Green Chemistry and Process Intensification

Green chemistry is the practice of designing chemical products and processes that reduce or eliminate hazard at the molecular and unit-operation level. Process intensification (PI) is the engineering corollary — shrinking equipment, eliminating intermediate storage, and pushing reactions to their thermodynamic and kinetic limits in continuous flow. The two disciplines converged in the 1990s and now define the way new pharmaceutical, agrochemical, and fine-chemical processes are commissioned. The 2024-2026 wave has shifted regulatory pressure (EPA TSCA reform, EU REACH, ICH Q13 continuous manufacturing guideline finalized 2023) onto API manufacturers to demonstrate quantitative greenness metrics in regulatory filings.

See also

• organic-chemistry-foundations
• inorganic-chemistry
• electrochemistry
• physical-chemistry
• polymer-chemistry
• medicinal-and-photo-chemistry
• reagent-and-reaction-catalog
• catalyst-instrumentation-and-monomers

---

The twelve principles of green chemistry

Anastas and Warner codified the framework in Green Chemistry: Theory and Practice (Oxford 1998) while at the EPA Office of Pollution Prevention and Toxics. The twelve principles are not a checklist but a hierarchy — design out hazard at the molecule first, design out waste at the process next, design out energy and feedstock vulnerability last.

1. Prevention. It is better to prevent waste than to treat or clean up waste after it has been created. Translates directly to the E-factor metric below.
2. Atom economy. Synthetic methods should be designed to maximize the incorporation of all materials used in the process into the final product. Trost 1991 Science 254:1471.
3. Less hazardous chemical syntheses. Wherever practicable, synthetic methods should use and generate substances that possess little or no toxicity. Targets phosgene, hydrazine, HF, methylating agents (dimethyl sulfate, methyl iodide), heavy-metal stoichiometric oxidants (CrO3, OsO4, MnO2 bulk), and azides.
4. Designing safer chemicals. Chemical products should be designed to preserve efficacy of function while reducing toxicity. Rational drug-design analogue for industrial chemistry — design out the toxophore, design in biodegradability.
5. Safer solvents and auxiliaries. Solvents drive 80-90% of mass in a typical fine-chemical process. See solvent selection guides below.
6. Design for energy efficiency. Energy requirements should be minimized; synthetic methods should be conducted at ambient temperature and pressure. Implicit case for catalysis over high-T high-P forcing.
7. Use of renewable feedstocks. Raw materials should be renewable rather than depleting whenever technically and economically practicable. Bio-based platform molecules: lactic acid, succinic acid, FDCA, levulinic acid, glycerol, isosorbide, 5-HMF.
8. Reduce derivatives. Unnecessary derivatization (protection/deprotection, blocking groups) generates waste and must be minimized.
9. Catalysis. Catalytic reagents (as selective as possible) are superior to stoichiometric reagents.
10. Design for degradation. Products should be designed so that at the end of their function they break down into innocuous degradation products. Counter-example: PFAS persistence.
11. Real-time analysis for pollution prevention. Analytical methodologies need to be further developed to allow for real-time in-process monitoring (PAT — process analytical technology; FDA 2004 guidance).
12. Inherently safer chemistry for accident prevention. Substances and the form of a substance used in a chemical process should be chosen to minimize the potential for chemical accidents.

Winterton’s Twelve More Principles of Green Engineering (Anastas-Zimmerman 2003 ES&T) extends the framework to unit operations: inherent rather than circumstantial safety, prevention over treatment, separations as design conditions, maximize efficiency of matter and energy use, output-pulled rather than input-pushed.

---

Atom economy and the E-factor

Atom economy

Atom economy (AE) is the molecular-weight ratio of desired product to total reagents incorporated stoichiometrically.

AE (%) = (MW_product / Σ MW_reactants) × 100

A Diels-Alder cycloaddition is 100% atom-economical: every atom from the diene and dienophile ends up in the cyclohexene. A Wittig olefination is poor: phosphonium ylide MW dwarfs the alkene product, and triphenylphosphine oxide leaves as waste (also costly to recover or dispose). Atom economy is an intrinsic property of the balanced equation — independent of yield, solvent, and workup. It points to the kind of chemistry to choose.

High-AE reactions: pericyclics (Diels-Alder, ene, [2+2], sigmatropic rearrangements), catalytic hydrogenation, isomerizations, additions across multiple bonds (hydroformylation, hydroboration, hydroamination), C-H activation, condensations that lose only water or small inert byproduct.

Low-AE reactions: classical Grignard (releases halide salt; uses 1 equiv reagent), Wittig (stoichiometric Ph3P=O), Mitsunobu (DIAD + PPh3 → hydrazine dicarboxylate + Ph3P=O — both equivalents waste), Appel (CCl4 + PPh3), peptide coupling with PyBOP or HATU (the urea/oxide byproducts are bulky).

The E-factor

Sheldon’s E-factor (Roger Sheldon, then Andeno/DSM; Chem Ind 1992; revisited Green Chem 2007 and 2017) is the mass ratio of waste to product, computed over the entire process:

E = total mass of waste / mass of desired product

Sheldon’s classic table (1992) by industry:

Sector	Annual production (t)	E-factor (kg waste / kg product)
Oil refining	10⁶ - 10⁸	~0.1
Bulk chemicals	10⁴ - 10⁶	< 1 - 5
Fine chemicals	10² - 10⁴	5 - 50
Pharmaceuticals	10 - 10³	25 - 100+

The pharma figure shocked the field in 1992 and remains the field’s central justification for greening. Water mass is sometimes excluded from E (the simple E-factor) and sometimes included (total E-factor). The latter is more honest for solvent-heavy aqueous workups.

Related metrics:

• PMI — process mass intensity = total mass in / mass product. PMI = E + 1. Adopted as the headline metric by ACS GCI Pharmaceutical Roundtable (Constable et al. 2007).
• Carbon efficiency = mass C in product / mass C in inputs.
• Reaction mass efficiency (RME) = mass product / mass of stoichiometric reagents (excludes solvents/workup).
• EcoScale (Van Aken 2006 Beilstein J Org Chem) — penalty-based scoring 0-100.
• iGAL (Yale Center for Green Chem 2018) — innovation Greener Aspiration Level; benchmarks new processes against best-in-class.

---

Solvent selection

Solvents typically account for 75-85% of mass in batch pharma and 50-80% of cumulative energy demand. Solvent selection guides quantify hazard and propose ranked greener alternatives.

GSK solvent guide

GSK Sustainable Solvent Guide (Henderson et al. 2011 Green Chem) scores 110 solvents on safety, health, environment, and life-cycle impact. Tiers: Recommended (water, EtOH, IPA, n-BuOH, EtOAc, iPAc, MeOAc, MIBK, anisole, sulfolane, 2-MeTHF), Usable (MeOH, acetone, MEK, toluene, heptane, MTBE), Problematic (THF, MeCN, DCM, DMSO, NMP, DMAc), Hazardous (DMF, dioxane, pyridine, benzene, CCl4, CHCl3, hexane, Et2O), Highly Hazardous (CS2, HMPA, benzene, CCl4).

CHEM21 solvent guide

CHEM21 (EU Innovative Medicines Initiative; Prat, Wells, Hayler, Sneddon, McElroy, Abou-Shehada, Dunn 2016 Green Chem 18:288) is the most-cited current guide. Three-color ranking on Safety, Health, Environment dimensions, plus combined hazard score and renewability flag. Includes ~50 widely used solvents plus emerging bio-based ones (Cyrene, 2-MeTHF, GVL, ethyl lactate).

Critical CHEM21 callouts:

• Avoid (red): hexane, DMF, NMP, dioxane, CHCl3, CCl4, DCM (for new processes; legacy use grandfathered), pyridine, Et2O, benzene.
• Use with caution (amber): THF, MeCN, toluene, DMSO, DMAc, DMPU, MTBE.
• Recommended (green): water, ethanol, IPA, n-butanol, ethyl acetate, isopropyl acetate, acetone, methyl ethyl ketone, anisole, sulfolane, 2-MeTHF.

Pfizer solvent guide

Pfizer’s internal guide (Alfonsi et al. 2008 Green Chem) groups solvents into Preferred / Usable / Undesirable. Sanofi, Astra-Zeneca, Merck, BMS, Lilly maintain analogous internal tools — increasingly aligned with CHEM21.

