1. Development of Homogeneous Catalysts for Conversion of Synthesis Gas into Fuel Ethanol
Michael V. Mundschau, Girish Srinivas, Erik W. Andersen
Steven D. Dietz, Brady J. Clapsaddle and Steven Gebhard
TDA Research, Inc.
4663 Table Mountain Drive
Golden, Colorado, U.S.A
2012 ACS Rocky Mountain Regional Meeting
17-20 October, Westminster, Colorado
Session on Catalysis, Westin Westminster Cotton Creek I
Thursday, October 18, 2012, 1:30-2:10 p.m. (76.)
Contact: Dr. M.V.Mundschau

2. Acknowledgements
● U.S. Department of Energy Office of Science Grant Award: DE-SC SBIR Phase I and Phase II Projects ● Conference and Session Organizers

3. Outline
● Introduction ● Background: Ethanol and Methanol from Syngas Heterogeneous Catalysts: High Loss to CH4 Thermodynamics: Syngas to Ethanol ● Rationale for Homogeneous Catalysis Approach Homogeneous Route Methanol to Acetic acid ● Results ● Future Work ● Conclusions

4. Introduction: Research Goals
● Develop a Homogeneous Catalytic Route for Production of Fuel Ethanol from Synthesis Gas (CO + H2) ● Minimize Production of Side-Products and Especially Low-Value Methane ● Maximize Selectivity towards Ethanol ● Design Routes for Economic Scale-up
Homogeneous Catalysis Syngas to Ethanol, Mundschau, Srinivas, Andersen, Dietz, Clapsaddle, Gebhard

5. U. S. Consumption of Fuel Ethanol >13. 2 Billion gal
U.S. Consumption of Fuel Ethanol >13.2 Billion gal. in 2012; EPA Mandate for 2013 is 13.8 Billion gal.
Ethanol is the primary oxygenate/antiknock agent in U.S. gasoline; most produced via fermentation of glucose from cornstarch: Consumes 30% of U.S. corn crop.
What to do if crop shortages become severe? USDA estimates 27% of U.S. corn lost in 2012. 5 5

6. Need Alternative Methods for Ethanol Production
RFS 1 Mandate – Met in 2012; 13.2 billion gallon estimate.
EPA Mandate for 2013: 13.8 billion gallons.
U.S. production of grain-derived ethanol has likely reached its upper limit.
RFS 2 Mandates – 36 billion gallons/year by 2022 (From cellulosic ethanol; lignin and biomass gasification?).
What to do about crop shortages in U.S. and worldwide? What to do if corn and food shortages worldwide push up grain prices? Ethanol producers are becoming concerned about availability of grain feedstock and price competition.
Prudent to seek alternative methods for production of fuel-ethanol and/or other oxygenates. 6 6

7. One Alternative: Ethanol Production from Synthesis Gas
● Syngas could be produced by gasification of lignin, whole biomass, coal, heavy oil, petroleum coke, oil shale, etc. or from steam reforming or partial oxidation of natural gas.
● Highly selective, stable catalysts are needed for economic conversion of synthesis gas into ethanol.
Adapted from Spivey and Egbebi, 2007 7 7

8. Heterogeneous Catalysis Syngas-to-Ethanol Not Economical Due to Poor Selectivity
● Only 3% conversion per pass—controls heat evolved Se l
ectivity ( m ol % )
Cata
yst M
eOH
Highe r A
coho s
Othe O x
ygenates CH 4
Other
HCs Co mm
ents
Cu/ Zn O/ 2 3
>99
<1 C o
erc i a Me t h n K
/Cu/ 69 - 97 16
0.5
4.5 10
Ter b u c /M /C / Cr 17 6 13 49 5 37 B S F
1920 28 56 19 26 21 39 D w,
Eca e , Ra ge ue
1.5
4.0 11 18 30 61 36 58 H g
, HCs R as ed
<0.5 33 64 34 sc er Tr
psc 20 70 80 ly
Adapted from Gerber, White, Stevens (2007) Mixed Alcohol Synthesis Catalyst Screening, PNNL
● Production of a commodity chemical such as ethanol is not
likely to be economical if a large fraction of syngas is lost to
formation of low-value methane.

