Interdisciplinary development of an overall process concept from glucose to 4,5-dimethyl-1,3-dioxolane via 2,3-butanediol

The study addresses the development of carbon-efficient, bio-hybrid fuel production routes by converting glucose to 4,5-dimethyl-1,3-dioxolane via 2,3-butanediol. Motivated by the need to reduce CO2 emissions in transportation, the work explores integrated process concepts combining microbial production of acetaldehyde, enzymatic transformations to acetoin and 2,3-butanediol, and chemo-catalytic steps to the cyclic acetal. The research question is whether such interdisciplinary routes can outperform or match a benchmark process (fermentative 2,3-butanediol production followed by chemocatalytic acetalization) in terms of specific energy demand and overall yield, and which parameters dominate performance. The purpose is to identify optimal catalyst combinations and solvent systems and to quantify bottlenecks impacting process efficiency.

Background work highlights cyclic acetals as promising bio-hybrid fuels with favorable combustion properties. Conventional fermentative 2,3-butanediol production is mature, but its downstream separation from aqueous broths is energy intensive, with distillation remaining the common though inefficient solution. Alternative separation and process intensification strategies have been proposed in literature, yet practical implementation often defaults to distillation. Prior studies demonstrate enzymatic cascades for vicinal diol synthesis in non-conventional micro-aqueous reaction systems (MARS) using organic solvents, which can reduce separation energy due to lower solvent enthalpy of evaporation. Acetaldehyde can be microbially produced but is toxic, necessitating in situ removal; gas stripping with absorption has been studied at lab scale. A ruthenium [Ru(triphos)(tmm)] catalyst has been reported for hydrogenation and acetalization reactions relevant to forming cyclic acetals from diols using CO2 and H2. These works inform the proposed hybrid routes and separation strategies.

• Experimental catalysis: Enzymatic ligation of acetaldehyde to acetoin using a ThDP-dependent lyase (PfBAL, whole lyophilized cells) and subsequent reduction to 2,3-butanediol using an alcohol dehydrogenase (LbADH) were conducted at 30 °C and 1 bar, both in aqueous medium and in MARS with cyclopentyl methyl ether (CPME). Aqueous yields: ligation 82.8%, reduction 60.3%. MARS yields: ligation 59.3%, reduction 56.9%. Isopropanol served as co-substrate for cofactor recycling, producing acetone. Enzyme inhibition by acetaldehyde required limiting feed concentration to 200 mM.
• Chemocatalysis: Homogeneous [Ru(triphos)(tmm)] with HNTf2 co-catalyst in CPME. Acetoin hydrogenation to 2,3-butanediol at 90 °C, 100 bar H2 for 16 h achieved full conversion. Acetalization of 2,3-butanediol with CO2/H2 (1:3) at 90 °C, 80 bar for 18 h yielded 4,5-dimethyl-1,3-dioxolane with 18% yield.
• Separation design: Acetaldehyde removed from fermenter off-gas via counter-current absorption in water and recovered by distillation (Henry constant H=2.59 bar for acetaldehyde in water; absorber designed for 5 theoretical stages). For MARS-based steps, extraction of acetoin or 2,3-butanediol from CPME into water was designed using COSMO-RS-predicted LLE (BP-TZVPD-FINE) and AspenPlus Hunter-Nash method; solvent-to-feed ratios: 0.14 (acetoin) with 7 stages, 0.035 (2,3-butanediol) with 5 stages, targeting 95% recovery. Phase separation and distillation of 4,5-dimethyl-1,3-dioxolane/water used COSMO-RS VLLE with experimental vapor pressure integration (boiling point 89.33 °C at 1 bar; heteroazeotrope ~77 °C; azeotrope ~57 mol% acetal/43 mol% water). Antoine parameters: A=10.4172, B=4813.9946, C=209.2139.
• Process modeling: Aspen Plus V11 with NRTL for non-ideality. Fermenter at 30 °C, 1 bar; acetaldehyde yield 0.4655 g/g_glucose (95% theoretical). Air flow based on P=5 g/L/h and vvm=1.5. Enzymatic steps modeled as stoichiometric reactors with yields matching experiments (aqueous: 0.8, 0.7; MARS: 0.7, 0.6). Chemocatalytic reactors modeled stoichiometrically with full conversion at stated T/P; downstream flash to remove H2 (1 bar, 30 °C). Distillation columns designed to 99 wt% purity via D/F specs and reflux optimization. Six process variants assessed: two routes (R1 enzymatic reduction; R2 chemocatalytic hydrogenation) crossed with three downstream options (DIST-AQ, DIST-ORG, EXT-DIST). Specific energy demand computed as total heat duty divided by product mass of 4,5-dimethyl-1,3-dioxolane; overall yields included conversion and separation yields.