Emerging green solvents

• 2-MeTHF. Derived from furfural (lignocellulose); higher boiling (80 °C) and less peroxide-prone than THF; immiscible with water → easy phase separation; greener Grignard solvent. Pennakem and Penn A Kem are major producers.
• CPME — cyclopentyl methyl ether. Zeon (Japan); bp 106 °C; low peroxide formation; biphasic with water; replaces THF, MTBE, dioxane in many cases.
• Cyrene (dihydrolevoglucosenone). Circa Group; bio-derived from cellulose; aprotic dipolar (replaces DMF, NMP); not a complete drop-in but increasingly used in SnAr, amide coupling, polymer dissolution.
• GVL — γ-valerolactone. Bio-derived from levulinic acid; biorefinery platform solvent.
• Ethyl lactate. Corn-derived; GRAS; replaces toluene and DCM in coatings.
• Limonene. Citrus peel co-product; replaces hexane in some extractions; food and personal-care.
• Anisole. Toluene replacement in suitable polarity range; CHEM21 recommended.
• Sulfolane (tetramethylene sulfone). High BP polar aprotic; stable; replaces DMF/NMP in some SnAr; refining heritage.

Ionic liquids

Ionic liquids (ILs) are salts molten below 100 °C, often at room temperature (RTILs). They have negligible vapor pressure, wide electrochemical windows (4-6 V), tunable polarity, and can dissolve cellulose, gases (CO2), and metal oxides.

Common cations: 1-alkyl-3-methylimidazolium ([Cnmim]+), N-alkylpyridinium, tetraalkylammonium, tetraalkylphosphonium. Common anions: [PF6]-, [BF4]-, [NTf2]- (bis(trifluoromethylsulfonyl)imide), [OTf]-, [OAc]-, [Cl]-, [HSO4]-, dicyanamide.

Use cases:

• Cellulose dissolution. [Emim][OAc], [Bmim]Cl — basis for Ioncell-F process (Aalto, Finland; pilot textile fiber 2019-2025).
• CO2 capture. [Bmim][PF6] and amine-functionalized ILs (e.g., [P66614][2-CNpyr] from Wang-Brennecke).
• Electrochemistry. Wide ECW supports reactive metal electroplating (Al, Li, Mg from ILs at room T).
• Biocatalysis. Lipases stable in some ILs; lower water activity enables esterification.

Caveats: imidazolium ILs are not always green — they can be toxic to aquatic organisms, biodegrade poorly, and their synthesis is reagent-intensive. CHEM21 treats most ILs as “case by case.” Recyclability is the practical criterion.

Deep eutectic solvents (DES)

Abbott (Leicester) 2003 — mixtures of hydrogen-bond donor and acceptor (HBA + HBD) with melting points far below either component. Choline chloride + urea (Reline, 1:2 molar, mp 12 °C) is the prototype. Bio-based DES (natural DES, NADES) use sugars, polyols, organic acids (citric, malic).

Examples:

• Reline (ChCl/urea 1:2). Mp 12 °C; conductivity 0.75 mS/cm at 25 °C; viscosity 750 cP. Metal oxide dissolution, electroplating.
• Ethaline (ChCl/EG 1:2). Mp -66 °C; lower viscosity; Cu electropolishing.
• Glyceline (ChCl/glycerol 1:2). Biomass valorization, biocatalysis.
• Maline (ChCl/malic acid 1:1). NADES for natural product extraction.

DES advantages: cheap, low toxicity, biodegradable, prepared by simple mixing. Disadvantages: high viscosity (especially without water), water can change the structure substantially (DES → aqueous solution at >50 wt% water; the Hammond critical point).

Supercritical CO2

scCO2 (Tc = 31 °C, Pc = 74 bar) is non-toxic, non-flammable, easily separated by depressurization, and tunable density/polarity. Industrial uses: decaffeination of coffee (HAG Kaffee 1978; Maxwell House licensed 1981; world’s largest scCO2 plant — Kraft Decaf, ~50,000 t/yr), hops extraction (Pfizer Folkestone; SS Steiner; >80% of US hops processed by scCO2), DuPont/Chemours fluoropolymer polymerization, dry cleaning (substituting perc). Limited solvent power for polar solutes — modifiers (5-10% EtOH, MeOH) extend range. Cross-link physical-chemistry for supercritical phase behavior.

---

Catalysis as a green technology

Catalysis replaces stoichiometric reagents with substoichiometric agents that are regenerated in each cycle, slashing waste. The Anastas-Trost-Sheldon “catalysis revolution” is the single biggest contribution to lower E-factors in fine chemicals since 1990.

Homogeneous transition metal catalysis

• Hydrogenation. Crabtree’s [Ir(cod)(PCy3)(py)]PF6 (1979) and Noyori’s Ru-BINAP (Nobel 2001) replaced stoichiometric NaBH4, LiAlH4, dissolving-metal reductions for asymmetric hydrogenation of ketones, enamides, β-ketoesters. Industrial application: (S)-naproxen, (S)-DOPA (Knowles Monsanto 1968; first commercial asymmetric catalysis; Nobel 2001), (S)-metolachlor (Syngenta — Iridium-Xyliphos catalyzed reductive amination produces ~10,000 t/yr; >70% global pre-emergent corn herbicide).
• Hydroformylation (oxo process). CO + H2 + alkene → aldehyde; Rh/PPh3 (Wilkinson; LP Oxo by Union Carbide-Dow now ~10 Mt/yr) replaced Co-based high-pressure original. Atom-economical, gas-phase H2/CO from syngas — though syngas itself is fossil-derived.
• Cross-coupling. Suzuki, Heck, Negishi (Nobel 2010) — Pd-catalyzed C-C bond formation. Pharma examples: rosuvastatin (Crestor; AstraZeneca; Sonogashira step), losartan (Merck; biaryl via Suzuki), valsartan, sildenafil. Aryl halide + ArB(OH)2 → biaryl, with KOH or K3PO4 base; Pd(OAc)2 / SPhos at 0.1-1 mol% is standard.
• Olefin metathesis. Grubbs catalysts I (1995, RuCl2(PCy3)2(=CHPh)), II (2003, with NHC ligand), Hoveyda-Grubbs (2002, chelating ether). Nobel 2005 (Chauvin-Grubbs-Schrock). Industrial: Materia/XiMo and BASF olefin metathesis for fragrance (Firmenich civetone synthesis), pheromones, biodiesel cross-metathesis (Elevance). RCM for macrocyclic drug candidates (Vaniprevir, MK-7009 from Merck).
• C-H activation. Direct functionalization of C-H bonds skipping prefunctionalization (halide, boronate). Sames-Hartwig-Yu palette of Pd, Ir, Rh, Co C-H activation chemistry. Boscalid (BASF fungicide) intermediate produced by Suzuki — but newer routes use C-H borylation.

Heterogeneous catalysis

• Pd/C, Pt/C, Rh/C, Ru/C — hydrogenation; nitro-to-amine; debenzylation. Suspended in EtOH, EtOAc, water with 1-50 bar H2 in batch autoclaves or continuously in trickle-bed reactors.
• Zeolites — shape-selective acid catalysis. ZSM-5 (Mobil 1975) for methanol-to-gasoline; mordenite, beta, USY for alkylation, isomerization. Cross-link inorganic-chemistry.
• Metal-organic frameworks (MOFs). Cu-BTC (HKUST-1), MIL-101(Cr), UiO-66 for adsorption and increasingly catalysis. See mof-cof-perovskite-catalog.
• Sulfated zirconia, niobic acid — superacid solid catalysts for esterification, alkylation; replace H2SO4.

Biocatalysis

The fastest-growing pillar of green chemistry. Engineered enzymes deliver near-perfect enantioselectivity at ambient T/p in water or biphasic media. Cross-link microbiology-foundations and protein-families-and-drug-targets.

• Lipases — Candida antarctica lipase B (CALB; Novozym 435 immobilized) — enantioselective acylation, kinetic resolution, transesterification. Industrial: ibuprofen kinetic resolution (R-enantiomer hydrolyzed selectively), CALB-mediated polyester synthesis.
• Ketoreductases (KRED). Asymmetric ketone reduction with NADPH cofactor. Codexis engineered KREDs deliver atorvastatin side chain (Pfizer Lipitor route).
• Transaminases. Asymmetric amine synthesis. Codexis-Merck sitagliptin (Januvia) transaminase, Savile-Janey-Mundorff-Moore-Devine-Krebber-Fleitz-Brands-Devine-Huisman-Collier 2010 Science 329:305. Rhodium-catalyzed reductive amination replaced by ATA-117 engineered transaminase; eliminated Rh, allowed water solvent, raised yield 13%, productivity 53%.
• Imine reductases (IRED), reductive aminases (RedAm). Direct ketone + amine → chiral amine in water.
• Ene reductases (Old Yellow Enzymes). Asymmetric alkene reduction.
• P450 monooxygenases. Selective C-H oxidation. Engineered P450 BM3 from Bacillus megaterium (Arnold Caltech; directed evolution Nobel 2018).
• Halogenases, methyltransferases (MTases), Friedel-Crafts acyltransferases — emerging tools.