9. Why Have Heterogeneous Catalysts Failed to be Economical for Syngas-to-Ethanol?
● Poor selectivity in part due to heterogeneous nature of catalysts and multiple active sites. ● Inadequate control of bed temperatures: non-uniform bed temperatures (>100°C) lead to poor selectivity; various products. ● Local surface temperatures are too-high; overcome kinetic barriers leading to side products especially low-value methane. ● Catalyst poisons from coal and biomass.
Homogeneous Catalysis Syngas to Ethanol Mundschau, Srinivas, Andersen, Gebhard

10. Syngas-to-Ethanol: Thermodynamics Favors Many Products over Ethanol if Kinetics Not Controlled
● Methane is especially favored by thermodynamics.
● Myriads of higher-chain oxygenated compounds and hydrocarbons (all of the Fischer-Tropsch products) as well as C, CO2, H2O are favored relative to ethanol.
● Must minimize overcoming kinetic barriers leading to thermodynamically favored side products.
● Low, uniform temperatures must be employed for ethanol synthesis to minimize rates of reaction leading to side products.
Adapted from Twigg, 1997

11. Ethanol from Synthesis Gas
SOME CHALLENGES:
● In ethanol synthesis, temperature rise can overcome kinetic barriers leading to CH4 and other side-products.
● Much more heat needs to be dissipated relative to the commercially successful synthesis of methanol:
2H2 + CO → CH3OH; ΔH°298 = kJ∙mol-1
3H2 + 2CO → CH3CH2OH + H2O; ΔH°298 = kJ∙mol-1
3H2 + 3CO → CH3CH2OH + CO2; ΔH°298 = kJ∙mol-1
● Not simple stoichiometry as with methanol synthesis—need an additional sink for oxygen such as H2O or CO2 which causes the loss of 1/3 of either the H2 or CO of the synthesis gas.
● CO2 as the oxygen sink may be preferred, better driving force: ΔG°→ CO2 = kJ∙mol-1; ΔG°→ H2O = kJ∙mol-1 Separation of water from ethanol is energy intensive; H2 expensive.

12. Why is Syngas-to-Methanol Commercially Successful at >99%Yield?
SOME REASONS FOR TECHNO-ECONOMIC SUCCESS:
● Uses an active and selective heterogeneous catalyst: Cu/ZnO
● Relatively low ΔH°298 = kJ∙mol-1 for overall reaction: CO + 2H2 = CH3OH (vs kJ∙mol-1 for ethanol + CO2)
●Heat is released in two steps on separate active sites:
Critical step: CO2 + 3H2 = CH3OH + H2O; ΔH°298= kJ∙mol-1
WGS step: CO + H2O = CO2 + H2; ΔH°298 = kJ∙mol H2O is consumed—no water-methanol separation issues
● Only 3% conversion per pass—controls heat release (vs. >10-15% reported in recent attempts in ethanol synthesis)
● Operation below 250°C (vs. 350°C for sulfided ethanol catalysts)
● Operation at modest pressure: bar ( psi)
● Reactor walls avoid methanation catalysts: Fe or Ni are NOT allowed; carbonyls known (1913) to transport Fe and Ni to catalyst beds

13. Why Attempt Homogeneous Catalysis for Conversion of Syngas into Ethanol?
● Homogeneous catalysts can be designed with uniform active sites offering high selectivity.
● Uniform temperature is easier to achieve and control in homogeneous liquid solutions—uniform temperature aids selectivity.
● Heterogeneous catalytic routes for conversion of synthesis gas into ethanol have failed commercially despite over 100 years of effort—problems unlikely to be solved soon.
● Poor selectivity by heterogeneous catalysis—15 to >25% of CO lost to low-value CH4.
Homogeneous Catalysis Syngas to Ethanol, Srinivas, Mundschau, Andersen, Gebhard

