Trickle-bed root culture bioreactor design and scale-up: growth, fluid-dynamics, and oxygen mass transfer

Trickle-bed root culture reactors are shown to achieve tissue concentrations as high as 36 g DW/L (752 g FW/L) at a scale of 14 L. Root growth rate in a 1.6-L reactor configuration with improved operational conditions is shown to be indistinguishable from the laboratory-scale benchmark, the shaker flask (mu=0.33 day(-1)). These results demonstrate that trickle-bed reactor systems can sustain tissue concentrations, growth rates and volumetric biomass productivities substantially higher than other reported bioreactor configurations. Mass transfer and fluid dynamics are characterized in trickle-bed root reactors to identify appropriate operating conditions and scale-up criteria. Root tissue respiration goes through a minimum with increasing liquid flow, which is qualitatively consistent with traditional trickle-bed performance. However, liquid hold-up is much higher than traditional trickle-beds and alternative correlations based on liquid hold-up per unit tissue mass are required to account for large changes in biomass volume fraction. Bioreactor characterization is sufficient to carry out preliminary design calculations that indicate scale-up feasibility to at least 10,000 liters.     

Study of hydrodynamics in microcarrier culture spinner vessels: A particle tracking velocimetry approach

Three-dimensional particle tracking velocimetry (3-D PTV), a modern, quantitative, visualization tool, has been applied to the characterization of the flow field in the impeller region of cell culture reactor vessels. The experimental system used here is a 250-mL microcarrier spinner vessel. The studies were conducted at three different agitation rates, 90, 150, and 210 rpm, corresponding to healthy, mildly damaging, and severely damaging shear intensities, respectively. The flow can be classified into three regions: a predominantly tangential (azimuthal) flow generated by the impeller; a trailing vortex region coming off the impeller tip; and a converging flow region close to the center of the vessel. The latter two are the regions of highest velocity gradients. Energy dissipation rates due to mean velocity gradients were also calculated to characterize the impeller stream. Local specific energy dissipation rates > 10,000 erg/(cm(3)sec) . have been measured. It is proposed that the critical regions for microcarrier culture damage due to impeller hydrodynamics are the trailing vortex region and the high energy converging flow region. Graphical representation of the mean velocity flow fields and the distribution of energy dissipation rates in the impeller region are also presented here. The merits of using the dissipation function (measure of specific energy dissipation rate) as a possible scale-up parameter are also discussed. (c) 1996 John Wiley & Sons, Inc.     

Cell growth and death rates as factors in the long-term performance, modeling, and design of immobilized cell reactors

The viable fraction of immobilized cells in a bioreactor may be critical in predicting long-term or steady-state reactor performance. The assumption of near 100% viable cells in a bioreactor may not be valid for portions of immobilized cell reactors (ICRs) characterized by conditions resulting in appreciable death rates. A mathematical model of an adsorbed cell type ICR is presented in which a steady-state viable cell fraction is predicted, based on the assumptions of no cell accumulation in the reactor and a random loss of cells from the reactor. Data on cell death rates, cell growth rates, and productivity rates as functions of temperature, substrate, and ethanol concentration for the lactose utilizing yeast K. fragillis were incorporated into this model. The steady-state reactor viable cell fraction as predicted by this model is a strong function of both temperature and ethanol concentration. For example, a stable 20% viable fraction of the immobilized cells is predicted in ICR locations experiencing continuous conditions of either 30 g/L ethanol at 40 degrees C, or 95 g/L ethanol at 25 degrees C. Steady-state ICR "plug flow" concentration profiles and column productivities are predicted at three operating temperatures, 20, 30, and 40 degrees C using two different models for ethanol inhibition of productivity. These profiles suggest that the reactor operating temperature should be low if higher outlet ethanol concentrations are desired. Three reactor design strategies are presented to maximize the viable cell fraction and improve long-term ethanol productivity in ICR's: (1) reducing outlet ethanol concentrations, (2) rotating segments of an ICR between high and low ethanol environments, and (3) simultaneous removal of the ethanol produced from the reactor as it is formed.     

