IITIAN1997 STUDIO

Advanced Separation Processes

Membranes, adsorbents and beyond.

Open the interactive studio →

Equilibrium and rate-governed separations, MF, UF, NF, RO, pervaporation, gas separation and electrodialysis, polarisation and fouling, adsorbents, isotherms, PSA and TSA, chromatography, supercritical extraction, reactive distillation and biofiltration from the ASTU CHE1818PE41 syllabus (2018-19 scheme).

What is inside

  • Reverse osmosis, polarisation and gas permeation
  • Adsorption isotherms and fixed beds
  • Chromatographic resolution
  • Supercritical extraction and reactive distillation

Syllabus: ASTU CHE1818PE41 Advanced Separation Processes, semester 8, 2018-19 scheme (programme elective 4). Official syllabus PDF.

Introduction to separation processes

Topics: Separation processes in chemical industries; categorisation of separation processes; equilibrium and rate-governed processes.

Why separations, and how to classify them

Separations take 40–70 % of the capital and operating cost of a typical chemical plant: feeds must be purified, reactor effluents split into product, recycle and waste, and products brought to specification. Every separation needs a separating agent — energy (heat in distillation, pressure in membranes, an electric field in electrodialysis) or mass (a solvent in extraction or absorption, an adsorbent, a membrane) — and a property difference between components.

A useful classification is by mechanism. Equilibrium-governed processes create a second phase and rely on the components distributing differently between phases at equilibrium: distillation (volatility), absorption and stripping (solubility), extraction (distribution coefficient), crystallisation, adsorption at equilibrium, ion exchange. Their limit is the separation factor at equilibrium, and they are staged. Rate-governed processes rely on different speeds of transport through a medium under a driving force: membrane permeation (gas separation, RO, dialysis), electrophoresis, ultracentrifugation, kinetic adsorption on carbon molecular sieves, thermal diffusion.

Mechanical separations (filtration, settling, centrifugation, cyclones) separate phases, not molecules. Choice among methods depends on the property difference available, the feed condition and scale, product purity and value, energy use and safety. Distillation dominates for volatile liquids; membranes win for desalination and small gas plants; adsorption for trace removal and air separation at small scale; chromatography for high-value biomolecules and isomers.

  • Separations: 40–70 % of plant cost.
  • Energy-separating vs mass-separating agents.
  • Equilibrium-governed: distillation, extraction, absorption.
  • Rate-governed: membranes, electrophoresis, kinetic adsorption.

Example: Classify: (a) desalination by RO, (b) air separation by PSA on zeolite 5A, (c) N₂ from air by PSA on carbon molecular sieve, (d) ethanol–water distillation.

  1. (a) Rate-governed: solution–diffusion through the membrane.
  2. (b) Equilibrium: N₂ adsorbs more strongly than O₂ on zeolite.
  3. (c) Rate (kinetic): O₂ diffuses faster into CMS pores.
  4. (d) Equilibrium: relative volatility.

Answer: Rate, equilibrium, rate, equilibrium.

Separations account for 40–70 % of plant cost. They exploit a property difference using an energy-separating agent (heat, pressure, field) or a mass-separating agent (solvent, adsorbent, membrane). Equilibrium-governed processes (distillation, absorption, extraction, crystallisation, equilibrium adsorption) are limited by phase equilibrium and are staged. Rate-governed processes (membranes, electrophoresis, kinetic adsorption) use different transport rates. Mechanical separations split phases.

  • Separation factor α_ij = (x_i/x_j)₁/(x_i/x_j)₂.
  • Mass-separating agents must be recovered and recycled.
  • Membranes avoid phase change and save energy.
  • Distillation is robust but energy-hungry.
  • Selection: property difference, scale, purity, energy.

Key formulas

  • α_ij = (x_i/x_j)_phase1/(x_i/x_j)_phase2
  • stage-wise: N from equilibrium + balances

Practice questions

Which is a rate-governed separation?

  1. distillation
  2. gas permeation through a membrane
  3. crystallisation
  4. liquid extraction
Show the answer

Answer: B. gas permeation through a membrane Different permeation rates.

An energy-separating agent is used in

  1. extraction
  2. absorption
  3. distillation
  4. adsorption
Show the answer

Answer: C. distillation Heat creates the vapour phase.

Separations typically account for what share of plant cost?

  1. under 5 %
  2. 40–70 %
  3. 100 %
  4. about 1 %
Show the answer

Answer: B. 40–70 % Often the largest cost.

Nitrogen PSA using carbon molecular sieve works by

  1. equilibrium selectivity
  2. faster diffusion of O₂
  3. boiling point difference
  4. magnetic forces
Show the answer

Answer: B. faster diffusion of O₂ Kinetic separation.

A mass-separating agent

  1. needs no recovery
  2. must usually be regenerated
  3. is always a gas
  4. is heat
Show the answer

Answer: B. must usually be regenerated Solvents and adsorbents are recycled.

