Process Design and Economic Analysis of Renewable Isoprene from

Feb 13, 2019 - Techno-economic analysis of a chemical process to manufacture ... Millisecond Pulsed Films Unify the Mechanisms of Cellulose Fragmentat...
0 downloads 0 Views 3MB Size
Subscriber access provided by University of Glasgow Library

Article

Process Design and Economic Analysis of Renewable Isoprene from Biomass via Mesaconic Acid Daniel Lundberg, David Lundberg, Kechun Zhang, and Paul J. Dauenhauer ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/ acssuschemeng.9b00362 • Publication Date (Web): 13 Feb 2019 Downloaded from http://pubs.acs.org on February 13, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Process Design and Economic Analysis of Renewable Isoprene from Biomass via Mesaconic Acid Daniel J. Lundberg1,2,*, David J. Lundberg1,2,*, Kechun Zhang1,2, Paul J. Dauenhauer1,2,† 1

Chemical Engineering and Materials Science, University of Minnesota, 421 Washington Ave. SE, Minneapolis, MN 55455 2 Center for Sustainable Polymers, University of Minnesota, 209 Smith Hall, 207 Pleasant Street SE, Minneapolis, MN 55455 *Authors contributed equally †Corresponding Author: [email protected] Keywords: Isoprene, Itaconic Acid, Mesaconic Acid, Hydrogenation, Dehydration, Techno-economic, Process Design

Abstract. Combined fermentation and thermocatalytic conversion of biomass to isoprene comprises a hybrid process to provide the key monomer in the manufacturing of renewable synthetic rubber. In this work, design and economic evaluation of a chemical process considers the three-step process chemistry: (a) fermentation of glucose to either mesaconic or itaconic acid, (b) catalytic hydrodeoxygenation of mesaconic or itaconic acid to 3-methyl-tetrahydrofuran, and (c) catalytic dehydra-decyclization of 3methyl-tetrahydrofuran to isoprene. Detailed reaction and separation systems were designed to maximize catalytic yield to isoprene and recover it with high purity. An economic sensitivity analysis identified hydrodeoxygenation and dehydra-decyclization catalytic selectivity as the critical opportunities for improving process economics. The process based on existing catalytic performance achieves a minimum sale price of isoprene (defined as the price which results in a project net present value of zero) of $4.07 kg1 ($1.85 lbm-1) at a scale of 100,000 metric tons yr-1 of mesaconic acid purchased at $1.00 kg-1. Six process enhancements based on improved future catalytic technology are considered, with several scenarios achieving a minimum sale price of isoprene below $2.50 kg-1 ($1.13 lbm-1).

ACS Paragon Plus Environment

Page 1

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 36

Introduction. Natural and synthetic rubber are critical soft material ingredients of automobile tires, adhesives, cements and numerous applications requiring rubbery materials such as hoses, conveyors, and seals. The natural form of rubber is a polymer obtained from the liquid latex of Hevea brasiliensis, the rubber tree of southeast Asia1,2, while newer technologies are producing rubber from the white latex of dandelions via genetic engineering3,4. Extraction of natural rubber is also being explored from other plants including guayule, goldenrod and sunflowers3. Naturally-occurring rubber consists of isoprene (2-methyl1,3-butadiene) monomers comprising the solid material, poly(cis-1,4-isoprene), which has high cis stereoselectivity5. Alternatively, ‘synthetic natural rubber’ is synthesized from the polymerization of fossilderived isoprene5 with tunable polymerization selectivity for cis-1,46, trans-1,4, or 3,4-polyisoprene7. Isoprene for synthetic rubber has been traditionally manufactured from fossil fuels including naphtha and gas oil as a minor byproduct in ethylene production8. Thermal cracking of larger hydrocarbons down to ethylene results in incomplete decomposition to a mixture of products; a small quantity of isoprene is separated and purified from this complex hydrocarbon stream via extractive distillation 9. This process manufactures approximately a million metric tons per year, but it requires the byproducts of the continued cracking of larger fossil-derived hydrocarbons for ethylene, which are currently being replaced with ethane crackers provided by the rapid advancement of shale gas in the United States of America10,11. Newer technologies to manufacture isoprene from small molecules such as isobutene and formaldehyde via Prins condensation12 or selective dehydration of renewable feedstocks13 will provide isoprene via independent, ‘on-purpose’ chemical processes. We have recently proposed a hybrid fermentation-thermocatalytic14 chemical process to manufacture isoprene from biomass-derived glucose as depicted in Figure 1. In the first step, glucose is fermented to a branched diacid such as mesaconic or itaconic acid. Subsequent hydrodeoxygenation of mesaconic acid proceeds via a cascade of reduction reactions to produce 3-methyl-tetrahydrofuran (3MTHF). In the last step, catalytic ring-opening dehydration of 3-MTHF yields isoprene in a vapor-phase packed bed reactor. Recently, an artificial metabolic pathway to branched mesaconic acid was designed and experimentally demonstrated. By screening and optimizing pathway enzymes such as glutamate mutase, 3methylaspartate ammonia lyase and reactivatase, the engineered E. coli produced mesaconic acid at a titer of 6.96 g/L directly from glucose15. In later work, further metabolic engineering by driving carbon flux and overexpressing sugar transporters led to the production of mesaconic acid with a titer of 23.18 g·L-1 and a yield of 0.465 g·g-1 glucose (64.4% of the theoretical maximum)16. In addition, it was demonstrated that mesaconic acid could be produced from lignocellulosic feedstock with high efficiency17. Diacids such as mesaconic acid or itaconic acid produced via fermentation undergo cascade hydrodeoxygenation designed to combine numerous reduction steps into a single, multi-phase reactor.

