Determination of liquid-liquid equilibria equilibria of a waterwateracetic acid-ethyl acetate ternary system Batiforra, Chester F. (Author) Department of Chemical Engineering University of the Philippines, Diliman Quezon City, Philippines
[email protected]
Camora, Christian Paul A. (Author) Department of Chemical Engineering University of the Philippines, Diliman Quezon City, Philippines
[email protected]
Abstract — The The purpose of this experiment is to generate a ternary binodal curve and to determine the plait point of an ethyl acetatewater-acetic acid system. Titration was the main process used in the experiment. Nine (9) solutions of different compositions of ethyl acetate in water were titrated with glacial acetic acid. Data obtained from this were used to generate the phase diagram. Three (3) solutions of different compositions of acetic acid, ethyl acetate, and water were placed in a separatory funnel to facilitate phase separation of an organic layer and aqueous layer. Each layer of each solution was titrated with standardized 0.85 M NaOH, resulting to the determination of the amount of acetic acid per phase, hence generating the tie lines needed for the extrapolation of the lines to determine the plait point. Using the UNIFAC model as the theoretical curve, the experimental binodal curve is generated, and in comparison, the experimental curve has very high deviations. On the other hand, the plait was determined to be at 15% EtOAc-45%H2O-40%HOAc by weight, with 2.296% error
from the UNIFAC model. Keywords: binary mixture, flash point, fire point
I.
Ordinario, Ma. Precious Lhyn (Author) Department of Chemical Engineering University of the Philippines, Diliman Quezon City, Philippines
[email protected]
Lerona, Paul Gabriel E. (Author) Department of Chemical Engineering University of the Philippines, Diliman Quezon City, Philippines
[email protected] the system. With this said, it has lower energy requirement compared to distillation (Shiveler & Koch, 2015). B. Plait Point Determination As stated earlier, liquid solutions portray immiscibility at least over a certain range of temperature and composition. For condition of constant pressure, binary LLE can be easily represented using solubility diagram (a plot of temperature and composition of the more volatile component). As shown on Figure 1, there are different classifications of LLE solubility diagram depending on the location of its equilibrium curve. Curve (a) represents the binodal curves that define an “island”. This portrays the composition of coexisting phase. On the other hand, as shown from the other curves, existence of lower consolute temperatures/ lower critical solution temperature and upper critical solution temperature is observed (Smith, Van Ness, Abbott & Swihart, 2005). LLE is possible bet ween these temperatures.
I NTRODUCTION
A. Liquid - liquid Equilibrium System Mixing of two different pure liquids that are partially miscible with each other will result to system with two liquid phases of different compositions. When this system comes to thermal, mechanical, and transfer equilibrium, the phenomenon is called as a liquid/ liquid equilibrium (LLE) (Smith, Van Ness, Abbott & Swihart, 2005). This is considered as a significant process in the field of separation technology. Unlike distillation which depends on the boiling point of substances, liquid – liquid liquid extraction uses the idea of relative solubilities in two immiscible liquids present in Figure 1. Solubility Diagrams of LLE Systems ("Section 1: Liquid Liquid Equilibria", 2013)
C. Ternary Liquid- Liquid Diagram As observed from the solubility diagrams earlier, it is quite evident that one of the liquid phases contains one of the components in a more concentrated form. Thus, a more detailed representation was used to exploit this idea in the process industry – via ternary liquid – liquid diagram. Due to inadequate difference in volatility of the components or existence of azeotrope, an alternative method is used by adding a third liq uid (solvent) that is partially immiscible with the original solution. This results to one of the components in the original mixture to become more concentrated due to its affinity with the added solvent ("Section 1: Liquid Liquid Equilibria", 2013) .
