Solvent Extraction Processes
The separation of components from a liquid mixture can be achieved by solvent extraction. It is a unit operation to separate one or more components from a liquid (or a solid) phase mixture by selectively dissolving it into a solvent.
In the case the feed mixture is a liquid, it is called liquid-liquid extraction.
This operation is possible thanks to the different affinity of components in a given solvent - thus solvent is chosen so that to maximize its affinity with respect to one (more) components of the mixture liquid (or solid) mixture to be separated from the rest of the liquid (solid) components.
That operation can be carried out both in a discontinuous or continuous single-stage and in a continuous multistage mode
It is a non-thermal separation process
CONTINUOUS MULTISTAGE MODE
RAFFINATEPhase rich in the component of the feed we don't want to extract
Solvent
Feed
ExtractPhase rich in solvent
In general we add the stream with the higher density at the top and the one with the lower density at the bottomIn that way we assure a countercurrent flow
BLOK FLOW DIAGRAM
A+B
Ffeed
Ssolvent
Extraction and phase separation
E
Extract (A+S+B)
Rraffinate (B+S+A)
Solvent Recovery
PEProduct A+B+S
Solvent Recovery
PRProduct B+A+S
Equilibrium
The system is characterized by (at least) 3 components and 2 phases
We want to be able to know the selectivity of component A with respect to component B
SA,B = KA / KB
KA = xAα / xAβ
KB = xBα / xBβ
> 6 Constrainte
The mass transfer is less efficient when we operate with liquids with respect to the mass transfer value for gas mixture
So, to decrease the number of stages we would have an high selectivity
For LLE, a g-g approach can be used (iso-activity conditions):
aAα = aAβ
aBα = aBβ
aSα = aSβ
γAαxAα = γAβxAβ
γBαxBα = γBβxBβ
γSαxSα = γSβxSβ
Σ xiα = 1
Σ xiβ = 1
V = 3 - 2 + 2 = 3
Fix T, p and then the last degree of freedom
Two solutions are obtained (one for phase α and the other for phase β)
Pressure does not have a strong effect on liquid-liquid equilibria
Increasing the T, the miscibility gap is reduced
Binodals
they represents the compositions in equilibrium
The points representing two mixtures in equilibrium are called conjugate points and the lines that connect two compositions in equilibrium with each other are called tie lines
The point where the binodals meet is called crit point
TUC: Upper Critical Solution Temperature
Above it the mixture is homogeneous
TERNARY SYSTEMS REPRESENTATION
Vertices are pure components
Segments (AB, BC, CA) are binary mixture
P is a ternary mixture - composition?
Segment BC - take the massole and we will find the composition of A
All the points belonging to the secret AU coming out from vertex A, are characterized by the common value of the zB/zc ratio
The mixing of mixtures R and S to obtain mixture H is shown. Point H on the segment RS divides it into parts proportional to the respective quantities, according to the lever rule.
I can have the following representation:
It is used a pseudo-binary representation, where the solvent is not a component (solvent-free component). The pseudo-binary fractions are defined as follows:
xA = XAα / (XAα + XBα)
xB = 1 - xA
yA = XAβ / (XAβ + XBβ)
yB = 1 - yA
on x axis I have the pseudo-binary composition of phase alpha
of phase beta
Solvent Selection
- High solubility towards the component to be extracted in order to limit solvent consumption
- High selectivity to extract only one component from the starting solution
- Low solubility in the raffinate to avoid solvent losses
- Low vapour pressure to reduce evaporative losses
- Low cost, non-corrosive property and no implications for safety and the environment
Single-stage operation: Mixer-settler
Ideal stage
We have a mixing tank where we feed F and S, then the mixture that comes out of the vessel enters into a liquid-liquid separator, where there is the separation between the light and heavy phases, obtaining the two products (raffinate and extract).
On that plot we can identify the points F and S.
When we mix F and S we obtain the mixture R. Since the material balance is linear in composition, the composition of the mixture will be on the straight line that connects F and S.
The compositions of E and R, hence the products are in equilibrium, are represented by two points located on the same tie line. That tie line has to pass through the point R, because of the material balance done around the separator.
By adjusting the solvent flow rate we will change the point R and so the composition of R and E.
Rmin - the minimum flowrate of S
Rmax - the maximum flowrate of S
Counter-current multistage operation
A mixer-settler system has insufficient performances and it has only one degree of freedom.
A multiple mixer-settler system has two degree of freedom (design) and one degree of freedom (setup).
Generally there is the specification on the extract (or raffinate) and also the inlet solvent flowrate is fixed.
Overall material balances
F + S = M = E1 + RNP
F xF,i + S xS,i = M xM,i = E1 y1,i + RNP xNP,i
For i = A, B, S
Material balances below stage n
Rn-1 + S = En + RNP
Rn-1 xn-1,i + S yS,i = En yn,i + RNP xNP,i
These equations can be re-written in the form: RNP - S = F - E1 = Δ
That point represents the intersection between the straight line connecting RNP and S and the straight line connecting F and E1
Considering now the material balances below N we can write:
RNP - S = RN-1 - EN = Δ [same point]
operating lines
So all the straight lines connecting the streams E and R between two stages pass through the points
Design Problem
- we represent F and S- we draw the straight line connecting the feed and to solvent point- we find the point l
- we fix (for example) the composition of E
From the overall material balance we find the composition of the raffinate, that is not in equilibrium with E
- To find the point Δ we connect the two straight line (S,RNP, F - E1)
We know E1 and to find R1 (in equilibrium with E1) we use the tie line. R1 and E2 are connected by the straight line passing through the point Δ, so with Δ and R1 points we find E2 -> E2 = R2, R2 and Δ = E3... We stop when we reach one specification and we find the number of stages
Raffing Problem
: the number of ideal stages available is unknownThere will be only one value of solvent flowrate S which ensures to achieve the desired output composition of the extract (or of the raffinate)