1. Thermodynamic modeling behavior of cellulose acetate / polyvinyl chloride blend membrane
preparation
Dr. Ayman El-Gendi
Dr. Heba Abdallah
Dr. Ashraf Amin
Good morning…..
-first I would like to thank the conference commity for the conference organization and for giving me this chance to present my work
I’m going to present my research work , entitled……..
This work has been carried out……at .NRC …. …with my colleagues Dr. Heba and Dr. Ashraf
National Research Centre, El Buhouth St., Dokki, Cairo, Egypt,,
,Chemical Engineering& Pilot Plant Department,
. E.mail:
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

2. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Outline:
Introduction
Objectives
Experimental work
Model Equations
Results
Conclusions
My presentation will follow this outline:
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

3. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Introduction
Polymer Membranes are widely used in different applications, high selectivity and high permeability are principal requirements.
Asymmetric Membranes:
to get High Jwater
to achieve liquid molecular separation
The phase inversion technique is the most popular membrane fabrication method, especially for commercial membrane manufacturing.
Polymeric Membranes:
high mechanical properties
good chemical stability
Tunable microstructures
There are different materials can be used for memb. Preparation as polymeric and ceramic membranes
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

4. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Objectives
- The goal of this work is to discuss the basis of solubility parameters, and their use in predicting polymer dissolution for polymers blends and solvents.
- The modeling results can be used to develop a better understanding of the experimental results and the precipitation process.
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

5. Asymmetric membranes preparation
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

6. Casting & Evaporation steps
Procedure overview
Polymer
Mixing
Solvent
Casting & Evaporation steps
Phase inversion step
3 steps involved:
MIXING
Casting & Phase inversion using a coagulation bath:
Many parameters can be adjusted and tuned:
- going from liquid thickness layer
- to POLYMER concentration & composition
- type of no solvent (‘Water, ), t°C of coagulation…
Very complex mechanism involving simultaneously mass transfer, heat transfer, Polymer Tg modification, irreversible processus and liquid phase equilibrium as well
Drying &
Drying procedure at RT
Thermal treatment
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec 6

7. Principales of phase inversion
gelification
liquid-liquid phase separation
homogeneous solution
+In the phase inversion process induced by immersion precipitation, a polymer casting solution, cast on a support is immersed in a coagulation bath containing a nonsolvent (as water). The solvent moves from polymer solution to nonsolvent while the nonsolvent penterates the casted polymer solution.
a- The mechanism of membrane formation has been reported by some researchers.
-the porosity of the sublayer (bottom layer) induced by liquid-liquid phase separation while the formation of pores may be ascribed to nucleation and growth of the dilute polymer phase
b-the toplayer is formed by gelation while liquid-liquid phase separation forms the sublayer of the membrane
Fig.: Ternary phase diagram containing; P: polymer; S: solvent; NS: non-solvent
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

8. SEM Control of membrane structure
Asymmetric membranes
phase inversion procedures
Preparation of several type of membranes:
Dense membrane to be used as reference for mechanical tests and permeation properties
Asymmetric Mb obtained from various set of conditions able to tuned separation properties:
- induction of a dense top layer thickness
- control of top layer thickness
- shape and extent of the porosity of the sub-structure
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

9. Membrane characterizations
SEM
Water content
Porosity
Mechanical test.
Performance test( as: Flux, Separation)
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec 9

10. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Mathematical Model
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

11. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
- Thermodynamic Model
The Flory–Huggins theory is chosen as a basis for the thermodynamic model since it is the most relevant theory for modeling the free energy of binary and ternary polymer mixtures.
The quaternary system has two low molecular weight component; as (S1 –S2), and two high molecular weight polymers (P1 and P2).
The diffusion model is used to investigate the immersion/precipitation process.
The phase behavior of polymer blends can be described by the Gibb’s free energy of mixing (ΔGmix)
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

12. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Model Equations - 1
-The model main equation is an equation for calculating the Gibbs free energy change for mixing a polymer with a solvent
(1)
Where:
: Change in enthalpy during the mixing process, J/mol
: Change in entropy during the mixing process, J/mol/K
T: Temperature, K
the critical value of the interaction parameter:
The Flory–Huggins theory is chosen as the basis for the thermodynamic model because it provides good predictions and satisfactory results; also it is the most relevant theory for modeling the free energy of binary and ternary polymer mixtures. The mathematical model equations were solved by MATLAB software (The Mathworks, Release 2009b) .
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

13. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Model Equations - 2
- The general expression of the extended Flory-Huggins model for quaternary polymer solutions
- Chemical potential was calculated taking the first derivative of the Gibbs free energy of mixing equation with respect to the mole fraction of each component
-- The mixing process will be a spon-atenous process when the change of the Gibbs free energy is negative. The change in Gibbs free energy was studied in the temperature range K.
- Chemical potential was used for calculating quaternary coexistence curve
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

14. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Model Equations – 3 Diffusion model of immersion precipitation step
Model assumptions
1. The reference system is chosen centered around the interface between coagulation bath and polymer film.
2. The interface of the casted blend membrane in the coagulation bath has an initial composition similar to the original coagulation bath (distilled water).
3. The polymers blend is considered as a matrix phase in which the two other components diffuse.
4. The polymer solution doesn't form a closed system, and the volume of the polymer solution is not constant.
5. According to diffusion path of solvent from polymer film and until the spinodal is reached, one phase system is assumed.
6. The mobility of the polymer is much slower compared to the low molecular weight components.
7. The concentration of components in the coagulation bath depends on the diffusion coefficients.
8. The fluxes are independent of the film thickness.
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

15. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Diffusion model of immersion precipitation
describes diffusion in the polymer solution (i) represents polymer
the diffusion of solvent
component (j) represents solvent
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
11/15/2018

16. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results – 1 Model verification: Effect of polymer blend composition(CA/PVC membrane)
Figure(1) illustates that the Gibbs free energy is approximately constant at volume fraction range ( ). The Gibbs free energy is found to range from -200 J/mol at 313K to -450 J/mol at 363K. Accordingly, polymer blend solution is homogenous down to the molecular level, in which the domain size is comparable to the macromolecular dimension; associated with negative value of the free energy of mixing,ΔGm ≈ ΔHm ≤ 0 and within the phase stability condition (ϐ2ΔGm/ϐ2φ)>0. That indicates there is no limitation of miscibility to two minimum compositions in addition the constant line indicates that phase separation can't occur during mixing of these blend polymer (PVC/CA) in this range of compositions
ΔG as a function of ϕCAat different temperatures
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

17. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results – 2 Effect of polymer blend composition
Figure (2) indicates that during mixing the chemical potential of CA at various temperatures decrease by increasing the composition of CA, that can lead to problems in blending between two polymers when 0.12 composition is used.
μCAas a function of ϕCAat different temperatures
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

18. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results - 3Verification of diffusion model of immersion precipitation
Figure illustrares the effect of time on the membrane formation during the coagulation process to form membrane; in this case equal volume fractions of PVC & CA are used (0.07), while the volume fraction of solvents were ϕNMP=0.15, ϕTHF=0.71. The model results indicate that the ratio of ϕCA/ ϕsolvents increased with coagulation time due to removal of mostly all the solvents from the polymeric solution to the bath, also the thickness of polymer solution on the coagulation plate decreased with time due to solvent removal, which leads to a concentrated polymer formation.
-where the polymer content is concentrated in the body of membrane and the solvent content outflows to the coagulation bath
ϕCA/ϕsolventas a function of time and thickness of membrane at T=298K, Initially; ϕCA= 0.07, ϕPVC= 0.07, ϕNMP=0.15, ϕTHF=0.71
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

19. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results - 4
Figure indicates similar model results when the ratio of ϕPVC/ ϕsolvents was used as a variable, which means that the coagulation rate is close to each other, so the best blend membrane can be formed.
ϕPVC/ϕsolventas a function of time and thickness of membrane at T=298K, Initially; ϕCA= 0.07, ϕPVC= 0.07, ϕNMP=0.15, ϕTHF=0.71
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

20. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results - 5
- The figures in this slide and next slide show the solvents 1nMP and THF concentrations. NMP and THF concentrations increased in the coagulation bath with time due to the outflow of solvents from the main casted polymer solution to the bath.
- At starting of immersion the solvent % in polymer solution sharply decreased with time due to the outflow of the solvent from polymer solution to the bath while it concentration increased with time in coagulation bath
ϕNMPas a function of time and thickness of membrane at T=298K, Initially; ϕCA= 0.07, ϕPVC= 0.07, ϕNMP=0.15, ϕTHF=0.71
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

21. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results - 6
ϕTHF as a function of time and thickness of membrane at T=298K, Initially; ϕCA= 0.07, ϕPVC= 0.07, ϕNMP=0.15, ϕTHF=0.71
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

22. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results - 7
Figure shows the formed membrane structure after 500 min in the coagulation bath which indicate that the upper selective layer was formed mostely from CA and the middle and bottom layer from PVC and the formd membrane thickness was 284 µm. By comparing the predicted composition with composition from experimental work, a similar shape of membrane cross section was produced, where the dense top layer was CA and the bottom layer was PVC as shown in next slide.
Representation of membrane constituents after 500 min of coagulation
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

23. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Results - 8
Cross section of PVC/CA membrane
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

24. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Conclusions - 1
A quaternary thermodynamic model was developed to describe a cases study of blend membrane formation using the phase inversion process. The model was verified and the following conclusions can be drawn from the present study:
The effect of polymer blend composition on the change of Gibbs free energy was observed, however the mixing was stable for all the polymers compositions at different solution temperatures .
The diffusion model was developed to study the effect of immersion/ precipitation time in the coagulation bath on the blend membrane formation.
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

25. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
Conclusions - 2
The model indicated that the solvent was segregated from polymer solution dissolving in the coagulation bath during precipitation time.
The volume fraction of solvent increased in the bath and the volume fraction of polymer solution decreased as a result of the membrane formation.
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec

26. Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec
THANK YOU
11/15/2018
Petrochemistry 2015 Atlanta USA ,30 NOV to 2 Dec