Methanol Feed Research Articles

Kinetic Characterization of Pt/Al2O3 Catalyst for Hydrogen Production via Methanol Aqueous-Phase Reforming

Compared to steam reforming, methanol aqueous-phase reforming (APR) converts methanol to hydrogen and carbon dioxide at lower temperatures, but also displays lower conversion rates. Herein, methanol APR is studied over the active Pt/Al2O3 catalyst under different operating conditions. Studies were conducted at different temperatures, pressures, methanol mass fractions, and residence times. APR performance was evaluated in terms of methanol conversion, hydrogen production rate, hydrogen selectivity, and by-product formation. The results revealed that an increase in operating pressure and methanol mass fraction had an adverse effect on the APR performance. Conversely, it was found that hydrogen selectivity was unaffected by the operating pressure and residence time for the methanol feed mass fraction of 5%. For the methanol feed mass fraction of 55%, hydrogen selectivity was improved by operating pressure and residence time. The alumina support phase change to boehmite as well as sintering and leaching of the catalytic particles were observed during catalyst stability experiments. Additionally, a comparison between methanol steam reforming (MSR) and APR was also performed.

Efficient production of itaconic acid from the single-carbon substrate methanol with engineered Komagataella phaffii

BackgroundAmidst the escalating carbon dioxide levels resulting from fossil fuel consumption, there is a pressing need for sustainable, bio-based alternatives to underpin future global economies. Single-carbon feedstocks, derived from CO2, represent promising substrates for biotechnological applications. Especially, methanol is gaining prominence for bio-production of commodity chemicals.ResultsIn this study, we show the potential of Komagataella phaffii as a production platform for itaconic acid using methanol as the carbon source. Successful integration of heterologous genes from Aspergillus terreus (cadA, mttA and mfsA) alongside fine-tuning of the mfsA gene expression, led to promising initial itaconic acid titers of 28 g·L−1 after 5 days of fed-batch cultivation. Through the combined efforts of process optimization and strain engineering strategies, we further boosted the itaconic acid production reaching titers of 55 g·L−1 after less than 5 days of methanol feed, while increasing the product yield on methanol from 0.06 g·g−1 to 0.24 g·g−1.ConclusionOur results highlight the potential of K. phaffii as a methanol-based platform organism for sustainable biochemical production.

Adaptive state feedback controller design for efficient biodiesel production under kinetic uncertainty

In this study, we optimize the quality and economic performance of biodiesel production through a scenario-based adaptive state feedback control framework. Our goal is to maximize the operational efficiency while ensuring the production of on-spec biodiesel, despite uncertainties in reaction kinetics. A first-principle model is employed to describe the process dynamics, with kinetic parameters estimated online using a moving horizon estimator (MHE). A pool of state feedback controllers is created via off-line nonlinear optimization and metaheuristic algorithm based on sampled kinetic scenarios. Then, the uncertainty space is partitioned into several clusters, each linked to an optimal controller from the pool. During online operation, the appropriate controller is selected by matching the estimated kinetic parameters to the corresponding cluster. Simulation studies on a semi-batch reactor demonstrate that manipulating the methanol feed flow rate and heat duty with our control approach significantly improves operational efficiency, reduces online computational time, and enhances robustness to kinetic uncertainties compared to model predictive control (MPC).

Design and optimization of reactive distillation for enhancing supercritical transesterification process to produce biodiesel

The present study explored the optimization of a reactive distillation column (RDC) for biodiesel production using supercritical transesterification (SCTE). The SCTE process at high pressure and temperature, is known for water-sensitive flexibility and catalyst-free operation, and has shown promising advantages in biodiesel production. The steady-state simulations were conducted using Aspen Plus, with two RDC configurations: RDC-1, employing a single feed of oil and methanol, and RDC-2, utilizing separate feeds. Surprisingly, high conversion rates were achieved at only 8.5 MPa pressure. The study aiming to identify an optimized RDC design, systematically examined the impact of design parameters such as reflux ratio, feed temperature, and the number of reactive stages on conversion and energy requirements. Internal column specifications and species flow through each stage were investigated to comprehend reaction and separation processes. Temperature analysis revealed that preheating to 380 °C significantly improved conversion and reduced reboiler heat duty. RDC-1, with 9 reactive stages, demonstrated greater conversion efficiency, while RDC-2, with 11 stages, exhibited better biodiesel separation. By optimizing reflux ratios, the study achieved remarkable conversions with enhanced methanol separation. Cost estimation revealed that RDC-1 promoted lower capital and operating costs, making it a preferred design and an efficient option for SCTE-based biodiesel production.

