Continuous production of

2,3-butanediol from whey permeate

using Klebsiella pneumoniae
immobilized in calcium alginate
H. K. L e e * a n d I. S. Maddoxt

Biotechnology Department, Massey University, Palmerston North, N e w Zealand

(Received 11 December 1985)

Cells o f Klebsiella pneumoniae were immobilized in calcium alginate gel and used in a packed column
reactor for the continuous production o f 2,3-butanediol from whey permeate. A maximum butanediol
productivity o f 2.3 g 1-1 h-1 was achieved at a dilution rate o f O. 77 h -1. The system proved to be
stable during a seven week period o f continuous operation. A low dilution rate (less than 0.3 h -1 )
should be maintained during start-up o f the reactor.

Keywords: Butanediol production; whey permeate; alginate immobilization

Introduction Materials and methods

There is increasing interest worldwide in the production of Organism
2,3-butanediol by fermentation using appropriate bacteria. Klebsiella pneumoniae NCIB 8017 was obtained from
This interest stems from the potential use of this product as the National Collection of Industrial Bacteria (Aberdeen,
a chemical feedstock or as a liquid fuel. 1 Early work on the Scotland, UK) and was maintained on slopes of nutrient
fermentation process has been well reviewed.: More re- agar (Oxoid Ltd, London, UK).
cently, processes have been described whereby butanediol
can be produced from starch, using Aeromonas hydro- Materials
phila; a hemicellulose sugars, using either Klebsiella pneu-
Rennet whey permeate was obtained from the New
moniae 4 or Bacillus polymyxa; s and whey, using either
Zealand Dairy Research Insitute (Palmerston North, New
K. penumoniae 6 or B. polymyxa. 7 Most investigators have
Zealand), prepared as described by Matthews. 11 For con-
utilized traditional batch fermentation technology with
tinuous fermentation experiments the permeate was supple-
freely suspended ceils, but some reports are available des-
mented with CaC12"6H20 (3 g 1-1). Sodium alginate was
cribing newer fermentation technologies. Thus, cells of
purchased from BDH Chemical Co. (Palmerston North,
Enterobacter aerogenes have been immobilized in K-carrag-
heenan for butanediol production from glucose, 8 while a New Zealand), while yeast extract and peptone were ob-
tained from Sigma Chemical Co. (St Louis, USA) and Difco
technique has been described for immobilizing A. hydro-
Laboratories (Detroit, USA), respectively. All media were
phila on titanium hydroxide for production from starch. 9
autoclaved at 121°C for 15 rain, while sodium alginate
Further, a continuous fermentation process incorporating
solutions were autoclaved at 121°C for 10 min.
cell recycle has been described for production from whey
permeate using B. p o l y m y x a . 1° Use of these newer tech-
Cell c u l t i v a t i o n
nologies results in improved reactor productivities and
thus cost savings. The purpose of the present work was to A medium containing (g 1-1): lactose (50), peptone (5),
investigate the continuous production of butanediol from yeast extract (5) and Kz HPO4 (2) was used to culture cells
whey permeate using cells of K. pneumoniae immobilized for immobilization. Inoculation was by direct transfer from
in alginate. Whey permeate was chosen as the substrate a slope into 100 ml medium contained in a 250 ml conical
because of its potential commercial significance, while flask. Incubation was at 30°C on a Lab-line Junior Orbit
K. pneumoniae appears to be a promising organism on this shaker (Lab-line Instruments Inc., Illinois, USA) operating
substrate.6 at 150 rev min-1 .

Cell i m m o b i l i z a t i o n
*Present address: School of Applied Science, Mara Institute of
Cells (18 h old) were harvested by centrifugation and
Technology, Shah Alam, Selangor, Malaysia suspended in sodium alginate (20 g 1=1) in the ratio of 1 ml
~'To whom correspondence should be addressed original culture to 1.25 ml sodium alginate solution. All

0141--0229/86/070409--03 $03.00
© 1986 Butterworth & Co. (Publishers) Ltd Enzyme Microb. Technol. 1986, vol. 8, July 409
Papers
operations were performed aseptically. The suspension was
extruded with the aid o f a peristaltic pump through a 1 ml
pipette into a solution of CaC12"6H20 (0.1 M) in 0.1 M
Tris-HC1 buffer, pH 7.0. The spherical beads, 3 mm dia-
01
8-
meter, were allowed to cure at room temperature for 2 h,
after which they were washed successively with buffer and
distilled water, and then placed in the fermentation equip- ,30 .
ment.
4- 20
=

