 K and 0.8 MPa, and
D

the hydration reaction equilibrium for DHM at 358-373 K and 0.8 MPa were determined to
acquire unreported NRTL parameters and reaction equilibrium constants.
TE

The effects of the feed ratios (solvent/reactant and DHM/water) on the total conversion
X were investigated in detail. In most cases, high concentrations of water and solvent will
promote hydration and increases the conversion of DHM. The conversions in heterogeneous
EP

reactions are not obviously higher than those in homogeneous ones, which means that the
biphasic reaction does not thermodynamically favor the hydration reaction. Acetone, which is
a better solvent for DHM and gave a higher equilibrium conversion, is a better solvent for
C

DHM hydration than 1,4-dioxane.

AC

Acknowledgement
Acknowledgement is made for the financial support from the National Natural Science
Foundation of China (No. 21106020, No. 21376053, No. 21476049) and the UNISPAR
Program of Fujian Province (No. 2014H6002).

Nomenclature
a activity
 ACCEPTED MANUSCRIPT

A1, B1 constants of the Othmer-Tobias correlation

A2, B2 constants of the Hand correlation
Ea activation energy of the reaction, J·mol-1
F total molar feed amount, mol
F0 initial total molar feed amount, mol
FO, FA molar amount of the organic and aqueous phases respectively, mol
∆gij NRTL binary interaction parameter, J·mol-1

PT
∆rGΘ standard Gibbs free energy of reaction, J·mol-1
∆rHΘ standard enthalpy of reaction, J·mol-1
Ka reaction equilibrium constant

RI
Kx reaction equilibrium constant based on molar fraction
Kγ reaction equilibrium constant based on activity coefficient

SC
N total number of the experimental tie lines
P pressure, MPa
Qa quotient of the activity of the product (DHMOL) to that of the reactants
(water and DHM)
Qx
U
quotient of the molar fraction of the product (DHMOL) to that of the
AN
reactants (water and DHM)
Qγ quotient of the activity coefficient of the product (DHMOL) to that of the
reactants (water and DHM)
universal gas constant, J·mol-1·K-1
M

R
R2 determination coefficient
Rr mass ratio of DHM to water
D

Rs mass ratio of solvent to reactant (DHM + water)

∆rSΘ standard entropy of reaction, J·mol-1·K-1
TE

t reaction time, min

T temperature, K
u standard uncertainty
EP

x molar fraction
X hydration conversion of DHM
x’ mass fraction
C

zi total molar fraction of the feed

AC

zi0 total molar fraction of the initial feed

Greek letter
γ activity coefficient
ν stoichiometric number
Superscripts
cal calculated value
exp experimental data
Subscripts
i, j = 1, 2, 3, 4
 ACCEPTED MANUSCRIPT

1 water
2 solvent
3 DHM
4 DHMOL
α organic phase
β aqueous phase

PT
References

RI
1 B. Wang, X. Ge, H. Zheng, T. Qiu, Y. Wu. Liquid-liquid equilibrium for the system water +
cyclohexene + cyclohexanol over the temperature range of (303.2 to 403.2) K. J. Chem. Eng.

SC
Data, 2010, 55(7): 2529-2531.
2 H. Utami, Sutijan, Roto, W. B. Sediawan. Liquid-liquid equilibrium for system composed
of α-pinene, α-terpineol and water. IJCEA, 2013, 4(1): 21-25.

U
3 T. Qiu, , X. Wang, H. Tian, Z. Huang. Liquid-liquid equilibrium for the system water +
AN
1,4-dioxane + cyclohexanol over the temperature range of 313.2-343.2 K. Fluid Phase
Equilib., 2012, 324(25): 28-32.
4 X. Li, K. Tamura. (Ternary liquid + liquid) equilibria for (water + acetone + α-pinene, or
β-pinene, or limonene) mixtures. J. Chem. Thermodyn., 2010, 42(11): 1400-1405.
M

5 H. Li, K. Tamura. Ternary and quaternary (liquid + liquid) equilibria for (water + ethanol +
α-pinene, + β-pinene, or + limonene) and (water + ethanol + α-pinene + limonene) at the
D

temperature 298.15 K. J. Chem. Thermodyn., 2006, 38(8): 1036-1041.

6 H. Li, K. Tamura. Ternary liquid-liquid equilibria for (water + terpene + 1-propanol or
TE

1-butanol) systems at the temperature 298.15 K. Fluid Phase Equilib., 2008, 263(2): 223-230.
7 T. Qiu, S. Li, S. Li, Y. Wu. Liquid-liquid phase equilibria of the ternary system of
water/1,4-dioxane/dihydromyrcene. Fluid Phase Equilib., 2009, 280(1-2): 84-87.
EP

8 X. Liu, Y. Wu, W. Chen, D. Wang, H. Zhou, W. Sun, X. Xu, F. Hou, Z. Huang, Z. Chen.
Preparation method of alpha-terpilenol: China, CN103044202B. 2013.
9 L. Lin, G. Yang, Z. Zhou, Z. Zhang. Production method of DHMOL: China,
C

CN102964215B. 2013.
10 S. Chen, J. Bao, W. Wang, D. Ying, Z. Xie, Z. Lan, D. Li. DHMOL fixed bed hydration
AC

continuous production method: China, CN102173978B. 2013.