Photobioreactor design: Mixing, carbon utilization, and oxygen accumulation

Photobioreactor design and operation are discussed in terms of mixing, carbon utilization, and the accumulation of photosynthetically produced oxygen. The open raceway pond is the primary type of reactor considered; however small diameter (1-5 cm) horizontal glass tubular reactors are compared to ponds in several respects. These are representative of the diversity in photobioreactor design: low capital cost, open systems and high capital cost, closed systems. Two 100-m(2) raceways were operated to provide input data and to validate analytical results. With a planktonic Chlorella sp., no significant difference in productivity was noted between one pond mixed at 30 cm/s and another mixed from 1 to 30 cm/s. Thus, power consumption or CO(2) outgassing limits maximal mixing velocities. Mixing power inputs measured in 100-m(2) ponds agreed fairly well with those calculated by the use of Manning's equation. A typically configured tubular reactor flowing full (1 cm diameter, 30 cm/s) consumes 10 times as much energy as a typical pond (20 cm deep flowing at 20 cm/s). Tubular reactors that flow only partially full would be limited by large hydraulic head losses to very short sections (as little as 2 m length at 30 cm/s flow) or very low flow velocities. Open ponds have greater CO(2) storage capacity than tubular reactors because of their greater culture volume per square meter (100-300 L/m(2) vs. 8-40 L/m(2) for 1-5-cm tubes). However, after recarbonation, open ponds tend to desorb CO(2) to the atmosphere. Thus ponds must be operated at higher pH and lower alkalinity than would be possible with tubular reactors if cost of carbon is a constraint. The mass transfer coefficient, K(L), for CO(2) release through the surface of a 100-m(2) pond was determined to be 0.10 m/h. Oxygen buildup would be a serious problem with any enclosed reactor, especially small-diameter tubes. At maximal rates of photosynthesis, a 1-cm tubular reactor would accumulate 8-10 mg O(2)/L/min. This may result in concentrations of oxygen reaching 100 mg/L, even with very frequent gas exchange. In an open pond, dissolved oxygen rises much more slowly as a consequence of the much greater volume per unit surface area and the outgassing of oxygen to the atmosphere. The maximum concentration of dissolved oxygen is typically 25-40 mg/L. The major advantage of enclosed reactors lies in the potential for aseptic operation, a product value which justifies the expense. For most products of algal mass cultivation, open ponds are the only feasible photobioreactor design capable of meeting the economic and operating requirements of such systems, provided desirable species can be maintained.     

Scale-down studies on the hydrodynamics of two-liquid phase biocatalytic reactors

The maintenance of constant interfacial area per unit volume is a key parameter for the successful scale-up of two-liquid phase bioconversion processes. To date, however, there is little published information on the hydrodynamics of such systems and a suitable basis for scale-up has yet to be defined and verified. Here we report power input and hydrodynamic data for a whole-cell bioconversion process using resting cells of Rhodococcus R312 to catalyse the hydration of a poorly water-soluble substrate 1,3-dicyanobenzene (1,3-DCB). Experiments were performed in geometrically similar 3-L and 75-L reactors, each fitted with a three-stage Rushton turbine impeller system. The two-phase system used comprised of 20% v/v toluene dispersed in 0.1 M aqueous phosphate buffer containing up to 10 g(ww) x L(-1) of resuspended biocatalyst and 20 g x L(-1) 1,3-DCB. The power input to the 3-L reactor was first determined using an air-bearing technique for both single-phase and two-phase mixing. In both cases, the power number attained a constant value of 11 at Re>10,000, while the measured power inputs were in the range 0.15-3.25 kW x m(-3). Drop size distributions and Sauter mean drop diameters (d(32)) were subsequently measured on-line in both reactors, using an in-situ light-backscattering technique, for scale-up on the basis of either constant power input per unit volume or constant tip speed. At both scales d(32) decreased with increasing agitation rate, while the drop size distributions obtained were log-normal. All the measured d(32) values were in the range of 30-50 microm, with the lowest values being obtained in systems with biocatalyst present. In all cases, constant power input per unit volume was found to be the most suitable basis for scale-up. This gave virtually identical d(32) values and drop size distributions at both scales. A number of correlations were also identified that would allow reasonable prediction of d(32) values for various agitation rates at each scale. While the results obtained are for a particular phase system, the scale-down methodology presented here would allow the rapid evaluation of other bioconversion processes in the 3-L reactor with a 25-fold reduction in scale. In this way, potential problems that might be encountered at the larger scale, such as the carryover of antifoam from the fermentation stage, could be quickly and efficiently identified.     