c
o
Fermentation ¢J

Batch fermentation was performed in 100 ml whey per- 2- to .9

°
meate contained in a 250 ml conical flask. Incubation was _J
at 30°C, either in a static mode with occasional manual
shaking or on a Lab-line Junior Orbit shaker operating at 5'0 16o t~o 2~o
260 3do
150 rev min -~ . Time, h
Continuous fermentation was performed in a packed I__ -I~ -I~ .I. I
column reactor using a glass column of 648 mm length 0.19 0.34 0.77 O. t t
I
x 24 mm internal diameter. After completely filling the Dilution r a t e , h - 1
column with the alginate beads the volume occupied was
approx. 150 ml, leaving a void volume of approx. 145 ml. Figure 1 Continuous production of butanediol f r o m whey per-
meate -- first experiment, o, Butanediol production (g I 1); e, lac-
The reactor was operated as described by Margaritis et al.12
tose consumption (g I-~ )
Thus, a cylindrical steel mesh (1 mm mesh size) was fitted
to the inside of the column which was then operated at an
angle of 10 ° to the horizontal with upward flow of feed
medium. This arrangement facilitated gas removal from the Results

reactor. At start-up, the column was filled with the alginate Initially, experiments were performed in batch fermentation
beads and whey permeate was then pumped in. The reactor to determine the effect of immobilization on butanediol
was operated in batch mode for 4 0 - 4 5 h prior to initiation production by the cells. At the same time, a comparison
of the continuous feed. Temperature was maintained at was made between incubating the flasks in a static mode
30°C by placing the reactor in a constant temperature room and under agitated conditions to assess whether the process
and the whey permeate feed medium was continuously sur- required aeration, as this would influence the choice of reac-
face-flushed with oxygen-free nitrogen gas. A particular tor for the continuous fermentation experiments. Cells
dilution rate was maintained until steady-state conditions from 20 ml original culture were harvested and either
were achieved, and the reactor productivity was calculated immobilized in alginate or used as freely suspended cells.
as the effluent butanediol concentration x dilution rate Maximum production of butanediol was observed after
(based on total column volume). 96 h fermentation and the results are summarized in Table
1. Acetoin and ethanol were produced in addition to
butanediol, with typical product ratios of 10:1.5:0.3,
A nalyses butanediol:ethanol:acetoin. The data demonstrate that
Acetoin, 2,3-butanediol and ethanol were determined immobilization in alginate is a feasible technique for this
using a gas liquid chromatograph (model GC 8A, Shimadzu fermentation process and confirm that butanediol produc-
Corporation, Kyoto, Japan) fitted with a flame ionization tion does not require aeration. Thus, a packed column reac-
detector and a column of 5% F F A P on Chromosorb. The tor was chosen for the continuous fermentation studies.
nitrogen carrier gas flow rate was 30 ml rain -~ . The column Experiments in continuous fermentation were initially
was operated at an initial temperature of 100°C which was attempted using whey permeate without any cation supple-
increased to 240°C at a rate of 32°C rain -x . Quantifica- mentation. However, this invariably led to disintegration
tion was performed using an internal standard o f isobutanol o f the alginate beads. To overcome this problem the whey
permeate was supplemented with CaC12'6H20 (3 g 1-1 ), a
and comparison with standard mixtures.
Lactose was determined as described by Ennis and concentration which prevented disintegration while having
no inhibitory effect on butanediol production. The results
Maddox. la
of two experiments are shown in Figures i and 2. In the
first experiment (Figure 1), the system was operated under
batch conditions for the initial 45 h, during which time a
Table 1 Butanediol production by free and immobilized cells in butanediol concentration of 2.2 g 1-t was obtained. The
static and agitated batch culture continuous feed was then started at a dilution rate of
0.19 h -1 and at steady state a butanediol concentration of
Cells Static culture Agitated culture 7.8 g 1-1 was observed, representing a productivity of
Buta nedi01 Butanediol
1.48 g 1-I h -1 . Further steady states were then achieved
(g I i ) yield (g I -I ) yield at dilution rates of 0.34, 0.77 and 0.11 h -1 , and observed
productivities were 1.39, 2.31 and 0.77 g 1-1 h -1 , respec-
Free 4.1 0.40 2.7 0.27 tively. Throughout the experiment the effluent pH value
Immobilized 8.7 0.50 6.5 0.30
remained approximately constant at pH 5.5.
In the second experiment (Figure 2) the system was
Results were obtained after 96 h fermentation. Yield is expressed as operated under batch conditions for 45 h, after which the
g butanediol/g lactose utilized continuous feed was started at a dilution rate of 0.29 h -1 .

