Sources and References

Primary texts — Shuler, M.L., Kargi, F., and DeLisa, M., Bioprocess Engineering: Basic Concepts, 3rd ed., Prentice Hall, 2017; Doran, P.M., Bioprocess Engineering Principles, 2nd ed., Academic Press, 2013.

Supplementary texts — Blanch, H.W. and Clark, D.S., Biochemical Engineering, 2nd ed., CRC Press, 1997; Bailey, J.E. and Ollis, D.F., Biochemical Engineering Fundamentals, 2nd ed., McGraw-Hill, 1986.

Online resources — MIT OCW 10.37 “Chemical and Biological Reaction Engineering”; NPTEL Bioprocess Engineering course; FDA Guidance documents on biopharmaceutical manufacturing; ICH Q6B and Q7 guidelines; EMA public summaries on GMP.

---

Chapter 1: Biological Systems for Production

1.1 Industrial Host Organisms

Microbial fermentation underpins products as varied as beer, insulin, and polylactic acid. Choice of host depends on product: bacteria (E. coli, Bacillus subtilis) grow rapidly and tolerate simple media; yeasts (Saccharomyces cerevisiae, Pichia pastoris) secrete folded proteins with some post-translational modification; filamentous fungi (Aspergillus niger) produce enzymes and organic acids; mammalian cells (CHO, HEK293) perform glycosylation required for therapeutic antibodies; insect cells (Sf9) support baculovirus expression; algae and cyanobacteria offer photosynthetic routes.

1.2 Product Categories

Primary metabolites (ethanol, lactic acid, amino acids) accumulate during growth. Secondary metabolites (antibiotics, pigments) appear in stationary phase. Recombinant proteins (insulin, erythropoietin, monoclonal antibodies) require engineered hosts with inducible expression systems. Whole-cell biocatalysts perform stereospecific transformations. Live biotherapeutics, vaccines, and even probiotics extend the product space.

1.3 Biosafety and Sustainability

Good Manufacturing Practice (GMP), biosafety levels (BSL-1 through BSL-4), containment, and regulatory oversight structure the industry. Sustainability considerations include renewable feedstocks (lignocellulosic sugars), water reuse, minimized solvent use, energy integration of fermentation with downstream processing, and valorization of byproducts.

Chapter 2: Cell Growth Kinetics

2.1 Monod Kinetics

For a limiting substrate \( S \), specific growth rate follows

\[ \mu = \mu_{max}\frac{S}{K_S + S}, \]

with maximum specific growth rate \( \mu_{max} \) and half-saturation constant \( K_S \). The form is structurally identical to Michaelis-Menten enzyme kinetics but is empirical at the population level.

2.2 Yield Coefficients

\( Y_{X/S} \) = mass of biomass produced per mass of substrate consumed; \( Y_{P/S} \) = product yield; \( Y_{X/O_2} \) for oxygen. Yields are not always constant; maintenance energy \( m_s \) and uncoupled metabolism shift yields with growth rate:

\[ q_S = \frac{\mu}{Y_{X/S}^{true}} + m_s. \]

2.3 Inhibition

Substrate inhibition (Haldane), product inhibition, and competitive substrates all modify growth. Haldane:

\[ \mu = \mu_{max}\frac{S}{K_S + S + S^2/K_I}. \]

Ethanol inhibition in yeast fermentation and lactate in mammalian cell culture are industrially important examples.

2.4 Death and Lysis

At very high cell densities or in stationary phase, viable cell concentration declines. Engineering models add a first-order death term: \( dX/dt = (\mu - k_d)X \).

Chapter 3: Bioreactor Modes of Operation

3.1 Batch

Biomass grows through lag, exponential, stationary, and death phases. Time evolution follows coupled ODEs:

\[ \frac{dX}{dt} = \mu(S)X, \quad \frac{dS}{dt} = -\frac{\mu(S)X}{Y_{X/S}} - m_s X. \]

Batch productivity is limited by turnaround (cleaning, sterilization, inoculation).

3.2 Fed-Batch

Substrate is fed during culture at a controlled rate; volume grows without withdrawal. Fed-batch allows high final cell density while avoiding substrate inhibition—dominant for industrial E. coli protein production, mammalian CHO antibody production, and penicillin fermentation.

For exponential feeding at desired specific growth rate \( \mu^* \):

\[ F(t) = \frac{\mu^* X_0 V_0}{S_F Y_{X/S}} e^{\mu^* t}. \]

3.3 Continuous (Chemostat)

Continuous feed and withdrawal at dilution rate \( D = F/V \). At steady state, \( \mu = D \) and

\[ S_{ss} = \frac{K_S D}{\mu_{max} - D}, \quad X_{ss} = Y_{X/S}(S_0 - S_{ss}). \]

Washout occurs when \( D > \mu_{max} \). Chemostats are unmatched for physiological studies and for some commodity products, but industrial adoption is limited by contamination risk and genetic drift.