indicator and potassium hydrogen phthalate. The materials and equipment needed are acid burette, base burette, 3 (250 mL) separatory funnels, 9 (100 mL) Erlenmeyer flasks, 3 (125 mL) Erlenmeyer flasks, 3 (250 mL) Erlenmeyer flasks, 250 mL beaker, stirring rod, 25 mL graduated cylinder, 50 mL graduated cylinder, small funnel, 4 (100 mL) beakers, 100 mL volumetric flask, 250 mL volumetric flask, glass spatula, watch glass, 2 (10 mL) pipettes, 2 (25 mL) pipettes, separatory funnel stand, iron rod and burette clamp. B. Construction of the Equilibrium curve In separate 100 mL flasks, the following mixtures were prepared Table 1: Mixtures for construction of equilibrium curve Mixture
Vol. EtOAc (mL)
Vol H2O (mL)
A
4.1
33.0
B
8.4
24.4
C
12.4
19.2
D
16.4
15.0
E
20.4
11.2
F
24.4
8.4
G
28.3
6.0
H
32.2
3.6
I
36.1
2.0
Figure 2. Ternary LLE Diagram ("Section 1: Liquid Liquid Equilibria", 2013)
In Figure 2, solvent C was added to preferentially extract A. Each vertex of the triangle represents a 100% mole fraction of the species indicated in each corner. Moreover, the side represent a mixture of two species indicated at the two ends of the side. The most significant part of this diagram is the dome DPE which corresponds to the two - phase region, while the area outside represents a ternary homogenous mixture. T he end points of line within the dome called tie lines corresponds to the extract and raffinate phases. In a binodal curve of a ternary phase diagram, t here exists a point in which the composition of the two phases are identical. This is called the plait point. This is comparable to the concept of azeotrope where the compositions of the liquid and vapor phases are the same. (Pr ice, 2003)The plait point is also called the isothermal critical-mixing-point (Widom, 1967) which implies that the mixture at the plait point has critical conditions at the conditions in which the diagram is plotted (OnePetro, 2015). In this experiment, the researchers aims to generate a ternary liquid-liquid equilibrium diagram from experimental data for a water acetic acid-ethyl acetate system and estimate the location of the plait point using experimental data. II.
METHODOLOGY
A. Materials and Equipment The reagents that were used in the experiment were, sodium hydroxide, glacial acetic acid, ethyl acetate, phenolphthalein
Using glacial acetic acid as titrant, each mixture w titrated until the disappearance of the two layers. This was done by rapidly agitating the mixture against a light source and determine if the two layers still form. The amount of glacial acetic acid used for mixture was recorded. C. Standardization of NaOH with KHP Using 1.0M NaOH pellets and a 100 mL volumetric flask, a 100-mL solution of NaOH was prepared. In a volumetric flask, approximately 0.3 g of potassium phthalate (KHP) was dissolved in 10 mL distilled water. After adding 1-2 drops of phenolphthalein indicator, the solution was titrated using 1.0M NaOH until the light pink endpoint. The average molarity of NaOH was recorded after conducting three trials. D. Plait-point determination. The following mixtures were prepared in a separatory funnel, ensuring that the stopcock is properly closed. The water was transferred first, followed by the ethyl acetate and the glacial acetic acid added last. Table 2: Mixtures for plait point determination Mixture
Vol. EtOAc (mL)
Vol H2O (mL)
Vol HOaC (mL)
AA
22.2
12.0
7.6
BB
22.2
15.2
4.6
CC
22.2
17.2
2.7
The separatory funnel was sealed and rocked gently to agitate the mixture. The stem was pointed upward and the stopcock was opened to release excess pressure. The stopcock was closed afterward. This was done until no significant excess pressure was observed. A 100-mL erlenmeyer flask was pre-weighed. From the top organic layer of mixture AA, 5-mL sample was collected and transferred to the erlenmeyer flask. The new weight was recorded. After adding 1-2 drops of phenolphthalein indicator, the solution is titrated using the standardized NaOH until the light pink endpoint. The amount of NaOH used was recorded. Repeat the procedure above for 2 trials, then for bottom aqueous layer and again for each mixture. III.