Theory questions

What factors govern the selection of a separation process? (5 marks)

Show the model answer

Feed conditions: flow rate (scale), composition, phase, temperature, pressure, and presence of solids or corrosive species. Product requirements: purity, recovery, state, value (high-value products justify costly methods such as chromatography). Property differences available: volatility, solubility, molecular size, charge, adsorption affinity, density. Separation factor and ease of separation: distillation is preferred if relative volatility exceeds about 1.05–1.1 and the components are stable; otherwise extraction, adsorption, membranes or crystallisation. Energy consumption and utilities; capital cost; need for and recovery of mass-separating agents; heat sensitivity of products (favouring membranes, vacuum or supercritical methods); safety and environmental constraints; maturity and reliability of the technology; possibility of hybrid processes.

Classify separation processes and distinguish between equilibrium-governed and rate-governed processes with examples. (10 marks)

Show the model answer

Separation by (1) phase creation using energy (distillation, evaporation, crystallisation, flash), (2) phase addition with a mass-separating agent (absorption, extraction, stripping), (3) a barrier (membranes: MF, UF, RO, gas permeation, pervaporation, dialysis), (4) a solid agent (adsorption, ion exchange, chromatography), (5) a force field or gradient (centrifugation, electrophoresis, thermal diffusion). Equilibrium-governed: the two phases approach equilibrium, and separation depends on the equilibrium distribution (relative volatility, K-values, isotherms); performance is improved by staging and reflux; examples distillation, extraction, equilibrium adsorption of N₂ on zeolite. Rate-governed: separation depends on different rates of transport through a medium; contact time and driving force matter; examples gas permeation (H₂ recovery), RO (water vs salt), kinetic PSA on carbon molecular sieve, dialysis. Some processes combine both (chromatography: equilibrium partition with dispersion).

Membrane processes and their applications

Topics: Microfiltration, ultrafiltration, nanofiltration, reverse osmosis, pervaporation, gas separation, dialysis and electrodialysis, liquid membranes, facilitated transport; concentration polarisation and fouling.

Pressure-driven membranes: from MF to RO

A membrane is a selective barrier. Pressure-driven processes are ranked by pore size: microfiltration (0.1–10 µm, 0.1–2 bar) removes bacteria and suspended solids; ultrafiltration (1–100 nm, 1–10 bar, defined by molecular-weight cut-off) retains proteins, colloids and emulsions; nanofiltration (~1 nm, 5–20 bar) passes monovalent ions but rejects divalent ions and small organics — water softening, dye removal; reverse osmosis (dense, 10–80 bar) retains almost all salts and is used for seawater and brackish-water desalination and ultrapure water.

Porous membranes (MF, UF) sieve particles, and flux follows Darcy/Hagen–Poiseuille: J = ΔP/(μR_m). Dense membranes (RO, gas separation, pervaporation) work by solution–diffusion: species dissolve in the polymer and diffuse across. For RO, water flux J_w = A(ΔP − Δπ) and salt flux J_s = B(C_m − C_p); rejection rises with pressure because water flux rises but salt flux does not. Osmotic pressure by van ’t Hoff: π = iCRT (seawater ≈ 27 bar).

Membranes are made of cellulose acetate, polyamide thin-film composites (RO, NF), polysulfone and PVDF (UF, MF) or ceramics. Modules: plate-and-frame, tubular (dirty feeds), spiral-wound (RO, NF; cheap, compact) and hollow fibre (gas separation, dialysers, very high area per volume). Cross-flow operation sweeps the surface and limits cake build-up; dead-end filtration is used for dilute feeds.

  • MF > UF > NF > RO in pore size, opposite in pressure.
  • Porous: sieving, J = ΔP/μR.
  • Dense: solution–diffusion, J_w = A(ΔP − Δπ).
  • Modules: spiral-wound, hollow-fibre, tubular, plate.

Example: Seawater (35 g/L NaCl) at 25 °C. Estimate its osmotic pressure and the minimum pressure for 50 % recovery.

  1. C = 35/58.44 = 0.599 mol/L = 599 mol/m³; i = 2.
  2. π = 2 × 599 × 8.314 × 298 = 2.97 × 10⁶ Pa ≈ 29.7 bar (real seawater ~27 bar).
  3. At 50 % recovery the brine is 70 g/L: π ≈ 59 bar; the pressure must exceed this at the outlet.

Answer: ≈ 30 bar feed, ≈ 60 bar at the brine end: plants run at 55–70 bar.