14. TDA Solution Syngas-to-Ethanol: Use Homogeneous Catalytic Route
RATIONELLE FOR USE OF HOMOGENEOUS CATALYSTS FOR PRODUCTION OF ETHANOL FROM SYNGAS ● Unlike heterogeneous catalysts with multiple active sites, homogeneous catalysts can be designed with a single, uniform active site—thus maximizing selectivity. ● Unlike heterogeneous catalysts with active sites depending upon exact route of fabrication, homogeneous catalysts self-assemble; forming identical, uniform coordination compounds in solution. ●Unlike heterogeneous catalyst crystallites containing millions of atoms—all releasing heat in exothermic reactions, homogeneous catalysts have one or few metal atoms—less heat released per active site; thus lowering probability of overcoming activation barriers leading to side products. ●Suspension in solution produces the ultimate slurry reactor and unsurpassed opportunity for heat exchange.

15. Precedence: Homogeneous Insertion Reactions: Methanol to Acetic Acid
● Homogeneous catalysis long used for industrial production of acetic acid from methanol by an insertion reaction: CH3OH + CO → CH3COOH.
● No (or negligible) H2 is required; low probability for formation of CH4.
● From World War II to 1980s, cobalt was used: [Co(CO)4]-1 /HCo(CO)4 possibly with additional ligands to reduce volatility.
● From 1980s, cobalt replaced by Rh and Ir; >99% yield, higher activity—lower P, T and reactor size.

16. Precedence: Homogeneous Catalytic Route Methanol to Acetic Acid at >99% Yield
● Industrially successful; scale-up proven: % of the world’s acetic acid produced with homogeneous catalysts: [Rh(CO)2I2]-1; [Ir(CO)2I2]-1
● Non-volatile ions are key
● Solvent: Acetic acid
● Insertion of CO: CH3OH + CO = CH3COOH
● ΔH°298 = kJ∙mol-1
● T: °C
● P: atm ( psi)
● Low molecular hydrogen; essentially zero methane
● Reactor walls: Inert ZrO2/Zr; Rh, Ir plate out on base metal
Adapted from Yoneda, Kusano, Yasui, Pujado, Wilcher, Appl.Catal.A 221(2001)253

17. Precedence: Celanese Syngas-to-Acetic Acid
● Celanese operates the world’s largest acetic acid production plant in Clear Lake, Texas.
● Homogeneous rhodium catalyst is used to form acetic acid from methanol by an insertion reaction: CH3OH + CO → CH3COOH; >99% yield.
● Methanol produced from synthesis gas: H2 + CO → CH3OH; >99% yield using a Cu/ZnO heterogeneous catalyst.
● Celanese just announced construction of a natural gas-to-syngas-to-methanol plant (Texas)1.3 million metric tons of methanol per year by July, What are they really up to?

18. Celanese: Syngas→Acetic Acid→Ethanol
After the start of the TDA Project, Celanese has announced:
● Construction of two syngas-to-ethanol plants in China; 1 million metric tons ethanol/year—each. Coal used as source of syngas-to-methanol.
● Construction of a similar plant in Indonesia for conversion of low-grade Indonesian coal into fuel ethanol, but could utilize part of Indonesia’s trillion ft3 of natural gas.
● New Texas plant: natural gas→synthesis gas→ methanol→acetic acid to ethanol??

19. Celanese: Syngas→Acetic Acid→Ethanol
● Natural gas to syngas: CH4 + H2O→3H2 + CO
● Syngas to methanol using heterogeneous Cu/ZnO: H2 + CO→ CH3OH; >99% yield.
● Methanol to Acetic Acid using homogeneous Rh catalyst: CH3OH + CO→ CH3COOH; >99% yield.
● Acetic Acid to Ethanol using Pt/Sn or other heterogeneous catalysts, according to Celanese patents:
CH3COOH + 2H2→ CH3CH2OH + H2O yield???—not published.

