Part A case study: catalytic car converter A pretext to speak about environmental catalysis (mobile sources) imc « Catalytic clean-up technologies » allow struggling against pollutions UCL Challenges for the catalysis Heterogeneous catalysis is a must to convert air pollutants in molecules not/less toxic for the environment 0 CO co2 NOx N2 SOx H2S S° o3^o2 Hydrocarbons -> C02 + H20 Volatile Organic Compounds (VOC) -> C02 + H20 without producing NOx (-» N2) / SOx (-» H2S) / Cl2 (-» HCl) for COV containing heteroatom (N, S, CI) Challenges for the catalysis and water ??? -> Organic molecules : hydrocarbons + fatty esters (oils : industrial + eatable) + dyes + solvents same approach as VOC and hydrocarbons in the air total oxidation -> Nitrates : = denitrification on (membrane) catalysts UCL Catalysis vs Trapping/Scrubbing Trapping / Scrubbing = Washing (for ex : on CaO milk) or Adsorption on active carbon or other porous solids (for ex : zeolites) Not everything is trappable : (CO, 03 = No) vs (NOx, SOx= Yes) Ad/bsorbing is not eliminating but only shifting a pollution !!! -> What to do when the adsorbant is saturated ??? (industrial dumping + risk of leaching /combustion + air pollution) vs destruction of pollutants by catalysis (for ex : dioxins) Replacement cost of adsorbants vs Starting investment for catalysis Automotive catalysis UCL Composition of exhaust gases gasoline engine Compound Vol % Compound Vol % CO 1.5 (0.1-6) o2 1.0 (0.2-2) HC (eq. CI) 0.5 (0.2-1) H2 0.4 NOx 0.15 (0.05-0.4) H20 10-12 S02 15-20 ppm co2 11-13 P 15-20 ppm N2 -> 10 times more CO than NOx Variations with tuning of the engine age of the vehicle and type of driving // age of driver UCL Composition of exhaust gases gasoline engine vs diesel Compound Vol % Compound Vol % CO 1.5 (0.1-6) 02 ^LO (0.2^2})^ HC (eq. CI) 0.5 (0.2-1) H2 0.4 NOx 0.15 (0.05-0.4) H20 10-12 S02 15-20 ppm co2 11-13 P 15-20 ppm N2 - to Diesel gasoline ratio CO, H: HC so: NOx C^~* t0 ^^Pr HCCI engines Particulates * to - (except HCCI engines --) Diesel = Much more 02 6-10% Before HDS of diesels -> Today = Homogeneous combustion engines (T° «) -> less NOx but more unburnt HC (under development) UCL Reactions to achieve !!! -> 3-ways Way 1 Way 2 Oxidation reactions : CO and hydrocarbons CO + Vi 0. HC+ O CO > C02 + H20 ex: C3H„ + 502—>3C02 + 4H20 Reduction reactions: NOx NO —*■ y2N Pt Pt Rh or Pd or Pd or Pd but Rh more selective to N2 UCL Reactions to achieve !!! -> 3-ways Way 1 Way 2 Oxidation reactions : CO and hydrocarbons CO + Vi 0. HC+ O ex: C3H8 + 50 CO Pt > C02 + H20 Pt + 3C02 + 4H20 -Way-3 // Way 1 Reduction reactions: NOx NO + CO NO + H, y2N2 + C02 1/2N2 + H20 Rh or Pd or Pd or Pd but Rh more selective to N2 UCL Reactions to achieve !!! -> 3-ways Oxidation reactions : CO and hydrocarbons CO + 'A 0. HC + 0- * CO. -> C02 + H20 ex: C3H8 + 502—>3C02 + 4H20 Too much 02 A lot of NO -> Complete oxidation of CO (not enough CO to reduce NO) Pt Pt ■ Reduction reactions: NOx NO + CO -" V2N2 + C02 Rh NO + H2 -► 1/2N2+H20 Not enough 02 -> complete reduction of NO -» a lot of CO and unburnt HC UCL Universite catholique de Lou vain Oxidation reactions : CO and hydrocarbons CO + Vz 0. HC+ 0, * CO. -> C02 + H20 ex: C3H8 + 502—>3C02 + 4H20 Too much 02 ^ A lot of NO -> Complete oxidation of CO (not enough CO to reduce NO) pt pt ■ Reduction reactions NOx NO + CO -" %N2 + co2 Rh NO + H2 -► 1/2N2+H20 Way 1 // Way 3 Key of 3-ways = tuning the ratio residual O / reductants in the effluent tuning the ratio air/fuel at the engine admission Not enough 02 -> complete reduction of NO a lot of CO and unburnt HC UCL Universite catholique de Lou vain too reducing mix too « rich » to oxidize CO and HC 14.2 14.5 14.8 Air / Fuel [g/g] 15.1 the window where to work !!! = mix stoechiometry of air/fuel = 14.5-14.7 too oxidant mix too « lean » to reduce NO UCL Lamba / Richness / Schlater A X = 14.6 Richness = 1/X > the motorist vision Schlater s= ox 202 + NO RED CO+H2 +3nCnH2n +(3n+l)CnH2n+1 the chemist vision ponderating by the number of O atoms effectively available (OX) and by the number of O atoms effectively needed (RED) UCL Universitě catholique de Lou vain CO C CD CO i_ O 00 c CD 00 73 c CO to c g co 00 E UJ CO N without ^ with catalyst Operating window 0.85 NO, with catalyst wMhout catalyst Hydrocarbons with catalyst ** ^Hydrocarbons without catalyst oxygen sensor * (X-sensor) 0.95 1.00 1.05 1.10 1.15 1 20 Same representation but expressed in emissions -> reverse curves ! The effect of the catalyst is clearly visible : 1° less emissions globally 2° transition around X = 1 is more marked X- value Figure 10.1. Emissions of CO, NOx and hydrocarbons along with the signal from the oxygen sensor as a function of the air/fuel composition; a = 1 corresponds to the air-to-fuel ratio of 14.7. Note that the three pollutants can only be converted simultaneously in a very narrow operating window of air-to-fuel ratios. UCL Universite catholique de Lou vain in the right lambda slot ? Variation of richness during driving = f(type of driving) • short term (1Hz) WwVVS WAV, A=1 fearful/stressed driver (grand mother ?) > young nervous ? driver releases long term (0.1 HZ) accelerator ^ > driver speeds up normal behavior ? (grand father ?) 1 push on accelerator every 10s UCL How to stay in the right lambda slot ? Variation of richness during driving = f(type of driving) • short term (1Hz) WwVVS WAV, driver releases long term (0.1 HZ) accelerator fearful/stressed driver (grand mother ?) young nervous ? Problem of NOx driver speeds up Problem of CO normal behavior ? (grand father ?) 1 push on accelerator every 10s UCL How to stay in the right lambda slot ? Variation of richness during driving = f(type of driving) • short term (1Hz) WwVVS WAV, long term (0.1 Hz) LAMBDA PROBE i=> CATALYST Expanding mat — Insulates, seals and provides an unbreakable I for the Stainless steel Lambda probe Measures the residual oxygen content in the exhaust gas Ceramic monolith Substrate for the catalytic noble "ltol Catalytic layer of J noble metal Washcoat —J Ceramic substrate —1 UCL The catalyst Cordierite Steel J Cordierite Wash-coat ceo2-Ai2o3 s^qyapgag Gas flow Noble metal particles Pt, Rh, Pd 30-50 Perpendicular view \ Axial view as most dispersed (nano) as possible !!! Pt and Rh particles - Ma muw •« monolith washcoat Figure 10.5. Monolith, washcoat and noble metal particles in an automotive exhaust catalyst. UCL A typical catalyst Three-way catalyst Support: Honeycomb Monolith 400 cpi (canals per sqi) Wall thickness: 0.15 mm Washcoat 20 wt-% of a porous support composed of CeOJAl203-La typically 12-20% Ce02 (today: CeCX is replaced by CeZi03) Metals Pt-Rh(^to 1.4 g L1 Today: PdlTfitwaysa^ded Conditions