Other membrane processes, polarisation and fouling

Gas separation uses dense polymers (polyimide, cellulose acetate, PDMS) or hollow fibres: H₂ recovery from ammonia purge gas and refinery streams, N₂ from air, CO₂ removal from natural gas. Selectivity α = P_A/P_B; the pressure ratio across the membrane also limits enrichment. Pervaporation passes a liquid feed against a vacuum: one component dissolves and evaporates through the membrane — used to dehydrate ethanol and other solvents beyond the azeotrope. Dialysis uses concentration differences to move small solutes (urea in haemodialysis); electrodialysis uses stacks of alternating cation- and anion-exchange membranes and an electric field to move ions out of the diluate — brackish water, whey demineralisation.

Liquid membranes are thin organic films (supported in a porous film or as emulsions) containing a carrier. Facilitated transport uses a carrier that reacts reversibly with the solute on the feed side and releases it on the strip side, giving very high selectivity — metal-ion recovery with extractants, CO₂ with amines, O₂ with haemoglobin-like carriers.

Concentration polarisation: rejected solute accumulates at the membrane surface; at steady state C_m/C_b = exp(J/k) (film model). It raises osmotic pressure, increases solute leakage and promotes scaling. In UF of proteins, C_m can reach a gel concentration, after which J = k ln(C_g/C_b) independent of pressure. Fouling (cake, pore blocking, scaling, biofouling) lowers flux irreversibly; control by pretreatment, high cross-flow velocity, turbulence promoters, back-flushing and periodic chemical cleaning.

  • Gas separation: H₂ recovery, N₂ from air, CO₂ removal.
  • Pervaporation breaks azeotropes.
  • Electrodialysis: ion-exchange membrane stacks.
  • C_m/C_b = exp(J/k); gel limit J = k ln(C_g/C_b).

Example: A UF unit concentrates whey protein (C_b = 10 g/L, C_g = 250 g/L) with k = 2 × 10⁻⁵ m/s. What is the maximum flux, and what if k doubles by raising cross-flow velocity?

  1. J = k ln(C_g/C_b) = 2e-5 × ln 25 = 6.44 × 10⁻⁵ m/s = 232 L/m²·h.
  2. Doubling k: 464 L/m²·h.

Answer: Flux is limited by mass transfer, so more cross-flow, not more pressure, raises it.

MF (0.1–10 µm), UF (1–100 nm, MWCO), NF (~1 nm, rejects divalent ions) and RO (dense, salts) are pressure-driven. Porous membranes sieve (J = ΔP/μR); dense ones follow solution–diffusion: J_w = A(ΔP − Δπ), J_s = BΔC, π = iCRT. Gas separation (selectivity, pressure ratio), pervaporation (dehydration), dialysis and electrodialysis (ion-exchange membranes, I = zFQΔC/Nξ), liquid membranes and facilitated transport. Concentration polarisation C_m/C_b = exp(J/k); gel layer J = k ln(C_g/C_b); fouling and its control.

  • Recovery = permeate/feed.
  • Rejection R = 1 − C_p/C_f.
  • Spiral-wound for RO; hollow fibre for gas separation.
  • Thin-film composite polyamide for RO.
  • Cleaning restores flux after fouling.

Key formulas

  • J_w = A(ΔP − Δπ)
  • π = iCRT
  • C_m/C_b = exp(J/k)
  • J_lim = k ln(C_g/C_b)
  • I = zFQΔC/(Nξ)

Practice questions

Which membrane rejects monovalent salts almost completely?

  1. MF
  2. UF
  3. NF
  4. RO
Show the answer

Answer: D. RO Dense RO membranes.

In RO, water flux is proportional to

  1. ΔP
  2. ΔP − Δπ
  3. Δπ
  4. 1/ΔP
Show the answer

Answer: B. ΔP − Δπ Net driving pressure.

Pervaporation is commonly used to

  1. desalinate seawater
  2. dehydrate ethanol beyond the azeotrope
  3. remove bacteria
  4. separate isotopes
Show the answer

Answer: B. dehydrate ethanol beyond the azeotrope Water selectively permeates.

Gel-limited UF flux increases with

  1. transmembrane pressure
  2. mass-transfer coefficient
  3. bulk concentration
  4. gel concentration decreasing
Show the answer

Answer: B. mass-transfer coefficient J = k ln(C_g/C_b).

Electrodialysis uses

  1. porous ceramic tubes
  2. alternating cation and anion exchange membranes
  3. activated carbon
  4. liquid carriers
Show the answer

Answer: B. alternating cation and anion exchange membranes Ion-exchange stacks.

Theory questions

Explain the solution–diffusion model for reverse osmosis and derive expressions for water flux and salt rejection. (10 marks)

Show the model answer

The dense membrane has no pores; water and salt dissolve in the polymer and diffuse down their chemical-potential gradients. Water: the chemical potential difference is proportional to (ΔP − Δπ), so J_w = A(ΔP − Δπ), A = D_wC_wV_w/(RTl). Salt: pressure has negligible effect on its chemical potential, so J_s = B(C_m − C_p), B = D_sK_s/l. Permeate concentration C_p = J_s/J_w. Rejection R = 1 − C_p/C_m = J_w/(J_w + B) ⇒ R increases with pressure, approaching 1. Recovery raises the brine concentration and Δπ, reducing J_w. Concentration polarisation raises C_m above C_b (C_m/C_b = e^(J/k)), lowering rejection. Temperature raises both A and B.

