IITIAN1997 STUDIO

Chemical Reaction Engineering-II

Selectivity, catalysts and real reactors.

Open the interactive studio →

Parallel and series reactions, adiabatic and optimum temperature progression, catalytic rate laws and packed beds, Thiele modulus and effectiveness factor, catalyst deactivation, RTD, tanks-in-series and dispersion models from the ASTU CHE241502 syllabus.

What is inside

  • Parallel and series reactions: maximise the product
  • Adiabatic operation and optimum temperature
  • LHHW rate laws, effectiveness factor, deactivation
  • RTD, segregation, tanks-in-series and dispersion

Syllabus: ASTU CHE241502 Chemical Reaction Engineering-II, semester 5, NEP 2020 (2024-25 onwards). Official syllabus PDF.

Design for multiple reactions

Topics: Introduction to multiple reactions; qualitative discussion of product distribution for parallel, series and series–parallel reactions; quantitative treatment; the Denbigh reactions and special cases; numericals.

Parallel reactions: concentration level decides the product split

With more than one reaction, reactor size is no longer the only concern: the distribution of products matters, because unwanted products waste feed and need separation. For parallel reactions A → R (desired, rate k₁C_A^a₁) and A → S (k₂C_A^a₂), the ratio of rates is r_R/r_S = (k₁/k₂)C_A^(a₁ − a₂). If a₁ > a₂, high C_A favours R: use a batch or plug-flow reactor, no inert dilution, high pressure for gases and avoid recycle. If a₁ < a₂, low C_A favours R: use a CSTR, dilute the feed or recycle product. If the orders are equal, the concentration level makes no difference; temperature (through the activation energies) or a selective catalyst must be used. When two reactants are involved (A + B), contacting patterns that keep each high or low (adding B slowly to a batch of A, side-feeding along a PFR) control selectivity.

Quantitatively, the instantaneous fractional yield φ = dC_R/(−dC_A) = r_R/(−r_A) changes with C_A. The overall fractional yield Φ = C_Rf/(C_A0 − C_Af) is the average of φ over the concentrations seen: in a PFR Φ = (1/(C_A0 − C_Af))∫φ dC_A; in a CSTR Φ = φ evaluated at the exit concentration. Plotting φ against C_A shows the area (PFR) or rectangle (CSTR) that gives C_R; for a falling φ curve the best scheme is a CSTR followed by a PFR.

  • r_R/r_S = (k₁/k₂)C_A^(a₁ − a₂).
  • a₁ > a₂: high C_A (PFR); a₁ < a₂: low C_A (CSTR).
  • φ = dC_R/(−dC_A); Φ_PFR = mean of φ; Φ_CSTR = φ(C_Af).
  • Equal orders: use temperature or catalyst.

Example: A → R (r = C_A²) and A → S (r = C_A), C_A0 = 1 mol/L, 90 % conversion. Compare C_R from a PFR and a CSTR.

  1. φ = C_A²/(C_A² + C_A) = C_A/(1 + C_A).
  2. PFR: C_R = ∫₀.₁¹ C/(1 + C) dC = [C − ln(1 + C)]₀.₁¹ = (1 − 0.6931) − (0.1 − 0.0953) = 0.302.
  3. CSTR: φ(0.1) = 0.1/1.1 = 0.0909 ⇒ C_R = 0.0909 × 0.9 = 0.082.

Answer: PFR 0.302 mol/L vs CSTR 0.082 mol/L: the higher-order desired reaction needs high concentration.

Series and series–parallel reactions; the Denbigh scheme

For reactions in series A → R → S (first order, k₁ and k₂), the intermediate R first builds up and then is destroyed. Its maximum depends on k₂/k₁ and on mixing. In a PFR or batch reactor, C_R/C_A0 = k₁(e^(−k₁τ) − e^(−k₂τ))/(k₂ − k₁), with a maximum (k₁/k₂)^(k₂/(k₂ − k₁)) at τ_opt = ln(k₂/k₁)/(k₂ − k₁). In a CSTR, C_R/C_A0 = k₁τ/((1 + k₁τ)(1 + k₂τ)), maximum 1/(√(k₂/k₁) + 1)² at τ_opt = 1/√(k₁k₂). Plug flow always gives more intermediate, because mixing fluid of different ages lets R that is already formed meet more reaction time.

Series–parallel reactions such as A + B → R, R + B → S (successive chlorination, nitration, ethoxylation) behave as series reactions with respect to A, R, S and as parallel reactions with respect to B. The product distribution depends on how much B is added and how A and B are contacted: adding B slowly to all of A, without mixing back the products, maximises R. The Denbigh scheme A → R → S with side reactions A → T and R → U combines both: R is maximised by keeping the ratio of desired to side rates high (concentration level from orders, temperature from activation energies) and stopping at the optimum time.

  • PFR: τ_opt = ln(k₂/k₁)/(k₂ − k₁), C_R,max = C_A0(k₁/k₂)^(k₂/(k₂ − k₁)).
  • CSTR: τ_opt = 1/√(k₁k₂), C_R,max = C_A0/(√(k₂/k₁) + 1)².
  • Mixing lowers the intermediate yield.
  • Series–parallel: series in A, parallel in B.

Example: A → R → S with k₁ = 0.5 min⁻¹, k₂ = 0.25 min⁻¹, C_A0 = 1 mol/L. Find the maximum C_R in a PFR and in a CSTR.

  1. PFR: τ_opt = ln(0.5)/(0.25 − 0.5) = 2.77 min; C_R,max = (0.5/0.25)^(0.25/(−0.25)) = 2^(−1) = 0.5 mol/L.
  2. CSTR: τ_opt = 1/√(0.125) = 2.83 min; C_R,max = 1/(√0.5 + 1)² = 1/2.914 = 0.343 mol/L.

Answer: PFR 0.50 mol/L at 2.77 min; CSTR 0.34 mol/L at 2.83 min.

Parallel reactions: concentration level controls the split (high C_A for the higher-order desired reaction, low for the lower-order); instantaneous φ and overall Φ; PFR vs CSTR vs combinations; contacting patterns for two reactants. Series reactions: intermediate has a maximum at an optimum time; plug flow beats mixed flow. Series–parallel and Denbigh reactions combine both behaviours; temperature acts through activation energies.