410 E n z y m e M i c r o b . T e c h n o l . 1 9 8 6 , v o l . 8, J u l y
 Butanediol production from whey permeate: H. K. Lee and I. S. Maddox
At steady state, the butanediol productivity was 0.15
g 1-1 h -1, rather lower than expected. On increasing the 0.4
dilution rate to 0.74 h -1 butanediol ceased to be produced.
/x A
These results suggested that during the startup of the con- A
tinuous reactor the dilution rate should not be too high.
2.5 -0.3
Consequently, it was reduced to 0.04 h -1, and at steady
state a productivity of 0.23 g 1-1 h -1 was observed. Sub-
sequent increases in the dilution rate to 0.36, 0.57, 0.86
-6
and 1.40 h -1 resulted in productivities of 0.97, 1.37, 2.0 0.2 >-
1.72 and 1.26 g 1-1 h-] , respectively. Finally, a dilution rate
of 0.14 h-1 was set, resulting in a productivity of "7,
d::
0.95 g 1-1 h-1. This dilution rate was maintained for a
"T.
period of six weeks, during which time the productivity 1.5 0.1
remained approximately constant.
A summary of the data obtained for butanediol produc-
tivity and yield at the various steady states is shown in "o
o 1.0 -0
Figure 3. Although there appear to be some discrepancies
in the productivity data, probably because the values were
obtained from two different experiments, the general trends
are clear. Thus, maximum productivity was observed in the 0.fi
dilution rate range of 0 . 7 7 - 0 . 8 6 h -1 . At the upper end of
this range, however, a decrease in yield was apparent. The
reason for the discrepancies between the two experiments
is not clear, but they may reflect differences in the bio- o I I I I
0 0.4 0.8 1.2 1.6
mass loadings of the alginate beads. No determinations were
Dilution rate, h 1
made of the loadings.
Figure 3 Butanediol productivity and yield at various steady
Discussion states in continuous fermentation. Productivity; o, first experiment;
e, second experiment; yield (g butanediol/g lactose utilized): z~ (first
The aim of this study was to investigate the continuous pro- experiment), • (second experiment). The order in which the various
duction of butanediol from rennet whey permeate using steady states were obtained is shown in Figures I and 2
immobilized K. pneumoniae. The reactor used was chosen
because of its suitability for anaerobic fermentation pro- batch fermentation on whey permeate, reported an over-
all productivity of 0.08 g 1-1 h -1 , compared to the maxi-
cesses and because of its reported ability to cope with the
m u m obtained in the present study of 2.31 g 1-1 h -1 . Chua
large volumes of gas that are often generated during such
et al., 8 using immobilized E. aerogenes in a continuous
processes. 12 During the experiments described above no
stirred tank reactor, reported a productivity of 0.75 g 1-1
problems were encountered with gas removal or blockage
h -1 from a glucose substrate, while Shazer and Speck-
resulting from excessive cell growth. Cells were continuously
man, 1° using a stirred tank reactor with cell recycle, achieved
lost from the alginate beads, as judged by microscopical
a productivity of 1.04 g 1-1 h -t from whey permeate using
examination of the effluent, but their numbers were almost
B. p o l y m y x a . Thus, the use of alginate-immobilized K.
certainly replenished. No attempts were made to determine
p n e u m o n i a e in a packed column reactor appears to be a
the biomass content of the beads. The system proved to be
promising technique for the production of butanediol from
stable during a total of seven weeks of continuous operation
whey permeate.
and examination of the beads after this time revealed that
their structural integrity was maintained.
The reactor productivities obtained in this study compare References
favourably with those reported by other workers. Lee and
Maddox, 6 using free cells of K. p n e u m o n i a e in traditional 1 Jansen, N. B. and Tsao, G. T. Adv. Biochem. Eng. 1983, 27,
83-99
8- 2 Prescott, S. C. and Dunn, C. G. Industrial Microbiology,
McGraw Hill, New York, 1959, pp. 399-427
3 Willets,A. Biotechnol. Lett. 1984, 6,263-268
4 Yu, E. K. C. and Saddler, J. N. Biotechnol. Lett. 1982, 4,
6 - -30 .
. 0
121-126
5 Laube, V. M., Groleau, D. and Martin, S. M. Biotechnol.
Lett. 1984, 6,257-262
~ 4. -2o ~ 6 Lee, H. K. and Maddox, I. S. Biotechnol. Lett. 1984, 6,815-
~ g
I~ "J 818
7 Speckman, R. A. and Collins, E. B. Appl. Environ. Microbiol
2- -lo 2u 1982, 43, 1216-1218
8 Chua, J. W., Erraslan, A., Kinoshita, S. and Taguchi, H. J.
Ferment. Technol. 1980, 58, 123-127
9 Willets,A. Biotechnol. Lett. 1985, 7, 261-266
sb ld0 1go 26o 2go 360 10 Shazer, W. H. and Speckman, R. A. J. Dairy Sci. 1984, 67,
Time, h supplement 1, 50-51
11 Matthews,M.E.,N. Z. J. Dairy ScL Technol. 1978, 13, 149-
I- - I I~ -I- -I~-I~I=-I = 156
0.29 0.74 0.04 0.36 0.570.86 1.4 0.14
12 Margaritis, A., Promod, K., Bajpai, P. and Wallace, B. Bio-
Dilution rate, h -1 technoL Lett. 1981, 3,613-618
Figure 2 Continuous production of butanediol from whey perme- 13 Ennis, B. M. and Maddox, I. S. Biotechnol. Lett. 1985, 7,
ate -- second e x p e r i m e n t . S y m b o l s as f o r Figure 1 601-6O6

E n z y m e M i c r o b . T e c h n o l . 1986, vol. 8, J u l y 411