What is concentration polarisation? Derive the film-model expression and explain how fouling is controlled. (10 marks)

Show the model answer

When solvent passes and solute is rejected, solute is convected to the surface at J·C and accumulates until back-diffusion balances it. Steady state in the boundary layer of thickness δ with complete rejection: JC − D dC/dx = 0 (flux of solute towards membrane = zero net) ⇒ dC/C = (J/D)dx ⇒ ln(C_m/C_b) = Jδ/D = J/k ⇒ C_m/C_b = exp(J/k). Effects: higher osmotic pressure at the wall, higher solute leakage, precipitation of sparingly soluble salts (CaSO₄, CaCO₃), gel formation in UF where J = k ln(C_g/C_b). Fouling: deposition of colloids, organics, biofilms and scale reduces flux over time. Control: feed pretreatment (coagulation, MF/UF before RO, antiscalants, pH adjustment, chlorination/dechlorination), higher cross-flow velocity and spacers (raise k), operating below critical flux, back-flushing, air scouring and periodic cleaning with acids, alkalis and enzymes, and choosing hydrophilic membranes.

Adsorption

Topics: Concepts and definitions; types of adsorbents, preparation and properties; adsorption isotherms and their importance; thermal swing, pressure swing and moving-bed adsorption.

Adsorbents and isotherms

Adsorption concentrates molecules from a fluid on the surface of a porous solid. Physical adsorption (van der Waals forces, ΔH ≈ 10–40 kJ/mol, reversible, multilayer possible) is used for separations; chemisorption (chemical bonds, 80–400 kJ/mol, monolayer, often irreversible) matters in catalysis. Adsorption is exothermic, so capacity falls with temperature and rises with pressure — the basis of thermal-swing and pressure-swing regeneration.

Commercial adsorbents: activated carbon (500–1500 m²/g, hydrophobic; made by carbonising coal, coconut shell or wood and activating with steam or CO₂ or with H₃PO₄/ZnCl₂) for organics, odours, solvent recovery and water treatment; silica gel (hydrophilic; drying gases and air); activated alumina (drying, fluoride removal); zeolites (crystalline aluminosilicates with uniform pores: 3A, 4A, 5A, 13X — molecular sieving, drying, N₂/O₂ separation, n-paraffin separation); carbon molecular sieves (kinetic N₂ generators); polymeric resins (antibiotics, phenols). Important properties: surface area (BET), pore size distribution, bulk density, crush strength, capacity and selectivity.

Isotherms give equilibrium loading q against concentration or pressure at constant temperature. Langmuir (monolayer of identical sites): q = q_mKC/(1 + KC); linear at low C, saturating at q_m. Freundlich (heterogeneous surface): q = K_FC^(1/n), n > 1 favourable. BET for multilayer gas adsorption, used to measure surface area. Favourable (concave) isotherms give sharp fixed-bed fronts; unfavourable ones spread out. Isotherms are fitted by linearised plots: C/q vs C (Langmuir), log q vs log C (Freundlich).

  • Physisorption reversible, ΔH 10–40 kJ/mol.
  • Carbon for organics, silica/alumina for water, zeolites for sieving.
  • Langmuir: q = q_mKC/(1 + KC).
  • Freundlich: q = K_FC^(1/n).

Example: Phenol on carbon follows Langmuir with q_m = 200 mg/g and K = 0.05 L/mg. How much carbon is needed in a single batch stage to reduce 1 m³ of 100 mg/L phenol to 10 mg/L?

  1. Equilibrium loading at C = 10: q = 200 × 0.05 × 10/(1 + 0.5) = 66.7 mg/g.
  2. Phenol removed = 1000 L × (100 − 10) = 90 000 mg.
  3. Carbon = 90 000/66.7 = 1350 g.

Answer: About 1.35 kg of carbon; two counter-current stages would need much less.

Fixed beds, TSA, PSA and moving beds

In a fixed bed, adsorbent near the inlet saturates first; a mass-transfer zone (MTZ) moves through the bed. The outlet concentration stays near zero until the MTZ reaches the exit (breakthrough time t_b), then rises along the breakthrough curve. The stoichiometric time t_s (where all capacity would be used if the front were sharp) is found from the area above the curve. Length of unused bed: LUB = (1 − t_b/t_s)L, which stays constant on scale-up for a constant-pattern front, so the design length = L needed for capacity + LUB.