  • Higher activation energy reactions are favoured by higher temperature.
  • Recycle and dilution lower C_A.
  • For series reactions avoid back-mixing.
  • Φ_PFR uses the area under the φ–C_A curve.

Key formulas

  • φ = k₁C_A^a₁/(k₁C_A^a₁ + k₂C_A^a₂)
  • C_R,PFR = ∫φ dC_A
  • C_R,max/C_A0 = (k₁/k₂)^(k₂/(k₂ − k₁))
  • τ_opt,CSTR = 1/√(k₁k₂)

Practice questions

For parallel reactions where the desired reaction has the higher order, prefer

  1. a CSTR
  2. a PFR
  3. dilution
  4. recycle
Show the answer

Answer: B. a PFR High C_A.

For A → R → S, the maximum C_R is higher in

  1. a CSTR
  2. a PFR
  3. both equal
  4. neither
Show the answer

Answer: B. a PFR No mixing of ages.

If the desired and undesired parallel reactions have equal orders, selectivity can be changed by

  1. concentration
  2. temperature or catalyst
  3. reactor volume
  4. pressure only
Show the answer

Answer: B. temperature or catalyst Concentration has no effect.

The instantaneous fractional yield is

  1. C_R/C_A0
  2. dC_R/(−dC_A)
  3. k₁/k₂
  4. X_A
Show the answer

Answer: B. dC_R/(−dC_A) Definition.

For a CSTR, the optimum τ for A → R → S is

  1. ln(k₂/k₁)/(k₂ − k₁)
  2. 1/√(k₁k₂)
  3. 1/k₁
  4. k₁/k₂
Show the answer

Answer: B. 1/√(k₁k₂) Mixed flow.

Theory questions

Discuss qualitatively how concentration level and reactor type affect product distribution in parallel reactions. (5 marks)

Show the model answer

For A → R (a₁) and A → S (a₂): r_R/r_S = (k₁/k₂)C_A^(a₁ − a₂). a₁ > a₂: keep C_A high — batch or PFR, no inerts, high pressure, avoid recycle; a₁ < a₂: keep C_A low — CSTR, inert dilution, low pressure, recycle; a₁ = a₂: reactor type does not matter, change temperature (if E₁ ≠ E₂) or catalyst. With two reactants A + B, choose contacting so each is high or low as required: both high — batch or PFR with both fed together; both low — CSTR or slow addition of both; one high, one low — add the low one gradually to a batch of the other or side-feed along a PFR.

Derive the maximum concentration of the intermediate for A → R → S in a PFR. (6 marks)

Show the model answer

dC_A/dτ = −k₁C_A ⇒ C_A = C_A0e^(−k₁τ). dC_R/dτ = k₁C_A − k₂C_R: linear first-order ODE; with C_R(0) = 0: C_R = C_A0k₁(e^(−k₁τ) − e^(−k₂τ))/(k₂ − k₁). Set dC_R/dτ = 0: k₁e^(−k₁τ) = k₂e^(−k₂τ) ⇒ τ_opt = ln(k₂/k₁)/(k₂ − k₁). Substituting: C_R,max/C_A0 = (k₁/k₂)^(k₂/(k₂ − k₁)). If k₁ = k₂ = k: τ_opt = 1/k and C_R,max = C_A0/e.

Temperature and pressure effects

Topics: Single reactions: equilibrium constants from thermodynamics, optimum temperature progression, adiabatic operations, non-adiabatic operations; multiple reactions: product distribution and temperature; numericals.

Equilibrium, adiabatic and non-adiabatic operation

Thermodynamics fixes how far a reaction can go. K = exp(−ΔG°/RT) at the standard temperature, and the van ’t Hoff equation d ln K/dT = ΔH_r/RT² carries it to other temperatures. For A ⇌ R (first order both ways) the equilibrium conversion is X_e = K/(1 + K). For exothermic reactions X_e falls as temperature rises; for endothermic reactions it rises. Pressure matters only for gas reactions with a change in moles (and inerts act like lower pressure).

In an adiabatic reactor all the heat of reaction stays in the stream: an energy balance gives a straight line on the conversion–temperature chart, X = c_p(T − T₀)/(−ΔH_r). For an exothermic reversible reaction this line runs into the equilibrium curve, which limits the conversion of a single adiabatic stage; higher conversion needs interstage cooling (staged adiabatic beds), cold-shot injection of feed, or a lower inlet temperature (at the cost of rate). For endothermic reactions the temperature falls as the reaction proceeds, so reheating between stages is used. Non-adiabatic reactors (jacketed vessels, multitubular reactors with coolant) follow a curved path set by the heat removed.

For multiple reactions, temperature changes the product distribution through the activation energies: raising T favours the reaction with the larger E. For parallel reactions with E₁ > E₂ use a high temperature; for E₁ < E₂ a low one. For series reactions A → R → S with E₁ > E₂, start hot (fast first step) and the optimum is to cool as R accumulates.

  • K = exp(−ΔG°/RT); van ’t Hoff for other T.
  • A ⇌ R: X_e = K/(1 + K).
  • Adiabatic line: X = c_p(T − T₀)/(−ΔH_r).
  • Higher T favours the reaction with higher E.

Example: For A ⇌ R, ΔH_r = −75 300 J/mol (constant) and K = 300 at 25 °C. Find K and X_e at 65 °C.

  1. ln K = ln 300 + (75 300/8.314)(1/338.15 − 1/298.15).
  2. = 5.704 + 9057 × (−0.0003967) = 5.704 − 3.593 = 2.111 ⇒ K = 8.26.
  3. X_e = 8.26/9.26 = 0.892.

Answer: K ≈ 8.3 and X_e ≈ 0.89 at 65 °C (0.997 at 25 °C).

Example: The same reaction is fed adiabatically at 30 °C with c_p = 250 J/(mol A·K). What is the most conversion one stage can give?

  1. Adiabatic line: X = 250(T − 30)/75 300 = (T − 30)/301.
  2. Find where it meets X_e(T): trial gives T ≈ 110 °C, X ≈ 0.265.