ACS Paragon Plus Environment

Page 2

Page 3 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Mesaconic or itaconic acid undergoes initial hydrogenation to methyl-succinic acid, followed by carboxylic acid reduction. Because the carbon-carbon double bond of either mesaconic or itaconic acid rapidly hydrogenates and processes identically, either feedstock is considered for this process. Subsequent cyclization produces methyl-gamma-butyrolactone (MGBL), which can be further reduced and ring opened to 1,4-methylbutanediol (MBDO)18. In the final step, ring-closing dehydration of the methyl-butanediol produces 3-methyl-tetrahydrofuran19,20, which is promoted by homogeneous or solid acid catalysts including carbon support21. Ring opening and reduction of lactones such as gamma-butyrolactone has previously been identified as a difficult catalytic reaction which can be promoted by the addition of Re to bimetallic catalysts22. In this case, the incorporation of Re into a Re-Pd/C catalyst at a 3.5:1.0 Pd/Re ratio maximized the overall rate of conversion to form 3-methyl-tetrahydrofuran from mesaconic acid21. In the third step, 3-methyl-tetrahydrofuran undergoes acid-catalyzed dehydra-decyclization to isoprene and byproduct water. Dehydra-decyclization of THF and related molecules has been demonstrated over a variety of solid acid catalysts and metal oxides. A ternary mixed oxide catalyst of V-Ti-P exhibited 70% selectivity to pentadiene from 2-methyltetrahydrofuran at conversion over 60%23. A phosphorouscontaining metal salt (sodium phosphate) achieved 31% yield of butadiene from tetrahydrofuran at 375 °C24,25. The use of silica-alumina (SiO2/Al2O3) to dehydrate 2-methyltetrahydrofuran achieved 68% yield of pentadienes at 623 K26. Alternatively, all silica zeolites containing phosphoric acid (P-Zeosils) exhibited high selectivity to butadiene (99%) from tetrahydrofuran27, pentadienes (85-99%) from 2methyltetrahydrofuran27, and isoprene from 3-methyltetrahydrofuran28. In this work, the combined three chemistries of glucose fermentation, mesaconic acid hydrodeoxygenation, and 3-methyl-tetrahydrofuran dehydra-decyclization will be incorporated into a conceptual process design with reaction, separation, and supporting process equipment. Techno-economic analysis includes determination of the costs associated with process chemistry and separation, and catalyst performance metrics are associated with economic potential. Major process parameters including the reaction and separation performance as well as chemical prices are utilized for a rigorous economic evaluation and optimization of the entire process. 2.0 Methods. A process was developed to assess the economic and industrial feasibility for the production of isoprene via sequential hydrodeoxygenation and dehydra-decyclization of mesaconic or itaconic acid. The designed process includes four unique operational blocks: (1) hydrodeoxygenation of mesaconic acid to 3-MTHF, (2) 3-MTHF purification, (3) 3-MTHF dehydra-decyclization to isoprene, and (4) purification of isoprene. A process flow diagram is shown in Figure 2, including an auxiliary block for glucose fermentation to mesaconic acid. Details of the fermentation block are not considered in the present techno-

