Wine Vinegar Production Using a Noncommercial 100-Litre Bubble Column Reactor Equipped with a Novel Type of Dynamic Sparger

Transcription

1 Wine Vinegar Production Using a Noncommercial 100-Litre Bubble Column Reactor Equipped with a Novel Type of Dynamic Sparger G. Fregapane, 1 H. Rubio-Fernández, 1 J. Nieto, 2 M.D. Salvador 1 1 University of Castilla-La Mancha, Faculty of Chemistry, Department of Food Technology, Ciudad Real, Spain; fax: ; gfregapane@qata-cr.uclm.es 2 Institute of Industrial Fermentation, CSIC, Madrid, Spain Received 7 July 1998; accepted 25 September 1998 Abstract: This paper describes batch and semicontinuous acetic acid fermentations for wine vinegar production carried out with Acetobacter pasteurianus, and an industrial strain using a noncommercial 100-L bubble column reactor equipped with a novel type of gas liquid dynamic sparger. Results showed acetification rates with this fermentor (i.e., an overall acetic acid productivity of 1.8 g/l/h and yield of 94%) similar to that of the Frings acetator and higher as compared to others fermentors in current industrial use in Spanish wine vinegar factories, and a linear relationship between overall productivity and k L a with different operating conditions and fermentation scales John Wiley & Sons, Inc. Biotechnol Bioeng 63: , Keywords: wine vinegar production; acetic acid fermentation; bubble column reactor; dynamic sparger INTRODUCTION Wine vinegar is considered to possess superior sensory qualities to other types of vinegar and is a highlyappreciated element in the diet of the Mediterranean countries (Carnacini and Gerbi, 1992; Gonzalez-Viñas et al., 1996). This product is of great interest in the Spanish region of Castilla-La Mancha, due to the high output there, accounting for more than 60% of the total wine vinegar produced by Spain. In 1996 the figure was about 28 million L (10% acetic acid). Industrial production of vinegar today is carried out mainly by submerged culture in semicontinuous processes, and although the Frings acetator is the most widely used reactor (Adams, 1985; Ebner and Follman, 1983) other types of fermentors are still employed in Spain (Llaguno and Polo, 1991). The development of acetic bacteria starter cultures (Mesa et al., 1996; Sokollek and Hammes, 1997) and a gas recirculation system (Gomez et al., 1994; Romero and Cantero, 1998) to reduce ethanol losses have been recently carried out, but these advances in vinegar fermentation technology did not improve acetification performances. Correspondence to: G. Fregapane New types of small scale fermentors for acetic acid fermentation have also been developed, in many cases using cell immobilization techniques to attain higher acetification rates than classical acetators (De Araujo and Santana, 1996; Toda and Asakura, 1994). However, the complexity of construction, combined with other technical and economical problems, has so far prevented industrial use of these reactors. The aim of this research project is to develop an acetator for industrial wine vinegar production that could compete with the currently used fermentors. To this end, a noncommercial 100-L bubble column reactor is under study, equipped with a novel type of gas liquid dynamic sparger (Nieto, 1985), which produces a fine bubble size distribution. It is known that acetic acid fermentation is strictly dependent on the supply of oxygen to the liquid phase and the choice of the gas liquid distribution system has a considerable effect on the performance of the reactor. Indeed, the productivity of industrial fermentation processes is often limited by the oxygen transfer rate. This paper describes batch and semicontinuous acetic acid fermentations using a noncommercial 100-L bubble column reactor equipped with a novel type of gas liquid dynamic sparger working in different operating conditions, and a study of the volumetric mass transfer coefficient (k L a) observed with this acetator. MATERIALS AND METHODS Strains The strain used in batch fermentation was Acetobacter pasteurianus (CECT 824, Spanish National Collection of Standard Cultures). It was maintained on an agar medium containing 50 g/l glucose (Sigma), 20 g/l yeast extract (Difco), 10 g/l peptone (Difco), and 20 g/l agar agar (Difco); ph was adjusted to 5.0 with NaOH. The strain was subcultured every 2 months. For semicontinuous fermentation, an industrial culture 1999 John Wiley & Sons, Inc. CCC /99/