Answer: About 26 % at 110 °C: interstage cooling is needed for more.

K from ΔG°, temperature dependence by van ’t Hoff; X_e = K/(1 + K) for A ⇌ R; exothermic reactions lose equilibrium conversion at high T. Adiabatic energy balance gives a straight line X = c_p(T − T₀)/(−ΔH_r); it is limited by the equilibrium curve; staged beds with intercooling or cold shot; non-adiabatic reactors follow curved paths. Temperature favours the reaction with the larger activation energy.

  • Adiabatic temperature rise = −ΔH_r/c_p.
  • Inerts lower the adiabatic temperature rise.
  • Endothermic: reheat between stages.
  • SO₂ and NH₃ converters are staged adiabatic beds.

Key formulas

  • ln(K₂/K₁) = −(ΔH_r/R)(1/T₂ − 1/T₁)
  • X_e = K/(1 + K)
  • X = c_p(T − T₀)/(−ΔH_r)
  • ΔT_ad = −ΔH_r/c_p

Practice questions

For an exothermic reversible reaction, raising temperature

  1. raises equilibrium conversion
  2. lowers equilibrium conversion
  3. has no effect
  4. makes K negative
Show the answer

Answer: B. lowers equilibrium conversion van ’t Hoff.

On an X–T chart the adiabatic operating line is

  1. a curve
  2. a straight line
  3. vertical
  4. horizontal
Show the answer

Answer: B. a straight line Energy balance.

Adding inert to an adiabatic feed

  1. raises the adiabatic temperature rise
  2. lowers it
  3. has no effect
  4. reverses the reaction
Show the answer

Answer: B. lowers it Larger c_p per mol A.

For parallel reactions with E₁ (desired) > E₂, use

  1. low temperature
  2. high temperature
  3. any temperature
  4. a catalyst only
Show the answer

Answer: B. high temperature Higher E favoured by T.

For A ⇌ R with K = 3, the equilibrium conversion is

  1. 0.25
  2. 0.5
  3. 0.75
  4. 0.9
Show the answer

Answer: C. 0.75 3/4.

Theory questions

Explain why a single adiabatic bed cannot give high conversion for an exothermic reversible reaction and how staged reactors overcome this. (5 marks)

Show the model answer

Adiabatic: the heat released raises the temperature along X = c_p(T − T₀)/(−ΔH_r). For exothermic reversible reactions X_e decreases with T, so the operating line meets the equilibrium curve at a modest conversion and the rate falls to zero there. Staging: stop each bed short of equilibrium, cool the stream (heat exchanger or cold-shot injection of cold feed), and react again in the next bed; the path zig-zags along the equilibrium curve towards higher conversion, approximating the optimum temperature progression (as in SO₂ converters with 3–4 beds and in ammonia converters).

How does temperature affect product distribution in multiple reactions? (4 marks)

Show the model answer

For competing reactions the ratio of rate constants k₁/k₂ = (A₁/A₂)exp(−(E₁ − E₂)/RT). If E_desired > E_undesired, high T increases the ratio (use high T, limited by equilibrium, materials, side reactions). If E_desired < E_undesired, low T is better (at the cost of a larger reactor). For series reactions A → R → S: if E₁ > E₂, start hot and cool as R builds (falling temperature progression); if E₁ < E₂, use as low a temperature as possible.

Choosing the right kind of reactor

Topics: Optimum operation of reactors; temperature progression for multiple reactions.

Optimum temperature progression and reactor choice

For a single reversible exothermic reaction, the rate at a given conversion is −r_A = C_A0[k₁(1 − X) − k₂X]. Raising T speeds both directions, but the reverse more (E₂ > E₁), so for each conversion there is a temperature giving the maximum rate. Setting ∂r/∂T = 0 gives E₁k₁(1 − X) = E₂k₂X, so T_opt = (E₂ − E₁)/(R ln[E₂k₂₀X/(E₁k₁₀(1 − X))]). The locus of T_opt on the X–T chart lies just below the equilibrium curve; following it from inlet to outlet (hot at first, cooler as conversion rises) gives the smallest reactor. For irreversible and endothermic reactions the optimum is simply the highest allowable temperature.

Choosing the reactor: for a single reaction the size is minimised by plug flow for positive orders, with a CSTR then a PFR for autocatalytic reactions; for multiple reactions the product distribution decides (PFR for series intermediates and high-order desired reactions, CSTR for low-order desired reactions). Temperature progression is achieved with staged adiabatic beds with interstage cooling, multitubular reactors with coolant (near-isothermal), cold-shot injection, or recycle of cooled product. Practical choice also weighs heat removal, safety (runaway), catalyst handling, mixing and cost.

  • T_opt from E₁k₁(1 − X) = E₂k₂X.
  • Optimum progression: hot inlet, cooler outlet.
  • Irreversible or endothermic: as hot as allowed.
  • Staged beds, multitubular, cold shot, recycle.

Example: For A ⇌ R with k₁ = exp(17.34 − 48 900/RT) and k₂ = exp(42.04 − 124 200/RT) (min⁻¹), find T_opt at X = 0.5 and X = 0.8.

  1. ln(E₂/E₁) = ln(124 200/48 900) = 0.932; ln k₂₀ − ln k₁₀ = 24.70.
  2. X = 0.5: denominator R(0.932 + 24.70 + 0) = 8.314 × 25.63 = 213.1 ⇒ T = 75 300/213.1 = 353 K (80 °C).
  3. X = 0.8: add ln(0.8/0.2) = 1.386 ⇒ 8.314 × 27.02 = 224.6 ⇒ T = 335 K (62 °C).

Answer: 80 °C at X = 0.5, falling to 62 °C at X = 0.8.

For reversible exothermic reactions the rate at fixed conversion has a maximum in temperature: T_opt from E₁k₁(1 − X) = E₂k₂X; the optimum progression runs from hot to cooler. For irreversible or endothermic reactions use the highest allowable T. Reactor choice combines size (PFR, CSTR, combinations), product distribution, heat removal, safety and cost; progression is approximated by staged adiabatic beds, multitubular reactors, cold shot or recycle.