Temperature 300to500°C may reach 900°C Space velocity 50000 to lOOOOOh-1 Volumic ratio: Catalyst/Engine capacity 0.8 to 1.5 j1.000.000 h-1 if considering only the wash-coat 1 -2% wt of the wash-coat UCL TWC^ Oxidation of CO 0) o E O) 0) c UJ CO + K O reaction coordinate Figure 10.6. Approximate energy diagram of CO oxidation on palladium. Note the largest energy barrier is the CO t O recombination. [Adapted from T. Engel and C. ErtlJ. Chem. Phys. 69 (1978) 1267.] -> CO + 1/2 o2 -> co2 Highly exothermic reaction (285 kJ/mole) -> On Pd : Activation energy = 100 kJ/mole for the r° CO*+0* (rate limiting step) vs Homogeneous phase : Activation energy = 500 kJ/mole for the dissociation Vz 02->0 (rate limiting step) ->-> The catalyst has modified the rate limiting step UCL TWC^ Oxidation of CO (0.5%CO + 0.5%O2) Turnover frequencies (second ~l) @ 250°C on bulk metals and Alumina -supported metals (dispersions in parentheses) Metal Pd Pt Rh unsupported 4.6 0.31 10.1 A1A-S imported 2.9 (41%) 0.9 (67%) 0.24 (7%) 0.10 (87%) 1.8 (57%) 0.4 (69%) -> R° faster on big particles than on small ones : dispersion ^ -> TOF ^ (mainly on Rh) -> Rh and Pd loose activity in presence of NO (not shown) Best catalyst = Pt NO = poison in UCL Universite catholique de Lou vain NO = poison !!! -> CO also (much more) 300 400 500 Temperature / K 600 700 Figure 10.7. CO? formation rate from CO and 02 over Rh(lll) and Rh(110) surfaces [Adapted from M. Bowker, Q. Cuo. and R.W. Joyner, Catal. Lett. 18 (1993) 119]. Note the similarity to the simple model used to describe the rate in Fig. 2.12. According Arrhenius : We should have Temp 71 -> Speed 71 Not the case: -> There is a speed maximum !!! UCL Universitě catholique de Lou vain NO = poison !!! -> CO also (much more) 300 400 500 Temperature / K 600 700 Figure 10.7. CO? formation rate from CO and 02 over Rh(lll) and Rh(110) surfaces [Adapted from M. Bowker, Q. Cuo. and R.W. Joyner, Catal. Lett. 18 (1993) 119]. Note the similarity to the simple model used to describe the rate in Fig. 2.12. According Arrhenius : We should have Temp 71 -> Speed 71 Not the case: -> There is a speed maximum !!! Low Temp : surface mostly covered by CO -> no place for O* ->-> no r° CO* + O* High Temp (> desorption temp of CO) surface covered by O* -> not enough CO* at the surface ->-> no r° CO* + O* UCL Universite catholique de Lou vain NO = poison !!! -> CO also (much more) 300 400 500 Temperature / K 600 700 Figure 10.7. CO? formation rate from CO and 02 over Rh(lll) and Rh(110) surfaces [Adapted from M. Bowker, Q. Cuo. and R.W. Joyner, Catal. Lett. 18 (1993) 119]. Note the similarity to the simple model used to describe the rate in Fig. 2.12. According Arrhenius : We should have Temp 71 -> Speed 71 Not the case: -> There is a speed maximum !!! Necessity to work at intermediate temperature at which coverage of the surface by CO and by O are comparable !!! -> Mechanism of Langmuir-Hinshelwood (cf Part lb - Section 6) UCL TWC^ Oxidation of CO because 02 -> 20 (l + KC0PC0 + K0P0^)2 Adsorption CO»>Adsorption 02 IfKcoPco»l-K0P0'^ Metal Pd Rh Support none Al203a none A120/ Ce02-Al205 none Al203a Ce02-Al203 m(02) +1.0 0 +1.0 f+L° 1 +1.0 +1.0 0 n(CO) +1.0 -1.0 1-0.9J V+0,3 J -1.0 -0.8 +0.2 Ea (kJmof1) 125s U08-133 50 125 104-125 84 117 92-113 104 Metal or Metal/alumine order-1 for the CO -> CO = inhibitor = CO adsorbs