The Codexis-Merck-sitagliptin paper (cited above) became the case study for industrial directed evolution. Frances Arnold’s Nobel 2018 in directed evolution of enzymes anchors the field.

Organocatalysis

List, MacMillan (Nobel 2021) — small organic molecules (proline, cinchona alkaloids, MacMillan imidazolidinones, thiourea catalysts of Jacobsen and Takemoto). No metal trace residues to worry about (key for API contamination limits — typical Pd, Rh, Ru limits 5-10 ppm). Cross-link organic-chemistry-foundations for enamine / iminium mechanisms.

Photocatalysis

Visible-light photoredox catalysis (MacMillan 2008; Yoon, Stephenson, Knowles, Doyle) uses Ru(bpy)3(PF6)2, Ir(ppy)3, Ir[dF(CF3)ppy]2(dtbbpy)PF6, or organic dyes (4CzIPN, eosin Y, acridinium salts) to generate radicals at room temperature from a wide range of substrates. Photocatalysis replaces high-energy UV chemistry and stoichiometric radical initiators (AIBN, peroxides). Now widely used in medicinal chemistry (Pfizer, Merck, GSK have dedicated photoredox screening platforms). Commercial flow photoreactors: Vapourtec UV-150, Corning Advanced-Flow G1/G3, Ehrfeld photoreactor cubes, HepatoChem Photoredox Box. Cross-link medicinal-and-photo-chemistry.

Electrocatalysis as a green driver

See electrochemistry-energy-storage. Electrochemistry replaces stoichiometric chemical oxidants and reductants with electrons. Examples:

• Kolbe electrolysis. Carboxylate decarboxylation/coupling — historically Adams pinacolone.
• Shono oxidation. α-Methoxylation of amines anodically — replaces NBS.
• Baran electrochemistry. Methylation, allylation, C-H functionalization on small scale to large pilot (Asymchem, Pfizer collaborations).
• Industrial. Chlor-alkali (NaCl → NaOH + Cl2 + H2; ion-exchange membrane cells displaced asbestos diaphragm and mercury cells), aluminum electrolysis (Hall-Héroult; 60% of EU green-electricity demand), adiponitrile via electrochemical hydrodimerization of acrylonitrile (Monsanto-Solutia ~300 kt/yr).

---

Continuous-flow chemistry

Flow chemistry conducts reactions in continuously fed reactors — tubes, chips, packed beds, CSTRs — rather than batch flasks. The 2007-onward boom in pharma flow chemistry rests on a few advantages over batch:

1. Heat and mass transfer. Surface-to-volume ratio scales as 1/r; for a 1-mm tube it is 4000 m²/m³ versus ~10 m²/m³ for a 1000-L batch reactor. Highly exothermic reactions (organolithiums, Grignards, diazotization, nitration) can be run safely without runaway risk.
2. Hazardous intermediates in small inventory. A continuous reactor at steady state holds only the inventory of one residence time; a 100 mL reactor at 10 mL/min holds 10 minutes of intermediate. Phosgene, HN3, HF, diazo, peroxide, organic azides — all viable on-demand.
3. Reaction telescoping. Multiple steps chained without isolating intermediates. Saves solvent swaps, evaporations, and crystallizations.
4. Precise residence time. Plug-flow gives tight RT distribution → narrower MWD in polymers, controlled formation of unstable intermediates.
5. Scale-up by numbering up. Numbering-up identical microreactors avoids scale-up effects in heat/mass transfer. Practical for fine chemicals; capital-heavy for bulk.

Hardware

• Tubular reactors — PFA, FEP, PEEK, SS316L tubing, coiled in oil baths or thermostatic blocks. Vapourtec R-series, Syrris Asia, Uniqsis FlowSyn, ThalesNano IceCube/CatCart/H-Cube, Chemtrix Plantrix.
• Microreactors and chips — etched silicon (Lonza FlowPlate, IMM, Mainz), silicon-glass (Dolomite Bio, Micronit), silicon carbide (Chemtrix Labtrix SOR, SiC for hard-grit corrosive media), high-pressure metal (Corning AFR G1/G3/G4).
• Packed-bed reactors. Heterogeneous catalyst (Pd/C, Raney Ni, polymer-supported reagent, immobilized enzyme); the H-Cube hydrogenator generates H2 in situ on a small Pd cartridge.
• CSTR cascades. Coflore ACR/ATR (AM Technology), Cambridge Reactor Design; mimic batch reactor environment but in flow. Useful for slurry-handling and crystallization-coupled chemistry.
• Falling film, gas-liquid contactors. Photocatalysis at scale (Corning Photo-G1 + LED arrays).
• Inline analytics (PAT). FlowIR (Mettler Toledo), ReactRAMAN, online HPLC autosamplers, in-line UV/vis, NMR (Magritek Spinsolve Carbon).

Industrial flow chemistry milestones

• Lonza Visp — phosgene generation and use in flow; multi-tonne capability.
• DSM/InnoSyn — Sphera continuous nitration; multiple API steps.
• Eli Lilly — continuous asymmetric hydrogenation of imines (Cole 2017 Science 356:1144); prexasertib end-to-end continuous synthesis (Cole et al. 2017).
• Novartis-MIT Continuous Manufacturing Center (2007-2017) — aliskiren end-to-end CM demonstration (Adamo et al. 2016 Science 352:61); reduced 200 days batch to ~2 days continuous, footprint cut by ~half.
• Pfizer-Cornell Portable On-Demand Manufacturing (Adamo 2016 Science) — refrigerator-sized continuous units producing diphenhydramine, lidocaine, diazepam, fluoxetine on-demand at clinical-trial quantities.
• Snapdragon Chemicals (now part of Cambrex) — flow process services for API manufacturers.
• Asymchem / Asymchem Crystal — large-volume flow API in China; flow hydrogenation, flow Grignard.
• Merck Rahway — continuous manufacturing of letermovir, sitagliptin intermediates.

FDA ICH Q13

ICH Q13 (Continuous Manufacturing of Drug Substances and Drug Products) — guideline finalized 2023, harmonizes US, EU, Japan, China regulatory expectations. Defines CM (continuous manufacturing) and HCM (hybrid). Addresses control strategy, batch definition (time-, mass-, or material-bounded), state of control, model-based release. Janssen Prezista (darunavir) was the first FDA-approved CM API (2016 PAS). Vertex Orkambi (ivacaftor/lumacaftor) tablet drug product was the first CM drug product (2015). Now ~15 CM-approved products globally.

---

Process intensification (PI)

PI is the engineering side of green chemistry — equipment that does more with less mass. Stankiewicz-Moulijn 2000 Chem Eng Prog coined the modern usage. Four PI pillars: structure (spatial — microchannel, packing), energy (thermal/electromagnetic/sonic intensification), synergy (combined functions — reactive distillation), time (alternating fields, oscillatory baffled reactors).

Microreactors

10-1000 µm channels; flow regime laminar (Re typically <100), mixing diffusion-limited. Mixing time ~ L²/D — for L = 100 µm and D = 10⁻⁹ m²/s → ~10 s. Active mixing (split-and-recombine, herringbone Stroock 2002 Science, hydrodynamic focusing) reduces to ms.

Limitations: clogging by solids/precipitates; fouling; throughput per chip modest (mL/min). Numbering-up rather than scaling-up.

Spinning disk reactor (SDR)

Disc rotating 200-3000 rpm; reactants pumped onto center; thin film (<1 mm thick) on disc surface, residence time ~ms-s. Used for fast exothermic reactions, polymerization (Boodhoo Newcastle, Jachuck), nanoparticle synthesis, biodiesel transesterification. Heat-transfer coefficients up to 14 kW/m²K (orders above stirred tank).

Oscillatory baffled reactor (OBR)

Tube with periodically spaced baffles; piston imposes sinusoidal flow. Plug-flow with batch-like mixing. NiTech (UK) commercialization; used for crystallization control, polymerization, slurry handling in flow. Lawton-Marshall-Carmichael 2013.