  • The T_opt line lies below the equilibrium curve.
  • A CSTR runs at a single temperature: poor for following T_opt.
  • Multitubular reactors suit highly exothermic reactions.
  • Runaway risk limits the inlet temperature.

Key formulas

  • T_opt = (E₂ − E₁)/(R ln[E₂k₂₀X/(E₁k₁₀(1 − X))])
  • −r_A = C_A0[k₁(1 − X) − k₂X]

Practice questions

For a reversible exothermic reaction, the optimum temperature progression

  1. rises along the reactor
  2. falls along the reactor
  3. is constant
  4. is random
Show the answer

Answer: B. falls along the reactor Hot first, cooler later.

For an irreversible endothermic reaction the best temperature is

  1. the lowest possible
  2. the highest allowable
  3. ambient
  4. the T_opt curve
Show the answer

Answer: B. the highest allowable Rate only rises with T.

Cold-shot cooling means

  1. cooling with brine
  2. injecting cold feed between beds
  3. vacuum cooling
  4. evaporative cooling
Show the answer

Answer: B. injecting cold feed between beds Direct mixing.

The optimum temperature at a given X maximises the

  1. equilibrium constant
  2. net reaction rate
  3. heat of reaction
  4. pressure
Show the answer

Answer: B. net reaction rate Definition.

A multitubular reactor with coolant gives nearly

  1. adiabatic operation
  2. isothermal operation
  3. batch operation
  4. recycle operation
Show the answer

Answer: B. isothermal operation Heat removal along the length.

Theory questions

Derive the optimum temperature for a reversible exothermic first-order reaction at a given conversion. (5 marks)

Show the model answer

Rate: −r_A = C_A0[k₁₀e^(−E₁/RT)(1 − X) − k₂₀e^(−E₂/RT)X]. At fixed X differentiate with respect to T: (E₁/RT²)k₁(1 − X) − (E₂/RT²)k₂X = 0 ⇒ E₁k₁₀e^(−E₁/RT)(1 − X) = E₂k₂₀e^(−E₂/RT)X ⇒ e^((E₂ − E₁)/RT) = E₂k₂₀X/(E₁k₁₀(1 − X)) ⇒ T_opt = (E₂ − E₁)/(R ln[E₂k₂₀X/(E₁k₁₀(1 − X))]). Since X/(1 − X) increases with X, T_opt decreases as conversion increases. It is a maximum because E₂ > E₁ for an exothermic reaction (ΔH = E₁ − E₂ < 0).

How would you choose a reactor system and temperature policy for (a) an irreversible endothermic reaction, (b) a reversible exothermic reaction, (c) series reactions A → R → S with E₁ > E₂? (5 marks)

Show the model answer

(a) Rate always increases with T and equilibrium does not limit (or improves with T): run as hot as materials, catalyst and side reactions allow; plug flow for positive order; supply heat (fired tubular reactor or reheating between adiabatic stages). (b) Follow the optimum temperature progression: high T at the inlet for rate, falling T as conversion rises to stay near equilibrium; implement with staged adiabatic beds and intercooling or cold shot, or a cooled multitubular reactor; plug flow. (c) Plug flow (no back-mixing) to maximise the intermediate; high temperature at first to speed A → R, then cooler as R accumulates to slow its conversion to S; stop at the optimum space time.

Catalysis and catalytic reactors

Topics: Catalysts; steps in a catalytic reaction; synthesising a rate law, mechanism and rate-limiting step; heterogeneous data analysis for reactor design; catalyst deactivation.

Catalysts, steps and rate laws

A catalyst speeds a reaction by providing a lower-energy path and is not consumed. It does not change the equilibrium. Heterogeneous catalysts are usually porous solids (metals on supports such as alumina, silica or carbon; zeolites; metal oxides; sulphides) with large internal surface areas (10–1000 m²/g). Their performance is described by activity, selectivity and stability; promoters improve them and poisons destroy them.

A gas–solid catalytic reaction goes through seven steps: (1) external diffusion of reactant from the bulk gas to the pellet surface, (2) internal (pore) diffusion into the pellet, (3) adsorption on an active site, (4) surface reaction, (5) desorption of products, (6) internal diffusion of products out, (7) external diffusion to the bulk. When diffusion is fast, one of steps 3–5 is usually rate-limiting and the others are at equilibrium.

Synthesising a rate law (Langmuir–Hinshelwood–Hougen–Watson): write each step with the site balance C_t = C_v + C_A·S + C_B·S + …; assume one step rate-limiting and the others at adsorption equilibrium (C_A·S = K_AP_AC_v); eliminate the unmeasurable surface concentrations. For A → B with surface reaction limiting (single site): −r′_A = kP_A/(1 + K_AP_A + K_BP_B). Adsorption-limiting and desorption-limiting mechanisms give different forms. Comparing the forms with initial-rate data (rate against total pressure or partial pressures) identifies the mechanism; e.g. a rate that rises and then levels off with P_A suggests single-site surface reaction, one that passes through a maximum suggests dual-site.

For design, rates are per mass of catalyst and the packed-bed design equation is W = F_A0∫dX/(−r′_A), integrated numerically when the rate law is complex; pressure drop (Ergun) can be included. Catalysts lose activity through sintering, fouling (coking) and poisoning; with activity a(t), −r′ = a(t)(−r′)_fresh. Separable deactivation such as first order a = e^(−k_dt) sets the run length; moving-bed and fluidised reactors regenerate continuously, and temperature–time trajectories hold conversion constant.

  • Catalyst changes rate, not equilibrium.
  • Site balance: C_t = C_v + ΣC_i·S.
  • Single-site surface-reaction limited: −r′ = kP_A/(1 + K_AP_A + K_BP_B).
  • W = F_A0∫dX/(−r′_A).
  • Deactivation: sintering, fouling, poisoning; a(t).

Example: A → B over a catalyst with −r′ = 0.05P_A/(1 + 0.5P_A + 2P_B) mol/(kg·s), P in atm, F_A0 = 10 mol/s, P_A0 = 2 atm. Find the catalyst weight for 60 % conversion.

  1. P_A = 2(1 − X), P_B = 2X.
  2. At X = 0: rate = 0.1/(1 + 1) = 0.05; at X = 0.6: rate = 0.04/(1 + 0.4 + 2.4) = 0.0105.
  3. W = 10∫₀^0.6 dX/(−r′) evaluated numerically (Simpson) ≈ 278 kg.