Regeneration: thermal-swing adsorption (TSA) heats the bed (hot gas or steam) to desorb; slow (hours), high capacity — drying of natural gas, solvent recovery with steam. Pressure-swing adsorption (PSA) adsorbs at high pressure and desorbs by depressurising and purging (Skarstrom cycle: pressurise, adsorb, blowdown, purge); fast (minutes), suited to strongly held gases — H₂ purification (99.999 %), O₂ and N₂ from air, CO₂ removal. Vacuum swing (VSA) desorbs below atmospheric pressure.

Moving-bed and simulated moving-bed (SMB) adsorption contact solid and fluid counter-currently, using the adsorbent more efficiently. Real moving beds suffer attrition, so SMB moves the inlet and outlet ports of a fixed series of columns instead (UOP Sorbex: Parex for p-xylene, Molex for n-paraffins; sugar separation; chiral drugs).

  • Breakthrough when MTZ reaches the outlet.
  • LUB = (1 − t_b/t_s)L.
  • TSA slow and high capacity; PSA fast.
  • SMB moves ports, not solids.

Example: A 0.3 m test bed breaks through at 120 min, with stoichiometric time 160 min. Find the breakthrough time of a 1 m bed at the same velocity.

  1. LUB = (1 − 120/160) × 0.3 = 0.075 m.
  2. Capacity time scales with length: t_s(1 m) = 160 × 1/0.3 = 533 min.
  3. t_b = t_s(L − LUB)/L = 533 × 0.925 = 493 min.

Answer: About 493 min (8.2 h).

Physisorption (reversible, exothermic, 10–40 kJ/mol) vs chemisorption. Adsorbents: activated carbon, silica gel, activated alumina, zeolites (3A/4A/5A/13X), carbon molecular sieves, polymer resins; properties: area, pore size, capacity, selectivity, strength. Isotherms: Langmuir q = q_mKC/(1 + KC), Freundlich q = K_FC^(1/n), BET. Fixed beds: mass-transfer zone, breakthrough curve, LUB = (1 − t_b/t_s)L. Regeneration by TSA (heat), PSA/VSA (pressure; Skarstrom cycle), moving and simulated moving beds.

  • Capacity falls with temperature.
  • Favourable isotherms give self-sharpening fronts.
  • PSA gives 99.999 % hydrogen.
  • Zeolite 4A dries; 5A separates n-paraffins.
  • SMB: Parex process for p-xylene.

Key formulas

  • q = q_mKC/(1 + KC)
  • q = K_FC^(1/n)
  • LUB = (1 − t_b/t_s)L
  • t_s = q₀ρ_bL/(u C₀) (sharp front)

Practice questions

Physical adsorption is

  1. highly endothermic
  2. exothermic and reversible
  3. always monolayer
  4. irreversible
Show the answer

Answer: B. exothermic and reversible Van der Waals forces.

The Langmuir isotherm assumes

  1. multilayer adsorption
  2. a monolayer of identical sites
  3. heterogeneous surface
  4. no saturation
Show the answer

Answer: B. a monolayer of identical sites Monolayer.

PSA regenerates the bed by

  1. heating
  2. lowering pressure and purging
  3. washing with acid
  4. replacing the adsorbent
Show the answer

Answer: B. lowering pressure and purging Pressure swing.

Which adsorbent is hydrophobic?

  1. silica gel
  2. activated carbon
  3. activated alumina
  4. zeolite 13X
Show the answer

Answer: B. activated carbon Carbon prefers organics.

LUB is

  1. bed length at saturation
  2. the unused part of the bed at breakthrough
  3. pellet diameter
  4. column diameter
Show the answer

Answer: B. the unused part of the bed at breakthrough Mass-transfer zone allowance.

Theory questions

Discuss different types of adsorption isotherms and their importance in design. (10 marks)

Show the model answer

Isotherms relate equilibrium loading to fluid concentration (or pressure) at constant temperature. Brunauer classified five types: I (Langmuir, microporous solids, monolayer), II (multilayer on non-porous solids, BET), III (weak adsorbate–surface interaction), IV and V (capillary condensation in mesopores with hysteresis). Equations: linear (Henry) q = KC at low concentration; Langmuir q = q_mKC/(1 + KC); Freundlich q = K_FC^(1/n); BET for multilayer; Toth and Sips for heterogeneous surfaces. Importance: (i) capacity at the feed concentration fixes the amount of adsorbent and bed size; (ii) the shape decides the front: favourable (concave) isotherms give constant-pattern, self-sharpening fronts, unfavourable give spreading fronts; (iii) the temperature dependence (isosteric heat) sets TSA conditions; the pressure dependence sets PSA cycles; (iv) selectivity between components from multicomponent isotherms (extended Langmuir, IAST).