Fluid flow pattern in upflow reactors for anaerobic treatment of beet sugar factory wastewater

Residence-time-distribution experiments for the fluid in a 30-m(3) pilot plant and a 200-m(3) prototype upflow reactor were performed by means of continuous injection of an LiCl solution as a tracer in the influent of the reactor and measurement of the response of this stimulus on several location in the reactor and in the effluent. In a similar way as described in an article published earlier, models have been developed by use of the measured data of the fluid flow pattern which consisted of region of ideal mixing, plug flow, dead space, and short circuiting. It appeared that the fluid flow patterns in the two reactors were to a large extent analogous. For the pilot plant, three-mixer models appeared to be appropriate while for the prototype reactor two-mixer models have been found. This differences was a result of the difference in the heights of the sludge beds in the reactors: 2-3 m in the pilot plant and only 0.4 m in the prototype reactor, a result of too small an amount of sludge. Another differences was that, due to large amount of mud in the prototype reactor, a region of dead space occurred in the models for the fluid flow pattern in this reactor. The dimension of the prototype reactor have been chosen according to several recommendations obtained from work with the pilot plant (e.g., scale-up should be done by increasing the cross section of the reactor; one influent point should be applied per 5 m(2) bottom surface). The results presented here clearly show the value of these recommendations.     

Oxygen transfer and scale-up in bioreactors using hydro-ejectors for gas-liquid contacting

The commonly used scale-up criteria are investigated for their applicability in the case of hydro-ejector reactors. In combination with the liquid jet momentum, which characterizes the hydro-ejector, a scale-up correlation with the oxygen transfer rate as scale-up criterion is proposed, independent of the type of hydro-ejector and the reactor configuration. The results with regard to the power input are compared with those of stirred tank and bubble column. Its competitiveness is at high power per volume input and above all in large scale reactors.     

Nitrate treatment effects on bacterial community biofilm formed on carbon steel in produced water stirred tank bioreactor

To better understand the impact of nitrate in Brazilian oil reservoirs under souring processes and corrosion, the goal of this study was to analyse the effect of nitrate on bacterial biofilms formed on carbon steel coupons using reactors containing produced water from a Brazilian oil platform. Three independent experiments were carried out (E1, E2 and E3) using the same experimental conditions and different incubation times (5, 45 and 80 days, respectively). In every experiment, two biofilm-reactors were operated: one was treated with continuous nitrate flow (N reactor), and the other was a control reactor without nitrate (C reactor). A Polymerase Chain Reaction-Denaturing Gradient Gel Electrophoresis approach using the 16S rRNA gene was performed to compare the bacterial groups involved in biofilm formation in the N and C reactors. DGGE profiles showed remarkable changes in community structure only in experiments E2 and E3. Five bands extracted from the gel that represented the predominant bacterial groups were identified as Bacillus aquimaris, B. licheniformis, Marinobacter sp., Stenotrophomonas maltophilia and Thioclava sp. A reduction in the sulfate-reducing bacteria (SRB) most probable number counts was observed only during the longer nitrate treatment (E3). Carbon steel coupons used for biofilm formation had a slightly higher weight loss in N reactors in all experiments. When the coupon surfaces were analysed by scanning electron microscopy, an increase in corrosion was observed in the N reactors compared with the C reactors. In conclusion, nitrate reduced the viable SRB counts. Nevertheless, the nitrate dosing increased the pitting of coupons.     

Determining design and scale-up parameters for degradation of atrazine with suspended Pseudomonas sp. ADP in aqueous bioreactors

This work investigated the kinetic parameters of atrazine mineralization by suspended cells of Pseudomonas sp. ADP in both shake flasks and spherical stirred tank batch reactors (SSTR). The degradation of atrazine and growth of Pseudomonas sp. ADP were studied. Experiments were performed at different temperatures and stirring speeds in both reactors at varying initial concentrations of atrazine. Cell growth and atrazine concentration were monitored over time, and a Monod model with one limiting substrate was used to characterize the kinetic behavior. Temperature, stirring speed, and reactor type were all found to significantly affect the regressed Monod parameters. At 27 degrees C and 200 rpm, for the shaker flask experiments, mu max and Ks were determined to be 0.14 (+/-0.01) h-1 and 1.88 (+/-1.80) mg/L, respectively. At 37 degrees C, mu max and Ks increased to 0.25 (+/-0.05) h-1 and 9.59 (+/-6.55) mg/L, respectively. As expected, stirrer speed was also found to significantly alter the kinetic parameters. At 27 degrees C and 125 rpm, mu max and Ks were 0.04 (+/-0.002) h-1 and 3.72 (+/-1.05) mg/L, respectively, whereas at 37 degrees C and 125 rpm, mu max and Ks were 0.07 (+/-0.008) h-1 and 1.65 (+/-2.06) mg/L. In the SSTR the kinetic parameters mu max and Ks at room temperature were determined to be 0.12 (+/-0.009) h-1 and 2.18 (+/-0.47) mg/L, respectively. Although the mu max values for both types of reactors were similar, the shaker flask experiments resulted in considerable error. Error analysis on calculated values of Ks were found to impact estimates in atrazine concentration by as much as two orders of magnitude, depending on the reactor design, illustrating the importance of these factors in reactor scale-up.     