Answer: About 278 kg: product adsorption (K_B large) makes the rate fall steeply.

Example: A catalyst deactivates as a = e^(−0.02t) (t in days). With fresh catalyst kτ′ = 3 in a packed bed (first order), when does conversion fall to 80 %?

  1. Fresh X = 1 − e^(−3) = 0.950.
  2. X = 0.8 needs kτ′a = −ln 0.2 = 1.609 ⇒ a = 0.536.
  3. t = −ln 0.536/0.02 = 31 days.

Answer: About 31 days on stream.

Catalysts lower activation energy without shifting equilibrium; porous supported catalysts; activity, selectivity, stability, promoters, poisons. Seven steps of a catalytic reaction; rate-limiting step with others at equilibrium; Langmuir–Hinshelwood rate laws from the site balance; identifying the mechanism from initial-rate data; PBR design W = F_A0∫dX/(−r′); deactivation by sintering, fouling and poisoning with activity a(t).

  • Strongly adsorbed products inhibit the rate.
  • Dual-site mechanisms give squared denominators.
  • Initial rates avoid product effects.
  • FCC regenerates coked catalyst continuously.

Key formulas

  • C_t = C_v + C_A·S + C_B·S
  • −r′_A = kP_A/(1 + K_AP_A + K_BP_B)
  • W = F_A0∫dX/(−r′_A)
  • a = e^(−k_dt)

Practice questions

A catalyst changes

  1. the equilibrium constant
  2. the rate of reaching equilibrium
  3. ΔH of reaction
  4. ΔG of reaction
Show the answer

Answer: B. the rate of reaching equilibrium Kinetics only.

In the LHHW approach, steps other than the rate-limiting one are assumed

  1. irreversible
  2. at equilibrium
  3. zero order
  4. diffusion controlled
Show the answer

Answer: B. at equilibrium Pseudo-equilibrium.

Coke deposition is an example of

  1. sintering
  2. fouling
  3. poisoning
  4. promotion
Show the answer

Answer: B. fouling Physical covering.

The packed-bed design equation gives

  1. volume
  2. catalyst weight
  3. pressure
  4. time
Show the answer

Answer: B. catalyst weight Rate per mass.

A rate that rises with P_A and then levels off suggests

  1. dual-site reaction
  2. single-site surface reaction or adsorption control
  3. gas-phase reaction
  4. zero order
Show the answer

Answer: B. single-site surface reaction or adsorption control Saturation of sites.

Theory questions

List the steps in a heterogeneous catalytic reaction and derive the rate law for A → B when the surface reaction is rate-limiting. (6 marks)

Show the model answer

Steps: external diffusion, pore diffusion, adsorption, surface reaction, desorption, pore diffusion out, external diffusion out. Mechanism: A + S ⇌ A·S (K_A), A·S → B·S (k_s), B·S ⇌ B + S (K_B). Surface reaction limiting: −r′ = k_sC_A·S. Adsorption equilibria: C_A·S = K_AP_AC_v, C_B·S = K_BP_BC_v. Site balance: C_t = C_v(1 + K_AP_A + K_BP_B). So −r′ = k_sK_AC_tP_A/(1 + K_AP_A + K_BP_B) = kP_A/(1 + K_AP_A + K_BP_B), with k = k_sK_AC_t.

Describe the types of catalyst deactivation and how they affect reactor operation. (5 marks)

Show the model answer

Sintering (ageing): loss of active surface by crystal growth or support collapse at high temperature; slow, often second order in activity. Fouling (coking): carbon or other deposits cover sites and block pores; can be rapid (FCC seconds) and reversible by burning. Poisoning: strong chemisorption of impurities (S, As, Pb, metals) on active sites; may be reversible or permanent; guard beds remove poisons. Effects: rate = a(t) × fresh rate; conversion falls with time on stream. Remedies: raise temperature gradually to hold conversion, periodic regeneration, moving or fluidised beds with continuous regeneration, feed purification, more stable catalysts.

Diffusion and reaction in catalyst pellets

Topics: Diffusion and reaction in spherical catalyst pellets; internal effectiveness factor; overall effectiveness factor; Thiele modulus; Weisz–Prater parameter.

Thiele modulus, effectiveness factors and the Weisz–Prater criterion

Inside a porous pellet the reactant diffuses through the pores with an effective diffusivity D_e = D_ABεσ/τ̃ (porosity, constriction, tortuosity) and reacts on the walls as it goes. A shell balance on a sphere for a first-order reaction gives D_e(1/r²)d/dr(r²dC/dr) = kC, with C = C_As at r = R and dC/dr = 0 at the centre. The solution is C/C_As = (R/r) sinh(φr/R)/sinh φ, where the Thiele modulus φ = R√(k/D_e) compares the surface reaction rate with the diffusion rate.

The internal effectiveness factor η = (actual rate)/(rate if the whole interior were at C_As) = (3/φ²)(φ coth φ − 1). For small φ (< 0.5) η ≈ 1: the whole pellet works. For large φ (> 5) η ≈ 3/φ: reactant is consumed in a thin shell near the surface, the rate is proportional to 1/R, and the apparent activation energy and order are distorted (apparent E ≈ E/2). External mass-transfer resistance (film around the pellet, coefficient k_c) adds in series; the overall effectiveness factor for a sphere is Ω = η/(1 + ηkR/(3k_c)), and −r = Ω k C_Ab.

The Weisz–Prater criterion checks for internal diffusion limitation using only measured quantities: C_WP = (−r′_obs)ρ_cR²/(D_eC_As) = ηφ². If C_WP ≪ 1 there is no internal limitation; if C_WP ≫ 1 it is severe. Remedies for diffusion limitation are smaller pellets, eggshell catalysts with the active metal near the surface, larger pores, or lower temperature.

  • φ = R√(k/D_e) (sphere, first order).
  • η = (3/φ²)(φ coth φ − 1); large φ: η ≈ 3/φ.
  • Ω = η/(1 + ηkR/(3k_c)).
  • C_WP = r_obsρ_cR²/(D_eC_As) = ηφ².
  • Strong diffusion control: apparent E ≈ E/2.