11 T. Qiu, B. Wang, H. Zheng, Y. Wang, Y. Wu. Novel method for producing DHMOL:
China, CN101921176B. 2013.
12 Y. Liu, Z. Zhou, G. Yang, Y. Wu, Z. Zhang. Kinetics study of direct hydration of
dihydromyrcene in a jet reactor. Ind. Eng. Chem. Res., 2010, 49, 3170-3175.
13 J. Chen, T. Xie, H. Zheng, S. Zhao, L. Wang. Liquid-liquid equilibrium for the ternary
system of water + acetone + DHMOL at 358 K to 368 K under 0.8 MPa. J. Chem. Eng. Data,
 ACCEPTED MANUSCRIPT

2015, 60, 1134-1138.

14 B. Wang, T. Qiu, S. Li. Liquid-liquid equilibrium for the system water + 1,4-dioxane +
2,6-dimethyloct-7-en-2-ol over the temperature range of (343.2 to 358.2) K. J. Chem. Eng.
Data, 2010, 55, 558-560.
15 B. Ghalami-Choobar, A. Ghanadzadeh, S. Kousarimehr. Salt effect on the liquid-liquid

PT
equilibrium of (water + propionic acid + cyclohexanol) system at T= (298.2, 303.2, and 308.2)
K. Chinese J. Chem. Eng., 2011. 4(19), 565-569.
16 G. Yang, Y. Liu, Z. Zhou, Z. Zhang. Kinetic study of the direct hydration of turpentine.

RI
Chem. Eng. J., 2011, 168, 351-358.
17 Y. Liu, Z. Zhou, G. Yang, Z. Zhang. Study on direct hydration of camphene to isoborneol
by cation exchange resins. Int. J Chem. React. Eng., 2011, 9, 1-16.

SC
18 M. Wang, Y. Hsieh. Kinetic study of dichlorocyclopropanation of 4-vinyl-1-cyclohexene
by a novel multisite phase transfer catalyst. J. Mol. Catal. A: Chem., 2004, 210, 59-68.
19 H. Panneman, A. A. C. M. Beenackers. Solvent effects in the liquid phase hydration of

U
cyclohexene catalyzed by a macroporous strong acid ion exchange resin. Chem. Eng. Sci.,
1992, 47, 2635-2640.
AN
20 H. Panneman, A. A. C. M. Beenackers. Solvent effects on the hydration of cyclohexene
catalyzed by a strong acid ion-exchange resin. 1. Solubility of cyclohexene in aqueous
sulfolane mixtrues. Ind. Eng. Chem. Res., 1992, 31, 1227-1231.
21 H. Panneman, A. A. C. M. Beenackers. Solvent effects on the hydration of cyclohexene
M

catalyzed by a strong acid ion-exchange resin. 2. Effect of sulfolane on the reaction knetics.
Ind. Eng. Chem. Res., 1992, 31, 1425-1433.
22 H. Panneman, A. A. C. M. Beenackers. Solvent effects on the hydration of cyclohexene
D

catalyzed by a strong acid ion-exchange resin. 3. Effect of sulfolane on the equilibrium

conversion. Ind. Eng. Chem. Res., 1992, 31, 1433-1440.
TE

23 Y. Liu, Z. Zhou, G. Yang, Y. Wu, Z. Zhang.. A Study of direct hydration of

dihydromyrcene to DHMOL using cation exchange resins as catalyst. Int. J. Chem. React.
Eng., 2010, 8(1), 1-15.
EP

24 I. V. Kozhevnikov, A. Sinnema, A. J. A. van der Weerdt, H. van Bekkum.. Hydration and

acetoxylation of dihydromyrcene catalyzed by heteropoly acid. J. Mol. Catal. A: Chem., 1997,
120, 63-70.
C

25 P. Botella, A. Corma, J. M. López Nieto, S. Valencia, M.E. Lucas, M. Sergio.. Selective

hydration of dihydromyrcene to dihydromyrcenol over H-beta zeolite. - influence of the
AC

microstructural properties and process variables. Appl. Catal. A: Gen., 2000, 203, 251-258.
26 P. M. Vardharaj, L. Arvind. Method for selective hydration of aroma olefins to alcohols in
continuous solid catalyst column reactor, PCT WO 2012/017447 A2. 2012.
27 J. Guilra, E. Ramírez, C. Fité, M. Iborra, J. Tejero. Thermal stability and water effect on
ion-exchange resins in ethyl octyl ether production at high temperature. Appl. Catal. A: Gen..
2013, 467: 301-309.
28 S. Li. On liquid-liquid equilibrium of dhydromyrcene hydration system. Master Thesis:
 ACCEPTED MANUSCRIPT

Fuzhou University, China, 2008.

29 L. Zhang. On hydration process of dihydromyrcene. Master Thesis: Fuzhou University,
China, 2008.
30 D. F. Othmer, P. E. Tobias. Tie line correlation. Ind. Eng. Chem., 1942, 34, 690-692.
31 D. B. Hand, Dineric distribution. J. Phys. Chem., 1930, 34(9), 1961-2000.

PT
32 H. Renon, J. M. Prausnitz. Local compositions in thermodynamic excess functions for
liquid mixtures. AIChE J., 1968, 1(14), 135-139.
33 J.M. sorensen, T. Magnussen, P. Rasmussen, A. Fredenslund. Liquid-liquid equilibrium

RI
data: their retrieval, correlation and prediction, Part II: Correlation. Fluid Phase Equilib. 1979,
3, 47-82.

SC
34 Y. Zeng, X. Chen, C. Guo, S. Zhao, H. Zheng, L. Wang. Measurement and correlation of
liquid-liquid equilibrium data for water + methanol + methyl acetate + potassium acetate.
Fluid Phase Equilib. 2013, 354, 319-325.

U
AN
M
D
TE
C EP
AC