ACS Paragon Plus Environment

Page 3

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 36

economic analysis; however, the purchase price of mesaconic acid reflects the costs associated with fermentative production. 2.1 Thermodynamic Simulation and Process Chemistry. The process was simulated using Aspen Plus (V10.0 Aspen Technology). The hydrodeoxygenation and 3-MTHF purification process blocks were simulated using the NRTL equation of state, and the dehydra-decyclization and isoprene purification process blocks were simulated using the Peng-Robinson equation of state. The RadFrac, RStoic, Flash2, Decanter, Heater, Pump and Compr Aspen Plus modules were used in the simulation. Economic accounting was completed using Microsoft Excel. 2.1.1 Mesaconic Acid Hydrodeoxygenation. Liquid phase hydrodeoxygenation of mesaconic acid was performed under 1000 psig hydrogen using a (10 wt % Pd-5 wt % Re)/C catalyst at 200 °C. Catalyst activity was selected directly from the work of Park et al. to be 2.06·10-3 mol gcat-1 hr-1, which measured 80% selectivity to 3-MTHF and 20% selectivity to 2-methyl-1-butanol (MBO) at complete conversion in a 5.0 wt% itaconic acid aqueous solution. This activity was measured via batch experiments, so a weight hourly space velocity of 2 h-1 will be used based on similar hydrodeoxygenation flow chemistries over platinum-group metal supported catalysts29. The experimentally measured carbon balance exceeded 90%, discussion of which is present in Section 5 of the Supporting Information. Catalyst on-line lifetime was approximated to be at least 24 hours21. It was assumed that short-term deactivation of the catalyst occurs primarily through coke formation and not sintering as the reaction temperature is below the Huttig temperature for both metals. Overall catalyst on-line lifetime was approximated to be five years, as compared to other supported palladium catalytic systems for high-pressure hydrogenation reactions which can typically last between five and ten years30. During operation, the reactant feed stream will be regularly switched between a parallel reactor, while the off-line reactor is regenerated by calcination with 450°C air for six hours31. It was observed that hydrogenation of the carbon-carbon double bond proceeded as the quickest step of the cascade reaction, such that mesaconic acid and itaconic acid experience identical catalytic kinetics and conversion. Based on this result, a high-pressure stirred-tank catalytic reactor (R-1 in Figure 2) was modeled for the complete conversion of mesaconic acid. In this paper, a 50 wt% mesaconic acid / 3-MTHF solution was used as the reaction mixture to prevent polymerization of mesaconic acid while also minimizing the total solvent recycled in the system. The use of 3-MTHF as a reaction solvent is discussed in Section 3.1.1. The liquid effluent of reactor R-1 was sent to a flash tank (F-1) operating at 1.0 bar and 10 °C and was flashed to remove 55% of the dissolved hydrogen in the liquid stream. The vapor stream out of flash tank F-1 contained negligible 3-MTHF and MBO and was purged. Recycle of hydrogen was not considered, as it would require a high-pressure compression system that would cost more to operate than the value of the purged hydrogen.

ACS Paragon Plus Environment

Page 4

Page 5 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

2.1.2. 3-MTHF Purification. The liquid stream from flash tank F-1 was sent to a decanter (D-1) at 60 °C and 1.0 bar. The decanter was integrated into an azeotropic distillation sequence to purify 3-MTHF from water: the minimum boiling azeotrope of 3-MTHF and water was reported at 75.8 °C and 75 wt% 3MTHF in 25 wt% water (compared to the normal boiling point of 3-MTHF of 89 °C).32 This sequence employed two distillation columns, C-1 and C-2, which were fed the organic-rich and aqueous-rich phases from decanter D-1, respectively, to obtain high purity streams of water and organic material in each column’s bottoms stream. The liquid distillate of both columns was returned to decanter D-1, and the vapor distillate of both columns was purged. The bottoms of column C-1 was sent to wastewater treatment to remove remaining organic impurities. The bottoms of column C-2 was fed to a subsequent distillation column (C-3) which removed the byproduct of MBO as bottoms at a purity of 99.9 mol%, appropriate for sale as a biofuel33,34. A purified stream of 3-MTHF was recovered as liquid distillate from column C-3, and a fraction of this stream was split and recycled to reactor R-1 to be used as reaction solvent. The vapor distillate from column C-3, containing mostly uncondensed hydrogen, was purged. Total recovery of 3MTHF through the azeotropic distillation sequence was greater than 99.7%, where most 3-MTHF was lost in the bottoms of column C-3. 2.1.3. Dehydra-Decyclization and Recycle of 3-MTHF. The fraction of purified 3-MTHF not recycled to reactor R-1 was heated, vaporized, and sent to the gas-phase dehydra-decyclization reactor (R2) operating at 300 °C where 3-MTHF was converted to isoprene over a phosphorous containing selfpillared pentasil zeolite catalyst (P-SPP). Catalyst performance for this reactor was selected directly from the work of Abdelrahman et al. to be 4.18·10-3 mol 3-MTHF gcat-1 hr-1 at a space velocity of ~1 s-1with 69.0% selectivity to isoprene at 16.6% conversion of 3-MTHF Major byproducts included piperylene (1,3pentadiene), 1,4-pentadiene, butene, and propylene at 12.9, 4.3, 11.6, and 1.3% selectivity, respectively. Formaldehyde was also produced in an equimolar ratio with butene and propylene. Again, experimentally measured carbon balance exceeded 90%. Catalyst on-line lifetime was approximated to be at least 24 h based on packed-bed flow reactor experiments28. The P-SPP catalyst can be regenerated after deactivation via coke formation by calcination with air at 550°C for 12 hours, while the feed is switched to a parallel reactor to maintain continuous operation of the process. Experimentally, P-SPP has been shown to maintain its activity and selectivity after three regeneration cycles between 24 hour periods of use, and phosphorouscontaining zeolites have been shown to maintain activity on-line for more than 10 days. Therefore, the overall catalytic lifetime was approximated as three months, requiring replacement of catalyst in both reactors every six months30,35,36,37. The effluent of reactor R-2 was cooled and sent through a packed bed adsorption column (A-1) containing 3A molecular sieves, which reduced the water content of the stream to 20 wt%.) as compared to typical industrial molecular sieve drying processes, such as breaking the aqueous azeotrope of ethanol (4.4 wt% water) or drying natural gas streams (