3.4 Perfusion and Recycle

Cell retention (by filtration, centrifugation, or settling) decouples dilution rate from cell specific growth rate. Perfusion achieves very high cell densities (10\(^8\) cells/mL) for continuous biopharmaceutical production and is advancing as an alternative to fed-batch.

Chapter 4: Transport in Bioreactors

4.1 Oxygen Supply

Aerobic cultures consume oxygen faster than its low water solubility (\( \sim 7 \) mg/L at 25 °C) can be replenished by batch-dissolved O\(_2\). Oxygen transfer follows

\[ OTR = k_L a (C^* - C_L), \]

where \( k_L a \) [h\(^{-1}\)] is the volumetric mass transfer coefficient. Correlations relate \( k_L a \) to power per unit volume \( P/V \), superficial gas velocity \( v_s \), and rheology.

\( k_L a \) measurement methods include dynamic gassing-out, sulfite oxidation, and off-gas analysis via mass balance on O\(_2\) and CO\(_2\).

4.2 Mixing and Rheology

Stirred-tank reactors use Rushton or pitched-blade impellers; power draw scales with \( N_p \rho N^3 D^5 \) (turbulent regime). Mixing time decreases with power input. Non-Newtonian broths (filamentous fungi, polysaccharide-producing bacteria) develop apparent viscosities several orders of magnitude above water, punishing oxygen transfer and demanding specialized impellers.

4.3 Heat Removal

Metabolic heat is substantial (\( \sim 10-20 \) kJ/g cells). Cooling coils or jackets remove this heat; the design equation is the usual \( Q = UA\Delta T_{lm} \). Large fermentors (100+ m\(^3\)) approach the limits of surface-area-to-volume heat transfer and require external cooling loops.

4.4 Scale-Up

No single scale-up criterion captures all phenomena. Common choices: constant \( P/V \), constant \( k_L a \), constant tip speed (shear), constant mixing time. Industrial practice uses one primary criterion and monitors the others. Scale-up of microbial cultures to 100 m\(^3\) is mature; mammalian cell culture to 20 m\(^3\) is state of the art.

Chapter 5: Dynamic Modeling and Simulation

5.1 ODE Models

The general unstructured model for fed-batch is

\[ \frac{d(VX)}{dt} = \mu V X, \quad \frac{d(VS)}{dt} = FS_F - q_S V X, \quad \frac{dV}{dt} = F. \]

Structured models add internal variables (RNA, key enzymes, intracellular metabolites); segregated models track subpopulations (age, viability). Flux balance analysis on genome-scale metabolic networks predicts steady-state fluxes under growth-optimization assumptions.

5.2 Monitoring and Control

Online sensors: pH, DO, temperature, agitation, off-gas CO\(_2\)/O\(_2\). Soft sensors estimate unmeasured states (biomass, product) from measured signals plus a model. Feedback control of DO (via agitation/aeration) and pH (via acid/base) is standard; advanced control integrates process analytical technology (PAT) for real-time release.

5.3 Numerical Simulation

Stiff integrators (BDF, Rosenbrock) handle widely separated timescales. Parameter estimation from batch/fed-batch data uses nonlinear least squares or Bayesian inference; identifiability analysis precedes estimation to confirm parameters are distinguishable.

Chapter 6: Downstream Processing

6.1 The DSP Train

Recovery starts with cell-broth separation (centrifugation, microfiltration), followed by cell disruption if the product is intracellular (high-pressure homogenization, bead milling), capture (adsorption, precipitation, extraction), intermediate purification (ion exchange, affinity chromatography), polishing (size-exclusion, hydrophobic interaction), and formulation (ultrafiltration/diafiltration, lyophilization).

6.2 Chromatography

Linear isotherm analysis:

\[ u_{band} = \frac{v}{1 + (1-\varepsilon)/\varepsilon \cdot K}. \]

Resolution combines selectivity, efficiency (number of theoretical plates \( N \)), and retention. For biological molecules, protein A affinity chromatography is indispensable for monoclonal antibodies: it captures from crude harvest with purity > 95% in a single step.

6.3 Filtration and Membrane Operations

Tangential flow filtration (TFF) uses cross-flow to limit cake buildup. Ultrafiltration separates by molecular weight cutoff (10–100 kDa), diafiltration exchanges buffer. Viral clearance filters (20 nm pore) are mandated for mammalian-cell-derived products.

6.4 Economics and Overall Process

Downstream costs often exceed upstream costs for high-value biopharmaceuticals, reaching 70% of total manufacturing cost for monoclonal antibodies. The engineering challenge is integrated: design upstream cultivation, clarification, and capture chromatography as one system, matching titer increases with purification capacity.