Example: A spherical pellet (R = 3 mm) has k = 2 s⁻¹ (per pellet volume) and D_e = 10⁻⁷ m²/s. Find φ and η, and the effect of halving the pellet size.

  1. φ = 0.003 × √(2/10⁻⁷) = 0.003 × 4472 = 13.4.
  2. η = (3/13.4²)(13.4 coth 13.4 − 1) = 0.0167 × 12.4 = 0.207.
  3. R = 1.5 mm: φ = 6.7, η = (3/44.9)(6.7 − 1) = 0.381; rate per pellet volume rises by 0.381/0.207 = 1.84.

Answer: φ = 13.4, η = 0.21; halving R nearly doubles the rate (η ≈ 3/φ regime).

Example: An observed rate gives C_WP = 37. What does this mean?

  1. C_WP = ηφ² ≫ 1.
  2. Severe internal diffusion limitation: the centre of the pellet is starved of reactant.

Answer: Strong pore-diffusion control; use smaller or eggshell pellets.

Shell balance in a spherical pellet gives the concentration profile; Thiele modulus φ = R√(k/D_e); internal effectiveness η = (3/φ²)(φ coth φ − 1), η ≈ 3/φ at large φ; overall effectiveness with external film Ω = η/(1 + ηkR/3k_c); Weisz–Prater criterion C_WP = ηφ² from observed rates; diffusion limitation halves the apparent activation energy; remedies: smaller pellets, eggshell catalysts.

  • D_e = D_ABεσ/τ̃.
  • Large φ: rate ∝ 1/R.
  • External limitation: rate independent of T, depends on velocity.
  • C_WP < 0.3 is usually safe.

Key formulas

  • φ = R√(k/D_e)
  • η = (3/φ²)(φ coth φ − 1)
  • Ω = η/(1 + ηkR/(3k_c))
  • C_WP = (−r′_obs)ρ_cR²/(D_eC_As)

Practice questions

A large Thiele modulus means

  1. no diffusion limitation
  2. strong pore-diffusion limitation
  3. external limitation only
  4. zero rate
Show the answer

Answer: B. strong pore-diffusion limitation Reaction ≫ diffusion.

For large φ in a sphere, η is approximately

  1. 1
  2. 3/φ
  3. φ
  4. 1/φ²
Show the answer

Answer: B. 3/φ Asymptote.

Under strong pore-diffusion control the apparent activation energy is about

  1. E
  2. 2E
  3. E/2
  4. zero
Show the answer

Answer: C. E/2 Half the true value.

The Weisz–Prater criterion uses

  1. the true rate constant
  2. the observed rate
  3. the equilibrium constant
  4. the RTD
Show the answer

Answer: B. the observed rate Measured quantities.

An eggshell catalyst places the active metal

  1. at the centre
  2. near the outer surface
  3. uniformly
  4. outside the pellet
Show the answer

Answer: B. near the outer surface Short diffusion path.

Theory questions

Derive the effectiveness factor for a first-order reaction in a spherical catalyst pellet. (6 marks)

Show the model answer

Shell balance: D_e(d²C/dr² + (2/r)dC/dr) = kC; C(R) = C_As, dC/dr = 0 at r = 0. Let ψ = C/C_As, λ = r/R, φ² = kR²/D_e: d²ψ/dλ² + (2/λ)dψ/dλ = φ²ψ; with y = ψλ, y″ = φ²y ⇒ ψ = sinh(φλ)/(λ sinh φ). Actual rate = flux at the surface × area = 4πRD_eC_As(dψ/dλ)|λ=1 = 4πRD_eC_As(φ coth φ − 1). Rate with no diffusion resistance = (4/3)πR³kC_As. η = ratio = 3(φ coth φ − 1)/φ². Limits: φ → 0: η → 1; φ large: η → 3/φ.

What is the Weisz–Prater criterion and how is it used? (4 marks)

Show the model answer

C_WP = (−r′_A,obs)ρ_cR²/(D_eC_As) = ηφ², the ratio of the observed reaction rate to a diffusion rate, all measurable (observed rate, pellet density and radius, effective diffusivity, surface concentration). If C_WP ≪ 1 (typically < 0.3 for first order), internal diffusion does not limit the rate and intrinsic kinetics can be measured; if C_WP ≫ 1 the reaction is severely diffusion-limited and smaller pellets or different catalyst design are needed.

Residence time distribution for chemical reactors

Topics: Modelling the real reactor as CSTR and PFR in series; predicting conversion and exit concentration; reactor modelling using the RTD; zero-parameter models.

Measuring and using the RTD

Real reactors depart from ideal plug flow or perfect mixing because of channelling, bypassing, dead zones and back-mixing. The residence time distribution E(t) describes how long fluid elements spend inside: E(t)dt is the fraction of the exit stream that has been in the reactor between t and t + dt. It is measured by injecting an inert tracer: with a pulse input, E(t) = C(t)/∫C dt; with a step input, the cumulative F(t) = ∫₀ᵗE dt = C(t)/C₀.

The mean residence time t_m = ∫tE dt equals the space time τ = V/v if there are no dead zones; a smaller t_m means stagnant volume. The variance σ² = ∫(t − t_m)²E dt measures spread; the skewness shows tailing. Ideal RTDs: PFR E(t) = δ(t − τ); CSTR E(t) = e^(−t/τ)/τ. Normalising with θ = t/τ gives E(θ) = τE(t) and σ_θ² = σ²/τ². Diagnosis: an early sharp peak indicates bypassing or channelling, a long tail indicates dead zones or recirculation.

With the RTD, a real reactor can be modelled as combinations of ideal ones: a CSTR and a PFR in series (the RTD alone cannot tell the order, which matters for non-first-order reactions), with bypass and dead volume fitted to the tracer curve. Zero-parameter models predict conversion from the RTD alone. The segregation model treats each fluid element as a little batch reactor: X̄ = ∫X_batch(t)E(t)dt; for first order X̄ = 1 − ∫e^(−kt)E dt, which is exact for linear kinetics. The maximum-mixedness model gives the other bound for non-linear kinetics.