Influence of media-packing ratio on performance of anaerobic hybrid reactors

In this study, the influence of media-packing ratio on the performances of anaerobic hybrid reactors (AHRs) at low, medium and high organic loading rates was evaluated by conducting COD profile, granulation and tracer studies. Four laboratory upflow anaerobic hybrid reactors, each with a total unpacked volume of 7.85 l, with varying packing depths, were operated at organic loading rates from 1 to 24 g COD/l d. The media-packing ratios were 75%, 60%, 40% and 20% of the total reactor height in the AHRs. Three types of soluble COD profiles along the reactor height were observed when the organic loading rate was gradually increased. When operated at 1 and 2 g COD/l d the COD profiles along the reactor height from bottom to top showed a plug-flow regime. From 4 to 12 g COD/l d the COD profiles were distorted in the reactors with 20%, 40% and 60% packing, while at 16 g COD/l d and above the COD profile indicated homogeneity in each reactor, suggesting a perfectly-mixed regime. The distorted COD profiles were considered to be caused by the non-ideal flow pattern prevalent in the reactors. The dead-space volume and the bypass flowrate due to short-circuiting were determined using the Cholette and Cloutier model. A `distortion index' (DI), which was calculated from the ratio of the average COD value of the sludge bed over the average COD value of the reactor, was used to describe distortion of the COD profile. The distortion index correlated well with the short-circuiting fraction.     

Scale-up and application of a cyclone reactor for fermentation processes

A cyclone reactor for microbial fermentation processes was developed with high oxygen transfer capabilities. Three geometrically similar cyclone reactors with 0.5?l, 2.5?l and 15?l liquid volume, respectively, were characterized with respect to oxygen mass transfer, mixing time and residence time distribution. Semi-empirically correlations for prediction of oxygen mass transfer and mixing times were identified for scale-up of cyclone reactors. A volumetric oxygen mass transfer coefficient k _L a of 1.0?s^?1 (available oxygen transfer rate with air: 29?kg?m^?3?h^?1) was achieved with the cyclone reactor at a volumetric power input of 40?kW?m^?3 and an aeration gas flow rate of 0.2?s^?1. Continuous methanol controlled production of formate dehydrogenase (FDH) with Candida boidinii in a 15?l cyclone reactor resulted in more than 100% improvement in dry cell mass concentration (64.5?g?l^?1) and in about 100% improvement in FDH space-time yield (300?U?l^?1?h^?1) compared to steady state results of a continuous stirred tank reactor.     

Air-bubbling, hollow-fiber reactor with cell bleeding and cross-flow filtration

Continuous asymmetric reduction of dyhydrooxoisophorone (DOIP) to 4-hydroxy-2,2,6-trimethylcyclo-hexanone (4-HTMCH) was achieved by a thermophilic bacterium Bacillus stearothermophilus NK86-0151. Three reactors were used: an air-bubbling hollow-fiber reactor with cell bleeding and cross-flow filtration, an air-lift reactor, and a CSTR with PAA immobilized cells. The maximum cell concentration of 11.1 g dry wt L(-1) was obtained in an air-bubbling hollow-fiber reactor, while in the other reactors the cell densities were between 3.5 and 4.1 g dry wt L(-1) The optimum bleed ratio was 0.1 at the dilution rate 0.3 h(-1) in the hollow-fiber reactor. The highest viable cell concentration was maintained in the dilution range of 0.4-0.7 h(-1) by a combination of proper cell bleeding and cross-flow filtration. The maximum volumetric productivity of 4-HTMCH reached 826 mg L(-1) h(-1) at the dilution rate 0.54 h(-1). This value was 4 and 2 times higher than those in the air-lift reactor and CSTR, respectively. The increasing viable cell concentration increased the volumetric productivity of 4-HTMCH. A cell free product solution was continuously obtained by cross-flow filtration.     