Improvement of the recombinant phytase expression by intermittent feeding of glucose during the induction phase of methylotrophic yeast Pichia pastoris.

For growth of methylotrophic yeast, glycerol is usually used as a carbon source. Glucose is used in some cases, but not widely consumed due to strong repressive effect on AOX1 promoter. However, glucose is still considered as a carbon source of choice since it has low production cost and guarantees growth rate comparable to glycerol. In flask cultivation of the recombinant yeast, Pichia pastoris GS115(pPIC9K-appA38M), while methanol induction point(OD600) and methanol concentration significantly affected the phytase expression, glucose addition in induction phase could enhance phytase expression. The optimal flask cultivation conditions illustrated by Response Surface Methodology were 10.37 OD600 induction point, 2.02h before methanol feeding, 1.16% methanol concentration and 40.36μL glucose feeding amount(for 20mL culture volume) in which the expressed phytase activity was 613.4 ± 10.2U/mL, the highest activity in flask cultivation. In bioreactor fermentation, the intermittent glucose feeding showed several advantageous results such as 68h longer activity increment, 149.2% higher cell density and 200.1% higher activity compared to the sole methanol feeding method. These results implied that remaining glucose at induction point might exhibit a positive effect on the phytase expression. Glucose intermittent feeding could be exploited for economic phytase production and the other recombinant protein expression by P. pastoris GS115.

Development of pure hydrogen generation system based on methanol steam reforming and Pd membrane

Methanol steam reforming (MSR) has been used for in-situ hydrogen production, while hydrogen purification remains a technical challenge to limit the applications of using hydrogen for fuel cells. This study presents a pure hydrogen generation system (PHGS) by integrating MSR and hydrogen purification modules, this system enables low-temperature reforming and high-temperature purification to achieve high yield and recovery of hydrogen. To investigate the combined effects of feedstock flow rates and pressure on MSR and purification, the performance of each module and overall PHGS system were tested. The results showed that PHGS can stably produce high purity hydrogen (>99.99%) under different conditions, the synergistic effects of two modules lead to hydrogen yield exceeding the numerical sum of individual modules. At methanol feed rate of 1.00 g/min and system pressure of 5.0 bar, the pure hydrogen flow rate, CO content, pure hydrogen yield, hydrogen recovery and thermal power output were tested about 1.70 ± 0.01 L/min, 2.4 ± 0.1 ppm, 81.1 ± 0.5%, 87.6 ± 0.5% and 350 W. The PHGS can deliver up to 500 W of thermal power and is suitable for small range-extended electric vehicles. Our work on the developed PHGS can provide a significant reference for future practical applications of systems combining MSR and membrane purification.

Enhancing methanol tolerance of Thermomyces lanuginosus lipase by rational design and biodiesel production through one-step feeding of methanol

The addition of a significant quantity of methanol solution during the biodiesel production results in the denaturation of lipase. To address this issue, this study proposed a novel approach to the preparation process by designing a methanol-tolerant Thermomyces lanuginosus lipase (TLL) using a computer-aided strategy. In silico prediction yielded seven mutants, out of which two mutants, E129Y and E134W, exhibited varying degrees of improved methanol tolerance and catalytic efficiency. The methanol tolerance and stability of E129Y and E134W were observed to increase by 10–30 % and 17–35 %, respectively. Additionally, E134W exhibited a 1.13-fold increase in catalytic efficiency compared to the wild type TLL, while E129Y showed a reduction of 12.05 %. Following a one-shot addition of methanol (molar ratio of 3:1 methanol to Cornus wilsoniana oil) for 72 h in the transesterification reaction, the yields of fatty acid methyl ester (FAME) for E129Y and E134W were 69 % and 85 %, respectively, with the yield of E134W increasing by 22 %. Structural analysis indicated that the E129Y mutation formed a single hydrogen bonding interaction, while E134W produced hydrophobic interactions with neighboring amino acids, thereby augmenting the structural stability of the mutations under high methanol concentration. This study has the potential to advance the commercialization of lipases in the biodiesel production industry.