R ESULTS AND DISCUSSION
A. Ternary plot construction Given a ternary (ethyl acetate-water-acetic acid) LLE system, nine different setups were made in which each triplet (x, y, z), where x,y,z ∈[0,1] and x+y+z=1, corresponding to t he mass fractions of each component per setup, forms the binodal curve in a ternary diagram. In this e xperiment, (x,y,z) represents ethyl acetate, water, and acetic acid mass fractions, respectively. This triplet was completed by cloud-point titration where the solution of (x,y) composition is titrated with z. In the first few drops, the formation of two layers in the solution was evident. Equilibrium is then reached when the interface between these layers is gone. The amount of z consumed corresponds to the third composition of the curve, thus completing the triplets. Figure 3 shows the experimental and the theoretical ternary diagram of the ethyl acetate-water-acetic acid system in terms of their mass fractions. The theoretical curve was obtained through the Dortmund Databank Software Package, in which the curve was based from the UNIFAC model under 1 bar, 28°C. From Figure 3, it can be observed that the experimental data has the same behavior as a typical binodal curve, however, the deviations were so high. These deviations were probably propagated from the experimenter’s mishandling of the solution and the buret’s uncertainty. Moreover, t hese deviations might also suggest unpredicted intermolecular attractions among the components that the UNIFAC model fails to account, given the atmospheric and geographical conditions. Also, the solutions that were provided might suggest that the stocked chemicals were not 100% pure, hence, contaminated, and that the water that was assumed to be distilled or 100% contains traces of carbonates, sulfates, and worse, polymers that might induce intermolecular attractions, suggesting that the solution is highly non-ideal. Lastly, there is an ulti mate bias on the experimenter’s perspective especially on the determination of the disappearance of the interface between the layers. Moreover, in the disappearance of the interface, it is regarded that the
Figure 3. Ternary LLE Diagram for ethyl acetate-water-acetic acid system
equivalence point is reached if a clear solution is obtained after the formation of a turbid solution. This turbid solution is caused by the for mation of micelles in the solution, wherein at the start of the titration, the stable binary solution was added, dropwise, with the titrant, the Gibbs energy of the solution increases, creating an unstable condition where the surface tension of the species increases, thus forming a two-phase solution. While at constant agitation and dropwise addition of the titrant, the layers become thick and with further addition, localized spherical layer separation, the micelle, forms and eventually these micelles decrease its size as it overcomes this energy barrier where the surface tension in the micelle membrane ultimately decreasing until it reaches equilibrium denoting the disappearance of interface. Apart from the various probable factors that caused the deviation, the theorecity of the model that was used also determines the fitness of data. From Figure 4, different models were compared against the UNIFAC curve and the experimental curve, and that the Predictive Soave-RedlichKwong (PSRK) curve was almost as close to the data. However, PSRK describes well the mixtures’ vapor densities but in terms of liquid density prediction, the deviations are too high. This suggests that UNIFAC will still be used as the theoretical model for the current LLE system, as it is more compatible with polar systems compared to other models and are nearly independent of temperature conditions. B. Plait point determination In the determination of plait point of a ternary system, it is important to take note of the composition of the two layers, the aqueous and the organic layers, of the system, and by the virtue of Gibbs phase rule, the degree of freedom is one, and in this context the system can be defined when the weight percent of HOAc is known. This method exploits the nonequilibrium condition of the system as the overall composition triplet (x,y,z) falls under the