  • Pulse: E = C/∫C dt; step: F = C/C₀.
  • t_m = ∫tE dt; σ² = ∫(t − t_m)²E dt.
  • CSTR: E = e^(−t/τ)/τ; PFR: δ(t − τ).
  • Segregation: X̄ = ∫X_batch(t)E(t)dt.
  • Early peak: bypass; long tail: dead zones.

Example: A pulse tracer gives C (g/m³) = 0, 1, 5, 8, 10, 8, 6, 4, 3, 2.2, 1.5, 0.6, 0 at t = 0–10, 12, 14 min. Find t_m, σ² and the conversion for a first-order reaction with k = 0.3 min⁻¹ by the segregation model.

  1. ∫C dt (trapezoidal) = 50.6 ⇒ E = C/50.6.
  2. t_m = ∫tE dt = 5.13 min; σ² = ∫(t − t_m)²E dt = 5.95 min².
  3. X̄ = 1 − ∫e^(−0.3t)E dt = 0.732.
  4. Compare: PFR with τ = 5.13 min gives 0.785, CSTR 0.606.

Answer: t_m ≈ 5.1 min, σ² ≈ 6.0 min², X ≈ 0.73.

Example: A real reactor is modelled as a CSTR (τ_s = 2 min) followed by a PFR (τ_p = 3 min). Find the conversion for a first-order reaction with k = 0.3 min⁻¹.

  1. CSTR: X₁ = kτ_s/(1 + kτ_s) = 0.6/1.6 = 0.375.
  2. PFR: (1 − X) = (1 − X₁)e^(−kτ_p) = 0.625 × e^(−0.9) = 0.625 × 0.4066 = 0.254.
  3. X = 0.746 (the same in either order for first-order kinetics).

Answer: X ≈ 0.75.

RTD E(t) from pulse (E = C/∫C dt) or step (F = C/C₀) tracer tests; mean residence time and variance; ideal RTDs of PFR and CSTR; normalised E(θ); diagnosis of bypassing and dead zones; modelling as CSTR + PFR combinations; zero-parameter segregation and maximum-mixedness models; the segregation model is exact for first-order reactions.

  • t_m < τ indicates dead volume.
  • For first-order reactions the RTD alone predicts conversion.
  • Order of CSTR and PFR matters for non-linear kinetics.
  • Maximum mixedness gives the other conversion bound.

Key formulas

  • E(t) = C(t)/∫C dt
  • t_m = ∫tE dt
  • σ² = ∫(t − t_m)²E dt
  • X̄ = ∫X_batch(t)E(t)dt
  • E_CSTR = e^(−t/τ)/τ

Practice questions

A pulse tracer test gives directly

  1. F(t)
  2. E(t)
  3. the rate constant
  4. the dispersion coefficient only
Show the answer

Answer: B. E(t) Normalised concentration.

For an ideal CSTR, E(t) is

  1. a spike at τ
  2. e^(−t/τ)/τ
  3. constant
  4. zero
Show the answer

Answer: B. e^(−t/τ)/τ Exponential decay.

A long tail in the RTD indicates

  1. bypassing
  2. dead zones or recirculation
  3. plug flow
  4. perfect mixing
Show the answer

Answer: B. dead zones or recirculation Slow exchange.

The segregation model is exact for

  1. second-order reactions
  2. first-order reactions
  3. all reactions
  4. zero-order reactions
Show the answer

Answer: B. first-order reactions Linear kinetics.

If t_m from the RTD is less than V/v, the reactor has

  1. bypassing only
  2. dead volume
  3. perfect mixing
  4. plug flow
Show the answer

Answer: B. dead volume Part of V is not used.

Theory questions

Explain how the RTD is measured with pulse and step tracer inputs and how t_m and σ² are calculated. (5 marks)

Show the model answer

Pulse: inject a small amount of tracer quickly at the inlet; measure exit concentration C(t); E(t) = C(t)/∫₀^∞C dt (the integral equals tracer mass/flow). Step: switch the inlet to a constant tracer concentration C₀ at t = 0; F(t) = C(t)/C₀ and E = dF/dt. Mean residence time t_m = ∫tE dt (equals V/v without dead volume); variance σ² = ∫(t − t_m)²E dt; integrals evaluated numerically (trapezoidal/Simpson). Tracer must be inert, easily detected, not adsorbed, with the same properties as the fluid.

Describe the segregation model and use it for a first-order reaction. (5 marks)

Show the model answer

The fluid is assumed to move as segregated globules that do not exchange material; each behaves as a batch reactor for its residence time t, so its conversion is X_batch(t), and the exit mixes globules according to E(t): X̄ = ∫₀^∞ X_batch(t)E(t)dt. For first order X_batch = 1 − e^(−kt) ⇒ X̄ = 1 − ∫e^(−kt)E(t)dt. For a CSTR (E = e^(−t/τ)/τ) this gives X̄ = kτ/(1 + kτ), identical to the CSTR equation, showing that for first-order reactions the degree of mixing does not matter, only the RTD.

Models for non-ideal reactors

Topics: One-parameter models: tanks-in-series model, dispersion model; two-parameter models: basic introduction.

Tanks-in-series and dispersion models

One-parameter models describe a real reactor with a single number fitted to its RTD. In the tanks-in-series (TIS) model the reactor is N equal CSTRs, each of space time τ/N; the RTD is E(θ) = N(Nθ)^(N−1)e^(−Nθ)/(N − 1)!, with σ_θ² = 1/N, so N = t_m²/σ². For a first-order reaction X = 1 − 1/(1 + kτ/N)^N; N = 1 is a CSTR and N → ∞ a PFR.

In the axial dispersion model, plug flow is overlaid by back-mixing described by a dispersion coefficient D_a, like diffusion: ∂C/∂t = D_a∂²C/∂z² − u∂C/∂z. The dimensionless group D_a/(uL) (the inverse of the Péclet number) measures the spread: 0 is plug flow, ∞ is mixed flow. For a closed vessel σ_θ² = 2(D_a/uL) − 2(D_a/uL)²(1 − e^(−uL/D_a)); for small dispersion σ_θ² ≈ 2D_a/uL, and N ≈ uL/(2D_a). For a first-order reaction in a closed vessel, X = 1 − 4a e^(Pe/2)/[(1 + a)²e^(aPe/2) − (1 − a)²e^(−aPe/2)], a = √(1 + 4Da/Pe), Da = kτ (Wehner–Wilhelm). Correlations give D_a for flow in pipes and packed beds.