2 obtained from a local vinegar factory was used. This was maintained on a liquid medium (30/70 wine/vinegar mixture) and stored in 20% (v/v) glycerol at 50 C. Growth Media Growth medium used in some batch fermentations (EYPS) contained 48 g/l ethanol, 2 g/l peptone, 5 g/l acetic acid, 0.8 g/l KH 2 PO 4, 0.35 g/l K 2 HPO 4 3H 2 O, 1.5 g/l (NH 4 ) 2 SO 4, 0.25 g/l MgSO 4 7H 2 O, 0.01 g/l NaCl, and 0.01 g/l MnSO 4 ; ph 3.5. A commercial white table wine (Airen variety) was used for semicontinuous and some batch acetifications. It was obtained directly from a local producer and had the following composition: 13 (%, v/v) (corresponding to 103 g/l ethanol), 4.7 g/l total acidity (as tartaric acid), 0.8 g/l volatile acidity (as acetic acid), ph 3.4, 0.5 g/l residual sugars (glucose and fructose), and 70 ppm SO 2 (that was eliminated bubbling through air). Acetator The body of the 100-L noncommercial bubble column reactor consisted of a transparent PVC cylinder with an internal diameter of 30 cm, useful liquid capacity of 70 L and a diameter/working height ratio of 1 to 5. Air and recirculating fermentation medium were fed in at the bottom of the column through a dynamic sparger using a centrifuge pump generally working at a flow rate of 12 L/min. The bioreactor was equipped with a novel type of gas liquid dynamic sparger, that produces a fine bubble size distribution. A cross-section diagram of this device is depicted in Fig. 1. The sparger consists of two circular stainless-steel plates (A, B), with an adjustable gap (C) between them and a series of holes (D; 30 holes, diameter 1 mm) exactly coinciding in each plate. The fine dispersion is obtained by the collision of the air flow (E) and the liquid stream (F), fed though the holes in the lower (A) and upper (B) plate respectively, that occurs in the region between the two plates and by the friction effect caused by the dispersion (air-liquid mixture) going through the plates. This low cost and ease of construction sparger was developed to produce a fine gas liquid or liquid liquid twophase dispersion, and this is the first described application to a reactor for acetic acid fermentation. More details of this system can be found in the corresponding patent (Nieto, 1985) and in a theoretical/experimental study of the properties of this sparger (Nieto, 1993). Air flow rates employed during fermentation ranged from 4.5 to 17.5 L/min, corresponding to a vvm (air flow rate/ reactor working volume ratios) of 3.7 and 15 h 1 respectively. The temperature of the fermenting broth was maintained at 30 C by circulating cool water through an internal stainless-steel coil. Two oxygen probes (Syland Scientific) inserted in the vessel and in the recirculation line were used to Figure 1. Cross-section diagram of the dynamic sparger. A, Lower plate; B, upper plate; C, adjustable gap; D, holes; E, air flow; F, liquid stream. monitor dissolved oxygen concentration as an easy way to follow the course of the fermentation over time. Experiments were carried out at least in triplicate. Reproducibility was good and mean values and standard deviation are shown in the tables. Batch Fermentation A 250-mL aerated shake flask with 100 ml of the EYPS medium was inoculated from a slant and incubated at 30 C and 200 rpm for 2 days. This flask was then used as an inoculum for a 6-L fermentor of the same growth medium at 30 C. When the acetic acid concentration reached 10 g/l, this culture was used for the inoculation of the 100-L reactor. Initial fermentation conditions were 48 ±1 g/l ethanol and 5.0 ± 0.5 g/l acetic acid. When wine was used for acetification it was diluted with distilled water to the adequate alcohol concentration (about 6 ). Semicontinuous Fermentation When an acetification cycle was completed, about 40% (28 L) of the fermented broth was discharged as end product vinegar, at an acetic acid concentration close to g/l. The same volume of fresh wine was then fed into the fermentor using a peristaltic pump working at a flow rate of 20 L/h and a new production cycle began. Initial fermentation conditions of the wine/vinegar mixture (40:60) were 47 ± 1 g/l ethanol and 70 ± 1 g/l acetic acid. 142 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 63, NO. 2, APRIL 20, 1999