Figure 4. UNIFAC a nd experimental curves versus other local composition models and equations of state
binodal curve, in which to achieve the liquid-liquid equilibrium condition, it splits into two phases to compensate for the instability of the system having a higher Gibbs energy and the dominating repulsive forces of each species with one another, creating a stable two-phase system. These two-phase condition, at a particular composition in the ternary diagram, forms a tie line when connected. Given the three different mixtures, AA, BB, and CC, their position in the ternary diagram should fall under the binodal curve, and to obtain the plait point, it should make use of the tie lines produced from the triplet composition of each phase. The method to determine the plait point utilizes the acid base titration, that starts first with the standardization of the NaOH with KHP. This standardization yields the relatively accurate molarity of the NaOH solution. In this experiment, the standardized NaOH molarity is 0.8503 M. Appendix shows the values for computing the molarity of NaOH. This known concentration of NaOH easily gives the a mount of acetic acid per phase after titration, given t he 1:1 molar ratio of HOAc-NaOH in the acid-base reaction. This gives the mass fraction of the HOAc per phase and can be plotted in the ternary diagram. Since the generated plot from previous section was highly deviated from the theoretical binodal curve, the mass fraction of the HOAc was projected onto the UNIFAC curve under the assumption that it follows the theoretical behavior. Figure 5 shows the red tie lines produced per mixture, where these tie lines are arranged from the bottom to top as AA, BB, and CC. The blue tie lines are the theoretical tie lines from the UNIFAC model generated through the Dortmun Data Bank Software Package. It can be observed that the obtained tie lines from the experiment follows the trend of the theoretical tie lines as if each tie line is being interpolated from its adjacent theoretical lines and that the lines are decreasing in length as it approaches the plait point. This shows that the results generated were accurate under the assumption that the weight percent of HOAc has the appropriate ethyl-acetate water weight percents
Figure 5. Tie lines produced after projecting the weight % of HOAc per phase of each mixture
as projected onto the binodal curve. Moreover, to determine the plait point, these red tie lines are extended to find the delta point, Δ, wherein after locating this point, a line will be extended tangentially to the binodal curve. Figure 6 shows the method of determining the plait point, and from the figure, the plait point is located at (0.15, 0.45, 0.40). From the UNIFAC data, the plait point is at (0.1566095, 0.4409335, 0.4024565) yielding a 2.296% error. This method also has consistent result as, while it only needed two tie lines to determine the plait point, the third line coincidentally had the same Δ point with the other two lines. Generally, this is not inconsistent with the results of the ternary diagram construction as this method has assumed that the weight percents of ethyl acetate and water are assumed to be exact ratios, suggesting that the ternary construction experiment has yet to observe more stringent discipline in obtaining an accurate result. Although, this does not eliminate the possible reason of chemical contamination in the reagents. IV.
ONCLUSIONS & R ECOMMENDATIONS
In an ethyl acetate-water-acetic acid ternary LLE sytem, the construction of ternary diagrams and the determination of plait point utilizes titrimetric techniques. In the construction of ternary diagram, it is important that the amount of ethyl acetate and water should be defined such that using the amount of the acetic acid, consumed after the observance of the formation of a turbid solution that eventually entails the disappearance of t he interface of two layers, a binodal curve can be generated and can be compared with the theoretical curve. I n this experiment, the experimental data was compared with the theoretical data generated using the UNIFAC model, the yielded result has high deviation and can be related to the experimenters mishandling of the solution, possible contamination of the chemicals, or the hardness of water used. Moreover, the use of a standardized 0.85M NaOH to determine the amount of acetic acid per phase