stronger on the metal and does not allow oxygen to adsorb UCL TWC^ Oxidation of CO because 02 -> 20 (l + KC0PC0 + K0P0^)2 Adsorption CO»>Adsorption 02 IfKcoPco»l+K0Po* Metal Pd Pt Rh Support none Al203a Ce02-Al203 none A120/ Ce02-Al205 none Al203a Ce02-Al203 m(02) +1.0 +0.9 0 +1.0 f+L° 1 +1.0 +1.0 0 n(CO) -1.0 -0.9 C+To) -1.0 1-0.9J V+0.3 J -1.0 -0.8 +0.2 Ea (kJmof1) 125 108-133 5o\ 125 104-125 84 117 92-113 104 Metal/ceria+alumina order +1 for the CO -> CO * inhibitor -> ceria brings new sites for the adsorption of oxygen !!! TWC^ Oxidation of CO Effect of the addition of ceria 0.1 I 10 02 concentration (vol %) A mixture « rich » in CO inhibits less Rh/Ce-Alumina than Rh/Alumina UCL TWC^ Oxidation of CO Effect of the addition of ceria CO CO - CO adsorption on metal (site #) - Oz adsortion on O vacancies of ceria (site *) - CO reaction with O species at the metal/support interface -> The adsorptions of CO and 02 are no more competitive because there are 2 distinct sites of adsorption for them UCL Universite catholique de Lou vain Effect of the addition of ceria CO -> The oxidation of CO happens thanks to O atoms from ceria (and not from the gas phase) Does it work without 02 in the reaction gas ? AUO,-CeO. CO adsorption on metal (site #) Oz adsortion on O vacancies of ceria (site *) CO reaction with O species at the metal/support interface UCL TWC^ Oxidation of CO because 02 -> 20 (l + KC0PC0 + K0P0^)2 Adsorption CO»>Adsorption 02 IfKcoPco»l+K0Po* Metal Pd Pt Rh Support none Al203a Ce02-Al203 none Al2o/ Ce02-Al205 none Al203a Ce02-Al203 m(02) +1.0 +0.9 CD +1.0 f+L° 1 +1.0 +1.0 0 n(CO) -1.0 -0.9 +i.o\ -1.0 1-0.9J V+0.3 J -1.0 -0.8 +0.2 Ea (kJmof1) 125 108-133 50 \ 125 104-125 84 117 92-113 104 With ceria, order 0 for 02 !!! = it works even without 02 in the gas !!! -> With ceria : it works more as MVK !!! UCL TWC^ Oxidation of CO Effect of the addition of ceria x> i VAVA x < i -^Ceria allows the system to get rid of the variations of richness of the gas to clean. o AI203-CeO. UCL TWC Oxidation of HC and alcools Light-off temperatures (50% conversion) Catalyst: commercial Pt-Rh/Ce02-Al203 n-alkanes T50 Alkenes, alkyne T50 Methane 515°C A Ethylene 205°C Ethane 435°C T Propene 185°C Propane 290°C Acethylene Q Hexane 195CC Light-off and T50 Aromatics T50 Alcohols T50 Benzene 205°C Methanol 195°C Toluene 220°C Ethanol 200°C O-Xylene 225°C Propanol 205°C Butanol 210°C TWC -> Oxidation of HC and alcools Light-off and T50 T50 : temperature at which catalyst has 50% of conversion T10, T90, etc... UCL TWC Oxidation of HC and alcools Light-off temperatures (50% conversion) Catalyst: commercial Pt-Rh/Ce02-Al203 n-alkanes T50 Alkenes, alkyne T50 Methane 515°C A Ethylene 205°C Ethane 435°C T Propene 185°C Propane 290°C Acethylene Q Hexane 195CC Aromatics T50 Alcohols T50 Benzene 205°C Methanol 195°C Toluene 220°C Ethanol 200°C O-Xylene 225°C Propanol 205°C Butanol 210°C In general : Alcohols easier to oxidize than alkanes -> alkanes more difficult than olefins -> Lighter alkanes -> more difficult (not true for alcohols) TWC -> Oxidation of HC and alcools Turnover frequencies (second_1) on unsupported and alumina-supported metals. Gas composition : 0.1 %HC + 1%02 -N2 HC CH4 C;H6 C3H8 C4H10 T;C Disp.