Explain thermal-swing and pressure-swing adsorption with a neat sketch of a two-bed PSA cycle. (10 marks)

Show the model answer

TSA: adsorb at low temperature, regenerate by heating the bed with hot purge gas or steam (capacity falls with T), then cool. Heating and cooling large beds take hours, so TSA suits strongly adsorbed species at low concentration: drying of natural gas and air before cryogenic separation, solvent recovery from air with steam regeneration of carbon. PSA: adsorb at high pressure, desorb by reducing pressure (capacity falls with partial pressure) and purge with part of the product. Skarstrom two-bed cycle: bed 1 pressurises with feed and adsorbs, delivering product, while bed 2 blows down to low pressure and is purged counter-currently with some product; then they switch. Steps take minutes; cycles are short, so the bed is small. Applications: H₂ purification from reformer gas (multi-bed, 99.99+ %), O₂ from air on LiX/13X zeolite (VSA), N₂ from air on carbon molecular sieve (kinetic), CO₂ capture. PSA needs weakly or moderately held impurities; TSA handles strongly held ones.

Chromatographic separations

Topics: Principles of chromatographic separation; criteria for effective separation; methodology and process design.

How chromatography separates and how to scale it

A pulse of mixture is carried by a mobile phase through a column packed with a stationary phase. Each solute partitions between the phases; one that spends more time in the stationary phase moves more slowly. Retention time t_R = t₀(1 + k′), where k′ = (t_R − t₀)/t₀ is the retention factor and t₀ the dead time. The selectivity α = k′₂/k′₁ measures how differently two solutes are held.

Peaks also spread by axial dispersion, eddy diffusion and finite mass-transfer rates. Column efficiency is measured by the number of theoretical plates N = 16(t_R/w)² = 5.54(t_R/w_½)² and the plate height HETP = L/N. The van Deemter equation HETP = A + B/u + Cu (eddy diffusion, axial molecular diffusion, mass-transfer resistance) shows an optimum velocity. Resolution R_s = 2(t_R2 − t_R1)/(w₁ + w₂); R_s = 1.5 is baseline separation. Purnell: R_s = (√N/4)((α − 1)/α)(k′/(1 + k′)) — selectivity is the most powerful lever.

Process (preparative) chromatography aims at throughput, not analysis: elution with overloaded columns, displacement chromatography, and continuous simulated moving bed. Types: adsorption, partition, ion exchange (proteins, sugars), size exclusion, affinity (antibodies on protein A), reversed-phase HPLC. Design steps: choose phase and mobile phase from analytical runs; find k′ and α; set N for the required R_s; choose particle size (smaller gives lower HETP but higher pressure drop); scale up at constant bed height and linear velocity by increasing diameter; check loading and cycle time.

  • k′ = (t_R − t₀)/t₀; α = k′₂/k′₁.
  • N = 16(t_R/w)²; HETP = L/N.
  • R_s = 2Δt_R/(w₁ + w₂); 1.5 = baseline.
  • Scale up: same bed height and velocity, larger diameter.

Example: Two peaks: t_R = 8.0 and 9.0 min, base widths 0.50 and 0.55 min, dead time 1.0 min. Find R_s, N and α. What length is needed for R_s = 1.5 if the column is 25 cm?

  1. R_s = 2 × 1.0/(1.05) = 1.90.
  2. N = 16 × (9.0/0.55)² = 4285.
  3. k′₁ = 7, k′₂ = 8 ⇒ α = 1.143.
  4. R_s ∝ √L: L = 25 × (1.5/1.9)² = 15.6 cm.

Answer: R_s = 1.9 with N ≈ 4300; 16 cm would just achieve baseline separation.

Chromatography: solutes partition between mobile and stationary phases and move at different speeds. Retention t_R = t₀(1 + k′); selectivity α = k′₂/k′₁. Band broadening: N = 16(t_R/w)², HETP = L/N, van Deemter HETP = A + B/u + Cu. Resolution R_s = 2Δt_R/(w₁ + w₂) = (√N/4)((α − 1)/α)(k′/(1 + k′)); R_s ≥ 1.5 for baseline. Types: adsorption, partition, ion exchange, size exclusion, affinity. Preparative scale-up at constant bed height and velocity; SMB for continuous separation.

  • Resolution grows with √N.
  • Selectivity is the strongest lever.
  • Optimum velocity minimises HETP.
  • Affinity chromatography purifies antibodies.
  • SMB = continuous counter-current chromatography.

Key formulas

  • N = 16(t_R/w)²
  • R_s = 2(t_R2 − t_R1)/(w₁ + w₂)
  • R_s = (√N/4)((α − 1)/α)(k′/(1 + k′))
  • HETP = A + B/u + Cu

Practice questions

Baseline resolution corresponds to R_s of about

  1. 0.5
  2. 1.0
  3. 1.5
  4. 5
Show the answer

Answer: C. 1.5 99.7 % separation.