Δ point
Figure 6. Tangential extrapolation of experimental tie lines to obtain the plait point of ethyl acetate-water-acetic acid system
exploits the 1:1 ratio of the acid-base reaction. The amount of acetic acid per phase, as projected onto the theoretical binodal curve, provides the necessary tie lines for the tangential determination of the plait point through the experimental tie li ne extrapolation and the location of the Δ point. In this experiment, the plait point was determined to be at (0.15, 0.45, 0.40) in the ternary diagram, indicating that the plait point can be achieved when the composition of the solution is 15% EtOAc-45%H2O40%HOAc by weight, with 2.296% error. This plait point implies that at this condition, the sol ution will not split into two phases. R EFERENCES Section 1: Liquid Liquid Equilibria. (2013). Nptel.ac.in. Retrieved 1 March 2018, from http://nptel.ac.in/courses/103101004/75 [2] Shiveler, G., & Koch, J. (2015). Design Principles for Liquid-Liquid Extraction. Aiche.org. Retrieved 1 March 2018, from https://www.aiche.org/resources/publications/cep/2015/no vember/design-principles-liquid-liquid-extraction [3] Smith, J., Van Ness, H., Abbott, M., & Swihart, M. (2005). Introduction to chemical engineering thermodynamics (7th ed.). McGraw Hill. [4] OnePetro. (2015). Ternary phase diagrams. Retrieved from PetroWiki: petrowiki.org/Ternary_phase_diagrams [1]
Price, R. M. (2003). Liquid Extraction. Retrieved from CBU Faculty and Staff: facstaff.cbu.edu/rprice/lectures/extract.html [6] Widom, B. (1967). Plait Points in Two- and ThreeComponent Liquid Mixtures. The Journal of Chemical Physics. [5]
APPENDIX Table 3 Properties of each component
Property Density (g/mL) Molar Weight (g/mol)
EtOAc 0.902 88.11
H2O 1 18.02
HOAc 1.05 60.05
Table 4 Ternary diagram construction
Setup
VEtOAc
VH2O
VHOAc
massEtOAc
massH2O
massHOAc
Total mass
A
4.1
33
1.5
0.041973
1.831299
0.026228
1.899499
B
8.4
24.4
7.2
0.085993
1.354051
0.125895
1.565939
C
12.4
19.2
8.3
0.126941
1.065483
0.145129
1.337553
D
16.4
15
8.7
0.16789
0.832408
0.152123
1.152422
E
20.4
11.2
7.6
0.208839
0.621532
0.132889
0.96326
F
24.4
8.4
6.5
0.249788
0.466149
0.113655
0.829592
G
28.3
6
4.5
0.289713
0.332963
0.078684
0.701361
H
32.2
3.6
1.5
0.329638
0.199778
0.026228
0.555644
I
36.1
2
0.15
0.369563
0.110988
0.002623
0.483174
AA
22.2
12
7.6
0.227266
0.665927
0.132889
1.026082
BB
22.2
15.2
15.2
0.227266
0.843507
0.265779
1.336552
CC
22.2
17.2
17.2
0.227266
0.954495
0.300749
1.48251
Table 4 c ontinuation
0.013808
%EtOAc 2.209663
%H2O 96.40954
%HOAc 1.380793
0.86469
0.080396
5.491435
86.46897
8.039593
0.094906
0.796591
0.108503
9.490563
79.6591
10.85034
D
0.145685
0.722312
0.132003
14.56846
72.23123
13.20031
E
0.216804
0.645238
0.137958
21.68044
64.52378
13.79579
F
0.301097
0.561901
0.137001
30.10972
56.19013
13.70015
G
0.413073
0.474739
0.112188
41.30726
47.47392
11.21883
H
0.593254
0.359543
0.047203
59.32537
35.95431
4.720313
I
0.764866
0.229706
0.005428
76.48659
22.97058
0.542831
AA
0.221489
0.649
0.129511
22.14891
64.89996
12.95114
BB
0.170039
0.631107
0.198854
17.0039
63.11071
19.88539
CC
0.153298
0.643837
0.202865
15.3298
64.3837
20.28649
Setup
mEtOAc
mH2O
mHOAc
A
0.022097
0.964095
B
0.054914
C
Table 5 Standardization of NaOH
Trial
massKHP (g)
Initial volume of NaOH, mL
Final Volume of NaOH, mL
Net Volume NaOH, mL
1
0.2926
5.1
6.8
1.7
2
0.2978
6.8
8.5
1.7
3
0.2952
6.9
5.2
1.7
Average
0.2952
1.7
Table 6 Plait point determination
Setup AA BB CC
Weight of the aliquot
NaOH used
mass HOAc
% HOAC
phase 1-organic
4.8195
21.3
1.087581
22.56627
phase 2- aqueous
4.9931
16.6
0.847599
16.9754
phase 1-organic
4.58395
0.9
0.045954
1.002501
phase 2- aqueous
4.7854
13.95
0.712289
14.88463
phase 1-organic
4.44
2.75
0.140415
3.16251
phase 2- aqueous
4.3385
10.825
0.552726
12.74003