Mathematical Modelling and Simulations of Active Direct Methanol Fuel Cell

A one dimensional isothermal model is proposed by modelling the kinetics of methanol transport at anode flow channel (AFC), membrane and cathode catalyst layer of direct methanol fuel cell (DMFC). Analytical model is proposed to predict methanol cross-over rate through the electrolyte membrane and cell performance. The model presented in this paper considered methanol diffusion and electrochemical oxidation at the anode and cathode channels. The analytical solution of the proposed model was simulated in a MATLAB environment to obtain the polarization curve and leakage current. The effect of methanol concentration on cell voltage and leakage current is studied. The methanol cross-over has the significant impact on cell performance. The presented model predicts higher leakage current with the increase of methanol feed concentration. The cell performance was predicted at 70°C and various methanol feed concentration. The proposed model was validated with the experimental polarization curve of active DMFC.

High performance production process development and scale-up of an anti-TSLP nanobody

Nanobodies (Nbs) represent a class of single-domain antibodies with great potential application value across diverse biotechnology fields, including therapy and diagnostics. Thymic Stromal Lymphopoietin (TSLP) is an epithelial cell-derived cytokine, playing a crucial role in the regulation of type 2 immune responses at barrier surfaces such as skin and the respiratory/gastrointestinal tract. In this study, a method for the expression and purification of anti-TSLP nanobody (Nb3341) was established at 7 L scale and subsequently scaled up to 100 L scale. Key parameters, including induction temperature, methanol feed and induction pH were identified as key factors by Plackett-Burman design (PBD) and were optimized in 7 L bioreactor, yielding optimal values of 24 °C, 8.5 mL/L/h and 6.5, respectively. Furthermore, Diamond Mix-A and Diamond MMC were demonstrated to be the optimal capture and polishing resins. The expression and purification process of Nb3341 at 100L scale resulted in 22.97 g/L titer, 98.7% SEC-HPLC purity, 95.7% AEX-HPLC purity, 4 ppm of HCP content and 1 pg/mg of HCD residue. The parameters of the scaling-up process were consistent with the results of the optimized process, further demonstrating the feasibility and stability of this method. This study provides a highly promising and competitive approach for transitioning from laboratory-scale to commercial production-scale of nanobodies.

Oxidative stress resistance prompts pyrroloquinoline quinone biosynthesis in Hyphomicrobium denitrificans H4-45

Pyrroloquinoline quinone (PQQ) is a natural antioxidant with diverse applications in food and pharmaceutical industries. A lot of effort has been devoted toward the discovery of PQQ high-producing microbial species and characterization of biosynthesis, but it is still challenging to achieve a high PQQ yield. In this study, a combined strategy of random mutagenesis and adaptive laboratory evolution (ALE) with fermentation optimization was applied to improve PQQ production in Hyphomicrobium denitrificans H4-45. A mutant strain AE-9 was obtained after nearly 400 generations of UV-LiCl mutagenesis, followed by an ALE process, which was conducted with a consecutive increase of oxidative stress generated by kanamycin, sodium sulfide, and potassium tellurite. In the flask culture condition, the PQQ production in mutant strain AE-9 had an 80.4% increase, and the cell density increased by 14.9% when compared with that of the initial strain H4-45. Moreover, batch and fed-batch fermentation processes were optimized to further improve PQQ production by pH control strategy, methanol and H2O2 feed flow, and segmented fermentation process. Finally, the highest PQQ production and productivity of the mutant strain AE-9 reached 307 mg/L and 4.26 mg/L/h in a 3.7-L bioreactor, respectively. Whole genome sequencing analysis showed that genetic mutations in the ftfL gene and thiC gene might contribute to improving PQQ production by enhancing methanol consumption and cell growth in the AE-9 strain. Our study provided a systematic strategy to obtain a PQQ high-producing mutant strain and achieve high production of PQQ in fermentation. These practical methods could be applicable to improve the production of other antioxidant compounds with uncleared regulation mechanisms.Key points• Improvement of PQQ production by UV-LiCl mutagenesis combined with adaptive laboratory evolution (ALE) and fermentation optimization.• A consecutive increase of oxidative stress could be used as the antagonistic factor for ALE to enhance PQQ production.• Mutations in the ftfL gene and thiC gene indicated that PQQ production might be increased by enhancing methanol consumption and cell growth.