% 400 350 250 225 Pd 65 0.030 0.0045 0.0014 16 I 0.31 \ 0.093 0.0072 0.0042 fail 5.4 3.6 0.25 0 19 Pt 87 - ^^^^ 0.16 6 0.0095 0.31 1.5 5 2 fail 0.017 10.0 Rh 57 0.0085 0.0095 0.0004 0.0004 7 0.017 0.011 0.0006 0.0004 foil 0.050 0.16 0.010 0.0076 -» Pt: best metal for C2-C4 -» but Pd : best for CH4 Rh always bad ! Dispersion TP Activity ^ (big particles more active) UCL TWC Oxidation of HC and alcools Effect of addition of ceria « Activity ratio » = activity of metal on Ce02- Al203 / activity of metal on Al203 Reaction Pd (0.15%) Pt (0.22%) Rh (0.15%) CH4 + 02 0.3 (400CC) (405 (50rp) 1 (500°C) C ^Hg "1" On 0.2 (350CC) 0.5 (250QC) (1(40 0tJ> -> Negative effect for Pd and Pt (ratio < 1) ->-> mainly true for Pt -> « Positive effect » for Rh (ratio > 1) TWC Oxidation of HC and alcools Effect of addition of ceria Metal Pd Pt Rh Support none A1203 Ce02-AK>3 none A1203 Ce02-Al203 none A1203 Ci02-Al203 m(02) 0 -0.1 -1 ■ 1 + 0.1 f 0 + 0.1 n (C^) -0.4 -0.6 + 1.2 + 2 + 0.5 \+0.5J + 0.4 Ea (kJ moľ1) 96 66-96 63 92 84-105 96 92 100 84 -> Pt: order for 02 = -1 vs order for C3H8 = +2 ^ adsorption 02 »> adsorption C3H8 ->-> opposite situation as for CO The surface is spontaneously more covered by 02 -> adding oxygen (via Ce02) diminishes further the place for C3H8 Consequence : activity of Pt in presence of ceria. UCL TWC Oxidation of HC and alcools Effect of addition of ceria Metal Pd Pt Rh Support none A1203 Ce02-AK>3 none A1203 Ce02-Al203 none A1203 Ci02-Al203 m(02) n (CjHJÚ Ea (kJ moľ1) 0 -0.4 96 -0.1 -0.6 66-96 i ^> 63 -1 + 1.2 92 84-105 ■ 1 + 2 96 + 0.1 + 0.5 92 f 0 \+0.5J 100 ■ + 0.1 + 0.4 U 84 -> Pt: adsorption 02 »> adsorption C3H8 -> Pd and Rh : smaller difference between orders for 02 and C3H8 ->-> effect of Ce02 less negative (or positive) via adsorption but additional activating effect via Eact ^ UCL TWC Oxidation of HC and alcools On Pd and Pt C3H8 On Rh C3H8 Constants of relative adsorption o CO c3H6 C3H6 02 CO -> Useful to predict the effect of ceria !!! UCL TWC^ Reduction of NO x Three main reactions: Side reactions (undesired) 1- Reduction b v CO ti NO + CO-> KN2 + C02 2- Reduction bv H, NO+H2 -» j + CO 5-Ammonia formation UCL TWC^ Reduction of NO X Reaction CO + NO Activity of Metal/Al203 catalysts 0.5 % NO + 2% CO Temperatures for a 50% conversion (TM) Ru, 205°C > Rh, 296 °C » Pd, 431°OPt, 471°C E * 100 kJ mol1 Relative activity (TOF) CRu, 550J Rh, 100 Pd, 1.7 Pt, 0.7 Kinetic orders 0 to - 0.4 in NO 0 to + 0.1 in CO Ru by far the best! but!!! in presence of 02 -> formation of Ru04 VOLATILE !!! One puts Rh in converter for its ability to activate CO+NO (Reminder: Rh is not efficient in oxidizing CO in the presence of NO) UCL TWC^ Reduction of NO x Reaction NO + H2 (°C) : comparison with NO + CO Catalysts: M/A1203 Catalyst NO + H2 NO + CO Pd 106 431 Pt 121 471 Rh 163 296 Ru 237 205 Reduction of NO easier with H2! -> Normal !? H2 more reducing Mainly on Pt and Pd -> Normal !? Pt and Pd activate easily Ho Pd and Pt: very active in NO reduction bv H, Rh and Ru: very active in NO reduction by CO -> reverse order of reactivity BUT ... (in the real system) ? UCL TWC^ Reduction of NO x Reaction NO + H TVl (°C) : comparison with NO + CO Catalysts: M/A1203 Catalyst NO + H2 NO+ CO NO + CO + H2 Pd 106 431 / 330V Pt 121 471 398 Rh 163 296 i 275 Ru 237 205 w temperature and order of reactivity close to those of NO+CO Pd and Pt: very active in NO reduction by H, Rh and Ru: very active in NO reduction by CO -> reverse