Reactive distillation, reactive extraction

Combine reaction + separation in one vessel. Eastman methyl acetate process (Agreda 1990 CHEMTECH) — replaced 9 distillation columns + reactor with 1 reactive distillation column. Esterification at boil, water removed as it forms, equilibrium driven forward. Halved capital cost; eliminated benzene azeotrope solvent.

MTBE, ETBE, cumene, isobutene oligomerization, transesterification, hydrolysis, hydration — all run reactively. Cross-link physical-chemistry for thermodynamics.

Ultrasound and microwave intensification

• Sonochemistry. Acoustic cavitation generates transient hotspots (~5000 K, ~1000 atm at bubble collapse). Activates Zn, Mg, Al powders; accelerates emulsion polymerization, particle synthesis. Hielscher and Sonotec systems at pilot scale.
• Microwave-assisted synthesis. Dielectric heating of polar solvent or substrate. Anton Paar Monowave, Milestone, CEM Discover, Biotage Initiator — laboratory scale dominant. Industrial scale-up debated (penetration depth ~few cm; throughput limits); continuous MW reactors (MicroSolution, Lambda) carve a niche.

Combined PI in pharma — sertraline case study

The Pfizer sertraline (Zoloft) redesign (Taber-Bender-Ricci 2002; Lipton 2006 Pfizer Sustainable Manufacturing Award) is the canonical industrial green-chemistry case. The original 1991 route used four solvents (toluene, THF, hexanes, MeOH), TiCl4 as a Lewis acid mediator, stoichiometric Pd/C hydrogenation, and produced ~144 kg solvent per kg product. The redesign:

• Consolidated to a single solvent — ethanol — for imine formation, asymmetric hydrogenation, salt formation, and recrystallization.
• Eliminated TiCl4 by relying on EtOH as proton donor (avoided HCl, Cl2 waste, TiO2 cake).
• Doubled the catalyst loading of Pd/C from 0.05 to 0.1 wt% but recovered to >99%.
• Atom-economical asymmetric hydrogenation step.

Outcome: solvent inventory cut 60%, total waste cut ~50%, doubled yield (37% → 78%), reduced Pd loading. Won the 2002 Presidential Green Chemistry Challenge Award.

Ibuprofen — the BHC/Boots/Hoechst-Celanese route

The original Boots synthesis (1961, six steps, AE ~40%) was redesigned by BHC (Boots-Hoechst-Celanese; Elango 1991, commercialized at Bishop, Texas, 1992). The new three-step route uses:

1. Friedel-Crafts acylation of isobutylbenzene with acetic anhydride catalyzed by anhydrous HF (recyclable; replaced AlCl3 which is stoichiometric).
2. Catalytic hydrogenation of the aryl ketone to the secondary alcohol.
3. Pd-catalyzed carbonylation with CO and methanol → carboxylic acid (replaced stepwise cyanohydrin chemistry).

Atom economy rose from ~40% to ~77%; HF is fully recycled. Won 1997 Presidential Green Chemistry Challenge Award. BASF runs an analogous Pd-carbonylation route; together with BHC they supply ~50% of global ibuprofen demand (~25,000 t/yr globally).

Sitagliptin — Codexis-Merck biocatalysis

Sitagliptin (Januvia, Merck DPP-4 inhibitor for type 2 diabetes; first approved 2006) replaced a high-pressure rhodium-catalyzed asymmetric reductive amination with an engineered transaminase (ATA-117 → CDX-RTA derived from Arthrobacter sp. via Codexis directed evolution). 27 rounds of mutagenesis introduced the substrate-binding pocket changes (Savile et al. 2010 Science). Process savings: eliminated Rh, eliminated H2 high-pressure equipment, switched solvent from MeOH/DCM to MeOH/water, raised volumetric productivity 53%, increased overall yield by 10-13%, reduced PMI ~50%. Won 2010 Presidential Green Chemistry Challenge Award. Now a textbook biocatalysis case.

---

Life-cycle assessment (LCA)

LCA quantifies environmental impact across cradle-to-gate or cradle-to-grave. ISO 14040/14044 standards. Four phases:

1. Goal and scope definition. Functional unit (e.g., 1 kg API; 1 kg auto-grade steel; 1 GJ delivered energy), system boundary, allocation rules.
2. Inventory analysis (LCI). Material and energy flows in/out of each unit process. Databases: ecoinvent (Swiss; >18,000 LCI datasets), GaBi (Sphera), Agri-footprint, USLCI.
3. Impact assessment (LCIA). Convert flows into damage categories. Methods: ReCiPe (Goedkoop-Heijungs), IPCC 2013/2021 GWP, ILCD, TRACI (US EPA), Eco-indicator 99 (legacy). Categories: GWP (kg CO2-eq), acidification (kg SO2-eq), eutrophication (kg PO4-eq), human toxicity (CTUh), ecotoxicity (CTUe), water depletion (m³), land use, ozone depletion, particulate matter.
4. Interpretation. Hotspot analysis; sensitivity; Monte Carlo uncertainty.

Software: SimaPro (PRé Sustainability), GaBi (Sphera), openLCA (free), Brightway2 (Python; academic).

Allocation methods

When a process produces multiple products (e.g., biorefinery: ethanol + DDGS + CO2), how do you split impact? Choices: mass allocation (simplest, but ignores value), economic allocation (price-based, volatile), system expansion (avoided burden — credit the byproduct’s displaced-product impact). ISO 14044 hierarchy: avoid allocation if possible, then physical, then economic.

Common pitfalls

• Burden shifting. Cutting GWP at the expense of eutrophication; e.g., switching to biofuels reduces fossil C but may increase agricultural N runoff.
• Truncation. Cradle-to-gate stops at factory output but might miss disposal impacts.
• End-of-life. Recycling credit accounting differs (cut-off method vs avoided-burden method).
• Aggregating GWP across timescales. GWP100 vs GWP20 — methane is 28× CO2 over 100 years but 84× over 20.

LCA is increasingly required in EU REACH dossiers, product environmental footprint (PEF), and corporate Scope 3 emissions reporting.

---

Industrial applications and case studies

Pregabalin (Lyrica, Pfizer)

Pregabalin (Lyrica; CNS GABA analog; epilepsy, neuropathic pain) was originally produced by classical resolution of racemic intermediate with mandelic acid — discarding 50% of mass as unwanted enantiomer. Pfizer’s enzymatic resolution (Martinez et al. 2008 Org Process Res Dev) uses Thermomyces lanuginosus lipase to hydrolyze (S)-ester selectively. Then the (R)-ester is racemized and recycled. Result: ~50% PMI reduction, eliminated chiral chromatography, halved waste. 2008 Presidential GCC Award.

Atorvastatin (Lipitor, Pfizer/Codexis)

The atorvastatin side chain — (R)-3,5-dihydroxy ester — was originally made by classical chemistry with two cryogenic steps and chiral auxiliaries. Codexis engineered three enzymes: KRED (ketoreductase), HHDH (halohydrin dehalogenase), and GDH (glucose dehydrogenase for NADPH recycle). Engineered enzymes turned over the substrate at 100 g/L. Yield jumped from ~40% to >90%, removed two solvents, cut waste >50%. Multiple PGCCA awards (2006, 2010).

Bristol-Myers Squibb Maxalt (rizatriptan)

Conversion of an SOCl2 / DCM step to a flow reactor handling the gas-evolution step safely and continuously; eliminated batch DCM, switched to recyclable solvent system, halved cycle time.

Vertex Orkambi / Trikafta tablets

First FDA-approved continuous drug-product manufacturing line (2015 Orkambi, 2019 Trikafta). Loss-in-weight feeders → continuous twin-screw blender → roller compaction → tablet press → coating → inline NIR for content uniformity. Eliminated multiple batch unit operations; halved footprint; days-not-months changeover.

Aspirin process modernization

Bayer’s traditional aspirin batch (1899 onwards) was redesigned in the 2010s by Bayer Sustainable Synthesis Center with reactive crystallization in continuous flow; eliminated multiple acetic anhydride recycle loops; halved energy demand per kg.

Cellulose dissolution in ionic liquids — Ioncell-F

Aalto University Helsinki + VTT spin-out — cellulose dissolved in [DBNH][OAc] ionic liquid, dry-jet-wet spun into Lyocell-style textile fiber from wood pulp or recycled cotton. Pilot plant 2019; commercial scale planned 2026-2028. Alternative to viscose (CS2 cancer hazard, sulfide effluent).

Green agrochemicals — flumioxazin (Sumitomo)

Sumitomo Chemical replaced toluene with cyclopentyl methyl ether (CPME) in flumioxazin synthesis; eliminated genotoxic intermediate exposure; doubled isolation yield; 2014 PGCCA Award.