Expression of recombinant human insulin precursor by Pichia pastoris in a 10 liter bioreactor

Recombinant insulin is a vital medicine for diabetic patients. This hormone is produced by microbes such as Pichia pastoris that carry the recombinant gene of a human insulin precursor (HIP). Large-scale protein production involves a bioreactor to promote the optimal condition for the yeast to express the protein target. In order to obtain a large amount of insulin, the cultivation of recombinant P.pastoris/pD902-IP carrying human insulin precursor gene in a bioreactor 10 Liter was developed. The isolate was cultivated in a half concentration of basal salt media for 124.5 hours. Induction of the protein was done by continual methanol feeding. The fermentation condition was set to have a temperature at 28°C, agitation at 300 rpm, aeration at 2 L/min and a pH value of around 5. Dry cell weight (DCW) was measured, and HPLC quantified the content of HIP, glycerol and methanol. This work’s DCW and HIP concentrations were 46.5 g/L and 928 mg/L, respectively. The results can be higher by increasing the number of cells in the culture or extending the cultivation time so that the HIP concentration may exceed 1 g/L.

Engineering the synthetic β-alanine pathway in Komagataella phaffii for conversion of methanol into 3-hydroxypropionic acid

BackgroundMethanol is increasingly gaining attraction as renewable carbon source to produce specialty and commodity chemicals, as it can be generated from renewable sources such as carbon dioxide (CO2). In this context, native methylotrophs such as the yeast Komagataella phaffii (syn Pichia pastoris) are potentially attractive cell factories to produce a wide range of products from this highly reduced substrate. However, studies addressing the potential of this yeast to produce bulk chemicals from methanol are still scarce. 3-Hydroxypropionic acid (3-HP) is a platform chemical which can be converted into acrylic acid and other commodity chemicals and biopolymers. 3-HP can be naturally produced by several bacteria through different metabolic pathways.ResultsIn this study, production of 3-HP via the synthetic β-alanine pathway has been established in K. phaffii for the first time by expressing three heterologous genes, namely panD from Tribolium castaneum, yhxA from Bacillus cereus, and ydfG from Escherichia coli K-12. The expression of these key enzymes allowed a production of 1.0 g l−1 of 3-HP in small-scale cultivations using methanol as substrate. The addition of a second copy of the panD gene and selection of a weak promoter to drive expression of the ydfG gene in the PpCβ21 strain resulted in an additional increase in the final 3-HP titer (1.2 g l−1). The 3-HP-producing strains were further tested in fed-batch cultures. The best strain (PpCβ21) achieved a final 3-HP concentration of 21.4 g l−1 after 39 h of methanol feeding, a product yield of 0.15 g g−1, and a volumetric productivity of 0.48 g l−1 h−1. Further engineering of this strain aiming at increasing NADPH availability led to a 16% increase in the methanol consumption rate and 10% higher specific productivity compared to the reference strain PpCβ21.ConclusionsOur results show the potential of K. phaffii as platform cell factory to produce organic acids such as 3-HP from renewable one-carbon feedstocks, achieving the highest volumetric productivities reported so far for a 3-HP production process through the β-alanine pathway.

Enhanced production of human interferon α2b in glycoengineered Pichia pastoris by robust control of methanol feeding and implications of various control strategies

Interferon α2b (IFNα2b), a prominent cytokine molecule, is bestowed with immunomodulatory, antiproliferative, and antiviral properties. Like any other recombinant protein produced in methylotrophic yeast Pichia pastoris, Human interferon α2b (huIFN α2b) production in P. pastoris depends on methanol utilization during the induction phase. This study investigates the impact of varying residual methanol concentrations on huIFNα2b productivity and assesses control strategies towards the tight control of residual methanol concentration in fed-batch experiments carried out in the fermentation calorimeter. Real-time monitoring of P. pastoris metabolism was facilitated by metabolic heat rate measurements and exhaust gas analyser. Optimal methanol concentration screening performed at different concentrations viz., 1.5, 3, 5, and 7 g/L, revealed the highest huIFNα2b productivity at 3 g/L (0.21 mg/g h). Methanol sensor provided input data for three different control strategies (on/off, proportional integral (PI), and model-based adaptive PI) to maintain optimal methanol levels (3 g/L). Model-based adaptive control resulted in the highest huIFNα2b yield, productivity (244.34 mg/L, 0.31 mg/g h), surpassing on/off (188.7 mg/L, 0.16 mg/g h) and PI (202.3 mg/L, 0.28 mg/g h) control strategies. The model-based adaptive PI control framework developed in this study could be employed for other heterologous protein production in P. pastoris.