order of reactivity BUT 1° inhibition NO+H2 by CO UCL Universite catholique de Lou vain I rci: comparison nlih NO + CO < alul>Als: M/Al,0, Catalyst NO + CO NO*CO+ H: Pd 106 431 330 Pt 121 471 398 Rh 163 296 275 Ru 237 205 210 IM .mil I'l: \ns ;hiIu- In No reduction b> II. Kh and Ku: very active in SO reduction I■ \ CO Ki-.uli.m NO ♦ II. ntronglv Inhibited l>> < O Reaction NO + H -> BUT 2° reduction of NO not to N2 in presence of H2 = NO+H2->NH3 (mainly on Pt and Pd) Selectivity at high conversion @480°C Gas composition: 1.5%NO + 4.5%CO + 4.5%H2 Rh remains the best most active most selective to N2 most stable Catalyst Conv. NO % Selectivities % NO->N2 NO^NH3 NO + CO NO + H2 Pd 94 26 9 /91\ • • • • Pt 94 23 8 • . 92 / • • •••• Rh 100 67 33 20 80 Ru 100 92 8 29 71 UCL TWC-> Further challenges ! 1 ° Improvement of engines conditions -> In order to diminish HC and C02 emissions . one must better burn the fuel (-> less unburnt) CO c CO c CO Further challenges ! 1° Improvement of engines conditions = one must work at A/F = 20 = with too much oxygen !!! -> a lot of NO in the exhaust gas 2° Diesel cars Better yield of diesel engines (less C02 formed) ^gasoline ^ ^diesel = less C (thus less energy) in 1 litre of gasoline BUT there is 6 to 10% of 0o at the exhaust!!! = price to pay to have less unburnt (better yield) -> as a counterpart, there is more NOx than for gasoline engines UCL Universitě catholique de Lou vain 1° Improvement of engines conditions = one must work at A/F = 20 = with too much oxygen !!! -> a lot of NO in the exhaust gas 2° Diesel cars Will I UVJ cL there is more NO X UCL Towards « NOx-trap » Concept (developed by Toyota -> « Toyota Process ») process in 2 steps during which the engine shifts successively from periods at lean regime (A/F ~ 20, 1 minute) to periods at rich regime (A/F ~ 10, 2 s). Towards « NOx-trap » Concept (developed by Toyota -> « Toyota Process ») TW converter Pt+Pd+Rh "NOx trap" catalyst TWC + Ba-Al203 "Lean" phase : Converter 0 HC, CO -> C02 + H20 NO -> NO, Converter © BaCO, -> Ba-nitrates "Rich" phase: converted Converter 0 HC, CO -> not totally converted and C02 NO ^N2 Converter © Ba-nitrates"^> Ba-carbonates+ (NO CO, HC + NOx -> C02 + H20 + N2 UCL Towards « NOx-trap » Concept (developed by Toyota -> « Toyota Process ») Lfiai-Oun range | StflfB UGm ataichondric range| RedbcelLÜx HC, CO. J? ftfBt3M#Jft P # nutaW O I Precwusrutis NO 02 R: Reducing agent* UCL Towards « NOx-trap » Concept (developed by Toyota -> « Toyota Process ») Why is working in cycles needed ? o 10 ~ 0 08 9^C I o. 06 5 0.04 I O 0.02 rich iean combustion 5 0 2 4 6 time (min) Figure 10.10. Principle of operation of NO* storage catalyst During lean combustion, NO rs oxidized to NO. and stored by BaO at barium nitrates. Once the getter Is saturated, a short nch excursion of the a \n-»\n. _ Converter '(), -> Ha-nitrale^ 'Rich" phase : Converter 0 IK . ( ()\-> no! tnlall) converted Converter © Ba-nitratls -» Ua-carhonates-fN, * catalysis = stoichiometric reaction UCL What else??? Automotive catalysis is also : - things about poisoning of catalysts : S, P, Si, Pb, etc (solution = heterogeneous catalytic HDS) - things about soots in diesel engines (solution = heterogeneous catalysis) - things about secondary catalytic reactions Between co-reactants present in exhaust gas = many other things to address !!!