3 The yield of the process was calculated as done by the vinegar industry, as the quotient of the final and initial total concentration, where the total concentration is defined by the sum of the alcohol (% v/v, ) and acetic acid (g/100 ml) concentrations (1 alcohol provides g/100 ml of acetic acid in this type of fermentation). Assays Samples of a few millilitres were withdrawn every 2 4 h (depending on total fermentation time) from the fermentor during the course of acetification and acetic acid, ethanol, ph and optical density were determined. Acetic acid was measured titrimetrically with phenolphthalein as an indicator. It was assumed that all broth acidity was due to acetic acid. Ethanol was enzymatically assayed by kit No (Boehringer-Manheim). The biomass was monitored by measuring the absorbance within a linear range at 575 nm. A previously prepared calibration curve was used to convert the absorbance readings into biomass dry weight. The volumetric oxygen-transfer coefficient (k L a) was determined using the gassing-in method (Sobotka et al., 1982). RESULTS AND DISCUSSION Volumetric Oxygen Transfer Coefficient (k L a) Volumetric oxygen transfer coefficient (k L a) was measured in the 100-L noncommercial fermentor at different air flow rates ranging from a vvm of 3.7 to 15 h 1 and with different gaps between the dynamic sparger plates (from 0.4 to 2 mm). The k L a improved at higher air flow rates and its relation with aeration was almost linear at all gap sizes; moreover the smaller the gap, the higher was the value of the oxygen transfer coefficient (Fig. 2). The highest k L a values were therefore obtained with a gap of 0.4 mm, and ranged from 75 h 1 (at a vvm of 3.7 h 1 ) to 170 h 1 (vvm of 15 h 1 ). Very small differences in the value of the volumetric oxygen transfer coefficient were observed when the gap was increased from 0.8 to 2 mm. It was not feasible to work with a gap of less than 0.4 mm due to a high increase in the pressure of the recirculating liquid. The volumetric oxygen transfer coefficient (k L a) was also measured at different flow rates of the recirculating liquid. The values obtained with a plate gap of 0.4 mm and a vvm of 7.5 h 1 were 85, 105, and 115 h 1 for rates of 8, 10, and 12 L/min respectively. In the range studied, the higher the recirculating liquid flow rate, the greater was the k L a obtained. In a previous study on the velocity distribution of the air liquid mixture with this dynamic sparger it was observed that the smaller the plate gap and the higher the recirculating liquid flow rate, the higher was the gas liquid Figure 2. Influence of the distance between the sparger plates on the volumetric oxygen transfer coefficient at different air flow rates., 0.4 mm;, 0.8 mm;, 1.2 mm;, 2 mm. k L a, volumetric oxygen transfer coefficient; vvm, air flow rate/reactor working volume ratio. mixture projection velocity and the better the bubble distribution in the fermentor body section (Nieto, 1993). Batch Fermentation Batch acetifications were carried out with A. pasteurianus in the 100-L noncommercial bubble column reactor using EYPS medium and white table wine prepared as indicated above. In all cases, the initial concentrations of ethanol and acetic acid were 48.0 ± 1.0 and 5.0 ± 0.5 g/l, respectively. Results of batch fermentations carried out with A. pasteurianus (at a vvm of 7.5 h 1, plate gap of 0.4 mm and recirculating liquid flow rate of 12 L/min) are reported in Table I. The overall acetic acid productivity was higher with white table wine than when EYPS medium was used, with values of 1.45 and 1.25 g/l/h, respectively, as already observed in a previous study using a 6-L fermentor (Fregapane et al., personal communication). Maximum acetification rates of 2.3 and 3.5 g/l/h using EYPS and wine respectively were obtained with an overall yield of over 93%. The time course of a batch acetic acid fermentation using EYPS medium is shown in Fig. 3. Following a short lag phase, the biomass stationary phase was reached in about 24 h of exponential growth and the final concentration of acetic acid was about 60 g/l. The dissolved oxygen concentration decreased to a minimum of 2 3 ppm during the period of FREGAPANE ET AL.: WINE VINEGAR PRODUCTION WITH A NOVEL DYNAMIC SPARGER 143

4 Table I. Results of batch fermentations carried out with A. pasteurianus using EYPS and white table wine media performed at an aeration of 7.5 h 1. a EYPS maximum acetification rate and until the beginning of the stationary phase, when it increased sharply. Semicontinuous Fermentation Wine AcH p (g/l/h) 1.25 ± ± 0.11 AcH p max (g/l/h) 2.3 ± ± 0.2 Total process time (h) 45 ± 1 37± 1 AcH production time (h) 38 ± 1 31± 1 Yield (%) 92.6 ± ± 0.6 Final AcH (g/l) 61 ± 3 60± 2 Final EtOH (g/l) 0.7 ± ± 0.6 a The plate gap was 0.4 mm, and the liquid flow rate was 12 L/min. Initial conditions: 48 ± 1 g/l EtOH, 5.0 ± 0.5 g/l AcH. AcH, acetic acid; AcH p, overall acetic acid productivity; AcH p max, maximum acetification rate; EtOH, ethanol. An industrial culture obtained from a local vinegar factory was used for semicontinuous fermentations because A. pasteurianus was unable to grow at those concentrations of acetic acid. Semicontinuous acetifications employing different plate gaps and recirculating liquid flow rates are shown in Fig. 4. According to the observed behaviour of the