20. Precedence: Texaco Homogeneous Catalyst Synthesis Gas to Oxygenated Products-1980s
Texaco homogeneous catalyst: Ru/Co = 3/1
Precursors: Ru3(CO)12 and Co(acac)3
Complex ions formed: [Ru(CO)3Br3]-1, [HRu3(CO)11]-1, [Co(CO)4]-1
Solvent: Ionic Liquid Tetrabutylphosphonium bromide
Produces distribution of
CH3OH
CH3CH2OH
CH3COOH
1-propanol, 2-propanol
ethylene glycol, acetates, and others
But very high CH4; >20% stated in patents
TDA Repeat of Texaco Method of J.F. Knifton Platinum Metals Rev. 29(2) (1985) Texaco publication was a starting point for present TDA work.

21. Separation: All Products and Intermediates Distill at <240°C from Non-volatile Solvent and Catalyst Ions
Ethanol °C (boiling points at 1 atm)
Methanol 65°C
1-propanol 98°C
2-propanol 82.5°C
1-butanol 118°C
Acetic acid 119°C
Methyl acetate 56.9°C
Ethyl acetate 77.1°C
Propyl acetate 102°C
Acetaldehyde 20.2°C
Dimethyl ether -24°C
Methane °C

22. TDA Now Investigating Homogeneous Catalysts: Varying Catalytic Metals, Ligands; Solvents, etc.
Much more selective to MeOH and EtOH
Initial goal: Improve selectivity of homogeneous catalysis vs. Texaco system.
Minimize production of methane; have already achieved < 1% CO loss to CH4.
TDA uses only non-volatile solvents and non-volatile ions of coordination compounds to allow ease of separation of volatile ethanol.
Above results used pressure: 2500 psig (170 barg); temperature: <220°C.
Molar ratio CO/H2 of 1/1; reaction time: 4 hours in a batch reactor.

23. Currently Fundamental Studies at Bench-Top Scale
Small batch reactor, inert inner walls, achieves pressures up to 3000 psi, a goal: use P <1000 psi.
Solvent added; precursors to homogeneous catalysts added.
Pressurized with bottled syngas using premixed H2/CO ratio.
Product gases sent to two GCs; one for hydrocarbons, second optimized for CO+CO2 analysis.
Liquids analyzed: third GC;MS-GC.
Analysis by FTIR, UV-Vis, NMR, thermogravimetric analysis.

24. A Continuously Stirred Reactor is also Used
Up to 3000 psi at 300°C.
Allows steady-state addition of syngas.
Allows steady-state removal of gas-phase products for analysis.
Liquid samples removed at pressure for analysis.

25. Some Results by Gas Chromatography Products of Catalyst “A”
CO/H2 Ratio: 1/1
2500 psi; 220°C;12 h
CH3OH 81.5 mol%
CH3CH2OH 14.0 %
1-propanol 0.3%
Methyl acetate 0.3%
Ethyl acetate 0.1%
IR no acetic acid
Acetaldehyde 1.3%
Methane 2.4 %

26. Turn-Over Frequency Shows that System is Definitely Catalytic
TOF= 125 mole/mole/h

27. Reaction Rates Comparable to Best Heterogeneous Catalysts but with Much Improved Selectivity
Rate: g CH3CH2OH/g active cat/h at 2500 psi; 220°C, from syngas; 2.4 mol% loss to CH4.
Best NREL: 0.24 g CH3CH2OH/g active cat/h at 1400 psi; but at 325°C and with >15 mol% loss to CH4.
The TDA rate decreased 51% by dropping pressure from 2500 to 800 psi.
Addition of 15 wt% methanol increases the rate 10 fold vs. conversion from syngas. Because the initial conversion of syngas-to-methanol appears not to be rate limiting, the added methanol may accelerate the reaction with acetic acid to form methyl acetate:
CH3COOH + CH3OH → CH3COOCH3 + H2O

28. Some Results by Gas Chromatography Products of Catalyst “A + B”
2500psi; 220°C;12h
CH3OH 65.5 mol%
CH3CH2OH 22.7 %
1-propanol 1.0% 1-butanol 0.4%
Methyl acetate 2.8% Ethyl acetate 0.9%
Acetaldehyde 3.8%
2-propanol 1.4%
Methane 1.3 %