Doubling column length multiplies R_s by

  1. 2
  2. √2
  3. 4
  4. 1
Show the answer

Answer: B. √2 R_s ∝ √N ∝ √L.

In the van Deemter equation the C term represents

  1. eddy diffusion
  2. axial diffusion
  3. mass-transfer resistance
  4. detector noise
Show the answer

Answer: C. mass-transfer resistance Grows with velocity.

Affinity chromatography relies on

  1. size differences
  2. specific biological binding
  3. boiling point
  4. charge only
Show the answer

Answer: B. specific biological binding E.g. protein A–IgG.

Preparative scale-up usually keeps constant

  1. column diameter
  2. bed height and linear velocity
  3. flow rate
  4. particle number
Show the answer

Answer: B. bed height and linear velocity Same chromatography, more area.

Theory questions

Explain the van Deemter equation and its significance. (5 marks)

Show the model answer

The plate height in a packed column is H = A + B/u + Cu, where u is the mobile-phase velocity. A (eddy diffusion) arises from multiple flow paths of different lengths through the packing; it depends on particle size and packing quality and is independent of velocity. B/u (longitudinal molecular diffusion) represents axial spreading of the band by diffusion along the column, important at low velocity and in gases. Cu (resistance to mass transfer) arises from finite rates of transfer between mobile and stationary phases, important at high velocity. The curve has a minimum H at the optimum velocity u_opt = √(B/C), giving maximum efficiency. Smaller particles reduce A and C, giving flatter curves and allowing faster separations (basis of HPLC and UHPLC), at the cost of higher pressure drop.

Explain the principle of chromatographic separation and define the criteria for effective separation. (10 marks)

Show the model answer

Mixture injected as a narrow band into a mobile phase flowing through a stationary phase; components distribute between phases with equilibrium constants K_i; a component’s velocity is u/(1 + k′_i), with k′ = K(V_s/V_m); differences in k′ produce different retention times. Effective separation needs: (1) retention: k′ between about 1 and 10 (too low, no separation; too high, long runs and dilute peaks); (2) selectivity: α = k′₂/k′₁ > 1, from the choice of stationary and mobile phases; (3) efficiency: large N (small HETP) from small uniform particles, good packing, optimum velocity (van Deemter); (4) resolution R_s = 2Δt_R/(w₁ + w₂) ≥ 1.5, related by Purnell’s equation R_s = (√N/4)((α − 1)/α)(k′/(1 + k′)). Also capacity and throughput for preparative work, low pressure drop, and recovery of solvent.

Special separation processes

Topics: Supercritical fluid extraction; microwave extraction; reactive distillation; biofiltration.

Supercritical, microwave, reactive and biological separations

A supercritical fluid is above its critical temperature and pressure; CO₂ (T_c = 31.1 °C, P_c = 73.8 bar) is non-toxic, cheap and non-flammable. Near the critical point its density — and hence solvent power — changes greatly with small changes in pressure or temperature, while its viscosity stays gas-like and its diffusivity is 10–100 times that of liquids. Extraction at high pressure, then release of pressure in a separator, precipitates the solute and recycles CO₂ without solvent residues. Applications: decaffeination of coffee and tea, hop extracts, essential oils and flavours, fat removal, pharmaceutical purification, cleaning of precision parts. Entrainers (ethanol) add polarity.

Microwave-assisted extraction heats moisture inside plant cells volumetrically and quickly; the pressure ruptures the cells and releases the contents into the solvent. It cuts extraction time from hours to minutes and reduces solvent use; used for essential oils, polyphenols, pigments and analytical sample preparation. Solvent must absorb microwaves (polar) or the matrix must contain water.

Reactive distillation combines reaction and separation in one column: products are removed as they form, driving equilibrium-limited reactions to completion, reaction heat helps vaporisation, and azeotropes can be avoided. Classic example: methyl acetate (Eastman), replacing a plant of about 11 units with one column; also MTBE, ETBE and esterifications. Constraints: reaction and distillation temperatures must overlap; catalyst packings are needed. Biofiltration passes polluted air through a moist bed of compost, peat or wood chips on which micro-organisms oxidise odours and VOCs (H₂S, mercaptans, solvents) to CO₂, water and biomass; cheap and effective for dilute, biodegradable streams; design by empty-bed residence time (15–60 s) and elimination capacity.

  • scCO₂: T_c 31 °C, P_c 73.8 bar; tunable density.
  • Microwave: volumetric heating ruptures cells.
  • Reactive distillation: Eastman methyl acetate.
  • Biofilter: microbes oxidise odours; EBRT 15–60 s.

Example: A biofilter must treat 5000 m³/h of air with an empty-bed residence time of 40 s and bed depth 1 m. Find the bed area and volume.

  1. V = Q × EBRT = (5000/3600) × 40 = 55.6 m³.
  2. Area = V/depth = 55.6 m².

Answer: About 56 m³ of media over 56 m².