Production and Purification of Soy Leghemoglobin from Pichia pastoris Cultivated in Different Expression Media

Plant-based meat alternatives, exemplified by Impossible Foods’ Impossible Burger, offer a sustainable, ethical substitute for traditional meat, closely mimicking the taste and appearance of meat by utilizing soy leghemoglobin (LegH), a 16 kDa holoprotein found in soy plants structurally similar to heme in animal meat. Cultivation medium plays an important role in bioprocess development; however, medium development or optimization can be labor intensive, and thus the use of previously reported media can be enticing. In this study, we explored the expression of recombinant LegH in Pichia pastoris in various reported cultivation media (BSM, BMGY, FM22, D’Anjou, BSM/2, and RDM) and using different feeding approaches (µ-stat and mixed feed with sorbitol). Our findings indicate that optimization techniques tailored to the specific process did not increase LegH yields, highlighting the need to investigate strain-specific strategies. We also utilized the collected process data to create and train a novel artificial neural network-based soft sensor for estimating cell biomass, relying solely on standard bioreactor measurements (such as stirrer speed, dissolved oxygen, O2 enrichment, base feed, glycerol feed, methanol feed, and reactor volume). This soft sensor proved to be robust and exhibited a strong correlation (3.72% WCW) with experimental data.

Development of a continuous fermentation process for the production of recombinant uricase enzyme by Pichia pastoris.

Recent studies in the biopharmaceutical industry have shown an increase in the productivity and production efficiency of recombinant proteins by continuous culture. In this research, a new upstream fermentation process was developed for the production of recombinant uricase in the methylotrophic yeast Pichia pastoris. Expression of recombinant protein in this system is under the control of the AOX1 promoter and therefore requires methanol as an inducing agent and carbon/energy source. Considering the biphasic growth characteristics of conventional fed-batch fermentation, physical separation of the growth and induction stages for better control of the continuous fermentation process resulted in higher dry-cell weight (DCW) and enhanced recombinant urate oxidase activity. The DCW and recombinant uricase activity enzyme for fed-batch fermentation were 79g/L and 6.8u/mL. During the continuous process, in the growth fermenter at a constant dilution rate of 0.025h-1 , DCW increased to 88.39g/L. In the induction fermenter, at methanol feeding rates of 30, 60, and 80mL/h, a recombinant uricase activity was 4.13, 7.2, and 0u/mL, respectively. The optimum methanol feeding regime in continuous fermentation resulted in a 4.5-fold improvement in productivity compared with fed-batch fermentation from 0.04u/mL/h (0.0017mg/mL/h) to 0.18u/mL/h (0.0078mg/mL/h).

Calcium hydroxy phosphate-functionalized graphene oxide/Nafion composite membrane for direct methanol fuel cell

A direct methanol fuel cell (DMFC) is one prominent alternative energy technology. However, there are many problems that prevent the commercialization of DMFC, such as methanol crossover, low power density, and degradation of the membrane. To resolve these issues, functionalized graphene oxide (GO) and calcium hydroxy phosphate (CHP) were incorporated into the Nafion membrane to make a composite membrane. The content of the CHP was varied to determine and further investigation involved spraying the same amount of CHP onto the Nafion membrane to compare with the composite membrane. The fillers were characterized using Fourier-transform infrared spectroscopy and X-ray Powder Diffraction. The ex-situ tests investigated water uptake, solubility, and chemical degradation, while the in-situ tests investigated electrochemical impedance (EIS), methanol permeability, fuel cell performance, and durability. Adding both 1 wt% of CHP and 0.05 wt% functionalized GO into the Nafion membrane, resulting in a 1.86-fold higher durability than that of recast Nafion. In addition, the maximum power density of the composite membrane was 1.34-fold higher (75.41 mW/cm2) than that of the recast Nafion (56.11 mW/cm2) at 1 M methanol and 70 °C. At 2 M methanol and 70 °C, the maximum power density of the composite membrane increased to 80.19 mW/cm2 but then decreased to 46.00 mW/cm2 when the methanol feed increased to 4 M, due to the large amount of methanol crossover.