29. Probable Catalytic Steps Based Upon Intermediates
Formation of methanol from syngas is a rapid step: H2 + CO→ CH3OH as evidenced by build-up of methanol.
Acetic acid and methyl acetate are intermediates: CH3OH + CO→ CH3COOH (CO insertion) CH3COOH + CH3OH → CH3COOCH3 + H2O CH3COOCH3 + H2 → CH3CH2OH + CH3OH; a slow step as evidenced by buildup of methyl acetate.
WGS step: CO + H2O = CO2 + H2.
Some direct reduction of acetic acid to ethanol: CH3COOH→ CH3CHO→ CH3CH2OH; last reduction is a slow step as evidenced by presence of acetaldehyde.
Some homologation to higher alcohols→higher acetates.
Can further improve catalysts to overcome rate-limiting steps; reduce residence time leading to CH4; higher alcohols.

30. Precursor TOF(molH2 ∙molcat -1∙h-1) T (°C) H4Ru4(CO)12 340 100
Precedence: Water-Gas Shift by Homogeneous Catalysts H2O + 2CO = H2 + CO2; ΔH°298 = kJ∙mol-1
Precursor TOF(molH2 ∙molcat -1∙h-1) T (°C)
H4Ru4(CO)
Ru3(CO)
Rh6(CO)
H4Os4(CO)
Ir4(CO)
Conditions: 25 bar CO (350 psi); aqueous trimethylamine +THF
●WGS reactions occur at exceptionally low temperatures vs. heterogeneous catalysts.
●Produces H2 and lowers the required ratio of H2/CO to 1/1.
●Consumes water that otherwise must be separated from ethanol; excess H2O can favor Fischer-Tropsch products.
Adapted from S. Werner, Dissertation, Ultra-Low Temperature WGS, Erlangen 2011

31. Can Convert Methanol into Ethanol; Eliminating 26
Can Convert Methanol into Ethanol; Eliminating 26.7% of Heat Needing Dissipation
15 wt% CH3OH dissolved into solvent
CO/H2=1/1; 2500 psi; 220°C
CH3OH 46.0 mol% CH3CH2OH 42.8 mol%
CH3COOCH3 4.4% CH3COOCH2CH3 3.3% n-propyl acetate 0.3%
1-propanol 1.4% CH3CHO 0.4%
CH4 1.7%

32. Conversion of Methanol into Ethanol
Methanol + syngas can be converted into ethanol.
Allows option of starting with methanol which can be made from syngas at >99% yield using Cu/ZnO.
Starting with methanol eliminates 26.7% of heat that needs to be dissipated.
Reaction rate increases 10 fold vs. direct conversion from syngas.
Methanol intermediate can be re-circulated to form more ethanol.

33. Ethanol can be Produced from Methanol at Lower Pressure—800 psi
Rate of reaction at 800 psi was about half that at 2500 psi
Used 15% CH3OH
CH3OH 65.4 mol%; CH3CH2OH 24.3 mol% 1-propanol 0.4% butanol 0.1%
Methyl acetate 1.9% Ethyl acetate 0.8%
Acetic acid 4.9%
Methane 2.2%

34. Select Data Showing Low Losses to CH4

35. Reproducibility of Homogeneous Catalyst System Using Different Catalyst Precursors

36. Catalysts Exposed to Air and Water Reform to Same Complex Ions Upon Reduction
Catalysts exposed to air and water exhibited no loss in activity or selectivity. Oxide precursors form the same complex ions as the carbonyl precursors.

37. Analysis and Characterization
Transmission FTIR (Fourier Transform Infrared) spectroscopy used especially for identification of catalytic-metal carbonyls. Shows that identical complex ions form from both carbonyls and oxides; shows that complex ions are stable at 1 atm.
Ultraviolet-visible (UV-Vis) spectroscopy used for analysis of catalytic-metal oxidation states and metal-to-ligand transitions
NMR (at Colorado School of Mines) used for further analysis.
TGA(Thermal Gravimetric Analysis) used to show stability of solvents to at least 220° (in inert or reducing gas); volatilization of products, and non-volatilization of catalysts.
Gas chromatography used—three instruments and columns (one for hydrocarbons, one for CO + CO2, one for ethanol and oxygenated compounds-ethanol industry column)
GC-Mass Spectroscopy (detection of dimethyl ether, etc.)