volumetric oxygen transfer coefficient (k L a), higher acetification rates were reached with lower sparger plate gaps and higher recirculating liquid flow rates. The best acetification performance was therefore obtained with a gap of 0.4 mm between the sparger plates and a fermentation medium recirculating rate of 12 L/min. Results obtained employing these conditions in semicontinuous acetifications at different air flow rates, are presented in Table II. The overall acetic acid productivity improved at higher aeration rates, ranging from 0.75 g/l/h at a vvm of 3.7 h 1 to 1.76 g/l/h at 15 h 1. The yield of the process was nearly 99% at lower air flow rates and about 94% at higher rates due to the different rate of ethanol loss during the course of the fermentation. Total process time decreased from 60.0 h with an air flow rate of 3.7 h 1 to25.5hat15h 1. The observed lag phase was very short and dependent on the aeration employed, it decreased from 6 h at thelowest rate to nil at the highest rate employed. The final concentration of acetic acid was about 115 g/l. Slightly higher values, about 120 g/l, could have been obtained, but this damaged the bacteria, resulting in a longer lag phase in the following cycle (data not shown). Figure 3. Batch fermentation with A. pasteurianus using EYPS medium. Plate gap 0.4 mm; recirculating liquid flow rate 12 L/min; air flow rate 7.5 h 1. AcH, acetic acid; DO, dissolved oxygen concentration. Figure 4. Influence of the sparger plate gap and recirculating liquid flow rate on the acetic acid productivity of semicontinuous fermentations. Plate gap: 0.4 mm (A, D, E), 0.8 mm (B), and 2.0 mm (C). Recirculating liquid flow rate: 12 L/min (A, B, C), 10 L/min (D), and 8 L/min (E). Aeration 7.5 h 1. AcH, acetic acid. 144 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 63, NO. 2, APRIL 20, 1999

5 Table II. Results of semicontinuous acetifications performed with an industrial strain employing different aeration and white table wine as fermentation medium. a Aeration, vvm (h 1 ) AcH p (g/l/h) 0.75 ± ± ± 0.02 AcH p max (g/l/h) 1.08 ± ± ± 0.06 Total process time (h) 60.0 ± ± ± 0.5 AcH production time (h) 54.0 ± ± ± 0.5 Yield (%) 98.6 ± ± ± 0.3 Final AcH (g/l) ± ± ± 0.8 Final EtOH (g/l) 10.4 ± ± ± 0.3 k L a (h 1 ) 75± ± ± 12 a The plate spacing was 0.4 mm, and the liquid flow rate was 12 L/min. Initial conditions: 47.0 ± 1.0 g/l EtOH, 70.0 ± 1.0 g/l AcH. k L a, volumetric oxygen-transfer coefficient; vvm, air flow rate/reactor working volume ratio; see Table I for other abbreviations. Acetic acid formation, biomass dry weight and acetic acid productivity during the course of three semicontinuous cycles employing different air flow rates (3.7, 7.5, and 15 h 1 ) are shown in Fig. 5. It is clear that the higher the air flow rate, the greater the acetic acid productivity is, with a consistent reduction of the total process time. The industrial process of wine vinegar production is carried out in cycles of 24 h, and depending on acetification rates achieved the percentage of fermenting vinegar unloaded as final product may vary. Under optimal operating conditions the Frings acetator, in reactors that range from 10 to 40,000 L, reach % acetic acid in 24 h, that corresponds to an overall productivity of g/l/h (Sokollek and Hammes, 1997). Other types of fermentors that are still common in Spain consisting basically of a static or dynamic sparger and an agitation system with or without liquid recirculation, achieve in the same period of time % ( g/l/h) in reactors up to 20,000 L. A total concentration of % and yield of 91 94% are obtained in all cases (Parras SA, Vulpi SL, El Encierro SL, Thor SA, and Mueloliva SA, personal communications). The overall productivity achieved with the acetator described in this study, that is still under optimisation, is similar to that of the Frings acetator and higher as compared to others fermentors in current use in Spanish wine vinegar factories. The process yield and the final product concentration are also similar to that obtained by the industry. This low cost and ease of construction sparger has got the advantage, as compared to the Frings turbine (aerator) and mechanically agitated reactors, of possessing no moving parts, and furthermore it is simple its assembly to fermentation tanks and to dismount for cleaning and maintenance purposes. Although industrial vinegar production has been the subject of many technical studies, hardly any advances in this field have been published in the scientific literature. This, together with pressing current demand for development of local technology, warrants further research in the field. The medium-term aim of this research