• Enzymatic performance: Aqueous medium gave higher ligation yield (82.8%) and reduction yield (60.3%) than MARS (59.3% ligation; 56.9% reduction). PfBAL is inhibited above 200 mM acetaldehyde.
• Chemocatalysis: Acetoin→2,3-BDO hydrogenation reached full conversion at 90 °C, 100 bar H2 in 16 h. 2,3-BDO→4,5-dimethyl-1,3-dioxolane acetalization at 90 °C, 80 bar (CO2/H2=1:3) yielded 18% product.
• Separation designs and data: Absorption of acetaldehyde in water (H=2.59 bar) designed with 5 stages; desorption via distillation recycles water. Extraction solvent-to-feed ratios minimized to 0.14 (acetoin) and 0.035 (2,3-BDO). 4,5-dimethyl-1,3-dioxolane forms a heteroazeotrope with water (~77 °C) enabling phase split and efficient distillation; pure compound boils at 89.33 °C (1 bar).
• Process comparison: Among six new concepts, R2.DIST-ORG (enzymatic acetoin in MARS, chemocatalytic hydrogenation, distillation from organic solvent) had the highest overall yield and lowest specific energy demand within the proposed routes, yet remained less efficient than the benchmark (direct 2,3-BDO fermentation + distillation + acetalization).
• Energy breakdown for best new concept (R2.DIST-ORG): 13.5% of specific energy demand from fermentation/acetaldehyde separation; 82.2% from separation of acetoin/2,3-BDO; 4.3% from acetalization and product purification.
• Benchmark for 2,3-BDO separation energy (distillation from water): mean 26.97 kJ/g (literature range ~24.74–28.68 kJ/g); model reproduced 27.5 kJ/g.
• Best-case scenario (assuming full enzymatic conversion and minimized solvent usage/stage counts): specific energy demand 42.33 kJ/g, requiring a 32.2% reduction to match benchmark.
• Sensitivity to enzyme stability: Increasing allowable acetaldehyde concentration in the enzymatic cascade to 600 mM (from current 200 mM) achieves break-even in specific energy demand versus benchmark.
• Additional figures: Acetoin and 2,3-BDO production/separation in aqueous solvents (DIST-AQ) had the highest energy use; DIST-ORG reduced energy but slightly reduced yields due to lower MARS conversions; EXT-DIST minimized energy but incurred yield losses in extraction.

The integrated bio- and chemo-catalytic routes demonstrate technical feasibility for producing 4,5-dimethyl-1,3-dioxolane from glucose, but the current implementations are outperformed by the benchmark process due to low concentrations and yields in the enzymatic steps. The dominant energy burden arises from separating dilute enzymatic products, amplified by aldehyde toxicity limiting acetaldehyde feed to 200 mM. While MARS facilitates downstream energy savings via lower solvent evaporation enthalpy and enables solvent continuity into chemocatalysis, enzyme activity penalties reduce overall yields. Using chemocatalysis for acetoin reduction (R2) avoids enzymatic yield losses and, when paired with distillation from organic solvent, gives the best trade-off among the new concepts. However, to become competitive, the enzymatic stage must tolerate higher acetaldehyde levels to boost product titers, thereby reducing separation energy. The analysis quantifies this requirement, identifying 600 mM acetaldehyde tolerance as a target for enzyme engineering. Overall, the study clarifies the critical parameters and offers a pathway to integrate catalysts and solvent systems effectively for carbon-efficient fuel synthesis.

The work develops and evaluates interdisciplinary process routes that combine microbial production of acetaldehyde, enzymatic synthesis of acetoin/2,3-butanediol, and homogeneous Ru-catalyzed transformations to 4,5-dimethyl-1,3-dioxolane. All conversion steps were experimentally validated, and separation strategies were designed and modeled. Among proposed concepts, the route using chemocatalytic hydrogenation (R2) and distillation from an organic solvent showed the best performance but still lagged the benchmark in energy efficiency. The main bottleneck is the enzymatic stage, specifically enzyme stability toward acetaldehyde, which limits feed concentration and thus product titers. A best-case analysis indicates that raising acetaldehyde tolerance to about 600 mM would enable parity with the benchmark specific energy demand. Future efforts should focus on enzyme engineering (stability enhancement via rational design, directed evolution, immobilization), process modes (e.g., fed-batch or continuous low-dose feeding), and further optimization of extraction/distillation schemes to reduce energy consumption and improve overall yields.

• Enzymatic steps are at an early development stage with significant inhibition by acetaldehyde, capping feed at 200 mM; conclusions rely on sensitivity analyses for higher tolerances.
• Several property datasets (e.g., ternary LLE for extraction) were predicted via COSMO-RS due to lack of experimental data, introducing model uncertainty.
• Absorption of acetaldehyde has been demonstrated mainly at lab scale; scale-up performance and solvent management assumptions may vary.
• Best-case scenarios assume full enzymatic conversion and minimized solvent use/ideal stage counts, which may not be fully attainable industrially.
• Chemocatalytic acetalization yield (18%) is currently modest; catalyst performance and selectivity (lack of stereoselectivity) could affect downstream separation and overall efficiency.
• Some process inputs (e.g., fermentation productivity, aeration rates) were assumed from analogous processes and may differ in practice.