Two-parameter models combine ideal regions with fitted volumes and flows: for example a CSTR with a dead zone (active volume fraction α) and a bypass (fraction β), or two CSTRs with exchange. Their parameters are found by fitting the tracer response and then used to predict conversion.

  • TIS: E(θ) = N(Nθ)^(N−1)e^(−Nθ)/(N − 1)!; N = t_m²/σ².
  • TIS first order: X = 1 − 1/(1 + kτ/N)^N.
  • Dispersion: D/uL; closed σ_θ² = 2D − 2D²(1 − e^(−1/D)).
  • Small dispersion: N ≈ 1/(2D/uL).
  • Two-parameter: bypass and dead volume.

Example: A reactor with τ = 5 min is represented by 3 tanks in series. Find the conversion for k = 0.3 min⁻¹ and compare with a dispersion model of D/uL = 0.15.

  1. kτ = 1.5; TIS: X = 1 − 1/(1 + 0.5)³ = 1 − 0.296 = 0.704.
  2. Dispersion: a = √(1 + 4 × 1.5 × 0.15) = 1.378; Wehner–Wilhelm gives X = 0.724.
  3. Limits: CSTR 0.600, PFR 0.777.

Answer: X ≈ 0.70 (TIS) and 0.72 (dispersion), between the CSTR and PFR limits.

Example: A tracer test gives t_m = 5.13 min and σ² = 5.95 min². How many tanks in series represent the reactor, and what is the closed-vessel dispersion number?

  1. N = t_m²/σ² = 26.3/5.95 = 4.4.
  2. σ_θ² = 5.95/26.3 = 0.226 = 2D − 2D²(1 − e^(−1/D)) ⇒ D/uL ≈ 0.13.

Answer: About 4–5 tanks; D/uL ≈ 0.13.

One-parameter models: tanks in series (N = t_m²/σ², X = 1 − 1/(1 + kτ/N)^N) and axial dispersion (D/uL from the variance; Wehner–Wilhelm conversion; small dispersion N ≈ uL/2D). Two-parameter models combine ideal reactors with bypass, dead volume or exchange, fitted to tracer data. All lie between CSTR and PFR limits.

  • N = 1: CSTR; N → ∞: PFR.
  • D/uL → 0: plug flow; → ∞: mixed flow.
  • Fit the model parameter from σ².
  • Two-parameter models capture bypass and dead space.

Key formulas

  • N = t_m²/σ²
  • X = 1 − 1/(1 + kτ/N)^N
  • σ_θ² = 2D − 2D²(1 − e^(−1/D)) (closed)
  • σ_θ² ≈ 2D/uL (small D)

Practice questions

In the tanks-in-series model, N is found from

  1. t_m only
  2. t_m²/σ²
  3. σ² only
  4. the rate constant
Show the answer

Answer: B. t_m²/σ² σ_θ² = 1/N.

A dispersion number D/uL of zero represents

  1. a CSTR
  2. plug flow
  3. a batch reactor
  4. bypass
Show the answer

Answer: B. plug flow No back-mixing.

For small dispersion, N tanks correspond to D/uL ≈

  1. N
  2. 1/(2N)
  3. 2N
  4. N²
Show the answer

Answer: B. 1/(2N) σ_θ² ≈ 2D/uL = 1/N.

A two-parameter model might include

  1. only N
  2. bypass and dead volume
  3. only k
  4. only τ
Show the answer

Answer: B. bypass and dead volume Two fitted quantities.

As N increases the conversion of a first-order reaction

  1. decreases
  2. increases towards the PFR value
  3. stays at the CSTR value
  4. becomes zero
Show the answer

Answer: B. increases towards the PFR value Approaches plug flow.

Theory questions

Explain the tanks-in-series model and derive the conversion for a first-order reaction. (5 marks)

Show the model answer

The reactor of space time τ is represented by N equal ideal CSTRs in series, each with τ_i = τ/N. N is chosen so the model RTD matches the measured spread: σ_θ² = 1/N ⇒ N = t_m²/σ². For first order, each tank: C_i/C_(i−1) = 1/(1 + kτ/N); over N tanks C_N/C₀ = 1/(1 + kτ/N)^N ⇒ X = 1 − (1 + kτ/N)^(−N). N = 1 gives the CSTR result; as N → ∞, (1 + kτ/N)^(−N) → e^(−kτ), the PFR result.

Describe the axial dispersion model and the meaning of the dispersion number. (5 marks)

Show the model answer

Flow is plug flow with superimposed axial mixing described by Fick-like dispersion: ∂C/∂t = D_a∂²C/∂z² − u∂C/∂z (+ reaction). D_a lumps molecular diffusion, turbulent eddies and velocity-profile effects. In dimensionless form the single parameter is D_a/(uL) = 1/Pe (vessel dispersion number): → 0 plug flow, → ∞ mixed flow. It is obtained from the RTD variance (closed vessel σ_θ² = 2D/uL − 2(D/uL)²(1 − e^(−uL/D)); for small D/uL < 0.01 the curve is nearly Gaussian with σ_θ² = 2D/uL) or from correlations of D/(ud) versus Re for pipes and packed beds. Conversion for first order follows the Wehner–Wilhelm solution.

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 Chemical Reaction Engineering-II Studio?

Selectivity, catalysts and real reactors. Parallel and series reactions, adiabatic and optimum temperature progression, catalytic rate laws and packed beds, Thiele modulus and effectiveness factor, catalyst deactivation, RTD, tanks-in-series and dispersion models from the ASTU CHE241502 syllabus.

What can I do in the Chemical Reaction Engineering-II Studio?

Parallel and series reactions: maximise the product; Adiabatic operation and optimum temperature; LHHW rate laws, effectiveness factor, deactivation; RTD, segregation, tanks-in-series and dispersion.

Is the Chemical Reaction Engineering-II Studio free?

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

Open the interactive studio →