Photocatalysis at scale — Merck-Eli Lilly

Both companies report flow-photocatalysis in process development chemistry — e.g., Merck’s continuous photoredox C-H methylation in Corning AFR G3 reactors with high-power 450-nm LED arrays. Productivity (g/h per L reactor) improvements vs batch ~100-fold thanks to thin reactor optical path.

---

Practical workflows and design heuristics

Green-chemistry retrosynthesis checklist

1. Choose the most atom-economical disconnection. Prefer cycloaddition, hydrogenation, isomerization over substitution or eliminative routes when possible.
2. Eliminate protecting groups. Use orthogonal selectivity (steric, electronic, catalyst-controlled) instead.
3. Use catalysis at every bond formation. Score catalytic vs stoichiometric for each step.
4. Select solvent from CHEM21 green tier. Same solvent across multiple steps reduces solvent swaps.
5. Run convergent rather than linear. Convergent synthesis halves the worst-case yield penalty.
6. Telescope steps without isolation wherever possible.
7. Quantify PMI per step; iterate worst step.
8. Plan recycling / recovery. Catalyst recovery, solvent recovery (distillation, membrane).
9. Evaluate continuous flow at every hazardous, exothermic, or air-sensitive step.

PI decision tree

• Hazardous intermediate (low LD50, explosive, gaseous reagent)? → Flow.
• Highly exothermic (>200 kJ/mol)? → Flow with active cooling (microreactor or CSTR cascade).
• Photochemistry or electrochemistry? → Flow (thin path; precise control).
• Multiphase (gas-liquid, liquid-liquid)? → Microstructured reactor or oscillatory baffled.
• Slow reaction (>4 h)? → Maybe batch — check PI to reduce volume.
• Crystallization-coupled or slurry? → CSTR cascade or OBR; batch acceptable.

Solvent-recovery economics

Rule of thumb: solvent makeup cost dominates fine-chemical OpEx. Recovery by distillation is typically >90% efficient for low-boiling solvents (EtOH, EtOAc, IPA, acetone) and 70-85% for higher-boiling (toluene, THF, DMF). Membrane separations (organic-solvent nanofiltration; Evonik Puramem, Borsig oNF-2) enable solvent recovery without thermal load — used in pharma for solvent swap and impurity rejection.

Cost of green vs cost of cleanup

A heuristic: every $1spentoninherentprocessimprovementsaves$5-10 in downstream waste treatment and disposal. The corollary — every batch process commissioned today will cost more to retire than a continuous green design.

---

Renewable feedstocks and the biorefinery

The biorefinery extends the petroleum-refinery model to lignocellulosic biomass — corn stover, wheat straw, miscanthus, switchgrass, woody chips, forestry residues, sugar bagasse. Biorefinery output: platform chemicals, fuels, energy, materials.

Top platform molecules (DOE 2004, updated 2010, 2017)

US Department of Energy “Top Value Added Chemicals from Biomass” identified twelve candidates whose chemistry allows multiple downstream products:

1. Lactic acid. Carbohydrate fermentation; building block for PLA polylactic acid bioplastic (NatureWorks 200+ kt/yr Blair, NE). Also acrylic acid, propylene glycol, lactic acid esters as green solvents.
2. Succinic acid. Glucose fermentation (Reverdia/Roquette; BioAmber acquired by LCY Biosciences; Myriant/Genomatica licensee). Replaces maleic anhydride in some polyester resins; route to 1,4-BDO, THF, GBL.
3. Levulinic acid. Acid-catalyzed hydrolysis of C6 sugars → 5-HMF → levulinic. Route to GVL solvent, MTHF, levulinate esters.
4. 5-HMF (5-hydroxymethylfurfural). Fructose dehydration; intermediate to FDCA → PEF (Avantium polyester replacing PET; first commercial 5 kt plant Delfzijl 2024).
5. FDCA (2,5-furandicarboxylic acid). Oxidation of 5-HMF. PEF (poly(ethylene furanoate)) — bio-PET analog with superior gas barrier; Avantium Synvina (was JV with BASF, now solo).
6. Glycerol. Co-product of biodiesel transesterification; oversupply post-2010 → glycerol-to-X chemistry boom: 1,3-PDO, 1,2-PG, epichlorohydrin (Solvay Epicerol; replaces propylene route), acrolein.
7. Sorbitol. Hydrogenation of glucose; intermediate to isosorbide, propylene glycol, 1,4-anhydroglucitol.
8. Isosorbide. Double dehydration of sorbitol; rigid bicyclic diol; polycarbonate (Mitsubishi Durabio used in iPhone touchscreen frames, automotive panels), polyester, polyurethane.
9. Itaconic acid. Aspergillus terreus fermentation; methyl-substituted unsaturated diacid; polyester, polyitaconate.
10. 3-HPA (3-hydroxypropionic acid). Engineered E. coli; route to acrylic acid (replaces propylene oxidation).
11. Aspartic acid. From L-aspartate; route to polyaspartates (biodegradable polycarboxylate replacement).
12. Xylitol. Hydrogenation of xylose; sugar alcohol; food, route to ethylene glycol.

Sugars from biomass

Cellulose hydrolysis is the bottleneck. Pretreatment (steam explosion, dilute-acid, AFEX ammonia fiber expansion, ionic-liquid, organosolv) opens lignocellulose; cellulases (Trichoderma reesei industrial cocktails — Novozymes Cellic CTec3, IFF Genencor Accellerase) hydrolyze cellulose to glucose. C5/C6 mixed sugars then fermented.

Lignin valorization remains the open problem — historically burned for process heat, but lignin holds ~30% of biomass carbon and has aromatic structure that could feed BTX/phenolics chemistry. Active research: Domsjö (Sweden), Lignol, Stora Enso LineoPro, Borregaard.

Bio-based polymers in commerce

• PLA (polylactic acid). NatureWorks Ingeo; 200 kt/yr; packaging, fiber, 3D printing.
• PHA (polyhydroxyalkanoates). Bacterial polyester from Cupriavidus necator or recombinant E. coli. Danimer Nodax, RWDC Solon, Newlight AirCarbon (methane-fed).
• Bio-PE, bio-PP. Braskem (Brazil) — ethanol → ethylene → PE; ~200 kt/yr Triunfo plant since 2010.
• Bio-PET. Coca-Cola PlantBottle (30% MEG bio-based from molasses ethanol via Toyota Tsusho); 100% bio-PET requires bio-PTA (terephthalic acid) — still pre-commercial.
• PEF (polyethylene furanoate). Avantium; 5 kt/yr 2024 Delfzijl; expansion to 50 kt planned.
• Biopolyurethanes. Bio-polyols from castor oil, soy oil; Cardura, Cardolite, Croda.
• Cellulose-based. Cellophane, viscose (still ~5 Mt/yr globally; CS2-based), Lyocell (NMMO-based; recyclable solvent), Ioncell-F (ionic-liquid-based; Aalto).

Process safety and inherently safer design (ISD)

Trevor Kletz (ICI; 1978 “Hazop and Hazan” book) framed ISD: design hazards out rather than control them.

Four ISD principles

1. Minimize. Reduce inventory of hazardous material. Flow chemistry exemplifies — 100 mL reactor vs 1000 L batch holding the same intermediate.
2. Substitute. Replace a hazardous material with a safer one. Water-based vs solvent-based; aqueous H₂O₂ vs peroxyacid.
3. Moderate. Use less-hazardous conditions or forms. Dilute solution vs neat; ambient T/P vs cryo or high pressure.
4. Simplify. Reduce equipment complexity; fewer valves/joints → fewer leak paths.

Bhopal as the textbook anti-case

The 1984 Union Carbide Bhopal disaster (3000+ immediate deaths, 200,000+ long-term injuries) released ~40 t methyl isocyanate (MIC) — a high-volume intermediate stored on-site because it was cheaper than buying as needed. ISD would have: produced MIC on-demand in flow (Minimize), used a less-hazardous intermediate (Substitute — phosgene + amine direct route in flow eliminates MIC altogether), produced at lower T (Moderate), removed multi-tank storage (Simplify). Modern carbamate synthesis (Eli Lilly, Bayer CropScience) uses flow MIC chemistry inline. Kletz: “What you don’t have can’t leak.”