Quantitative measurements of mixing intensity in shake-flasks and stirred tank reactors. Use of the Mixmeter, a mixing process analyzer

A new mixing probe has been developed which measures the motions of the fluid during mixing as pressure fluctuations and converts the measurements into a mixing signal (MS). The MS is the root mean square (RMS) pressure fluctuation in the 1-64Hz range as determined by a sensitive pressure sensor and a digital signal processor specifically designed for the purpose. The MS is a measure of the actual mixing flow of the fluid rather than a measurement of the input motions or energies into the reactor system (e.g. RPM, torque or power). In other studies, the MS has been measured as a function of mixing speed in numerous sized reactors from 10 to 1000l, and provides consistent and reproducible measurements. The MS increases monotonically as a function of mixing speed, with a change of slope corresponding to the transition from laminar to turbulent mixing regimes. Maps of MS as a function of location in the reactor are useful in understanding stirred tank reactor design and performance. Quantitative measurements of mixing are especially useful during process development as a tool to increase the success of scale-up during the transition from process development to manufacturing. Measurements at a fixed location in a given reactor are useful in understanding changes in mixing that occur during the course of a given process, and are useful in manufacturing situations where validated documentation of lot-to-lot consistency of mixing is required (e.g. pharmaceutical manufacturing). In addition, the probe has been used to measure mixing in vessels with vibrational mixers with similar results. The probe has been successfully used in feedback loops to control either mixing speed or vibrational mixing amplitude in order to maintain constant mixing of the fluid during processing. With this system it is possible to maintain constant mixing over a wide range of fluid volumes in a given reactor, and, for instance, to compensate for changes in viscosity throughout the course of the process. Adaptations of this system for the measurement of mixing in shake-flasks is described in this paper.     

Ambruticin S production in air-lift and stirred-tank bioreactors

Summary Using the method of equi-inocular synchronized comparative fermentation (EISCF) the cultivation of Sorangium cellulosum So ce 10 and production of the polyketide antibiotic ambruticin S was compared in stirred-tank and air-lift reactors of different geometry. This method requires that inocula originate from the same pre-culture and cultivation parameters are synchronized to similar values. Similar ambruticin yields were obtained from both reactor systems provided that the concentration of dissolved oxygen (DO) was maintained above a certain value (ca. 40%). For cultivation of S. cellulosum it is the DO level rather than the oxygen transfer rate which presents the proper criterion for scale-up and comparative reactor studies. Offprint requests to: W.-D. Deckwer     

Correlation of cotyledonary node shoot proliferation and somatic embryoid development in suspension cultures of soybean (Glycine max L. Merr.)

Suspension cultures of soybean were initiated from hypocotyl or cotyledon callus tissue of several soybean genotypes. When these were grown on L2 medium with 0.4 mg/liter 2,4-D several genotypes produced numerous embryoids while others produced only a few such structures. Due to internal anatomy, no embryoid developed into a complete plant. A genotype's propensity to form normal appearing embryoids was correlated with the ability to proliferate shoots at the cotyledonary node on a medium with benzylaminopurine as determined in previous testing.Abbreviations 2,4-D 2,4-Dichlorophenoxyacetic acid - BAP Benzylaminopurine - L2 Phillips and Collins (1979) legume medium     

Granule development and performance in sucrose fed anaerobic baffled reactors

Two 90 L anaerobic baffled reactors were used to study the granulation of sludge and the effect of the organic loading rate and NaHCO3/COD ratios on reactor performance. Furthermore, it was determined whether an anaerobic baffled reactor would promote phase separation and if additive of bentonite or granular active carbon was capable of enhancing granule formation. In order to minimize feed variations, and have a totally biodegradable substrate, a synthetic sucrose substrate was used. Granulation was achieved in both reactors within 75 days. However, the granules from the granular active carbon amended reactor appeared earlier and were larger and more compact. The reactors were maintained at a hydraulic retention time of 20 h during performance study stage. The results showed that when organic loading rate were changed from 2.15 to 6.29 kg COD m(-3)day(-1), chemical oxygen demand (COD) removal was not decreased (91-93%), but a slight increase in effluent COD was observed. It was found that the COD removals were generally good (87-92%) and had not obviously change with the decreasing NaHCO3/COD ratios. From the bacterial distribution and the concentration of volatile fatty acids in four compartments, it was concluded that a separation of phases occurred within the anaerobic baffled reactors.     