38. Infrared Spectroscopy for Characterization of Homogeneous Catalysts
TDA homogeneous catalysts are non-volatile anion complexes with carbonyl and other ligands
The carbonyl anions are stable even at 1 atm; high pressures otherwise needed to stabilize molecular carbonyls are not necessary
Carbonyls give very large signals in range cm-1
Carbonyl peak positions of transition metals are well established in the literature-allowing definitive peak assignments
Ethanol intermediates: acetic acid, acetates, acetaldehyde, etc. with C=O are also identified
O-H stretching vibration of ethanol is easily detected

39. Future Work: TDA Routes to Scale-Up —Bubble Column
Precedence: Bubble columns are used in large scale homogeneous catalytic reactions by rhodium complexes in hydroformylation: propene + syngas→ products.
Volatile products recovered; non-volatile solvent and homogeneous anion complexes do not evaporate.
Reactor wall design must prevent plating out of catalyst metals (Rh in hydroformylation) and dissolution of (catalytic) wall materials into solvent.
Heat of reaction is controlled to minimize formation of methane; uniform reactor temperature improves selectivity.
Residence time controlled to minimize methane and growth of longer-chain Fischer-Tropsch products.
Design must consider control of CO diffusion to minimize Fischer-Tropsch chain growth.

40. Preliminary Bubble Column Design
Initial Design calls for simple sparging of gas into solvent mixture from reactor bottom
Using standard stainless steel tubing sizes and fittings allows quick reactor aspect ratio changes
Stainless parts can be silicon coated – including sintered parts – to minimize corrosion and wall effects
The various bubble distribution methods can then be easily added and changed quickly:
Sieve trays
Packing
Shafts
Static Mixing
Homogeneous mixture simplifies reactor :
No differing particle sizes to get create differing reaction zones
Fouling of sparging surface is minimized
Industry expertise would allow quicker reactor design finalization and improved reaction results. v

41. Preliminary Bubble Column Design
Gas Controls
Cold Trap #1 N2
Condensers and traps will collect products and solvent entrained in the gas flow.
Condensing Coil H2 CO
Initial plans call for simplified Bubble Column reactor
The many variables necessitate numerous experiments to achieve the best design
Industry expertise can provide accelerated results
The ability to quickly change reactor parameters is important at this early stage will be valuable
Parameters and geometry can be changed easily to compare their effects on the reaction
Real-time gas analysis provides needed information on reaction status allowing parameter changes for current experiment optimization.
Cold Trap #2
Off-gas to GC and gas analyzers
Pressure Control Valve

42. Future Work Optimize catalysts; vary ligands further.
Further lower operating pressure.
Refine analysis of diffusion limitations.
Further quantify dimethyl ether and additional side products.
Work on product separation issues; re-circulation of methanol and other intermediates.
Work on heat transfer, viscosity issues.
Optimize residence time.
Perform techno-economic analysis; identify major cost factors and minimize.
Work with partners to identify routes to scale-up.

43. Future Work
Can an efficient, one-step synthesis of ethanol from synthesis gas be optimized, or is it better to separate and optimize each intermediate step?
1) H2 + CO → CH3OH; Cu/ZnO, psi, 250°C >99% yield with proven commercial catalysts.
2) CH3OH + CO → CH3COOH; psi, °C >99% yield
3) CH3COOH + CH3OH → CH3COOCH3 + H2O, distill 56.9°C
4) CH3COOCH3 + 2H2 → CH3CH2OH + CH3OH Use Cu/ZnO, 220°C, 500 psi, recycle CH3OH

44. Conclusions
● Homogeneous catalysis provides new lower-temperature routes for conversion of synthesis gas into ethanol. ● Rate of ethanol production equal to best NREL heterogeneous catalyst. ● Production of methane <1 mol %. ● System far from optimized. ● Routes to scale-up need to be addressed.
Homogeneous Catalysis Syngas to Ethanol, Srinivas, Mundschau, Andersen, Gebhard,