Process hazard analysis

• HAZOP (hazard and operability). Structured deviation-keyword walk-through (no flow, more flow, reverse flow, etc.). Identifies plausible failure modes.
• LOPA (layer of protection analysis). Quantitative — IPL (independent protection layers), PFD (probability of failure on demand), SIL (safety integrity level 1-4).
• DOW Fire and Explosion Index, Mond Index. Hazard scoring.
• QRA (quantitative risk assessment). Frequency × consequence; fatality risk per year.
• Bow-tie analysis. Causes → top event → consequences with barriers.

Reactive chemistry hazards

Adiabatic temperature rise ΔT_ad = ΔH_rxn / (m c_p). If ΔT_ad pushes T past decomposition threshold of a reagent/product → runaway. Calorimetry (DSC, ARC accelerating rate calorimeter, Mettler RC1, METTLER Reaction Calorimeter) characterizes ΔH_rxn, onset T, T₂₄ (24-h ARC adiabatic decomposition T).

Stoffel/Brogli classification:

• Class 1: T_p (process T) < T_D24 (24-h ARC onset T); no runaway risk.
• Class 2-3: requires cooling or quench within hours.
• Class 4-5: minutes; high attention.
• Beyond Class 5: requires inherently safer redesign — too risky to operate.

Flow chemistry typically drops Class by 1-2 due to reduced inventory and superior heat transfer.

Green analytical chemistry

The analytical lab consumes solvent (HPLC, GC, prep), generates electronic waste, and burns energy. Green Analytical Chemistry (GAC; Galuszka-Konieczka-Migaszewski 2012) defines twelve principles paralleling Anastas-Warner:

1. Direct analysis (no derivatization).
2. Smaller sample.
3. In-situ measurement (real-time PAT).
4. Integrated sample preparation.
5. Automation, miniaturization.
6. Avoid derivatization.
7. Avoid waste generation.
8. Multi-analyte methods.
9. Reduce energy use.
10. Renewable reagents.
11. Eliminate toxic reagents (HF, chromate, cyanide for cleanup).
12. Operator safety.

Metrics: AGREE tool (Pena-Pereira 2020), GAPI, AMVI. UHPLC vs HPLC: smaller particles, narrower bore, sub-2 µm columns reduce solvent ~5-10× per analysis. Greener mobile phases: ethanol or supercritical-CO2 chromatography replacing acetonitrile.

Industrial enzymes at scale

Cross-link to biocatalysis section above. The global industrial enzyme market is ~$10 B/yr (2024) with the following major segments:

• Detergents (~40% — proteases for laundry/dish, lipases, amylases, mannanases). Novozymes/IFF Genencor/AB Enzymes/DSM Mibelle competition. Subtilisin BLAP (Bacillus lentus) is the workhorse protease.
• Food and beverage (~25%): amylases for starch hydrolysis, glucose isomerase for HFCS, lactase, rennet (chymosin from engineered K. lactis or A. niger; replaces calf rennet for 90% of global cheese), pectinase for fruit juice.
• Textiles (~10%): cellulases for denim biostoning, catalases for peroxide removal.
• Animal feed (~10%): phytase (releases phosphate from phytate), xylanase, β-glucanase.
• Biofuel (~5%): cellulases for cellulosic ethanol — never fully commercial despite POET-DSM, Abengoa, Granbio attempts.
• Pharma (~5%): biocatalysis for chiral intermediates.

Process: submerged fermentation in 200-500 m³ stirred tanks; high titers (100+ g/L enzyme) achieved by engineered Aspergillus, Trichoderma, Bacillus secretion + medium optimization + strain engineering. Downstream: cell removal, ultrafiltration, polishing chromatography, granulation/encapsulation.

Microwave and ultrasound process intensification — extended

Microwave heating

Dielectric heating of polar molecules; energy absorption ∝ ε” loss factor of solvent + substrate at 2.45 GHz (industrial-scientific-medical band). Penetration depth in water ~1-2 cm at 2.45 GHz; in nonpolar solvents (toluene, hexane) negligible. Useful for accelerating polar reactions, but throughput-limited by penetration depth.

Polar-tag strategies (graphite, silicon carbide passive heaters, ionic-liquid additives) extend microwave applicability to nonpolar systems by introducing a strong absorber. Microwave-flow combinations (Milestone flowSYNTH, Monowave-Flow) couple intensified heating with continuous processing.

Reported rate enhancements often attributed to “non-thermal microwave effects” remain controversial — Kappe and others argue careful temperature measurement shows enhancements are explained by superheating and rapid bulk heating, not specific microwave-molecule interactions.

Sonochemistry mechanisms

Acoustic cavitation: ultrasound at 20-1000 kHz creates pressure oscillations exceeding solvent tensile strength → microbubbles → grow over 100s of cycles → implode adiabatically. Bubble-collapse temperatures ~5000 K and pressures ~1000 atm in transient hot-spot (Suslick 1990 Science). Drives:

• Sonolysis of water. OH• and H• radicals; analog of radiolysis.
• Mechanochemical activation. Cleaning, dispersion, emulsification.
• Particle synthesis. Acoustic cavitation nucleation; sub-100-nm metal oxides, alloys.
• Activation of solid reagents. Zn, Mg dust activation for Reformatsky and Grignard.
• Sonoluminescence. Light emission from cavitation — exotic mechanism.

Industrial sonochemistry: Hielscher UIP1000hd, UIP4000hd flow-cell reactors at kW-scale; food processing (homogenization, emulsification), cleaning, biofuel transesterification, mineral processing.

Atom-efficient C-H functionalization

Direct arylation

Pd-catalyzed coupling of Ar-X with another Ar-H — bypassing pre-functionalization (no need for boronic acid or stannane). Atom-economical: only HX byproduct vs Suzuki’s HX + B(OH)3. Fagnou and Lautens pioneered (2007-2010); now applied widely in pharma. Selectivity by directing group (pyridine, amide, carboxylate) or electronic bias.

Iridium-catalyzed borylation (Hartwig-Ishiyama-Miyaura)

Ar-H + B₂(pin)₂ → Ar-B(pin) at ~80 °C with [Ir(cod)OMe]₂ + dtbpy ligand. Regioselectivity sterically controlled (less hindered C-H). Combined with Suzuki coupling → telescoped borylation-coupling without isolating boronate. Tolerates ketones, esters, halides, nitro.

Rh-, Ru-, Co-catalyzed amination and oxidation

White-Chen Fe-catalyzed C-H oxidation (Pmplexed iron catalyst with bulky tetradentate ligand); selective oxidation of strong C-H bonds with H2O2 oxidant.

C-H amination (Du Bois Rh nitrene chemistry, Davies dirhodium) enables direct N-installation onto C-H bonds; useful for synthesizing N-heterocycles avoiding multistep approaches.

Continuous crystallization

Crystallization is the workhorse downstream step in 80% of API processes — and often the limiting step for green credentials. Continuous crystallization (vs batch) improves polymorph control, particle-size distribution narrowness, productivity.

Equipment classes:

• MSMPR (mixed-suspension mixed-product-removal) cascade. Series of CSTRs at progressively lower T or higher anti-solvent; CMU Continuous Crystallization Center designs.
• PFC (plug-flow crystallizer). Tubular; tight residence-time distribution; small crystal size.
• Oscillatory baffled crystallizer (NiTech). Plug-flow-like with batch mixing; slurry-tolerant.
• DTBC (draft-tube baffled crystallizer). Industrial classical, mostly batch but continuous variants.

PAT integration (FBRM particle-size, ATR-FTIR concentration, Raman polymorph) closes the loop for online control. Janssen darunavir, Vertex ivacaftor/lumacaftor demonstrated end-to-end continuous including crystallization.

Industrial gases and on-demand chemistry

Multiple historically batch-stocked hazardous gases now produced on-demand at point of use:

• Phosgene (COCl2). CO + Cl2 over activated carbon at 100-200 °C. Lonza Visp, BASF, Covestro on-demand generators feed isocyanate / carbamate / chloroformate steps. Bayer Bhopal redesign exemplar.
• Diazomethane (CH2N2). Highly explosive; generated in flow from N-methyl-N-nitroso-urea or Diazald and used immediately. Vapourtec gas-membrane generators; Lonza, Bayer pharma use.
• Hydrogen cyanide (HCN). Andrussow or BMA process (Degussa) from CH4 + NH3 + O2 over Pt-Rh; on-demand for adiponitrile, methionine, acrylonitrile.
• Ozone (O3). O2 + corona discharge; used immediately. Water treatment, sterilization, organic oxidation (ozonolysis).
• Hydrogen. Steam methane reforming + WGS (gray H2) or electrolysis (green H2). On-site generation eliminates tube-trailer logistics and inventory.