project is to develop a new acetator for industrial wine vinegar production which will compete with currently-used fermentors, such as the Frings acetator; that can offer an alternative choice to new vinegar plants and to factories that need to have their reactors enhanced. Further investigation is currently being carried out in order to optimise the acetator and the acetic acid fermentation. In particular, process costs, operating conditions and fermentation scale are taken into account. Relationship between k L a and Process Productivity The relationship between the process productivity and the volumetric oxygen transfer coefficient (k L a) is shown in Fig. 6. A linear correlation between k L a, measured in different operating conditions, and the overall acetification rate for semicontinuous processes, was observed with both the 100-L acetator and a 6-L airlift fermentor (a scaled-down version of the larger reactor) at least in the k L a range of h 1. Figure 5. Acetic acid formation, acetic acid productivity and biomass dry weight during the time course of three semicontinuous cycles carried out employing different air flow rates (3.7, 7.5, and 15 h 1, respectively). AcH, acetic acid; AcH p, acetic acid productivity. CONCLUSIONS The highest productivity values were obtained with the 100-L noncommercial bubble column reactor equipped with FREGAPANE ET AL.: WINE VINEGAR PRODUCTION WITH A NOVEL DYNAMIC SPARGER 145

6 A linear relationship between the overall acetic acid productivity and the k L a was observed with different working conditions and fermentation scales. The authors thank the following wine vinegar factories for the valuable information provided: Parras, S.A. (Toledo), Vulpi, S.L. (Toledo), El Encierro, S.L. (La Rioja), Mueloliva, S.A. (Córdoba) and Thor, S.A. (Madrid). References Figure 6. Relationship between overall acetic acid productivity and volumetric oxygen transfer coefficient (k L a) in semicontinuous acetifications., 100-L acetator;, 6-L fermentor. Aeration, vvm (h 1 ): 1.8 (A); 3.7 (B, F); 7.5 (C, G, H, I, J, K); 15 (D, L); and 30 (E). Sparger plate gap (mm): 0.4 (F, G, J, K, L); 0.8 (I); and 2.0 (H). Recirculating liquid flow rate (L/min): 8 (G); 10 (J); and 12 (F, H, I, K, L). AcH p, acetic acid productivity. a novel type of dynamic sparger, employing a sparger plate gap of 0.4 mm and a recirculating liquid flow rate of 12 L/min. An overall acetic acid productivity of 1.8 g/l/h and yield of 94% for an aeration of 15 h 1 were attained in semicontinuous fermentations. Results showed an acetification rate with the 100-L reactor similar to that of the Frings acetator and higher than is achieved industrially in some Spanish wine vinegar factories. Adams MR Vinegar. In Wood BJB, editor. Microbiology of fermented foods. New York: Elsevier Applied Science Publishers. p Carnacini A, Gerbi V Le vinaigre de vin: Un produit mediterraneen. Vitic Enol Sci 47: De Araujo AA, Santana MHA Continuous oxidation of ethanol with Acetobacter cells immobilized in denser particles of gel matrix. J Food Sci Technol India 33: Ebner E, Follman H Acetic acid. In: Dellweg H, editor. Biotechnology. Weinheim: Verlag Chemie. p Gomez JM, Romero LE, Caro I, Cantero D Application of a gas recirculation system to industrial acetic fermentation processes. Biotechnol Tech 8: Gonzalez-Viñas MA, Salvador MD, Cabezudo MD Taste group thresholds and sensory evaluation of spanish wine vinegars. J Sensory Studies 11: Llaguno C, Polo MC El vinagre de vino. 1st edition. Madrid, Spain: Consejo Superior de Investigaciones Cientificas. Mesa MM, Caro I, Cantero D Protocol for the preparation and development of Acetobacter aceti inoculum in industrial acetic acid fermentations. Chem Biochem Eng Q 10: Nieto J Dispositivo para la obtención de dispersiones gas/líquido o líquido/líquido, de aplicación en los procesos de transferencia de materia. Patente Española no Nieto F Estudio de la distribución de velocidades en un Jet en Ranura Circular. Tesis de Licenciatura. Madrid, Spain: Universidad Complutense de Madrid. Facultad de Ciencias Químicas. Romero LE, Cantero D Evaluation of ethanol evaporation losses in acetic acid fermentations. Bioprocess Eng 18: Sobotka M, Prokop A, Dunn LJ, Einsele A Review of methods for the measurement of oxygen transfer microbial systems. In: Tsao GT, editor. Annual reports on fermentation process. London: Academic Press. p Sokollek SJ, Hammes WP Description of a starter culture preparation for vinegar fermentation. Syst Appl Microbiol 20: Toda K, Asakura T Acetic acid production by Acetobacter aceti in a silicone tube bioreactor. Biotechnol Lett 16: BIOTECHNOLOGY AND BIOENGINEERING, VOL. 63, NO. 2, APRIL 20, 1999