Direct production of diethyl carbonate from ethylene carbonate and ethanol by energy-efficient intensification of reaction and separation

This study presents a novel process for producing diethyl carbonate and ethylene glycol directly from ethylene carbonate and ethanol through reactive distillation. Previous studies have mainly focused on investigating two reactive distillation systems with a pressure swing azeotropic separation for producing diethyl carbonate. By integrating the two separated reactive distillation systems into a single flowsheet, an energy-efficient process is enabled. Especially, the need for separating the azeotrope mixture is eliminated, and the primary feed of methanol is not required due to the reaction stoichiometry. Two design alternatives have been proposed based on the reactive distillation configuration and catalyst employed. The first design significantly reduces the amount of equipment needed by utilizing two reactive distillation columns, while the second design offers high energy savings, as all reactions occur in a single reactive distillation column. Compared to the conventional process of producing individual dimethyl and diethyl carbonate, both design alternatives result in total energy consumption reductions of 12% and 24%, respectively, as well as total annual cost reductions of 12% and 14%, respectively. Furthermore, both design alternatives demonstrate a reduction in CO2 emissions by 15% and 11% respectively. Not only do these results demonstrate their economic viability, but they also indicate their effectiveness in CO2 mitigation. From an industry perspective, the two proposed design alternatives can be effectively implemented for the direct production of diethyl carbonate and mono-ethylene glycol by intensifying the existing dimethyl carbonate production process.

Co-feeding strategy alleviates hypoxic stress in large-scale bioreactor for optimal production of secretory recombinant proteins in Pichia pastoris.

The loss of mixing efficiency inherent to bioreactor process operated at large-scale yields to the formation of concentration gradient and thus to heterogeneous culture conditions. For processes operated with methanol feeding, P. pastoris faces oscillatory culture conditions that significantly affect the cell ability to produce secretory recombinant proteins at high yield. Extended cell residence time in microenvironments of high methanol concentration and low oxygen availability that are typically found in the upper part of the bioreactor near the feeding point, triggers the unfolded protein response (UPR) and thus impairs proper protein secretion. Methanol co-feeding with sorbitol was shown herein to reduce the UPR response and to restore productivity of secreted protein.

High methanol tolerant proton exchange membranes based on novel coupling-type sulfonated poly(phenylquinoxaline) for direct methanol fuel cells

Coupling-type sulfonated poly(phenylquinoxaline) (c-SPPQ) bearing both side chain and main chain sulfonic acid groups were synthesized through the post-sulfonation process. The proton exchange membranes prepared with c-SPPQ showed good performance, such as low methanol permeability, high methanol tolerance and good mechanical properties due to the acid-base interaction between the sulfonic acid and phenylquinoxaline groups. The dimensional change of c-SPPQ-3 in the length direction was 33% at 30 °C in 45 wt% methanol solution, demonstrating the stability of membrane electrode assembly against shearing force. In contrast, NR212experienced significant swelling. The methanol flux of c-SPPQ-3 was 45 mg cm−2 h−1 for at 30 °C with a methonal concentration of 10 wt%. This was almost 10 times lower than that of NR212. With a 20 wt% methanol feed concentration at 75 °C, the maximum power density of c-SPPQ-3 was 88 mW cm−2, 1.8 times higher than that for NR212 (48 mW cm−2). This was attributed to the reasonably high proton conductivity and low methanol permeability of c-SPPQ-3. When the methanol feed concentration was further increased to 50 wt%, the maximum power density of c-SPPQ-3 slightly decreased to 60 mW cm−2, while that of NR212 was negligible. With such superior performance, c-SPPQ shows excellent promise for meeting the requirements of direct methanol fuel cells.

Hybrid Deep Modeling of a GS115 (Mut+) Pichia pastoris Culture with State–Space Reduction

Hybrid modeling workflows combining machine learning with mechanistic process descriptions are becoming essential tools for bioprocess digitalization. In this study, a hybrid deep modeling method with state–space reduction was developed and showcased with a P. pastoris GS115 Mut+ strain expressing a single-chain antibody fragment (scFv). Deep feedforward neural networks (FFNN) with varying depths were connected in series with bioreactor macroscopic material balance equations. The hybrid model structure was trained with a deep learning technique based on the adaptive moment estimation method (ADAM), semidirect sensitivity equations and stochastic regularization. A state–space reduction method was investigated based on a principal component analysis (PCA) of the cumulative reacted amount. Data of nine fed-batch P. pastoris 50 L cultivations served to validate the method. Hybrid deep models were developed describing process dynamics as a function of critical process parameters (CPPs). The state–space reduction method succeeded to decrease the hybrid model complexity by 60% and to improve the predictive power by 18.5% in relation to the nonreduced version. An exploratory design space analysis showed that the optimization of the feed of methanol and of inorganic elements has the potential to increase the scFv endpoint titer by 30% and 80%, respectively, in relation to the reference condition.