Pharmaceutical solvent recovery economics

Solvent recovery typically requires distillation. Recovery rate determines net solvent demand:

• First-pass distillation: ~85-92% recovery for low-bp solvents.
• Multi-stage with ABE / pressure swing: ~95-98%.
• Membrane-assisted solvent recovery (organic-solvent nanofiltration; Evonik Puramem, Borsig oNF-2): ~99% recovery, low energy.

Even at 90% recovery, ~10% of solvent is virgin makeup per cycle. Net solvent mass for a 100-kg/yr API at PMI 200 → 20,000 kg solvent total; with 90% recovery → 2,000 kg/yr makeup + 18,000 kg/yr recovered. Recovery boilers consume ~0.5-1 GJ per ton of solvent recovered.

Solvent-recovery integration (heat pumps, MVR mechanical vapor recompression, divided-wall columns) cuts energy 30-50%. Modern API plants justify recovery investment within 1-2 year payback for solvents priced ≥$2/kg.

Photochemistry at scale

Flow photoreactors enable scale-up of light-driven chemistry — historically very difficult in batch because absorption follows Beer-Lambert (light penetration ~1-10 mm in dense reaction mixtures).

Hardware:

• Vapourtec UV-150. Coil of FEP tubing around UV-light source (medium-pressure Hg, LED arrays).
• Corning Advanced-Flow G1/G3 Photo. Microstructured silicon-carbide reactors with integrated LED panels.
• EhrFeld photoreactor cubes. Modular CHIP-based.
• HepatoChem PhotoRedox Box. Lab-scale screening (~100 mg).
• DSM Lightning, Pfizer Flow Photoreactor. Custom large-scale.

Industrial uses: vitamin D synthesis (industrial Roche, BASF), [2+2] cycloadditions in fragrance (artemisinin precursor — Seeberger continuous flow Singenta), photoredox in API process development (now standard at Merck, Pfizer, GSK). Productivity gains 10-100× vs equivalent batch.

LED revolution (2010+) replaced low-efficiency UV lamps with high-power monochromatic LEDs (365, 395, 450 nm). Energy efficiency 30-50% (LED → photon) vs 10-20% (lamp); precise wavelength selection enables selective excitation of catalyst over substrate. Hg lamps banned under Minamata Convention (2017+) accelerating LED adoption.

Process analytical technology (PAT)

ICH Q8/Q9/Q10 quality-by-design framework relies on PAT — real-time, in-process measurement → real-time release. Major PAT tools:

• NIR (near-infrared) spectroscopy. Online via fiber-optic probes; quantifies water, drug content, blend uniformity in tableting. Bruker MPA, Foss DS3000, Sentronic SentroPATPharma.
• Raman. Selective for polymorph, crystal form, API identity; non-invasive through glass/plastic; Kaiser-Endress Optic, Marqmetrix.
• FBRM (focused-beam reflectance measurement, METTLER PARSUM). Particle-size and -count in crystallizers; chord-length distribution.
• PVM (particle vision microscopy, METTLER). Real-time imaging.
• ATR-FTIR (ReactIR, METTLER). Concentration in liquid phase.
• Online HPLC, online LC-MS. Periodic sampling and analysis at minute timescale.
• Online NMR (Magritek Spinsolve, Bruker BenchMark). Compact magnets for in-process reaction monitoring.

PAT enables tight control loops: closed-loop crystallization with FBRM feedback adjusts anti-solvent or T to maintain target crystal size; closed-loop reaction with ReactIR adjusts feed to maintain target conversion.

Hazardous-waste minimization case studies

Pharma waste profile

Typical small-molecule API process generates (per kg API):

• 50-100 kg solvent
• 10-50 kg aqueous waste with dissolved salts and organics
• 1-5 kg solid waste (catalyst residues, filter aids, spent reagents)
• ~100-1000 kg CO2 (cradle-to-gate, ex-fermentation routes)

CRO/CMO sustainability metrics

Major contract development manufacturing organizations (Lonza, Catalent, Patheon-Thermo, Cambrex, Recipharm, WuXi STA, Asymchem) report PMI and E-factor per process; client choice increasingly considers these metrics. PMI <50 considered excellent for fine chemicals; <30 for late-stage commercial API.

Pollution prevention recognition

US EPA Presidential Green Chemistry Challenge Awards (1996-present); ACS GCI Pharmaceutical Roundtable (2005-present) industry consortium; UK Royal Society of Chemistry Green Chemistry Group; EU SusChem platform. Categories: academic, small business, industrial. Past winners include the case studies covered above.

Green chemistry in agrochemicals

Pesticide active ingredients face stricter green-chemistry scrutiny than pharma APIs because environmental release is intentional.

Replacing organochlorines

DDT (banned 1972 US), aldrin, dieldrin, chlordane, lindane phased out 1970-1990s under Stockholm Convention on persistent organic pollutants (POPs). Replaced by:

• Pyrethroids (synthetic mimics of natural pyrethrin from chrysanthemum). Permethrin, deltamethrin, lambda-cyhalothrin. Photostable, low mammalian toxicity, but resistance widespread.
• Neonicotinoids (Bayer imidacloprid 1991; Syngenta thiamethoxam, clothianidin). Highly selective for insect nicotinic acetylcholine receptors. EU banned outdoor use 2018 over bee toxicity; restricted in US.
• Diamides (Syngenta cyantraniliprole, DuPont chlorantraniliprole). Ryanodine receptor activators; very low mammalian toxicity.
• RNA interference — Greenlight Biosciences double-stranded RNA sprays target insect mRNA; first EPA registration 2023 for Colorado potato beetle.

Replacing organophosphates

Chlorpyrifos, parathion, dichlorvos — acetylcholinesterase inhibitors with high mammalian neurotoxicity. EPA banned chlorpyrifos for food use 2021; progressive global phase-out.

Sustainable herbicides

Glyphosate dominance (Roundup, Monsanto/Bayer; ~750 kt/yr globally) under pressure: IARC 2A classification (2015), litigation, weed resistance. Alternatives: glufosinate (Bayer Liberty Link), bromoxynil, mesotrione, sulfonylureas (low-dose herbicides), benzobicyclon. Bio-herbicides (Marrone Pro Granular) emerging.

Bio-based agrochemicals

Bacillus thuringiensis (Bt) toxins — 100+ years of safe use; engineered into corn/cotton (transgenic crops). Spinosad (fermentation product of Saccharopolyspora spinosa; Dow AgroSciences). Avermectin (Streptomyces avermitilis fermentation; ivermectin for veterinary; abamectin agricultural). Pheromone confusion (mating-disrupters, Suterra).

Green polymer chemistry

Bio-based monomers

• Lactide → PLA. Lactic acid from corn dextrose; ring-open polymerization to PLA. ~1 Mt/yr capacity globally (NatureWorks Ingeo 200 kt/yr Blair NE; TotalEnergies Corbion ~100 kt/yr Thailand; COFCO; Henan Jindan).
• 2,5-Furandicarboxylic acid (FDCA) → PEF. Bio-replacement for PET; superior O₂/CO₂ barrier; Avantium 5 kt/yr Delfzijl 2024; ~50 kt expansion planned 2026-2027.
• Bio-PE, bio-PP. Braskem sugarcane-ethanol to ethylene (200 kt/yr Triunfo Brazil); bio-PP via metathesis (Braskem partnership).
• Bio-MEG. UPM, India Glycols, Avantium plant-MEG (from CMF), Toyota Tsusho sugar-MEG.
• Bio-isosorbide → polycarbonate. Mitsubishi Durabio (15+ kt/yr Tsukuba Japan); replaces BPA in optical-grade PC.
• Furfural → various. Acid hydrolysis of pentose-rich biomass; route to GVL, furfuryl alcohol, MTHF, FDCA.

Biodegradable polymers

• PLA. Marine-biodegradable under composting conditions; not at ambient seawater rate.
• PHA (polyhydroxyalkanoates). Bacterial polyester; truly marine-degradable. Danimer Nodax (PHA, 30 kt/yr Bainbridge GA), Mango Materials (methane-PHA), Newlight AirCarbon (now AirCarbon Initiative), CJ BIO PHACT, RWDC Industries Solon. Cost premium ~3-5× polyethylene.
• PBAT (polybutylene adipate-co-terephthalate). Petroleum-derived but biodegradable; BASF Ecoflex, Novamont Mater-Bi. Often blended with PLA.
• PBS (polybutylene succinate). Mitsubishi BioPBS; succinic acid bio-derived.
• Cellulose acetate. Eastman, Daicel; cigarette filters, photographic film, glasses frames. Photo-degradable; marine fate disputed.