Supercritical fluid extraction (scCO₂, T_c 31.1 °C, P_c 73.8 bar): tunable density, gas-like transport, solvent-free products; decaffeination, essential oils, hops. Microwave extraction: volumetric heating ruptures cells; fast, less solvent. Reactive distillation: reaction + separation in one column, shifts equilibrium, uses reaction heat (methyl acetate, MTBE). Biofiltration: micro-organisms on moist media oxidise odours and VOCs; sized by empty-bed residence time and elimination capacity.

  • Pressure reduction precipitates the extract.
  • Entrainers raise scCO₂ polarity.
  • Reactive distillation needs overlapping T windows.
  • Biofilters need moisture, nutrients, pH control.
  • Microwave needs polar solvent or moist matrix.

Key formulas

  • EBRT = V_bed/Q
  • Elimination capacity = Q(C_in − C_out)/V

Practice questions

The critical temperature of CO₂ is about

  1. −78 °C
  2. 31 °C
  3. 100 °C
  4. 374 °C
Show the answer

Answer: B. 31 °C 31.1 °C, 73.8 bar.

A major advantage of scCO₂ extraction is

  1. no solvent residue in the product
  2. very low pressure
  3. high viscosity
  4. flammability
Show the answer

Answer: A. no solvent residue in the product CO₂ vents off.

Reactive distillation is most useful for

  1. irreversible gas reactions
  2. equilibrium-limited liquid reactions
  3. solid reactions
  4. nuclear reactions
Show the answer

Answer: B. equilibrium-limited liquid reactions Removing products shifts equilibrium.

Biofilters remove pollutants by

  1. adsorption only
  2. microbial oxidation
  3. condensation
  4. membrane sieving
Show the answer

Answer: B. microbial oxidation Biodegradation.

Microwave extraction speeds up because

  1. it cools the matrix
  2. internal heating ruptures cells
  3. it raises pressure to 100 bar
  4. it uses UV light
Show the answer

Answer: B. internal heating ruptures cells Volumetric heating.

Theory questions

Describe supercritical fluid extraction with a flow diagram and give applications. (10 marks)

Show the model answer

Above T_c and P_c a fluid is neither gas nor liquid; its density (and dissolving power) can be adjusted by pressure and temperature, while viscosity is low and diffusivity high, so extraction is fast. CO₂ is preferred: T_c = 31.1 °C (gentle on heat-sensitive materials), P_c = 73.8 bar, non-toxic, non-flammable, cheap, leaves no residue. Process: liquid CO₂ is pumped to 150–350 bar and heated above T_c; it passes through the extractor packed with solids (coffee beans, hops, seeds); the loaded fluid is expanded through a valve into a separator at lower pressure where solvent power falls and the extract precipitates; CO₂ is condensed and recycled. Modifiers (ethanol, methanol) increase polarity. Applications: decaffeination, hop extract for brewing, essential oils, spices, omega-3 oils, nicotine removal, pharmaceutical crystallisation (RESS, SAS), textile dyeing, dry cleaning, soil remediation. Drawbacks: high-pressure equipment cost, batch operation for solids.

What is reactive distillation? Discuss its advantages and limitations with an example. (5 marks)

Show the model answer

Reaction and distillation in one column, usually with a solid catalyst in structured packing (catalytic distillation). Advantages: continuous removal of products drives equilibrium-limited reactions (esterification, etherification) to high conversion; heat of reaction vaporises liquid, saving energy; azeotropes can be bypassed; fewer units and lower capital cost; better selectivity by removing intermediates. Example: Eastman’s methyl acetate process — methanol and acetic acid fed at different points react in the middle section; methyl acetate leaves overhead, water at the bottom; one column replaced a reactor plus about eight distillation and extraction units. Also MTBE/ETBE from isobutylene and alcohols. Limitations: reaction and separation must work in the same temperature and pressure range; catalyst life and replacement; complex design and control; not suitable when reactants are much more volatile than products in the wrong way.

Notes, examples and questions were written for this studio; the prescribed books are listed in the studio for reference.

Frequently asked questions

What is the Advanced Separation Processes Studio?

Membranes, adsorbents and beyond. Equilibrium and rate-governed separations, MF, UF, NF, RO, pervaporation, gas separation and electrodialysis, polarisation and fouling, adsorbents, isotherms, PSA and TSA, chromatography, supercritical extraction, reactive distillation and biofiltration from the ASTU CHE1818PE41 syllabus (2018-19 scheme).

What can I do in the Advanced Separation Processes Studio?

Reverse osmosis, polarisation and gas permeation; Adsorption isotherms and fixed beds; Chromatographic resolution; Supercritical extraction and reactive distillation.

Is the Advanced Separation Processes Studio free?

Yes. It is free and runs in your browser with no sign-up.

Open the interactive studio →