Characterization of flow conditions in 2 L and 20 L wave bioreactors® using computational fluid dynamics

Characterization of flow conditions is of great importance to control cell growth and cell damage in animal cell culture because cell viability is influenced by the flow properties in bioreactors. Alternative reactor types like Wave Bioreactors® have been proposed in recent years, leading to markedly different results in cell growth and product formation. An advantage of Wave Bioreactors® is the disposability of the Polyethylenterephthalet‐bags after one single use (fast setup of new production facilities). Another expected advantage is a lower shear stress compared to classical stirred‐tank reactors, due to the gentle liquid motion in the rocking cellbag. This property would considerably reduce possible cell damage. The purpose of the present study is to investigate in a quantitative manner the key flow properties in Wave Bioreactors®, both numerically and experimentally. To describe accurately flow conditions and shear stress in Wave Bioreactors® using numerical simulations, it is necessary to compute the unsteady flow applying Computational Fluid Dynamics (CFD). Corresponding computations for two reactor scales (2 L and 20 L cellbags) are presented using the CFD code ANSYS‐FLUENT®. To describe correctly the free liquid surface, the present simulations employ the Volume of Fluid (VOF) method. Additionally, experimental measurements have been carried out to determine liquid level, flow velocity and liquid shear stress, which are used as a validation of the present CFD simulations. It is shown that the obtained flows stay in the laminar regime. Furthermore, the obtained shear stress levels are well below known threshold values leading to damage of animal cells. © 2009 American Institute of Chemical Engineers Biotechnol. Prog., 2010     

Degradation of diesel oil in soil using a food waste composting process

We investigated the simultaneous degradation of diesel oil in soil and the organic matter in food waste by composting in 8 l reactors. Using a 0.5 l/min air flow rate, and 0.5-1% diesel oil concentrations together with 20% food waste, high composting temperatures (above 60°C) were attained due to the efficient degradation of the food waste. Petroleum hydrocarbons were degraded by 80% after only 15 days composting in the presence of food waste. In a 28 l reactor scale-up experiment using 1% oil, 20% food waste and 79% soil, removal efficiencies of petroleum hydrocarbons and food waste after 15 days were 79% and 77%, respectively.     

Biomass production of hairy roots of Artemisia annua and Arachis hypogaea in a scaled‐up mist bioreactor

Hairy roots have the potential to produce a variety of valuable small and large molecules. The mist reactor is a gas phase bioreactor that has shown promise for low‐cost culture of hairy roots. Using a newer, disposable culture bag, mist reactor performance was studied with two species, Artemisia annua L. and Arachis hypogaea (peanut), at scales from 1 to 20 L. Both species of hairy roots when grown at 1 L in the mist reactor showed growth rates that surpassed that in shake flasks. From the information gleaned at 1 L, Arachis was scaled further to 4 and then 20 L. Misting duty cycle, culture medium flow rate, and timing of when flow rate was increased were varied. In a mist reactor increasing the misting cycle or increasing the medium flow rate are the two alternatives for increased delivery of liquid nutrients to the root bed. Longer misting cycles beyond 2–3 min were generally deemed detrimental to growth. On the other hand, increasing the medium flow rate to the sonic nozzle especially during the exponential phase of root growth (weeks 2–3) was the most important factor for increasing growth rates and biomass yields in the 20 L reactors. A. hypogaea growth in 1 L reactors was µ = 0.173 day^?1 with biomass yield of 12.75 g DW L^?1. This exceeded that in shake flasks at µ = 0.166 day^?1 and 11.10 g DW L^?1. Best growth rate and biomass yield at 20 L was µ = 0.147 and 7.77 g DW L^?1, which was mainly achieved when medium flow rate delivery was increased. The mist deposition model was further evaluated using this newer reactor design and when the apparent thickness of roots (+hairs) was taken into account, the empirical data correlated with model predictions. Together these results establish the most important conditions to explore for future optimization of the mist bioreactor for culture of hairy roots. Biotechnol. Bioeng. 2010;107: 802–813. © 2010 Wiley Periodicals, Inc.