Mechanical recycling

PET, HDPE, PP commodity recycling; downgraded quality after each cycle. Closed-loop bottle-to-bottle PET (Coca-Cola, Indorama). Limitations: contamination (food residue, label adhesive), color, additive incompatibility.

Chemical recycling

• Glycolysis of PET. Ethylene glycol depolymerizes PET to BHET monomer at 200-240 °C; re-polymerized to virgin-quality. Indorama, Loop Industries (now Lyondell-Loop), Eastman methanolysis (rebooted 2022 with $1B investment in Kingsport TN PET methanolysis plant).
• Methanolysis of PET. Eastman, IBM, Loop.
• Carbios enzymatic depolymerization. Engineered cutinase (LCC variant; Tournier-Marliere 2020 Nature) depolymerizes PET to monomers in hours. 13 kt/yr demo plant Clermont-Ferrand France 2025; commercial 50 kt 2027.
• Pyrolysis to feedstock. Mixed plastic → naphtha-equivalent; Plastic Energy, Brightmark, Encina, Cyclyx (Lyondell partner). Output feeds steam crackers as “circular feedstock.”
• Polyurethane chemical recycling. Dow Renuva (glycolysis), Covestro Evocycle.

Vitrimers

Leibler 2011 Science — polymers with associative dynamic covalent bonds; below T_v solid like thermoset, above flowable like thermoplastic. Combine reformability (recyclability) of TP with strength (and chemical resistance) of TS. Mallinda, Cyclica, Adesso Advanced Materials commercial scale-up early.

Solvent selection in practice

Quick reference (CHEM21)

Green-recommended (single-shot use):

Solvent	Bp °C	Notes
Water	100	Best when chemistry tolerates; expensive recovery
Ethanol	78	Versatile; food-grade; recovery 90%+
iPA	82	Cleaner than ethanol for some workups
n-Butanol	118	Phase-separating with water
Ethyl acetate	77	Extraction; recovery 85%
iPropyl acetate	89	Like EtOAc; lower water solubility
Methyl acetate	57	Volatile; rapid stripping
Acetone	56	Universal; recovery moderate
MEK	80	Like acetone; less aqueous miscible
Anisole	154	Higher-bp; toluene replacement
Sulfolane	287	Polar aprotic; recyclable; high-T stable
2-MeTHF	80	Greener THF; biphasic with water

Amber (use with care):

Solvent	Bp °C	Issue
THF	66	Peroxide forms; substitute with 2-MeTHF, CPME
Acetonitrile	82	Cyanide formation if heated dry; limited supply
Toluene	110	Aromatic; restricted (CMR) but workable
DMSO	189	High bp; sulfide odor on contamination
MTBE	55	Groundwater contamination history; suspect

Red (avoid for new processes):

Solvent	Bp °C	Issue
Hexane	69	Neurotoxic (n-hexane → 2,5-hexanedione metabolite)
DMF	153	Reproductive toxicity; REACH restriction 2023
NMP	202	Reproductive toxicity; REACH restriction 2020
Dioxane	101	Suspected carcinogen; peroxide-forming
DCM	40	Suspected carcinogen; ozone-depleting potential
CHCl₃	61	Hepatotoxic; suspected carcinogen
Pyridine	115	Toxic; persistent odor
Et₂O	35	Peroxide; volatile; explosive
Benzene	80	Carcinogen; restricted globally

Carbon capture and utilization (CCU)

Green-chemistry adjacency: capturing CO₂ from industrial flues or air and converting it into useful chemicals or fuels.

CO₂ capture

• Amine absorption. MEA (monoethanolamine), MDEA, piperazine. Industrial standard but energy-intensive (regeneration ~3-4 GJ/t CO₂). Carbon Clean Solutions, Aker Carbon Capture, Mitsubishi Heavy Industries KS-1/KS-21, Shell CANSOLV.
• Solid sorbents. Amine-functionalized silica, MOFs (Mg-MOF-74, HKUST-1), zeolites, activated carbon. Carbon Engineering, Climeworks, Global Thermostat — direct air capture (DAC).
• Membrane. Polymeric and inorganic; lower CapEx but lower selectivity at low CO₂ concentration.
• Cryogenic. Energy-intensive but high-purity output.
• Chemical-looping. Metal oxide (Fe₂O₃, CuO) reduced by fuel, re-oxidized by air; intrinsic CO₂ separation.

CO₂ utilization

• Direct use. Beverages, greenhouse enrichment, supercritical solvent. Modest demand (~250 Mt/yr).
• Mineralization. Concrete curing (CarbonCure, Solidia), aggregate sequestration (Heirloom, CarbiCrete). Permanent carbon storage in cement/concrete matrix.
• Polymers. Polycarbonates via cyclic-carbonate route (Covestro Cardyon, Saudi Aramco Converge); aromatic polycarbonate replacements.
• Fuels and chemicals. Methanol (CRI George Olah Plant Iceland), methane, dimethyl carbonate, formic acid, ethanol via thermo- or electro-catalysis.
• E-fuels (synfuels). Sasol, Topsoe, Honeywell UOP — Fischer-Tropsch on CO₂-derived syngas. Aviation/maritime SAF (sustainable aviation fuel) push.

Policy levers

US 45Q tax credit ($85/t for CCS, $180/t for DAC under IRA); EU ETS at €70-100/t CO₂ (2024); UK CCS Cluster funding £20B over 20 years; Norway Longship CCS project; Japan METI green hydrogen / blue ammonia roadmap.

Industrial decarbonization

Cement (8% global CO₂)

• Clinker substitution. Blended cements with fly ash (waste from coal — declining supply), slag (steel byproduct), limestone, calcined clay (LC3 — Scrivener UNI Lausanne; very promising).
• Alternative binders. Geopolymers (alkali-activated aluminosilicate; Heidelberg Materials Eco-Crete, EnviroBlend), CO₂-cured binders (Solidia silicate cement).
• CCS on cement plants. Norcem Brevik (Heidelberg; Norway) first full-scale cement-plant CCS, commissioning 2025-2026.

Steel (7% global CO₂)

• Hydrogen DRI. Direct reduced iron with H₂ instead of CO/coke. HYBRIT (SSAB-LKAB-Vattenfall Sweden), H2 Green Steel, ArcelorMittal Sestao Spain. EAF (electric arc furnace) downstream.
• Electrolysis (Boston Metal molten oxide electrolysis). Direct Fe from Fe-oxide; no carbon intermediates.
• CCS on integrated steel mills. Tata IJmuiden, Drax (UK), ArcelorMittal multiple sites.

Chemicals

• Green NH₃. Yara, CF Industries, Ovako, NEOM. Water electrolysis → H₂ + N₂ from air → Haber-Bosch.
• Green H₂. See electrochemistry-energy-storage for electrolyzer chemistry.
• Green MeOH. CRI Iceland; Topsoe / Yara projects.
• Electric-arc steam cracker. BASF + SABIC + Linde demonstration (Ludwigshafen 2024) — eliminates fossil-fuel combustion in cracking heat.

---

Further reading

• Anastas, P.T., Warner, J.C. — Green Chemistry: Theory and Practice. Oxford 1998 — the canonical 12-principles text.
• Sheldon, R.A. — “The E-factor at 30: a passionate plea for sustainability metrics” (2023 Green Chem 25:1704) — three-decade retrospective from the metric’s originator.
• Plutschack, M.B., Pieber, B., Gilmore, K., Seeberger, P.H. — “The hitchhiker’s guide to flow chemistry” Chem Rev 2017, 117:11796 — comprehensive review on flow chemistry hardware and use cases.
• Constable, D.J.C., Curzons, A.D., Cunningham, V.L. — “Metrics to ‘green chemistry’” Green Chem 2002, 4:521 — GSK’s foundational PMI paper.
• Sneddon-Hayler-Prat-Wells (2016) “CHEM21 selection guide of classical and less classical solvents” Green Chem 18:288 — the industry-aligned guide.
• Bornscheuer, U.T., Huisman, G.W., Kazlauskas, R.J., Lutz, S., Moore, J.C., Robins, K. — “Engineering the third wave of biocatalysis” Nature 2012, 485:185 — vision-paper for engineered enzymes.
• Federsel, H.-J. — Asymmetric Catalysis on Industrial Scale, Blaser-Schmidt (Wiley) — case studies of metolachlor, sitagliptin, and more.